Contents

PREFACE xiii

PART ONE CYCLE THE HYDROLOGIC 1 CHAPTER lntroduction

1.1 t.2

r.3
t.4 1.5 1.6 I.7 1.8

3 Hydrology Defined 3 A Brief History 5 The Hydrologic Cycle 5 The Hydrologic Budget 11 HydrologicModels 11 HydrologicData 12 Common Units of Measurement to Environmental Problems Application of Hydrology

t2

2 CHAPTER Precipitation 1 5
2.1 2.2 2.3 2,4 2.5 2.6 2.7
15 Water Vapor 17 Precipitation Distribution of the Precipitation Input 27 Point Precipitation 29 Areal Precipitation 34 Precipitation ProbableMaximum 36 Grossand Net PreciPitation

vi

coNTENTS

CHAPTER 3 Interception Depression and Storage

3.1 3.2 3.3 Interception 40 Throughfall 44 Depression Storage 45

40

CHAPTER 4 Infiltration 52
4.I 4.2 4.3 4.4 4.5 4.6 4.7 4.8 4.9 4.10 Measuring Infiltration 53 Calculation of Infiltration 53 Horton's Infiltration Model 57 Green-AMPT Model 64 Huggins-MonkeModel 67 Holtan Model 68 Recoveryof Infiltration Capacity 69 Temporal and Spatial Variability of Infiltration Capacity SCS Runoff Curve Number Procedure 73 76 @Index

70

CHAPTER 5 Evaporation Transportation 82 " and

5.1 5.2 5.3 5.4 5.5 5.6 5.7 Evaporation 86 EstimatingEvaporation 86 EvaporationControl 95 Transpiration 95 TranspirationControl 100 Evapotranspiration 100 EstimatingEvapotranspiration

103

CHAPTER 6 Streamflow 111

6.1 6.2 6.3 6.4 6.5 6.6 6.7 6.8 DrainageBasinEffects 111 The Hydrograph 11,2 Units of Measurement Streamflow for 113 Measuringand RecordingStreamflow 113 Measurements Depth and Cross-Sectional of Area II4 Measurement Velocity of lI4 RelatingPoint Velocity to Cross-Sectional Flow Velocity The Slope-AreaMethod for DeterminingDischarge II7

115

oONTENTS Vii

PART TWO AND fT'IEASUREMENTS MONITORING 121 HYDROLOGIG 7 CHAPTER DataSources Hydrologic
7.1 7.2 7.3 7.4

123

I23 GeneralClimatologicalData 123 Precipitation Data 124 StreamflowData Evaporationand TranspirationData

I24

8 CHAPTER 126 fnstrumentation

8.1 8.2 8.3 8.4 126 Introduction 127 HYdrologicInstruments 135 TelemetrySYstems 135 RemoteSensing

9 CHAPTER Networks 144 Monitoring The Purposeof Monitoring 9.r

9.2 9.3 9.4

144

I45 SpecialConsiderations I47 in Uie of ComPuters Monitoring 147 Networks Hydrological-Meteorlogical

PART THREE WATERHYDROLOGY151 SURFACE 1O CHAPTER andthe Catchnient Runoff

10.1 tO.2 10.3 IO.4 10.5

153

153 and Watersheds, DrainageBasins Catchments, 155 Affecting Runoff Basin Characteristics RudimentaryPrecipitation-RunoffRelationships 164 166 StreamflowFrequencyAnalysis 168 Forecasting StreamflQw

1 CHAPTER 1 Hydrographs 171

11.1 Il.2 11.3 HYdrograPhs 171 Streamflow FactorsAffecting HydrographShape ComPonents 174 HydrograPh 172

viii

ooNTENTS lI.4 11.5 11.6 Il.7 BaseFlow Separation I77 HydrographTime Relationships Time of Concentration I82 BasinLae Time I82 181

CHAPTER 12 UnitHydrographs 188

l2.I 12.2 12.3 12.4 12.5 12.6 Unit Hydrograph Definition 188 Derivation of Unit Hydrographsfrom StreamflowData 190 Unit HydrographApplications by Lagging Methods I94 S-Hydrograph Method 198 The Instantaneous Unit Hydrograph 201 SyntheticUnit Hydrographs 205

CHAPTER 13 Hydrograph Routing

13.1 13.2 13.3

234

HydrologicRiver Routing 235 HydrologicReservoirRouting 245 Hydraulic River Routing 248

CHAPTER 14 SnowHydrology
I4.l I4.2 I4.3 I4.4 14.5 14.6

265

Introduction 265 Snow Accumulation and Runoff 267 Snow Measurements and Surveys 268 Point and Areal Snow Characteristics 269 The SnowmeltProcess 271. SnowmeltRunoff Determinations 284

CHAPTER 15 Urbanand SmallWatershed Hydrology 309

15.1 15.2 i5.3 15.4 Introduction 309 PeakFlow Formulasfor Urban Watersheds 311 PeakFlow Formulasfor Small Rural Watersheds 33I Runoff Effects of Urbanization 344

CHAPTER 16 Hydrologic Design

16.l 16.2

359

Hydrologic DesignProcedures 360 Data for HydrologicDesign 363

CONTENTS

IX

16.3 16.4 16.5 t6.6 16.7 16.8

HydrologicDesign-Frequency Criteria DesignStorms 373 Critical EventMethods 391 Airport DrainageDesign 400 Designof Urban Storm Drain Systems FloodplainAnalysis 409

365

402

PART FOUR GROUNDWATER HYDROLOGY 425 CHAPTER 17 Groundwater, Soils,and Geology

l7.l I7.2 I7.3 I7.4 I7.5 I7.6

427

Introduction 427 Groundwater Flow-General Properties 429 Subsurface Distribution of Water 429 GeologicConsiderations 430 Fluctuationsin GroundwaterLevel 433 Groundwater-Surface Water Relations 433

CHAPTER 18 Mechanics Flow of

18.1 t8.2 18.3 18.4 18.5 18.6

435

Hydrostatics 435 GroundwaterFlow 436 Darcy's Law 436 Permeability 438 Velocity Potential 440 HydrodynamicEquations 441 r8.7 ' Flowlines and EquipotentialLines 18.8 BoundaryConditions 447 18.9 Flow Nets 449 1 8 . 1 0 VariableHydraulic Conductivity 1 8 . 1 1 Anisotropy 452 18.t2 Dupuit's Theory 453

444

451

CHAPTER 19 Wellsand Collection Devices

19.1 19.2 19.3 19.4 19.5

460

Flow to Wells 460 SteadyUnconfinedRadial Flow Toward a Well 461 SteadyConfined Radial Flow Toward a Well 462 Well in A Uniform Flow Field 463 Well Fields 465

CONTENTS

19.6 I9.7 19.8 I9.9 19.10 19.ll 19.12

The Method of Images 466 UnsteadyFlow 467 Leaky Aquifers 4'13 Partially PenetratingWells 473 Flow to an Infiltration Gallery 473 Saltwater Intrusion 474 GroundwaterBasin Development 475

CHAPTER 20 Modeling Regional Groundwater Systems

20.I 20.2 20.3 20.4 20.5 RegionalGroundwaterModels 481 Finite-DifferenceMethods 484 Finite-ElementMethods 493 Model Applications 494 GroundwaterQuality Models 500

481

PART FIVE HYDROLOGIC MODELING

505 5O7

CHAPTER 21 Introduction Hydrologic to Modeling

2l.I 2t.2 21.3 21.4

HydrologicSimulation 508 Groundwater Simulation 509 Hydrologic Simulation Protocol 524 Corps of EngineersSimulation Models

526

CHAPTER 22 Synthetic Streamflows 535

22.I 22.2 SyntheticHydrology 536 Serially DependentTime SeriesAnalysis 539

CHAPTER 23 Continuous Simulation Models

23.1 23.2

548

Continuous Streamflow SimulationModels 549 ContinuousSimulation Model Studies 570

CHAPTER 24 Single-Event Simulation Models

24.1 24.2 24.3

594
597

StormEventSimulation 594 Models FederalAgency Single-Event Storm SurgeModeling 625

CONTENTS Xi

CHAPTER 25 Urban Runotf Models Simulation

25.1 25.2 25.3

630

Urban StormwaterSystemModels 63I Urban Runoff Models Compared 659 Vendor-DevelopedUrbanStormwaterSoftware 663

PART SIX METHODS STATISTICAL

669 671

26 CHAPTER Probability Statistics and

26.1 26.2 26.3 26.4 26.5 26.6 26.7 26.8 26.9

2 6 .r0

RandomVariablesand StatisticalAnalysis 672 Concepts Probability of 673 ProbabilityDistributions 676 Moments of Distributions 681 Distribution Characteristics 682 Types of Probability Distribution Functions 685 ContinuousProbabilityDistributionFunctions 685 Bivariate Linear Regression and Correlation 690 Fitting Regression Equations 692 Regression Correlation Applications 697 and

CHAPTER 27 Frequency Analysis

27.1 27.2 27.3 27.4 27.5 27.6 27.7

708

FrequencyAnalysis 708 GraphicalFrequencyAnalysis 709 FrequencyAnalysis Using FrequencyFactors RegionalFrequencyAnalysis 7I9 Reliability of FrequencyStudies 730 FrequencyAnalysis of Partial Duration Series Flow Duration Analysis 737

7Il

734

APPENDICES 751 INDEX 757

Preface

New federalthrusts,the growis Watermanagement taking on new dimensions. public sentimentregardingenvironmentalprotecing list of global iisues,and strong tion havebeen the principal driving forces. and development manageIn the early yearsof the 20th century,waterresources on water supplyand flood control' Today,these ment were focuied almostexclusivd issuesare still important, but protecting the environment,ensuring safe drinking -and compete equally for recriatioinal experiences water, and providing aesthetic for public is pressing an environmentallyconscious attentionand funds.Furthermore, practices,with fewer structural compomanagement greaterreliance on improved to ients, to solve this nuiion'. water problems.The notion of continually striving replacedby one of husbandingthis preciousnatural provide more water has been resource. There is a growing constituencyfor allocatingwater for the-benefitof fish and wildlife, for protection-of marshesand estuary areas' and for other natural system uses.But estimatingthe quantitiesof water neededfor environmentalprotectionand is for maintainingand/or restoringnatural systems difficult, and there are still many of are ,putt", and our understanding the complexinteracunknowns.Scilntific data is of an scales rudimentary.Indeed,this is a critical issue' tions inherentin ecosystems sincethe quantitiesof water involvedin environmentalprotection can be substantial of and competitionfor thesewatersfrom traditional water usersis keen' The nations decisionsregardingnatural systems-decisions that are the world are facing major urgencyassociladenwith significantectnomic and social impacts.Thus there is_an and of ecologicsystems of their hydrologic a atedwith developing betterunderstanding components. Water policies of the future must thereforetake on broader dimensions'More and mustbe placedon regionalplanningand management, regionalinstituemphasis mustbe practicedat, Watermanagement be devised. this tions to accommodate muJt Land useand wateruseplanningmustbe more andbetween,all levelsof government. tightly coordinatedas well'

XIV

PREFACE

Water scientistsand engineersof tomorrow must be equipped to addressa diversity of issuessuch as: the design and operation of data retrieval and storage systems; forecasting;developingalternativewater use futures; estimatingwater requirementsfor natural systems; exploringthe impactsof climate change;developing more efficient systems applyingwaterin all water-usingsectors; for and analyzingand designingwater managementsystemsincorporating technical, economic, environmental, social, legal, and political elements. knowledgeof hydrologicprinciplesis A a requisitefor dealingwith suchibsues. This fourth edition hasbeendesigned meet the contemporaryneedsof water to scientistsand engineers. is organizedto accommodate It studentsand practitioners who are concernedwith the development,management,and protection of water resources. The format of the book follows that of its predecessor, providing material for both an introductory and a more advanced course. Parts One through Four provide the basicsfor a beginninglevel course,while Parts Five and Six may be used for a more advancedcourseon hydrologicmodeling. This fourth edition has been updated throughout, and many solved examples havebeen added.In addition, new computer approaches have been introduced and problem-solvingtechniques includethe use of spreadsheets appropriate.New feaas turesof eachchapterincludean introductory statement contentsand,at the concluof sion of the chapter,a summaryof key points. Many sourceshave been drawn upon to provide subject matter for this book, and the authorshope that suitable acknowledgment has been given to them. Colleagues and students recognizedfor their helpful commentsand reviews,parare ticularly the following reviewers. Gert Aron, ThePennsylvania StateUniversity JohnW. Bird, Universityof Nevada-Reno IstvanBogardi, Universityof Nebraska RonaldA. Chadderton, VillanovaUniversity RichardN. Downer,Universityof Vermont Bruce E. Larock, Universityof Califurnia-Davis Frank D. Masch,Universityof Texas-San Antonio Philip L. Thompson, FederalHighwayAdministration A specialnote of thanks is due to Dr. John W. Knapp, President the Virginia of Military Institute,coauthorof previouseditionsof this book, for his pastcontributions andvaluableguidance. WarrenViessman. Jr. Gary L. Lewis

PARTONE

CYCLE THE HYDROLOGIC

L.

Chapter1

lntroduction

I Prologue
The purPoseof this chaPteris to: . , . . Define hydrology. . earth science' -,.1L Give a brief niJiory of the evolution of this important Statethe fundamentalequationofhydrology' decision trow ffiofogic principle, "urib" appliedto supplement Demonstrate management' for support systems water and environmental

DEFINED 1.1 HYDROLOGY

Hydrologyisanearthscience'Itencompassestheoccuffence'distribution,moveA knowledgeof hydrologyis fundamenr, and propertiesof the watersof the earth. mentaltodecisionmutingp,o.",,e,*he,ewaterisu"ompon"nto.f.th.esystemof inextricably linked' and it is important concern.water and environmentalissuesare toclear$understandhowwaterisaffectedbyandhowwateraffectsecosystem maniPulations'

1.2 A BRIEFHISTORY
focusedtheir i Ancient philosophers flows Production of surfacewater oc"ur."n"e of water in variousstag to from the seato the atmosPhere t faulty'l early speculationwas often reservoirsth of large subterranean is interestingto note, however'tha
*suoeriornumbersindicatereferencesattheendofthechapter.

CHAPTER1

INTRODUCTION

Greek aqueducts both conveyance on crosssectionand velocity.This knowledgewas lost to the Romans,and the proper relation betweenarea,velocity, and rate of flow remainedunknownuntil Leonardoda Vinci rediscovered duringihe it Italian Renaissance. During the first century s.c. Marcus Vitruvius, in Volume 8 of his treatise De Architectura Libri Decem (the engineer'schief handbookduring the Middle Ages), setforth a theory generallyconsidered be the predecessor modern to of notionsof the hydrologiccycle. He hypothesized that rain und ,no* falling in mountainousareas infiltrated the earth's surface and later appearedin the lowlands as streamsand springs. In spiteof the inaccurate theoriesproposed ancienttimes,it is only fair to state in that practical applicationof varioustry-orotogic principleswas often carried out with considerable success. example,about4000 s.c. u du- was constructed For acrossthe Nile to permit reclamation of previously barren lands for agricultural production. Severalthousandyears later a canal to convey fresh water from Cairo io Suezwas built. Mesopotamian towns were protecteduguinrt floodsby high earthenwalls. The Greek and Roman aqueducts and early Chineseirrigation and flood control works were also significantprojects. Nearthe endof the fifteenth century the trend towarda more scientific approach to hydrology based on the observationof hydrologic phenomena becameevident. Leonardo da Vinci and Bernard Palissyindepende-ntly reachedan accurateunderstandingof the watercycle.They apparentlybised theii theoriesmore on our"*iion than on purely philosophical reasoning. Nevertheless, until the seventeenth century it seemsevident that little if any effort was directed toward obtaining quantitative measurements hydrologicvariables. of The adventof what might be calledthe "modern" science hydrology of is usually considered beginwith the studiesof suchpioneersasPerrault,Mariotte, to and Halley in the seventeenth century.r'a Perraultobtainedmeasurements rainfall in the Seine of River drainagebasin over a period of 3 years. Using these and measurements of runoff, and knowing,thedrainage areasize,he showeJthat rainfall was adequate in quantity to accountfor river flows. He also mademeasurements of evaporati,on and capillarity. Mariotte gaugedthe velocity of flow of the River Seine. Recordedvelocities were translatedinto termsof dischirgeby introducingmeasurements of the river crosssection'The English astronomer Halley measured rate of evaporation the the of Mediterranean Seaand concludedthat the amountof water evaporated was sufficient to accountfor the outflow of rivers tributary to the sea.Measurements suchas these, althoughcrude,permitted reliable conclusions be drawn reggrding to the hydrologic phenomena being studied. brth numerousadvances hydraulic theory in zometer,the Pitot tube, Bernoulli's theorem, ples.8 perimental hydrology flourished.Significant ydrology and in the measurement surface of water. Such significantcontributionsas Hagen-Poiseuille'scapillary flow equation, Darcy's law of flow in porous media, und th" Dupuit-Thiem well formula were evolved'e-lrThe beginningof systematicstream guoling can also be traced to this period' Although the basis for modern hydrology wui tirrr established the ninein ,)

BUDGET 1.4 THE HYDROLOGIC

of teenth century, much of the effort was empirical in nature. The fundamentals In or not yet beenwell established widely recognized. the early physicalhydtotogyhad of years of tle twJntieth ""niury the inadequacies many earlier empirical formulabegan to known. As a result, interestedgovernmentalagencies tions becamewell From about 1930to 1950,rational programsof hydrologicresearch' their own develop analysis began to ieplace empiricism.3 Sherman's unit hydrograph, Horton's to infiltration theory, und Th"it's nonequilibrium-approach well hydraulicsare outof the great progressmade'r2-'o standingexamples Since 1930 a theoreiical approachto hydrologicproblemshas largely replaced methods of ttre past. Advancesin scientific knowledgepermit a less sophisticated ' ofthe physicaibasisofhydrologic relations,and the adventand better understanding in of continueddevelopnient high-speeddigital computershavemadepossible, both mathematicalmanipulationsthat would iense,extensive a practical and an economic havebeen overwhelmingin the past. For a more compiehensivi historical treatment, the reader is referred to the and their co-workers'1'2'4'5'15 works of Meinzer,Jonls, Biswas,

CYCLE 1 . 3THE HYDROLOGIC

the The hydrologiccycle is a continuousprocessby which water is transportedfrom exist' The to the landind back to the sea.Many subcycles oceansto the atmosphere precipitation over land beforereturnevaporationof inlan-dwater and its subsequent The driving force for the global watertransportsystem example. ingio the oceanis one Note that is providedby the sun,which furnishesthe energyrequiredfor evaporation' through the cycle; for example,sea changesduring passage the water quality also water is convertedto fresh water through evaporation' The completewater cycle is global in nature. world water problems require Practical studieson regional,national,internitional, continental,and global scales.16 that the total supply of fresh water availableto the earth is significanceof the fact has limited and very small compared with ihe salt water content of the oceans at the receivedlittle attention.Thus watersflowing in one country cannotbe available u's' same time for use in other regions of the world. Raymond L' Nace of the are resources a global problem with thatoowater GeologicalSurvey has aptly sta=ted hydrologistsare obligated to cope with problems requiring local roots."tu Mtdern developing definition in varying scalesof oider of magnitudedifference.In addition, weather must receive careful attention, since climatological techniquesto contiol water changesin one area can profoundly affect the hydrology and therefore the of resources other regions.

BUDGET 1 . 4 THE HYDROLOGIC

the Because total quantity of water availableto the earth is finite and indestructible, subsysthe global hydrolojic ,yrt"* may be lookedupon as closed.Open hydrologic system' For and temsare abundantlhowever, theseare usuallythe type analyzed' any ' to can be developed accountfor the hydrologiccomponents' a water budget

CHAPTER1

INTRODUCTION

FiguresI'I,I.2, and 1.3 showa hydrologic budgetfor the coterminous United States, conceptualized a hydrologiccycle,andthe distributionof a precipitationinput, respectively. Thesefiguresillustrate the components the water cycle with which a of hydrologist concerned. a practicalsense, is In somehydrologicregionis dealtwith and a budgetfor that region is established. Suchregionsmay be topographicallydefined (watersheds river basinsare examples), and politically specified(e.g- couniy or city limits), or chosenon some other grounds. Watersheds drainagi tasins are the or easiest deal with sincethey sharply define surfacewater boundaries. to Thesetopographically determinedareasare drainedby a river/streamor systemof connecting rivers/streamssuch that all outflow is discharged through a single outlet. Unfortunately,it is often necessary deal with regions that are not well suitedto tracking to hydrologiccomponents. theseareas, hydrologist For the will find hydrologicbudgeting somewhatof a challenge. The primary input in a hydrologicbudgetis precipitation.Figures 1.1-1.3 illustrate this. Someof the precipitation (e.g.,rain, snow,hail) may be interceptedby trees,grass,other vegetation, and structuralobjectsand will eventuallyreturn to the atmosphere evaporation. by Onceprecipitationreaches ground,someofit may fill the (becomedepression depressions storage), part may penetralethe ground (infiltraie) to replenishsoil moisture and groundwaterreservoirs,and some may become surface runoff-that is, flow over the earth's surfaceto a definedchannelsuchas a stream. Figure 1'3 showsthe dispositionofinfiltration, depression storage, surface and runoff.

and vegetation

'1.'

*r-d{i;

Consumptive use 100bgd

bgd = billion gallons per day

Figure 1.1 Hydrologic budgetof cotermiriousunited States.(U.S. Geologicalsurvey.)

BUDGET 7 1.4 THEHYDROLOGIC

Clouds and water vaPor Clouds and water vaPor

" 1 ) t t l l ,
p\--

ffi

--1vtivv
P P P P P P

t t
E' evaporation;P' Figure 1.2 The hydrologiccycle: ?, transpiration; 'surfac-e groundwater flow; and I' runoff; G, R, p.Sqipi*i"tt inflltration.

t Precipitation inPut (hyetogaph)

'l,r\
t SSeamflow (hyclrograPh)

input' of Figure 1.3 Distribution precipitation

CHAPTER1

INTRODUCTION

water enteringthe ground may take severalpaths.Somemay be directly evaptransferfrom the soil to the surfaceis maintained.This can easily oratedif adequate a high groundwater table (free water surface) is within the limits of occur where usingsoil moistureor groundwaVegetation capillary transportto the ground surface. by can also transmit infiltrated water to the atmosphere a processknown tei directly transpiration.Infiltrated water may likewise replenishsoil moisture deficiencies as and enter storageprovided in groundwaterreservoirs,which in turn maintain dry weather streamflow.Important bodies of groundwaterare usually flowing so that inflltrated water reachingthe saturated,on" muy be transportedfor considerable' Groundwatermovement is subject, of course,to distancesbefore it is discharged. physicaland geologicalconstraints. will eventually evaporateor infiltrate the ground I Water storedin depressions minor channels(gullies, rivulets, and the : runoff otti-ut"ty reaches surface.Surface an and rivers,and finally reaches ocean.Along the course major streams like), flowsto evaporationand infiltration can also occur. of a stream, The foregoing discussion suggeststhat the hydrologic cycle, while simple in complex.Pathstaken by particlesof water precipiconcept,is actually exceedingly are numerousand varied before the seais reached.The time scale tated in any arca minutes,days,or years. may be on the order of seconds, basedon theprocessesillusgeneralhydrologicequationcan be developed A 1.2 and 1.3. ConsiderFig. 1.4. In it, the hydrologicvariablesP, E, T, trated in Figs. R, G, and l are as definedin Fig. 1.2. Subscriptss and g are introduced to denote R, respectivd. For example, vectorsoriginatingaboveand belowthe earth's surface,

["
Earth's surface Surface channels

R2

Level of plastic rock . (no water below this level)

Figure 1.4 Regionalhydrologiccycle.

BUDGET 1.4 THE HYDROLOGIC

signifies groundwaterflow that is effluent to a surface streamoand E, represents areas.Letter S stands evaporationfrom surfacewaterbodiesor other surfacestorage as specified A hasa lower boundarybelow regionunderconsideration The for storage. which water will not be found. The upper boundary is the earth's surface.Vertical boundsare arbitrarily set asprojectionsof the peripheryof the region.Remembering in, that the water budgetis a balancebetweeninflows, outflows,and changes storage, whereall values statements, be translatedinto the following mathematical Fig. 1.4can are given in units of volume per unit time: 1. Hydrologicbudgetabovethe surface P+R1 -RrIRr-E"-7,2. Hydrologicbudgetbelow the surface I + Gt- G2- Rr- E, - 4: AS, (1.2)

1:AS"

(1.1)

3. Hydrologicbudgetfor the region (sum of Eqs' 1.2 and 1.3) p - (Rr- R,) - (E" + E) - (r" + Tr) - (Gr- G,) : a(S, + ss), (1.3) refer If the subscripts droppedfrom Eq. 1.3 sothat letterswithout subscripts are valuesof surfaceflow, undergroundflow, evaporation, to total precipitation and net the transpiration,and storage; hydrologicbudgetfor a regioncan be written simply as p_R-G_E_T:LS (1.4)

whereterms For This is the basicequationof hydrology. a simplifiedhydrologicsystem G, E, and Z do not apply, Eq. 1.4 reducesto p-R:AS (1.5)

of Equation 1.4 is applicableto exercises any degreeof complexity and is therefore all hydrologicproblems. basic to the solution of The difficulty in solvingpractical problemslies mainly in the inability to measure or estimateproperly the various hydrologic equation terms. For local studies, often are made,but on a global scaleqqantificationis usuallycrude. reliableestimates locatedthroughoutan area.Surface by Precipitationis measured rain or snowgauges various devicessuchas weirs, flumes,velocity meters, using flows can be measured of locatedin the rivers and streams the area.Under goodconditions and depthgauges percent or more accufate,but large floods cannot be these measurementsare 95 Soil directly by current methodsand dataon sucheventsare sorelyneeded. measured usingneutronprobesand gravimetricmethods;infiltration moisturecan be measured can be deterrnined locally by infiltrometers or estimated through the use of of precipitation-runoff data. Areal estimates soil moistureand infiltration are generare The extentandrate of movementof groundwater usually illy very crude,however. are dataon quantitiesof groundwater determine, and adequate difficult to exceedingly for of the geologyof aregion is essential groundwater Knowledge not alwaysavailable. estimatesif they are to be more than just rough guides.The determinationof the

1O

CHAPTER1

INTRODUCTION

quantities of water evaporatedand transpired is also extremely difflcult under the are of Most estimates evapotranspiration the presentstateof developmentiof science. or masstransfermethods, empirpans,energybudgets, obtainedby usingevaporation ical relations.A predicamentinherent in the analysisof large drainagebasinsis the fact that rates of evaporation,transpiration,and groundwatermovementare often to assumed be highly heterogeneous. that it can The hydrologicequationis a usefultool; the readershouldunderstand the magnitude and time distribution of be employed in various ways to estimate hydrologicvariables.An introductory exampleis given here,and otherswill be found throushout the book. EXAMPLE 1.1 The average 20 wabrshedreceived in. of precipitation. In a given year,a 10,000-mi2 measured the river drainingthe areawasfound to be 700 cfs (cubicfeet in rate of flow and of per second). Make a rough estimate the combinedamountsof waterevaporated the region during the year of record. transpiredfrom Solution. Beginningwith the basic hydrologicequation P - R - G - E - Z : A S can and and sinceevaporation transpiration be combined, ET:P-ft-G-AS (1.6)

(r.4)

{4
t:

.+ v' a'E-;.

The term EZ is the unknown to be evaluatedand P and R are specifled.The equation thus has flve variables and three unknowns and cannot tre solved without additional information. In order to get a solution, two assumptionsare made. First, since the in drainageareais quite large (measured hundredsof squaremiles),a presumpdivide is probadivide (boundary)follows the surface tion that the groundwater In bly reasonable. this casethe G componentmay be consideredzero. The vector R, existsbut is included in R. The foregoingassumptionis usually not valid forsmall areasand mustthereforebe usedcarefully.It is alsopresupposed that AS : 0, thus implying that the groundwaterreservoir volume has not changedduring the year. For such short periods this assumptioncan be very inaccurate,evenfor well-wateredregionswith balancedwithdrawalsand good is potentials.In arid areaswheregroundwater beingmined (AS consisrecharge Neverin supposition many cases. it tently negative), would be an unreasonable qualified by is the theless, assumption made here for illustrative purposesand sayingthat pastrecordsof waterlevelsin the areahaverevealedan approximate Hydrology is not an exact science,and reaconstancyin groundwaterstorage. are sonablewell-foundedassumptions required if practical problemsare to be solved. Using the simplificationsjust outlined, the working relation reducesto ET:P_R

'I1 1.6 HYDROLOGIC DATA which canbe solveddirectly.First, change into inchesper yearsothat the units R are compatible:
_ ft3
r \ t

1
. ,

sec

sec

area (m n-l

yt

II

^ :R,in.

R _

7 0 0 x 8 6 ; 4 0 0 x 3 6 5 x 1: 2
104x (5280)'

0.95in.

Therefore, : 20 - 0.95 : 19.05in./yr. ET The amountof evapotranspiration the year in questionis estimatedto for be 19.05in. This is admittedlya crudeapproximation could serveasa useful but guide for water resources planning. ll

1.5 HYDROLOGIC MODELS

Hydrologic systemsare generally analyzedby using mathematicalmodels.'These modelsmay be empirical, statistical,or foundedon known physicallaws.They may be usedfor suchsimplepurposes determiningthe rate of flow that a roadwaygrate as mustbe designed handle,or they may guidedecisions to aboutthe bestway to develop a river basinfor a rnultiplicity of objectives. The choiceof the modelshouldbe tailored to the purposefor which it is to be used.In general,'thesimplestmodel capableof producinginformation adequate deal with the issueshouldbe chosen. to Unfortunately,most waterresources systems practicalconcernhavephysical, of social,political, environmental, and cannotbe andlegaldirnensions; their interactions in exactly described mathematicalterms. Furthermore,the historical data necessary for meaningful hydrologic analysesare often lacking or unreliable. And when one considersthat hydrologic systems generallyprobabilistic in nature, it is easyto are understand that the modeler'stask is not a simpleone.In fact, it is often the casethat of the best that can be hoped for from a model is an enhancedunderstanding the systembeing analyzed.But this in itself can be of great value, leading,for example, to the implementationof datacollectionprogramsthat canultimately supportreliable modelingefforts. For the most part, mathematicalmodels are designedto describethe way a system's elements respondto sometype of stimulus(input). For example,a model of 'a groundwater system might be developed demonstrate effectson groundwater to the storageof various schemes pumping. Equations 1.1 and L2 are mathematical for modelsof the hydrologicbudget,and Figure 1.3 can be considered pictorial model a of the rainfall-runoff process. later chapters, variety of hydrologicmodelswill be In a presented and discussed. Thesemodelsprovidethe basisfor informed watermanagement decisions.

DATA 1.6 HYDROLOGIC

evaporation;soil Hydrologic dataarc neededto describeprecipitation; streamflows; moisture; snow fields; sedimentation;transpiration;infiltration; water quality; air, s9i!, and water temperatures; and other variablesor componentsof hydrologicsys-

12

CHAPTER 1

INTRODUCTION

tems. Sources of data are numerous,with the U.S. Geological Survey being the primary one for streamflow and groundwaterfacts. The National Weather Service (NOAA or National Oceanicand AtmosphericAdministration)is the major collector and of meterologicdata.Many other federal,state,and local agencies other organizations also compile hydrologicdata. For a completelisting of theseorganizationssee Refs.3 and 17.

OF 1 . 7 COMMON UNITS MEASUREMENT

cubic Streamandriver flowsare usuallyrecordedascubic metersper second(m3/sec), (sec-ft);groundwater flowsand watersupplyflo\i/s feetper second(cfs),or second-feet are commonly measuredin gallons per minute, hour, or day (gpm, gph, gpd), or millions of gallonsper day (mgd); flowsusedin agricultureor relatedto water storage are often expressedas acre-feet (acre-ft), acre-feet per unit time, inches (in.) or per centimeters(cm) depth per unit time, or acre-inches hour (acre-in./hr). Volumesare often given as gallons,cubic feet, cubic meters,acre-feet,secondfoot-days,and inchesor centimeters.An acre-footis equivalentto a volume of water (cfs-day,sfd) is the 1 ft deep over 1 acre of land (43,560 ft3). A second-foot-day accuinulated volumeproducedby a flow of 1 cfs in a24-hr period.A second-foot-hour (cfs-hr) is the accumulatedvolume produced by a flow of 1 cfs in t hr. Inches or centimeters depthrelate to a volume equivalentto that many inchesor centimeters of useful it of water over the areaof concern.In hydrologicmassbalances, is sometimes to note that 1 cfs-day : 2 acre-feetwith sufficient accuracyfor most calculations. Rainfall depthsare usually recordedin inchesor centimeterswhereasrainfall rates are given in inches or centimetersper hour. Evaporation,transpiration, and infiltration rates are usually given as inchesor centimetersdepthper unit time. Some usefulconstantsand tabulatedvaluesof severalof the physicalpropertiesof water are given in Appendix A at the end of the book.

'

PROBLEMS 1.8 APPLICATIONOF HYDROLOGYTO ENVIRONMENTAL

It is true that humanscannot exist without water; it is also true that water, mismanaged,or during times of deficiency(droughts),or times of surplus(floods),can be life threatening.Furthermore,there is no aspectof environmentalconcernthat doesnot relate in someway to water. Land, air, and water are all interrelatedas are water and all life forms. Accordingly, the spectrum of issuesrequiring an understandingof is hydrologicprocesses almost unlimited. the As waterbecomes more scarceand as competition for its useexpands, need for improved water managementwill grow. And to provide water for the world's food production, recreational expandingpopulation, new industrial developments, and other purdemands,and for the preservationand protection of natural systems poses, will becomeincreasinglyimportant for us to achievea thoroughunderstandit with which we must contend.This is the ing of the underlyinghydrologicprocesses policymakers,lawyers, planners, engineers, challengeto hydrologists, waterresources economists,and others who must strive to see that future allocationsof water are sufficient to meet the needsof human and natural svstems.

PROBLEMS

13

r summary
distribution' movethe Hydrology is the scienceof water. It embraces occurrence, an sense, account-"nt, urii propertiesof the watersof the earth. In a mathematical of inputs,outputs,and waterStofages a regionsothat a history ing may be madeof the of water movementfor the region can be estimated' the hydrologic After reading this chapter you should be able to understand You shouldalso budgetand make a simpleu".ouniing of water transportin a region' be used to facilitate have gained an undersiandingof trow hydrologic analysescan for processes water resourcessystems' designand management

PROBLEMS
1.1.. One-half inch of runoff resultsfrom a stormon a drainagearea
of 50 mi2.Convertthis amount to acre-feetand cubic meters. surfacearea of t.2. Assume you afe dealing with a vertical walled reservoir having a will it take to occurs:How many hours of 1.0 m3/sec 500,000 m' and that anlnflow raise the reservoirlevel bY 30 cm? time is 15 acre-ft and 1.3. consider that the storageexistingin a river reachat a reference tie reach is 500 cfs and the outflow from the reachis at the sametime the inflow to 650cfs.onehourlater,theinflowis550cfsandtheoutflowis630cfs.Findthe meters' changein storageduring the hour in acre-feetand in cubic walled reservoir was t.4. During a24-hr time period, the inflow to a 500-acre vertical or fall in the sameinterval, evaporationwas 1 in. was there a rise 100 cfs. During in inchesand centimeters' How muchwasit? Give the answer waterelevation? surface areais 3000 acres' 1.5. The annualevaporationfrom a lake is 50 in. If the lake's surface daity evaporationrate in acre-feetand in centimeters? what would beiire time requiredto raise 1.6. A flow of 10 cfs entersa 1-mi2vertical walledreservoir.Find the levelbY 6 in. the reservoir Iftheaverageannualrunoffis5l02cfsand areaof4511mi2. t.7. Adrainagebasinhasan for losses the areain the averalerainfall is 42.5 in.,estimatethe evaportranspiration you think this estimateis? 1 year. Iiow reliable do Determinethe storage in 1.8. The storage a reachof a river is 16.0acre-ft at a given time. of inflow and outflow during the hour are rates (u"r"-f""tj t hr later if the average 700 and 650 cfs, resPectivelY. areafor 3 days' (a) L.9. Rain falls atataverage irrtensity of 0.4 in./hr over a 600-acfe the feetper second;(b) determine 3-day rate Determinethe average ofrainfau in cubic volumeofrainfallinacre-feet;and(c)determinethe3-dayvolumeofrainfallininches of equivalentdepth over the 600-acrearea' 100 acre-ft/day'Deter1.10. The evaporationrate from the surfaceof a 3650-acrelake is (feet) in the lake during a 365-dayyearifthe inflow to the lake mine the depthchange is25.2cfs.1s the changein lake depth an increaseor a decrease? if acre-feet the drainage to 1.11. One and one-half inchesof runoff areequivalent how many : 43,560 ft"') areais 25-mi2?lNote: I acte rate of how many cubic feet L.12. one-half inch of rain per day is equivalentto an average many metersper second? p". ,".ona if the areais 500 acrei?How

14

cHAPTER INTRoDUCTIoN 1

REFERENCES
1. P. B. Jones,G. D. Walker,R. W. Harden,and L. L. McDaniels,"The Development the of Scienceof Hydrology," Circ. No. 60-03, TexasWater Commission,Apr. 1963. 2. W. D. Mead, Noteson Hydrology. Chicago:D. W. Mead, 1904. 3. Ven Te Chow (ed)., Handbook of Applied Hydrology. New York: McGraw-Hill,1964. 4. O. E. Meinzer,Hydrology,Vol. 9 of Physicsof the Earth.New York: McGraw-Hlll, 1942. Reprintedby Dover, New York, 1949. 5. P. D. Krynine, "On the Antiquity of Sedimentationand Hydrology," Bull. Geol. Soc. Am. 70. l7 2I - l7 26(1960\. 6. RaphaelG. Kazmann,Modern Hydrology.New York: Harper & Row, 1965. 7. H. Pazwoshand G. Mavrigian, "A Historical Jewelpiece-Discovery of the Millennium Bull. 16(6), 1094-1096(Dec. 1980), Hydrologic Works of Karaji," WaterResources 8. Hunter Rouseand Simon Ince, History of Hydraulics, Iowa Institute of Hydraulic Re- :. search,State University of Iowa, 1957. in 9. G. H. L. Hagen, "Ueber die Bewegungdes Wassers engen cylindrischenRohren," gendorf Ann. Phys. Chem.16, 423- 442(1839). Pog s V. 10. Henri Darcy, Les fontaines publiques de la ville de Dijon. Paris:. Dalmont, 1856. des 11. J. Dupuit, Etudesthdoriqueset practiques sur le mouvement eaux dans les canauxs ed. dtcouvertset d travers les terrainspermdables,2nd Paris:Dunod, 1863. 12. L. K. Sherman,"Stream Flow from Rainfall by the Unit-Graph Method," Eng. NewsRec.108(1932). 13. R. E. Horton, "The Role of Infiltration in the Hydrologic Cycle," Trans.Am. Geophys. Union 14, 446- 460(1933). l 14. C. V. Theis,"The RelationBetweenthe Lowering of the PiezometricSurfaceand the Rate Union 16, and Duration of a Well Using Ground WaterRecharge,"Trans.Am. Geophys. 519-524(1935\. 15. Asit K. Biswas,"Hydrologic EngineeringPrior to 600 s.c.," Proc. ASCE J. Hyd. Div., Proc. Paper5431,Vol. 93, No. HY5 (Sept.1967). 16. RaymondL. Nace,"WaterResources: GlobalProblemwith Local Roots,"Environ. Sci. A Technol. 1(7) (July i967). 17. D. K. Todd (ed.), The WaterEncylopedia.New York: Water Information Center, 1970.

Chapter 2

Precipitation

r Prologue
The purposeof this chapteris to: ' Define precipitation, discussits forms, and describeits spatial and temporal attributes. ' Illustrate techniques estimatingareal precipitation amountsfor specific for storm eventsand for maximum precipitation-generating conditions. Precipitation replenishes surfacewater bodies,rbnewssoil moisture for plants, and recharges aquifers.Its principal forms are rain and snow.The relative importanceof theseforms is determinedby ttre climate of the area under consideration. many In parts of the westernUnited States,the extentof the snowpackis a determiningfactor relative to the amountof waterthat will be availablefor the summergrowing season. In more humid localities, the timing and distribution of rainfall are of principal concern. Precipitatedwaterfollows the pathsshownin Figs. r.2 and,1.3. some of it may be intercepted,evaporated, infiltrated, and becomesurfaceflow. The actual disposition depends the amountof rainfall, soil moistureconditions,topography, on vegetal cover soil type, and other factors Hydrologic modeling and water resourcesassessments dependupon a knowledgeof the form and amountof precipitation occurringin a region of concernover a time period of interest.

2.1 WATERVAPOR
The fraction of watervapor in the atmosphere very small comparedto quantitiesof is other gases present,but it is exceedingly important to our way of life. Precipitationis derived from this atmosphericwater. The moisture centent of the air is also a significantfactor in local evaporation processes. Thus it is necessary a hydrologist for to be acquainted with waysfor evaluatingthe atmospheric watervapor contentand to understand thermodynamiceffects of atmospheric the moisture.l

16

2 CHAPTER

PRECIPITATION

Under most conditions of practical interest (modest ranges of pressureand point is excluded),water vapor essentemperature,provided that the condensation tialiy obeys the gas laws. Atmospheric moisture is derived from evaporationand transpiration,the principal sourcebeing evaporationfrom the oceans.Precipitation the overthe United Statescomeslargelyfrom oceanicevaporation, watervaporbeing circulation system. primary atmospheric over the continentby the transporated humidity are relatedbasicallyto condiof Measures watervapor or atmospheric occurring over a level surfaceof pure water. tions of evaporationand condensation Considera ilosed systemcontaining approximatelyequal volumesof water and air If maintainedat the sametemperature. the initial condition of the air is dry, evaporaA tion takesplace and the quantity of water vapor in the air increases. measurement proceeds,pressurein the of pressurein the airspacewill reveal that as evaporation of in of because an increase partial pressure the watervapor (vapor airipaceincreases pressure the overlying air equalsthe of preJsure). Evaporationcontinuesuntil vapor leaving(evaporatof of surfacevapor pressure measure the excess water molecules [a At this point, evaporationceases, ing from) the water surfaceover thosereturning]. is of and if the temperatures the air spaceand water are equal,the airspace saidto be of closed,the equilibriumwould not saturated.If the containerhad beenopeninstead Somecomhavebeenreached,and all the water would eventuallyhaveevaporated. absohumidity are vapor pressure, moistureor of monly usedmeasures atmospheric humidity, and dew point temlute humidity, specifichumidity, mixing ratio, relative Perature.

Water Amount of Precipitable

. Estimatesof the amount of precipitation that might occur over a given region with favorable conditions are often useful. These may be obtained by calculating the extendingup from the earth's amountof water containedin a column of atmosphere althoughit cannotall be This quantity is known as theprecipitable water 14{ surface. Precipitable water is usually by removed from the atmosphere natural processes. in expressed centimetersor inches. An equationfor computingthe amountof precipitablewater in the atmosphere can be derived as follows. Considera column of air having a squarebase 1 cm on a side.The total water masscontainedin this column betweenelevationzero and some height z would be

W:

J^

r p*dz

(2.1)

humidity and IVis the depthof precipitablewaterin centimewherep. : the absolute into ters. The integral can be evaluatedgraphically or by dividing the atmosphere uniform specifichumidities,solvingfortheseindividually, and layersof approximately amountof precipitablewater for the then summing.Figure 2.1 illustratesthe average continentalUnited Statesup to an elevationof 8 km.2

Variations and Temporal Geographic

water vapor varieswith location and time. ThesevariaThe quantity of atmospheric tions may be attributed mainly to temperatureand sourceof supply considerations. The greatestconcentrationscan be found near the ocean surfacein the tropics, the

2.2 PRECIPITATION

17

Sault Ste.Mtrie Portled

0;7
VCT -NJ 0.8

0.9 1.0

0.8 b.z 1'o o.d 1.1 r.2

t.J

Bromsville

Figure 2.L Mean precipitablewater, in inches,to an elevationof 8 km. (U.S. WeatherBureau.)2

generallydecreasing concentrations with latitude, altitude, and distanceinland from coastalareas. About half the atmospheric moisturecanbe found within the first mile abovethe earth's surface.This is becausethe vertical transport of vapor is mainly through convectiveaction,which is slight at higher altitudes.It is also of interestthat there is not necessarily water vapor over a any relation betweenthe amount of atmospheric regionand the resultingprecipitation.The amountof water vapor containedover dry that over considerablymore areasof the Southwest,for example,at times exceeds precipitation while humid northern regions,eventhough the latter areasexperience the former do not.

2.2 PRECIPITATION
Precipitation is the primary input vector of the hydrologiccycle. Its forms are rain, snow,andhail andvariationsof thesesuchasdrizzle and sleet.Precipitationis derived from atmospheric water, its form and quantity thus being influencedby the action of Atmoother climatic factors such as wind, temperature,and atmosphericpressure. sphericmoistureis a necessary not sufficient condition for precipitation. Contibut nental air masses usuallyvery dry sothat mostprecipitationis derivedfrom moist are maritime air that originates overthe oceans. North America about50 percentof the In evaporated water is taken up by continental air and movesback againto the sea.

18

2 CHAPTER

PRECIPITATION

Formationof PreciPitation
to are Two processes considered be capableof supportingthe growth of dropletsof (dropletsfrom about 500 to 4000 p'min diameter)to overcomeair sufficient mass fall and consequently to the earth asprecipitation.Theseare known as the resistance process' processand the coalescence ice crystal their process one by which the small cloud dropletsincrease is The c^oalescence with other droplets through collision. Water droplets may be size due to contact to as considered falling bodiesthat are subjected both gravitationaland air resistance at equilibrium (terminal velocities) are proportional to the effects. Fall velocities more quickly squareof the radius of the droplet; thus the larger dropletswill descend ones. As a result, smaller droplets are often overtakenby larger than the smaller droplets,and the resulting collisionstend to unite the drops, producingincreasingly largir particles.Very large drops (order of 7 mm in diameter)break up into smallof processand producesomewhat a chain effect. dropletsthat repeatitre coalescence large raindropsmay be producedto generatesignificant In this *unn"r, sufficiently precipitation. This processis ionsidered to be particularly important in tropical regionsor in warm clouds. An important type of growth is known to occurif ice crystalsand waterdroplets down to about -40'C- Under temperatures are found toexist togetherat subfreezing theseconditions,certain particles t saltsserveasfreezingnucleisothat theseconditionsis higher over the t occurson the surface condensation particle sizedistributionsde' uneven with otherparticles.This is considet mechanism. The artificial inducementof precipitation has been studied extensively,and nuclei supthat condensation thesestudiesare continuing.It has been demonstrated induceprecipitation.The ability of humansto ensurethe producplied to cloudscan location or timing has not yet been iion of precipitation or to control its geographic attained,however' with the prospects problemsare associated Many legal as well as technological ..rain-makiig" processes. interesthereis the impacton hydrologicestimates that Of of partially controlled artificial precipitation might have. Many uncontrolled oi onty as naturally occurringLydrologicvariablesare considered statisticalvariatesthat are or distributedwith a random component.If the distribueither randomty distrlUuted tion or time seiiesof the variablecanbe modeled,an inferenceasto the frequencyof occurrenceof significanthydrologiceventsof a given magnitude(suchas precipitation) can be made.If, however,artificial controls are usedand if the effectsof these may prove to be totally unreliable cannot be reliably predicted,frequencyanalyses tools.

TyPes Precipitation
and of Dynamic or adiabaticcoolingis the primary cause condensation is responsible transport of air massesis a for most rainfall. Thus it can be seen that vertical requirementfor precipitation.Precipitationmay be classifiedaccordingto the condi-

'1 2.2 PRECIPITATIOI'|9 of vertical air motion. In this respect,the three major categories tions that generate precipitation type are convective, orographic, and cyclonic. precipitation typical of the tropicsand is is Convective Precipitation Convective brought about by heatingof the air at the interfacewith the ground. This heatedair quantities expands with a resultantreductionin weight.During this period,increasing of water vapor are taken up; the warm moisture-ladenair becomesunstable;and pronouncedvertical currents are developed. Dynamic cooling takes place, causing condensation precipitation.Convectiveprecipitationmay be in the form of light and are showersor stormsof extremelyhigh intensity (thunderstorms a typical example). precipitation resultsfrom the mechanical Orographic Precipitation Orographic lifting of moist horizontal air currentsover natural barriers suchas mountainranges. This type of precipitation is very common on the West Coast of the United States where moistureladen air from the Pacific Oceanis interceptedby coastalhills and mountains.Factorsthat are important in this processinclude land elevation,local slope,orientation of land slope,and distancefrom the moisture source. In dealingwith orographicprecipitation,it is commonto divide the regionunder study into zonesfor which influencesasidefrom elevationare believedto be reasonably constant.For eachof thesezones,a relation betweenrainfall and elevationis maps(seeSection2,5). for developed usein producingisohyetal of is with Cyclonic Precipitation Cyclonicprecipitation associated the movement differregions.Thesepressure regionsto low-pressure from high-pressure air masses encesare createdby the unequalheatingof the earth's surface. Cyclonicprecipitationmay be classifiedasfrontal or nonfrontal. Any barometric low canproducenonfrontal precipitationasair is lifted throughhorizontalconvergenceof the inflow into a low-pressure area. Frontal precipitation results from the lifting of warm air over cold air at the contact zone between air masseshaving colder If are different characteristics. the air masses moving so that warm air replaces warm air, the front is known asawarmfront; if , on the otherhand,cold air displaces air, the front is saidto be cold.If the front is not in motion,it is saidto be a stationary front. Figure 2.2 illustratesa vertical sectionthrough a frontal surface.

through a frontal surface. Figure 2.2 Vertical cross-section

20

CHAPTER PRECIPITATION 2

Thunderstorms
convectivestorms,which are Many areasof the United Statesare subjected severe to generallyidentifiedasthunderstorms of because their electricalnature.Thesestorms, although usually very local in nature, are often productive of very intenserainfalls that are highly significantwhen local and urban drainageworks are considered. with intense associated Thunderstorm cells develop from vertical air movements in surface heatingor orographiceffects.Thereare threeprimary stages the life history Theseare the cumulusstage,the mature stage,andthe dissipating of a thunderstorm. stage.Figure 2.3 illustrateseachof thesestages. All thunderstorms beginascumulusclouds,althoughfew suchcloudseverreach the stage of developmentneededto produce such a storm. The cumulus stageis characterized strongupdrafts that often reachaltitudesof over 25,000ft. Vertical by wind speeds upperlevelsare often as greatas 35 mph. As indicatedinFig.2.3a, at there is considerable horizontal inflow of air (entrainment)during the cumulusstage. of This is an important elementin the development the storm, as additional moisture is provided.Air temperatures inside the cell are greaterthart thoseoutside,as indicatedby the convexity of the isothermsviewed from above.The number and size of The the water dropletsincreaseasthe stageprogresses. duration ofthe cumulusstage is approximately i0-15 min. and the The strong updrafts and entrainmentsupport increasedcondensation in development waterdropletsand ice crystals.Firrally, whenthe particlesincrease of size and number so that surfaceprecipitation opcurs,the storm is said to be in the mature stage.In this stage strong downdrafts are created as falling rain and ice crystals cool the air below. Updraft velocities at the higher altitudes reach up to of Downdraft speeds over20 mph are 70 mptrin the early periodsof the maturestage.

El

o o F

(a)

(b)

(c)

cell. (Departstages a thunderstorm of Figure 2.3 Cumulus,mature,and dissipating ment of the Army.)

2.2 PRECIPITATION 21

usual aboveabout 5000 ft in elevation.At lower levels,frictional resistance tendsto decrease downdraft velocity. Gusty surfacewinds move outwardfrom the region the of rainfall. Heavyprecipitationis often derivedduring this preiod, which is usuallyon the order of 15-30 min. In the final or dissipating stage, downdraftbecomes the predominant until all the air within the cell is descending and being dynamically heated.Since the updraft ceases, mechanismfor condensation the ends and precipitation tails off and ends.

PrecipitationData
Considerabledata on precipitation are available in publications of the National WeatherService.a's Other sources includevariousstateand federal agencies engaged in water resourceswork. For critical regional studiesit is recommendedthat all possibledata be compiled; often the establishmentof a gauging network will be (seealso Chapters necessary 7-9).

pitation Variabiity Preci I

Precipitationvariesgeographically, temporally, and seasonally. Figure 2.4 indrcates the mean annualprecipitation for the continentalUnited States,while Fig. 2.5 gives an exampleof seasonal differences.It should be understoodthat both regional and temporal variationsin precipitation are very important in water resources planning and hydrologicstudies. For example,it may be very important to know that the cycle of minimum precipitationcoincides with the peakgrowing season a particular atea, in or that the periodofheaviestrainfall shouldbe avoided scheduling in certainconstruction activities. Precipitation amounts sometimesvary considerablywithin short distances. Recordshaveshowndifferences 20 percentor more in the catchof rain gauges of less that2Oft apart.Precipitationis usuallymeasured with a rain gaugeplacedin the open so that no obstacle projects within the inverted conical surfacehavingthe top of the gaugeas its apexand a slopeof45'. The catchofa gauge influenced the wind, is by which usually causes low readings.Variousdevicessuchas Nipher and Alter shields havebeendesigned minimize this error in measurement. to Precipitationgauges may be of the recording or nonrecordingtype. The former are requiredif the time distribution of precipitation is to be known. Information about the featuresof gaugesis readily available.3 Because precipitationvariesspatially,it is usuallynecessary usethe datafrom to severalgaugesto estimatethe average precipitation for an area and to evaluateits reliability (seeChapter27). This is especiallyimportant in forestedareaswhere the variation tendsto be large. Time variations in rainfall intensity are extremely important in the rainfallrunoff process, particularly in urban areas(see g. 2.6a). The arealdistributionis also Fi significantandhighly correlatedwith the time history of outflow (seeFig. 2.6b).These considerations discussed greaterdetail in following chapters. are in

c) o U)
H

()

o E t r 9

.n

EF

C)
d

.v)
C) k

h0
A

C)

()

n"i
c.)

()
6

s
o E o

e<

(.)
o
H 6

U
F o

t^
0,) $ 6l

bo

,t,
,Ff

/i E

.;

.t

.o
k P)

t z

c)
H
'ir @h{ 6d F O

* \ - r (
E 1

H 0)

e
>r

\
F Fi
j;i.
oooooo aFhs {6d^O ./

/
,,-C)

t:
F o 5 k> 6

ts

EU)
q . -

$,-,\ g( J: qi
F

O r +
! H

5 >
!2-

i < F "

. E
-

,E( ; \ * \ i \
F h i -

,/

'r
"

9999qa9
h

\ --II--LItr--I-----LIL,]

lh '' LH : ,,l..15 r+: H= 5

R- Y F =

H S

au)

: R

t J

oooo t s

h+6Nio o
ooooo *odio

Ntr
o L ?1 ^

CHAPTER2

PRECIPITATION

- 3

'

sec) Time(x 102 (a)

-/'?:;"
E [u Y.s.

Isohyets-lines of equal rainfall depth

-"@l5
Ashd 1.56

I o R.H.R. i 3.50

t.e+ in.a'1t :.zs

1.55 upl.

^ 1<

'k '& &.1 *'uffi

Dsn. 1.08
(b)
F,,**

Figure 2.6 (a) Rainfall distribution in a convective storm June 1960, , Baltimore, Maryland. (b) Isohyetal pattern, storm of SeptemberL0, 195'1 Baltimore, Maryland. O, recording rain gauge.

INPUT 2.3 DISTRIBUTION THE PRECIPITATION OF

25

2.3 DISTRIBUTION THE PRECIPITATION OF INPUT

Total precipitation is distributedin numerousways.That interceptedby vegetation and treesmay be equivalent the total precipitationinput for relatively small storms. to Once interception storageis fllled. raindrops begin falling from leavesand grass, where water storedon thesesurfaces eventuallybecomesdepletedthrough evaporation. Precipitationthat reaches ground may take severalpaths. Some water will the fill depressions and eventually evaporate;some will infiltrate the soil. Part of the infiltrated watermay strike relatively imperviousstratanearthe soil surfaceand flow parallel to it as interflow until an outlet is reached. approximately Other portions may replenishsoil moisturein the upper soil zone, and someinfiltrated water may reach reservoirthat sustains weatherstreamflow. The component the of the groundwater dry precipitation input that exceeds local infiltration rate will developa film of water the Detention depths on the surface(surfacedetention)until overlandflow commences. varying from I to 1j in. for various conditions of slope and surfacetype havebeen reported.3 Overland flow ultimately reaches defined channels and becomes streamflow. Figure 2.7 ilhxtrates in a generalway the dispositionof a uniform storm input in to a natural drainage basin.Although suchan input is not to be expected nature,the indicated relations are representativeof actual conditions. Modifications resulting from nonuniform stormswill be discussed they arise. as In Fig. 2.7anote that the storm input is distributeduniformly over time /o at a rateequalto i (dimensionally into equalto LT '). This inputis dissected components I, through lo, the sum of which is equal to I at any time r. Figure 2.7b illustratesthe manner in which infiltrated water is further subdividedinto interflow, groundwater, and soil moisture. Figure 2.7c showsthe transition from overland flow supply into The mechanics theseprocesses be treatedin detail in later sections. strearhflow. of will The nature of the curvespresenteddepictsthe generalrunoff process.It should be realized,however,that actual graphsof infiltration and/or other factorsversustime might appear quite different in form and relative magnitudewhen comparedwith of theseillustrations becatrse the effects of nonuniform storm patterns,antecedent conditions,and other factors. The rate and areal distribution of runoff from a drainagebasin are determined by a combination of physiographic and climatiOfactors. Important climatic factors includethe form of precipitation (rain, snow,hail), the type of precipitation (convective, orographic,cyclonic),the quantity and time distributionof the precipitation,the characterischaracterof the regionalvegetativecover,prevailingevapotranspiration factorsof significance tics, and the statusofthe soil moisturereservoir.Physiographic soil includegeometricpropertiesof the drainagebasin,land-usecharacteristics, type, geologicstructure,and characteristics ofdrainage channels(geometry,slope,roughness,and storagecapacity). Large drainagebasinsoften react differently from smalleroneswhen subjected to a precipitationinput. This can be explainedin part by suchfactorsas geologicage, characterrelativeimpact ofland-use practices,sizedifferential,variationsin storage istics, and other causes. Chow definesa small watershedas a drainagebasin whose

CHAPTER2

PRECIPITATION

(a/

\
I I

Soil moisture \a Int".flot

Mechanics of surface runoff

(c)

Figure 2.7 The runoff process:(a) dispositionof precipitation, (b) componentsof infiltration, and (c) dispositionof overland flow supply.

27 PRECIPITATION 2.4 POINT ofhigh-intensity, shortdo characteristics not filter out (1) fluctuationscharacteristic practices.6 this basis,small On (2) the effectsof land management duration storms;or 100 mi2. A large basin is one in which basinsmay vary from lessthan an acre up to channelstorageeffectively filters out the high frequenciesof imposedprecipitation and effectsof land-usepractices.

2.4 POINTPRECIPITATION
at Precipitation eventsare recordedby gauges specificlocations.The resulting data p"rmit determinationof the frequency and charactei of precipitation eventsin the vicinity of the site. Point precipitation data are used collectively to estimateareal designstorm variability ofrain and snowand are alsousedindividually for developing in Design storms are discussed for characteristics small urban er other watersheds. detail in Chapter 16. Point rainfall data are usedto deriveintensity-duration-frequencycurvessuch as those shown in Fig. 2.8. Such curves are used in the rational method for urban in storm drainagedesign(Chapter 25); thek constructionis discussed Chapter27 'ln the applyingthe rational method,a rainfall intensityis usedwhich represents average duration.The frequencychosen intensity of a storm of given frequencyfor a selected should reflect the economics of flood damage reduction. Frequenciesof up to 100 yearsare commonlyusedwhereresidentialareasare to be protected.For highervaluedistricts and critical facilities,up to 500 yearsor higherreturn periodsare often Local conditionsand practicenormally dictatethe selectionof thesedesign selected. criteria. (ExecutiveOrder 11988,Floodplain Management,I97 7).

\
00-yr frequency 50-yr frequency I -20-yr frequency f, 10-yrfrequency
7 , 5-Yr,frequencl

\ \

t\
B.
a

\ \

\1

tl=.t

120
Duration (min)

curves Figure 2.8 Typical intensity-duration-frequency for Baltimore, Maryland, and vicinity.

28

CHAPTER2

PRECIPITATION

Figure 2.9 Four quadrants surrounding precipitation stationA.

It is occasionallynecessary estimatepoint rainfall at a given location from to recordedvaluesat surroundingsites.This can be doneto completemissingrecordsor precipitation to be used at the point of interest. The to determinea representative National WeatherServicehasdeveloped procedurefor this which hasbeenverified a on both theoreticaland empiricalbases.T Considerthat rainfall is to be calculatedfor point A in Fig. 2.9. Establisha set of axes running through A and determinethe absolutecoordinatesof the nearest surroundingpoints B, C, D, E, and F. These are recorded in columns 3 and 4 of precipitationat A is determined a weightedaverage the as of Table 2.L The estimated points.The weightsarereciprocalsof the sumsof the squares AX and AY; of other five that is, D2 : LX2 + LYz, and W : llDt. The estimatedrainfall at the point of interestis given by I (P x W)/> I{. In the specialcasewhere rainrall is known in only two adjacentquadrants(e.g.,I and II), the estimateis given asI (p x lV). This has the effect of reducing estimatesto zero as the points move from an area of
TABLE 2.1 DETERMINATION POINT RAINFALLFROM DATA AT NEARBY OF GAUGES

t1)
Point

(2) Rainfall
(in.)

(3) AX

(4) AY

(5) (D')

(6) wx103

(7) PxWx103

B C D E F Sums

r.60 1.80 1.50 2.00 r .

4 1 3 3 7 0

2 6 2 3 2

20 37 1,3 18 8

50 27.O 76.9 55.6 125.0

,TT3

80.0 48.6 115.4 111,.2 2t2.5 567.7

*Note.'Estimatedprecipitation(P) at A = 567.7 /334.5; P = 1.70 in.

_l

29 PRECIPITATION 2.5 AREAL to precipitation to one with no records.This is considered be the most logical proceresult will alwaysbe lessthan the The dure for handlingthis unusualcase.7 estimated greatestand greaterthan the smallestsurroundingprecipitation. For specialeffects suchas mountain influences,an adjustmentprocedurecan be applied.

2.5 AREALPRECIPITATION
it For most hydrologicanalyses, is important to know the areal distributionof precipitation. Usually, averagedepths for representativeportions of the watershedare determinedand used for this pwpose. The most direct approachis to use the arithare metic average gaugedquantities.This procedureis satisfactoryif gauges uniof is formily distributedand the topography flat. Other commonly usedmethodsare the at method.The reliability of rainfall measured one isohyetalmethodand the Thiessen gaugein representing average depth over a surroundingareais a function of (1) the area,(2) the sizeof the the distancefrom the gaugeto the centerof the representative (e.g.,stormeventversus (4) area,(3) topography, the natureofthe rainfall ofconcern For meanmonthly), and (5) local stormpatterncharacteristics.8 more information on errorsof estimation,the readershouldconsultRefs.7 and 8. Chapter27 alsocontains a discussion areal variability of precipitation. of Figures 2.10 and 2.11 illustrate how the measuredrainfall at a single gauge in with change ( 1) the relativeposition rainfall over a watershed relatesto the average of the gaugein the watershedand (2) the time period over which the averageis In calculated. the first caseit is clearthat the more centralthe gaugelocation,the more area, providing for will match the average a representative closely its observations that areal averages not that the regionis not too large.Figure 2.11 shows, surprisingly,

o bo o >
J

o > o

2, *,
d 9 1

b 9
d 9 n

3.=

B.E

/
k l O

,.1:i'j
1 2 3

(n

Storm rainfall at one gauge in inches (a)

Stormrainfall at one gaugein inches (b)

average Figure 2.L0 Errors resultingfrom useof a singlegaugeto estimatewatershed area is 0.75 mi2 and (giuge location effect, Soil ConservationService),(a) Watershed areais 0.75 mi2 and gaugeis 4 mi outsidethe gaugeis near the center. (b) Watershed watershed boundary.

30

CHAPTER2

PRECIPITATION
d bo
6 o 2 o

o, bI) o >

EE 2
5 . =

b 9

a
d

220

'a
d r l

,/
10
U)

v)
1 ) {

Storm rainfall at one gauge in inches

Stormrainfall at one gaugein inches

(a)

(b)

average estimatewatershed Figure 2.L1 Errors resultingfrom use of a singlegaugeto Service).(aiWatershedareais 5'45 mi2 and the ConserVation (time period effect, Soil gaugeisontheboundary.(b)Watershedareais5.45In|zandthegaugeisonthe boundarv. to conform more closely over long time periods, in this case one year, may be expected storm event' This suggests to a single guog" uu"ruge than those for an individual with both space and time that the Oerlgn-of guuging networks should be tempered considerations.

lsohyetalMethod
are the isohyetal of The two principal methodsfor determiningareal averages rainfall basedon interpolation method and the Thiessenmethod. The isohyetal method is in surveyingand between gauges.It closely resemblesthe calculation of contours to plot the^rain gauge -upptng."Tf" first step in developingan isohye?1..-up is (Fig' 2'I2)' Next' an locationson a suitablemap and to reJord the rainfall amounts at selectedincreinterpolation betweengaugesis performed and rainfall amounts to thenconnected form mentsare plotted.tdenticaldepthsfrom eachinterpolationare of the weightedaverage is isohyets(lines of equal rainfall depth)' The areafaverage isohyets'The isohyetal depthsbetweenisohyets,that is, the meanvaluebetweenthe precipitation over an method is the most accurateapproachfor determiningaverage attentionto topographic area,but its proper oserequir#a skilled analystand careful the represenand other tactori that impact on areal variability. Figure 2. 13 illustrates tationofamajorstormeventinNorthCarolinabyanisohyetalmap.

ThiessenMethod
In method' this is arealrainfallaverages theThiessen of method calculating Another ascenters' using subareas raingauges into is the procedure area subdivided polygonal Thiessen depth' average the in estimiting watershed as are Thesubareas used weights for suitable is not in as are diagrams constructed shown Fig. i.t+. fnis procedure

2.5 AREAL PRECIPITATION

31

Average

precipitation
for area A4 is 4.25 in.

----x------*
A2

-*!fl1
--^ 3 in'
Average preciPitation = 2 A i P i \ i for entire basin

(c)

and map:(a)locateraingauges of Figure 2.12 Construction an isohyetal and gauges; (c) plot isohyets' between (b) pdt values; interpolate network is fixed of mountainousareasbecause orographicinfluences'The Thiessen are if any gauges mustbe reconstructed io, u glu"o gaugeconfiguation, and polygons relocated.

AccuracY
locatior\of the Irrespectiveof the methodusedfor estimatingareal precipitation,the tothe point of applicationof the estimate guu* orra in derivingthe estimaterelative may be lbcattiesovertical distances In mustbe takeninto consideration. mountainous are horizontal spacings For gentlelandscapbs, -ft" irnpottant than horizontal ones.

32

CHAPTER PRECIPITATION 2
a . !.1 a.l - o !

oo>
I
I

I ti-

a ( h w t

o 'r-r

r'.

-a)

!?+

? z

+ r Q

r E
I f.\

o ( g 9 ' - s
o ( )

c R o >

\
s

sz' g
d . a

c..l &

c]

(f)x
!-l +i

N i 4

e-H

TH

+t Figure 2.14 Constructionofa Thiessen diagram:(a) connectrain gaugelocations;(b) draw perpendicularbisectors;and (c) calculate Thiessenweights l,er, ,qr, A3). (d) A completed network. the most important. when a precipitation gauging network is to be developed, both spacing and arrangement of gaugei must be considered.

Average depth entire over = watershed

EXAMPLE 2.1 $yen 1rrcdrainagearea of Fig. 2.r5 and the rainfall data displayedin column 3 of Table2.2, calculatethe average rainfall over the areausing 1aj tne arithmetic mean, and (b) the Thiessenpolygonweighting system.

Figure 2.15 Thiessendiagram for Example 2.1.

34

CHAPTER 2

PRECIPITATION

TABLE 2.2 DATA AND THIESSEN POLYGON FOR EXAMPLE 2,1. CALCULATION (1) GaugeNo.

(2) "/" Area

(3)
Precip.-in 1.56

(4)

(2) x (3) 0.08 0.12 0.10

0.44

2 3

2.95
3.44

3 t5
ll

2.91
4.17 4.21

0.46 0.80
0.11 0.17 0.81

6
7 8 9

19
4

2.'| 2.45
3.88 3.98

21
6 5

10
ll

0.24 0.13
3.45

2.51

Total

100

Solution. falling within the areaboundary.They includegauges a. Identify thosegauges yields an 1, 4 through 6, 8, and 9. Averagingthe valuesfor thesesix gauges estimatedmean areal rainfall of 3.20 inches. b. Followine the Thiessen method as described in Section 2.5, construct polygonsusingtrianglesto connectgaugepoints. Thesepolygonsare shown with each on Fig. 2.15. Calculatethe percentof the total area associated weightedaverage gaugeandrecord asin column2 of Table2.2.The Thiessen is obtainedby multiplying the valuesin column 2by the yaluesin column 3. The Thiessenaverage computedas 3.45 inchesof rainfall. The use of a is (Table 2.2) facilitates computations and aids in organizing spreadsheet data. lr

MAXIMUM PRECIPITATION 2.6 PROBABLE

rainThe probablemaximum precipitation (PMP) is the critical depth-duration-area which would result from a storm containing fall relation for a given areaand season probable.e Such storm events conditionsconsidered the most critical meteorological are used in flood flow estimatesby the U.S. Corps of Engineersand other water of conditions are basedon analyses The critical meteorological resourcesagencies. air-mass properties(effectiveprecipitablewater,depthof inffow layer,wind, temperature, and other factors),synoptic situationsduring recorded stormsin the region, and topography, season, locationofthe area.The rainfall derivedistermedprobable theory and maximumprecipitation sinceit is subjectto limitations of meteorological data and is basedon the most e_ffectiwcqmbination of factors controlling rainfall

) {)
o q)
k C)
H

9.'T\):

.o >

E
o

(.)
k

U
o
o L*

- N r

C)

,r)
ri s c'l

N k

o o
H

H g

\o ol I
b0 It

q6

,/

) {)
o q)
k C)
H

9.'T\):

.o >

E
o

(.)
k

U
o
o L*

- N r

C)

,r)
ri s c'l

N k

o o
H

H g

\o ol I
b0 It

q6

,/

36

CHAPTER2

PRECIPITATION 1000 800 600 400

$ zoo
E
9?

sr o u 9 9 t
4 4 0 20 10 l0

1oo

20 30 40 50 6Q 70 80

90 100 110 120 130 140 150

Percentageof 200 miz, 24-hr values

Figure 2.17 Seasonal variation, depth-area-duration relations; percentage to be applied to 200 ni2-24 hr probable maximum precipitation values for August in Zone 6. (U.S. Department of Commerce,National WeatherService.)

intensity.e earlierdesignation "maximumpossible An of precipitation" is synonymous. Additional information on PMP is given in Chapter 16. The seasonalvariation of PMP is important ip the designand operation of multipurposestructures and in floodingconsiderations may occurin combination that with snowmelt. In both of these cases,annual probable maximums might be less important than seasonal maximums.Figures2.16 and,2.r7 display 24-hr pMp for the eastern half of the United Statesfor 200-ffi2 watersheds during the month of August (similar figuresare availablefrom the National WeatherService).

2.7 GROSS AND NETPRECIPITATION

The net (excess) precipitation that contributesdirectly to surfacerunoff is equivalent to the gross precipitation minus lossesto interception, storm period evaporation, depression storage, and infiltration. The relation betweenexcess precipitationP" and grossprecipitation P is thus P":P.-Ilosses (2.2)

where the lossesinclude all deductions from the grossstorm input. The paths that water precipitatedover an areamay take can be represented by flow diagramsof the type given in Fig. 1.3 and by equationsof the form ot&q. 2.i. Modelssuchastheseare the basisfor most hydrologicinvestigations, muchbf the and contentof this book is devotedto the conceptualization individual components of of the various hydrologicprocesses and to synthesizing thesecomponentsinto holistic representations hydrologicevents. of

PROBLEMS 37

r Summary
for precipitation is the sourceof fresh waterreplenishment the planetEarth. Too much In prosperityand disaster. betweenthese canmeanthe differencebetween or too little with a frequency are the normal precipitation eventsthat are experienced extremes position and topographicfeatures' and intensity relatedmainly to geographic After reading this chaptei you should understandthat both the timing and considamountof precipitition occuiring over anareaareimportant and that thereis areal erable g"olrapttic variability in precipitation. You should be able to estimate process simplehydrologic precipiLtiin amountsfrom gaugedata and conceptualize valuesofprecipitation for a region shed It shouldbe recognizedihataverage models. uses, some light on the quantiiy of water that might be made availablefor various are of while a i<nowledge tne time-aistribution and time-dispositionof precipitation and shortage' plans for periods of excess management requisitesfor developing PROBLEMS 2.1.

"r)

estimatethe amount of rainfall for gaugeX' average, artdc compute the rainfall for gaugeX in Problem 2.1 if the storm readingsat A, B, 'werc respectively' 3.7,4.I, and4'8 in., in 2.3. compute the meanannualprecipitationfor the watershed the following figureusing The the arithmetic mean, the'itri"*"n polygon method, and the isohyetalmethod.

38

CHAPTER2 PRECIPITATION gaugereadings gauges for A-K, respectively, are: 29.79,34.97,25.6,24.2i,24.6, 42.61,42.35,15.51,39.99, 43.04,and28.41. 2.4. Computethe meanannualprecipitationfor the watershed the figurefor Problem2.3 in using the arithmetic mean and the Thiessen polygonmethod.The gaugereadingsfor gauges A-K, respectively, are: 28.1, 33.7, 25.6, 23.9, 24.6, 40.1, 41.3, 37.2, 38.7, 41.1,and29.3. 2.5. The chart from a rain gaugeshown in the sketchrepresentsa record that you must interpret.Find the average rainfall intensity (rate)between6 e.r'1. noon on August and 10. Find also the total precipitation on August 10 and August 11.
0.30 o.25 0.20 0.15 0.10 0.05 0
Midnight 6 a.m.

Aus. l 0

Ais..I7

6 p.m. Midnight 6 a.m. Time

Noon

6 p.m. Midnight

2.6. Refer to the chart of Problem 2.5. Calculate the rainfall intensity for the period between6.q..1u. noon on August 11. Would you considerthis to be a period of and intenserainfall? 2.7. Use the map of Fig. 2.6 andfrom it constructa setof Thiessen polygons. Using these, estimatethe mean rainfall for the region. 2.8. A meandraft of 100mgd is producedfrom a drainageareaof 200 mi2.At the flow line the reservoiris estimated cover 4000 acres.The annualrainfall is 37 in., the mean to annualrunoff is l0 in., and the mean annual lake evaporationis 30 in. Find the net gain or lossin storage.Computethe volume of water evaporated. How significantis this amount? 2.9. A meandraft of 380,000m3/dayis producedfrom a drainageareaof 330 km2.At the flow line, the reservoiris estimatedto cover about 1600hectares. The annualrainfall is 96.5 cm, the meanannualrunoff is 22.8 cm, and the meanannuallake evaporation is 77.1 cm. Find the net gain or loss in storageand compute the volume of water evaporated. Calculatevolumes m3. in 2.10. Drainage areas within each of the isohyetal lines for a storm are tabulated for a watershed.Use the isohyetal method to determinethe averageprecipitation depth within the basin for the storm. Make a conceptualsketch.

lsohyetalinterval(in.)

Area (acres) 2700 1900 1000 0

0-2
a i

4-6 6-8

2.11. ReworkProblem2.10 if the valuesin the secondcolumn of the table are2,500, 2,100 1,200, and 300, respectively,

REFERENCES 39 2.12. Discusshow you would go about collectingdata for analysisof the water budgetof a region.What agencies would you contact?What other sourcesof information would you seekout? 2.13. For an areaof your choice,plot the meanmonthly precipitation versustime. Explain how this fits the pattern of seasonal water usesfor the area.Will the form of precipitation be an important consideration?

REFERENCES
I

2. 3.
A

5. 6.
,1

8. 9.

Tennessee Valley Authority, "Heat and Mass TransferBetweena Water Surfaceand the Atmosphere,"Lab. Rep. No, 14, TVA EngineeringLab. Noiris, TN, Apr. 1972. A. L. Shands,"Mean PrecipitableWater in the United States," U.S. WeatherBureau, Tech.PaperNo. 10, 1949. R. K. Linsley, Jr., M. A. Kohler, and J. L. H. Paulhus,Applied Hydrology. New York, McGraw-Hill, 1949. D. W. Miller, J. J. Geraghty,and R. S. Collins, WaterAtlas of the United States.Port Washington, NY Water Information Center, 1963. U.S. WeatherBureau,Tech. Papers1-33. Washington, DC: U.S. GovernmentPrinting Office. Ven Te Chow (ed.), Handbook of Applied Hydrology. New York, McGraw-Hill,l9&. Staff, Hydrologic Research Laboratory, "National WeatherServiceRiver ForecastSystem ForecastProcedures,"NOAA Tech. Mem. NWS HYDRO 14, National Weather Service,SilverSpring,MD,Dec. 1972. V. Mockus, Sec.4, in SCSNational EngineeringHandbookon Hydrology,Washington, DC: Soil ConservationService, A:ug.1972. J. T. Riedel, J. F. Appleby, and R. W. Schloemer,"SeasonalVariation of the Probable Maximum Precipitation East of the 105th Meridian for Areas from 10 to 1000 Square Miles andDurations 6,12,24,and48 Hours,"Hydrometeorological of Rept.No. 33, U.S. WeatherBureau,Washington, D.C., 1967.

Chapter3

Interception Depression and Storage

r Prologue
The purposeof this chapteris to: . Define interception and depression storqge. ' Define the roles theseabstracting play in affectingthe amountof mechanisms precipitatedwater ultimately availablefor other distribution. ' Providesomeapproaches estimatingthe quantitiesof waterinterceptedand to storedin depressions during precipitation events. Figure 1.3indicatesthe pathsthat precipitatedwatermay take asit reaches earth. the The first encountersare with interceptingsurfacessuch as trees,plants, grass,and structures. Waterin excess interceptioncapacitythen beginsto filI surfacedepresof sions.A film of water alsobuilds up overthe ground surface. This is known as surface detention.Once this film is of sufficientdepth, surfaceflow toward definedchannels commences, providing that the rate at which water seeps into the ground is lessthan the rate of surfacesupply.This chapterdealswith the first two mechanisms which by the grossprecipitation input becomestransformedinto net precipitation.

3.1 INTERCEPTION
Part ofthe stormprecipitationthat occursisintercepted vegetation otherforms by and ofcover on the drainage area.Interception be defined that segment can as ofthe gross precipitationinput which wets and adheres aboveground to objectsuntil it is returned to the atmospherethrough evaporation.Precipitation striking vegetationmay be retained on leavesor blades of grass,flow down the stemsof plants and become stemflow,or fall off the leaves becomepart of the throughfall. The modifying effect to that a forest canopycan haveon rainfall intensity at the ground (the throughfall) can be put !o practical use in watershed management schemes. The amountof waterintercepted a function of ( 1) the stormcharac (2) the is ter, species, and density age, ofprevailingplantsandtrees,and (3) the season ofthe year. Usually about 10-20 percentof the precipitationthat falls during the growing season

41 3.1 INTERCEPTION is interceptedand returned to the hydrologiccycle by evaporation.Water lossesby closedforest stands-as much as interceptionare especiallypronouncedunderdense has 25 percentof the total annualprecipitation.Schomaker reportedthat the average is annualinterceptionlossby Douglasfir standsin westernOregonand Washington about24 percent.l A lO-year-oldloblolly pine plantation in the South showedlosses pine forestsin Califoron a yearlybasisof approximately14 percent,while Ponderosa nia were found to intercept abort 12 percent of the annual precipitation. Mean interception lossesof approximately 13 percent of gross summer rainfall were reported for hardwoodstandsin the White Mountainsof New Hampshire.Additional information given in Table 3.1 includes some data on interception measurements obtainedin Maine from a mature spruce-firstand, a moderatelywell-stockedwhite and gray birch stand,and an improved pasture.r Lull indicatesthat oak or aspenleavesmay retain as much as 100 drops of For tree, interception storageon the order of 0.06 in' of wdtter.z a well-developed retentionof about precipitationcould thereforebe expected the basisof an average on CATCH PRECIPITATION OFSTANDARD TABLE WEEKLYAVERAGE 3,1 IN LOCATED RAIN WEATHER BUREAU.WPE GAUGES U,S.
A SPRUCE-FIR STAND, A HARDWOOD STAND, AND A PASTUREDURINGTHE WINTER OF 1965_1966 Weekly average precipitationcatch (in. ot equivalent rain) Percent interception by forest cover

Measuring date'

Spruce-fir 0.24 1.01 1.01 1.41 0.55 0.66 0.20 0.36 Trace 0.25 1.38 0.05 0.29, 0.76 0.17 0.86 0.76 0 0.73 10.69

Birch

Pasture 0.39 L45 r.36 t.79 0.87 1.08 0.26 0.61 Trace 0.59 1.96 0.06 Trace 0.98 0.22 t.45 0.97 0 1.27 15.31

Spruce-fir 38 30 26 2l 37 39 23 4l 58 30 l7 22
ZJ

Birch l5 l4 10 8 7 l2
A

tl/9t65 11tL6/65 ru23t65 12tr0t65" 12n7 t65 12/30t65 lt4t66 1tr2t66 U18t66 U25t66 2tr/66 2t8t66 2n1t66 2n5t66 2t21t66 3t2t66 3t7t66 3t15t66 3t29t66 Total

0.33 r.25 r.23 1.65

0.81 0.95 0.25 0.55 Trace 0.58 l.9l 0.07 0.02 0.81 0.22
l -za

10 2
J

t6 t7 0 I6 13 1l 9.5

0.84 0 1.13 13.83

4l 22 43 30.2

" The period betweenmeasuringdatesis 7 days,exceptwhen precipitation occurred on the seventhday. In this event, measurgmentwas postponeduntil precipitation ceased. bMeasurementswere delayeduntil a method was devisedto melt frozen precipitation on the site. 'This measurementin the spruce stand was the result of foliage drip during a thaw from previously intercepted snow. "The Effect of Forest and Pastureon the Disposition of Precipitation," Source: Aftet C, E. Schomaker, Maine Farm Res.(July 1966).

42

oHAPTER rNrER;;-;;rr;-.r,o* s

"ro*ou=

20 drops per leaf. For light showers(where gross precipitation P < 0.01 in.) 10t percentinterceptionmight occur, whereas showers for whereP > 0.04 in., losses in the rangeof 10-40 percentare realistic.3 Figure 3.1 illustratesthe generaltime distribution pattern of interceptionloss intensity.Most interceptionlossdevelops during the initial stormperiod and the rate of interception rapidly approacheszero thereafter.l-6Potentialitorm interception losses be estimated usins2'3,6 can bv Zi:,S+KEt

(3.1)

where . L, : the volume of water intercepted(in.) S : the interceptionstorage that will be retainedon the foliage againstthe forcesof wind and gravity (usuallyvariesbetween 0.01 and 0.05 in.) K : the ratio of surfaceareaof interceptingleaves horizontal projection:, to of this area E : the amountof waterevaporated hour duringthe precipitationperiod ' per (in.) r : time (hr)

i= it+i2+4+i4

Figure 3.1 Disposition of rainfall input in terms of interception, depression storage,infiltration, and overlandflow.

3.1 INTERCEPTION 43 Equation3.1 is basedon the assumption that rainfall is sufficientto fully satisfy the storageterm S. The following equationwas designed accountfor the rainfall to amountT-e
L;:S(1 -e-P/s)+KEt

(3.2)

whereP : rainfall and e is the baseof natural logarithms.Note in Eqs. 3.1 and 3.2 that the storm time duration t is given in hours, while ,L,, S, and E are commonly measured in. or mm. in It is important to recognizethat forms of vegetationother than trees can aiso interceptlarge quantitiesof water. Grasses, crops,and shrubsoften haveleaf-areato ground-arearatios that are similar to thosefor forests.Table3.2 summarizes some observations that havebeen madeon crops during growing seasons on a variety and of grasses. Interceptedamountsare aboutthe sameasthosefor forests,but sincesome of thesetypes of vegetationexist only until harvest,their annualimpact on interception is generallylessthan that of forested areas. Precipitationtype, rainfall intensityand duration,wind, and atmospheric conditions affecting evaporationare factors that serve to determineinterception losses. Snow interception,while highly visible,usually is not a major loss sincemuch of the interceptedsnowfall is eventuallytransmittedto the ground by wind action and melt. Interceptionduring rainfall eventsis commonly greaterthan for snowfall events.In both cases, wind velocity is an important factor. The importanceof interceptionin hydrologicmodelingis tied to the purposeof the model. Estimates of loss to gross precipitation through interception can be significantin annual or long-term models,but for heavy rainfalls during individual storm events,accountingfor interceptionmay be unnecessary. is important for the It modeler to assess carefully both the time frame of the model and the volume of precipitation with which one must deal. TABLE3.2 OBSERVED PERCENTAGES INTERCEPTION OF BY
VARIOUSCROPSAND GRASSES'

Vegetation type
Crops Alfalfa Corn Soybeans Oats Grassesb Little bluestem Big bluestem Tall panic grass Bindweed Buffalo grass Blue grass Mixed species Natural grasses

(%) Intercepted

Comments

36 16 15 7 50-60-.1

s7 l s 7 f 17 l )
31 17
)A

Water applied at rate of ] in. in 30 min

Pdor to harvest

t4-t9

"Valuesroundedto nearest percent.Data for table were obtainedfrom Refs.2,4, and 5, 'Grass heightsvary up to about 36 in.

44

CHAPTER3

INTERCEPTION AND DEPRESSION STORAGE

but Equations3.1 and 3.2 canbe usedto estimatetotal interceptionlosses, for to ofindividual storins,it is necessary dealwith the areal variability detailedanalyses however. are for Generalequations estimatingsuchlosses not available, of suchlosses. Most researchhas been related to particular speciesor experimentalplots strongly associated with a given locality. In addition, the lossfunction varieswith the storm's experimentaldata are available,the nature of the varianceof character.If adequate interceptionversustime might be inferred. Otherwise,common priictice is to deduct volumeentirelyfrom the initial period of the storm(initial abstraction). the estimated EXAMPLE 3.1 by Using the following equationsdeveloped Horton6for interceptionby ash and oak trees, estimatethe interception loss beneaththesetrees for a storm having a total precipitation 1.5 in. of Solution 1. For ashtrees.

L ; : 0 . 0 1 5+ 0 . 2 3 P : : 0.015 0.23(1.5) 0.36 + in.

2. For oak trees,

L;:0.03+0.22P : : 0.03+ 0.22(1.5) 0.36 in. rl

3.2 THROUGHFALL
for A numberof relationships estimatingthroughfall for a variety of foresttypeshave tt been developed.n Deiermining factors for throughfall quantities include canopy wind velocity, and total leaf area,numberand type of layersof vegetation, coverage, rainfall intensity.The arealvariability ofthesefactorsresultsin little or no throughfall in somelocationsand considerable throughfall in others.In general,prediction equaareaand coverasprime of tions for throughfall mustincludemeasures canopysurface variables.An example of a throughfall relationship for an easternUnited States hardwoodforest follows.l2 For the growing season Tn: 0.901P 0.O3ln For the dormant season T n : 0 . 9 I 4 P- 0 . 0 1 5 n where ?1,: throughfall (in.) P : total precipitation (in.) n : number of storms

(3.3) (3.4)

3.3

DEPRESSION STORAGE

45

3.3 DEPRESSION STORAGE

Precipitationthat reaches ground may infiltrate, flow over the surface, become the or trappedin numeroussmall depressions from which the only escape evaporationor is infiltration. The natureof depressions, well astheir size,is largely a funition of the as original land form and local land-usepractices.Because extremevariability in the of nature of depressions and the paucity of sufficient measurements, generalized no relation with enoughspecifiedparameters all cases feasible.A rational model for is can, however,be suggested. Figure 3.1 illustratesthe dispositionof a precipitationinput. A studyof it shows that the tate at which depression storageis filled rapidly declinesafter the initiation of a precipitationevent.Ultimately, the amountof precipitation goinginto depression storage will approachzero,given that thereis alargeenoughvolume of precipitation to exceedother lossesto surfacestoragesuch as inflltration and evaporation.Ultimately, all the water stored in depressions will either evaporateor seep into the ground.Finally, it shouldbe understood that the geometryof a land surface usually is complex and thus depressions vary. widely in size, degreeof interconnection,and contributingdrainagearea.In general,depressions may be looked upon as miniature reservoirsand as suchthey are subjectto similar analytical techniques. According to Linsley et a1.13 volume of water storedby surfacedepressions the at any given time can be approximated using
Y:Sd(l -e-kP")

r? 5)

where V : the volume actually in storageat sometime of interest S, : the maximum storagecapacityof the depressions P" : the rainfall excess (grossrainfall minus evaporation,interception,and infiltration) k : a constantequivalentto l/So The valueof the constantcan be determined considering by that if P" : 0, essentially all the water will fill depressions and dv/dp" will equal one. This requires that k : r/Sa. Estimatesof s, may be securedby making samplefleld measurements of the areaunder study.Combiningsuchdata with estimates P" permits a determinaof tion of V. The mannerin which Vvarieswith time must still be estimated depression if storagelossesare to be abstracted from the grossrainfall input. One assumptionregarding dVldt is that all depressions must be full before overlandflow supply begins.Actually, this would not agreewith reality unlessthe locationsof depressions were gradedwith the largestonesoccurring downstream. If the depression storagewere abstracted this manner, the total volume would be in deducted from the initial storm period suchas shownby the shadedareain Fig.3.2. Such postulateshave been used with satisfactory results under special circumstances.ra Depression storage intensitycan alsobe estimated usingEq. 3.5. If the overland flow supplyrale oplus depression storage intensityequali - /, wherei is the rainfall intensity reachingthe ground and/is the infiltration rate, then the ratio of overland flow supply to overlandflow plus depression storagesupply can be proved equal to

i - f

(3.6)

46

CHAPTER INTERCEPTION DEPRESSION 3 AND STORAGE

o b0 o

!t o
d

1
Time (min)

16

20

Figure3.2 Simple depression storage abstraction scheme. This expression can be derivedby adjudging c : i - f - o (3.7) i - f i - f and noting that o is equal to the derivativeof Eq. 3.5 with respectto time. Then )

o:fiso1t_e-kP")

(3.8)

u : (Soke-kg#
It was shown that k : 1/S, so that

(3.e)

u : ,-o'"d!"
dt

(3.10)

precipitationP, equalsthe grossrainfall minus infiltrated water, and since The excess the derivativewith respectto time canbe replacedby the equivalent intensity(i * f), the intensity of depression storagebecomes
o:e-or.(i-f)

(3.11)

i - f

(i-f)*G-f)e-."" i - f

(3.r2)

o.125

o.25

0.315

0.50(turf)

Mass overlandflow and depression storagesupply ( P - F)

0.0625
All depressions filled before overland flow supply begins -=
:\ on

0.0938

0.125(pavements) 1.00

----l

9)E o " 7

RO

0.80
bl I

l-

Exponential relationship a

o.7o I
-\ I ,

F i 6 0 6
o

-lP-F)tS, " = - - e I I

i,t

tl

OGEE sumrnation the of probability curve standard

n5n a

o b!

B
o E

{
.: 40

9 d J U

o o0

s2
o

6zo t
!)

I
E

50

100

150

200

Mean depth as a percentage of overall depth of depression storage

Figure 3.3 Depth distributioncurve ofdepressionstorage. Enter graph from top, readdown to selected curve, and project right or left as desired.(After Tholin and Kiefer.r5) and o f)(I - e-kP")

i * f

:(i

i - f
| g-kP"

(3.13) (3.r4)

Figure 3.3 illustratesa plot of this function versusthe massoverlandflow and depression storage supply(P - F), whereF is the accumulated massinfiltrationl5 and

48

3 CHAPTER

AND DEPRESSION STORAGE INTERCEPTION

P is the grossprecipitation.In the plot meandepthsof 0.25 in. for turf and 0.0625in. Maximum depthswere 0.50 and0.125in., respectively. were assumed. for pavements The figure also depictsthe effect on estimatedoverlandflow supplyrate, which model. Three modelsare shown storage is derivedfrom the choiceof the depression are all depressions full before overlandflow that in the figure: the flrst one assumes with a mean depth of 0.25 in., the figure begins.For a turf area having depressions showsthat for P - F valueslessthan 0.25 in., thereis no overlandflow supply,while for P - F valuesgreater than 0.25 in., the overlandflow supply is equal to i - f . will be greaterthanzero. Tholin For the exponentialmodel (Model 2), c always a relation betweenthosepreviouslymentionedis that and Kiefer haverecommended A of likely more representative fully developedurban areas.15 cumulative normal in and is also described Fig. 3.3 probability curve was selected this representation for (Model 3). are storage deductions usually madefrom the first part of the storm Depression as illustrated in Fig. 3.2. The amount to be deductedis a function of topography, this loss During major storms, groundcover,and extentand type of land development. storage guidelines estimatingdepression for to is often considered be negligible.Some studiesof experimentaland other watersheds. basedon losseshavebeen developed Values for depressionstoragelossesfrom intense storms reported by Hicks are Tholin andKieferhaveused 0.15in. for loam,and0.10in. for c1ay.16 0.20 in. for sand, Studies of and 0.0625in. forpavements.ls valuesof 0.25 in. inpervious urban areas by Viessmanyieldedthe information shownin four small imperviousdrainageareas storageloss is highly correlatedwith slope.This is Fig.3.4, where mean depression

0.15

I
Ei U.IU

2 Slope(70)

Figure 3.4 Depression storage loss versus slope for four ra) impervious drainageareas.(From Viessman.

PROBLEMS

49

oo

Antecedent rainfall during preceding 30 min

o

Time (min)

Figure3.5 Depression storage intensity versus for animpervious time area.(After Turner.l7)

easilyunderstood, sincea given depression hold its maximum volumeif horizonwill tally oriented. Using very limited data from a small, paved-streetsection, Turner devised curvesshownin Fig. 3.5.17 the Other sources datarelatedto surface of storase are available the literature.2,18,1e in

r Summary
Accountingfor the dispositionof precipitation is an important part of the hydrologic modelingprocess.Two abstractions from the precipitation input, intercepiion, and depression storagewere coveredin this chapter. Interception lossesduring the courseof a year may be substantial, but during intensestorms,they may be sufficiently small to neglect.Precipitationtype, rainfall intensity and duration, wind, and atmospheric conditions affeiting evaporationare factorsthat serveto determineinterceptionlosses a given foresi standor ground for coverconfiguration.Interceptionduring rainfall eventsis commonly greaterihan for snowfall events. Depressionstoragedeductions occur early in a storm sequence and they are a function of topography,ground cover, and extent and type of land development. During major storms,this loss is often considered be negligible. to

PROBI.EMS
3.1. UsingFig. 3.2, estimatethe volume of depression storage a 3-acrepaveddrainage for area' Statethe volume in cubic feet and cubic meters.Convert it to equivalentdep;h over the area in in. and cm.

50

CHAPTER3

STORAGE AND DEPRESSION INTERCEPTION

of 3.2. Estimatethe percentage the total volume of rainfall that is indicatedas depression in storage Fig. 3.2. annual precipitationfor your state,estimatethe annual amountof 3.3. Using the average interceptionloss. 3.4. Refer to Fig. 2.4 and estimate the annual interception lossesin lllinois, Florida, California, and New Mexico. How good do you think theseestimatesare?In which Why? In which of thesestateswould the do estimates you havethe most confldence? water budgetbe most affectedby interception? of 3.5. Using Fig. 3.4, estimatethe percentage rainfall that would be lost to depression for storage a l0-acre parking lot havinga mean slopeof 1 percent.Repeatfor a slope of 3 percent.Using the total rainfall volume determinedin Problem 3.2, estimatethe storageloss for both slopes.Stater equivalentdepth over the area of the depression in mm and in. depths 3.6. Refer to Fig. 3.3 and estimatethe ratio of overlandflow supplyto overlandflow and storage supplyif the areais turf, the OGEE summationcurve is the model, depression storageis (a) 75 percent and (b) 125 percent. mean depth of depression and the 3,7. Explain how a relation such as that given in Fig. 3.3 could be used in a simulation model of the rainfall-runoff process' 3.8. Using Eqs. 3.3 and 3.4, estimatethe throughfall in in. for 28 in. of rainfall during the growing season(21 events), and 17 in. of rainfall during the dormant season(13 events). 3.9. Using Horton's equationsgiven in Example 3.1, estimatethe interceptionlossesby ash and oak trees for a storm having a total precipitation of 1.33 in'

REFERENCES
"The Effect of Forestand Pastureon the Dispositionof Precipitation," 1. C. E. Schomaker, Maine Farm Res.(July 1966). 2. ven Te chow (ed.), Handbook of Apptied Hydrology. New York: McGraw-Hill,1964. New York: McGraw-Hill' 1948. 3 . JosephKittredge, ForestInfluences. "Interception of Rainfall by HerbaceousVegetation,"Science86(2243), i + . O. n. Ctark, "Resultsof the Mountain Home Rainfall Interceptionand Inflltration Project 5. J. S. Beard, on Black Wattle, 1953-1954," J. S. Afr. Foresty Assoc'27,72-85(1956)' "Rainfall Interception," Monthly WeatherRev.47,603-623(L9I9)' 6. R. E. Horton, "A Note on the InterceptionLoss Equation," J. Geoplrys. Res.65' 38507. R. A. Meriam, 1 (1 9 6 0 ) . 385 Council, on 8. D. M. Gray (ed.),Hand.book the Principles of Hydrology.National Research WaterInformation Center,Inc., 1973. Canada,Port Washington: and J. L. Thames,Hydrology and the 9. K. N. Brooks, P. F. Folliott, H. M. Gregersen, Ames, IA: Iowa StateUniversity Press/Ames,1991. of Watersheds. Management "The InterceptionProcess."In Prediction in Catchment Hydrology,National 10. G. J. Blake, Symposiumon Hydrology,eds.T. G. Chapmanand F. X. Dunin, MelbourneAust. Acad. S c i . , 5 9 - 8 11 9 7 5 . , "Throughfall in PlantedStandsof Fourth SouthernPines 11. F. A. Roth, II, and M. Chang, Bulletin 17' 880-885(1981) in Species EastTexas,"WaterResources

59r-s92(r937).

REFERENCES 51 "canopy and Litter Interceptionby Hardwoodsof Eastern J. D. Helvey and J. H. Patric, Resour. Res.l, 193-206(1965)' United States,"Water New York: R. K. Linsley, Jr., M. A' Kohler, and J. L. H. Paulhus,Apptied Hydrology' McGraw-Hill, 1949. "A Linear Model for synthesizingHydrographs Small Drainage for warren viessman,Jr., at paperpresented the Forty-eighthAnnual Meetingof the AmericanGeophysical Areas," D.C., APr. 1967. Union, Washington, "The Hydrology of Urban Runoff," Trans. ASCE 125, A. L. Tholin una C. J. Kiefer, ( 1 3 0 8 -1 3 7 91 9 6 0 ) . ..A Method of Computing Urban Runoff ,', Trans' ASCE |09, I2L,7 W. I. Hicks,

Aue. 1966.

Chapter4

lnfiltration

Prologue
The purposeof this chapteris to: . Defineinfiltration. ' Indicatethe role infiltration playsin affectingrunoffquantities and in replenishing soil moistureand groundwaterstorages. ' Presentmodelsfor estimatinginflltration and provide examples how they of can be used. Infiltration is that processby which precipitation movesdownwardthrough the surface of the earth and replenishes soil moisture, rechargesaquifers,and ultimately supportsstreamflows during dry periods.Along with interception,depression storage, and stormperiodevaporation, determines availability,if any,of the precipitation it the input for generatingoverland flows (Fig. 1.3). Furthermore,infiltration rates influence the timing of overland flow inputs to channelizedsystems. Accordingly, infiltration is an important componentof any hydrologicmodel. The ratef at which infiltration occursis influencedby suchfactorsas the type and extent of vegetalcover, the condition of the surfacecrust, temperature,rainfall intensity,physicalpropertiesof the soil, and water quality. The rate at which wateris transmittedthrough the surface layeris highly dependent on the condition of the surface.For example,inwashof fine materialsmay seal the surfaceso that infiltration ratesare low evenwhen the underlyingsoils are highly permeable. After water crosses surfaceinterface,its rate of downwardmovement the is controlled by the transmissioncharacteristics the underlying soil profile. The of volume of storageavailablebelow ground is also a factor affecting infiltration rates. Considerable research infiltration hastakenplace,but considering infinite on the combinationsof soil and other factors existing in nature, no perfectly quantified generalrelation exists.

OF 4.2 CALCULATION INFILTRATION 53

4.1 MEASURING INFILTRATION Commonly usedmethodsfor determininginfiltration capacityare hydrographanalyas Infiltrometersareusuallyclassified rainfall simulators sesandinfiltrometer studies. In the former, arlificalrainfall is simulatedover a small test plot or flooding devices. ofrainfall andrunoff, with considerfrom observations and the inhltration calculated storageand surfacedetention.lFlooding infiltrometers are ation given to depression usually rings or iubes insertedin the ground. Water is applied and maintainedat a required' made of the rate of replenishment constantlevel and observations over havethe advantage infiltration basedon hydrographanalyses Estimatesof relating more directly to prevailing conditions of precipitation and infiltrometers of field. However,they areno betterthan the precisionwith which rainfall andrunoff are Of measured. partitular importancein suchstudiesis the areal variability of rainfall. Several meth;ds have been developedand are in use. Reference1 gives a good descriptionof thesemethods.

OF 4.2 CALCULATION INFILTRATION

from the applicationof reported averInfiltration calculationsvary in sophistication types and vegetalcoversto the useof differential equations ageratesfor specificsoil porousmedia.For small urban areasthat g6verningthe flow of wateiin unsaturated warranted'On input, more precisemethodsare sometimes iespondiapidly to storm or to peak flow production from prolongedstorms,average large waterlhedssubject may be adequate. values representative The infiltration pto"".r is -omplicated at best. Even under ideal conditions in (uniform soil propertiei andknown fluid properties),conditionsrarely encountered Accordingly,therehasbeenconsiderdifflcult to characterize. is practice,the process Most of theseeffortshaverelatedto the developabtestudyof the infiltration process. and (2) the solution of equationsbasedon field observations ment of i1) empirical mechanicsof saturatedflow in porous media.l'2 equationsbasedon the Later in this chapter,severalcommonly usedinfiltration modelsare discussed. a As a prefaceto that discussion, brief descriptionof the infiltration processfollows' principal factors affecting infiltration and points out some of the It reviews the by problemsencountered hydrologicmodelers. with an idealcase,onein which the soil is homogeneous Webeginour discussion profile and all the pores are directly interconnectedby capillary throughout the that the rainfall is uniformly distributedovef the purru!"r. Furtheimore,it is assumed may be chatactetized theseconditions,the infiltration process ur"u of "on"ern. Under the major influencing factors are therefore soil type and as one dimensional and moisturecontent.3 through which the The soil type characterizes sizeand numberof the passages the moisturecontentsetsthe capillary potentialand relative the watermustflow while conductivity of the soil. Capillary potential is the hydraulic head due to capillary forces. capillary suction is the sameas capillary potential but with oppositesign.

54

CHAPTER4

INFILTRATION

9 aoo
Ei o

. 300 -.
d

2oo

0.1

0.2

0.3

0.4

0.5

Moisture content, 0 (vol/vol)

conducFigure 4.1 Typical capillary suction-relative (AfterMein andLarson.e.) tivity-moisture content relation. Capillary conductivity is the volume rate of flow of water through the soil under a gradient of unity (dependent soil moisture content). Relative conductivity is the on capillary conductivity for a specifiedmoisturecontent divided by the saturatedconductivity. Figure 4.1 illustratesthe relations amongthesevariables.Note that at low moisturecontents,capillary suctionis high while relative conductivityis low. At high moisture contentsthe reverseis true. With this background,an infiltration event can be examined.Consider that rainfall is occurringon an initially dry soil. As shownin Fig. 4.l,the relative conductivity is low at the outsetdue to the low soil moistureconditions.Thus, for the water to move downwardthrough the soil, a higher moisture level is needed.As moisture builds up, a wetting front forms with the moisturecontentbehindthe front beinghigh (essentially saturated) and that aheadof the front being low. At the wetting front, the capillary suctionis high due to the low moisture content aheadof the front. At the beginning of a rainfall event, the potential gradient that drives soil moisture movementis high because wetting front is virtually at the soil surface. the Initially, the infiltration capacity is higher than the rainfall rate and thus the and more water infiltration rate cannot exceedthe rainfall rate. As time advances entersthe soil, the wetting zone dimensionincreasesand the potential gradient is reduced.Infiltration capacitydecreases until it equalsthe rainfall rate. This occursat Figures4.2 and4.3 illustrate the time the soil at the land surfacebecomessaturated.

4.2

OF CALCULATION INFILTRATION

55

Moisturecontent.d

I
proflledevelopment a constant with rainfall Figure4.2 Typicalmoisture rate. these conditions. Figure 4.2 showshow a moisture profile might developwhen a rainstormofconstant intensityoccurs.In the diagramthe soil moistureat the surface value at the top is shownto rangefrom its initial value at the top left to its saturated right. Thus in moving downwardon the left-handside of the diagram,one can trace the downward progressionof the wetting front for varying levels of soil moisture

92

::

Time, r Figure 4.3 Infiltration rate versustime for a given rainfall (After Mein and Larson.e) intensity.

56

CHAPTER4

INFILTRATION

at contentat the land surface. Figure 4.3 indicatesthat until saturationis reached the surface, infiltration rate is constantand equalto the rainfall applicationrate at the the surface. Point 4, apoint that coriesponds the time at which saturationoccursat At to the surface,the infiltration rate beginsto proceedat its capacity rate, the maximum rate at which the soil can transmit water acrossits surface.As time goes on, the infiltration capacity continues to decline until it becomesequal to the saturated conductivity of the soil, the capillary conductivity when the soil is saturated.This ultimate infiltration rate is shown by the dashedline to the right of K" in Fig. 4.3. Of particular interest is the determinationof Point 4 on the curve of Fig. 4.3. This is the point at which runoff would beginfor the conditions specifiedabove.It is also the point at which the actual infiltration rate/becomes equal to the infiltration capacityratefo ratherthan the rainfall intensityrate i. The time of occurrenceof this point depends, a given soil type, on the initial moisture content and the rainfall for rate. The shapeof the infiltration curve after this point in time is also influencedby thesefactors. Another factor that must be reckonedwith in the infiltration processis that of hysteresis. Fig. 4.1 it can be seenthat the plot of capillary suctionversussoil In moistureis a loop. The curve is not the samefor wetting and drying of the soil. The curves shown on the figure are the boundary wetting and boundary drying curves, curves applicableunder conditions of continuouswetting or drying. Betweenthese curves, an infinite number of possiblepaths exist that dependon the wetting and problemhavebeen to drying history of the soil. A numberof approaches the hysteresis reportedin the literature.3 was basedon an ideal soil. The illustration of the infiltration processpresented Natural soils are Unfortunately,suchconditionsare not replicatedin natural systems. highly variable in composition within regions and soil cover conditions are also far-ranging. Becauseof this, no simpleinfiltration model can accuratelyportray all the conditionsencountered the fleld. The searchhas thus beenfor modelsthat can in of be called upon to give acceptable estimates the rates at which infiltration occurs durine rainfall events. Mein and Larson have describedthree generalcasesof infiltration associated with rainfall.3The first caseis one in which the rainfall rate is lessthan the saturated conductivity of the soil. Under this condition, shownas (4) in Fig. 4.4, runoff never this condition occurs since all the rainfall infiltrates the soil surface.Nevertheless, sincethe level of soil moisture mustbe recognized continuous in simulationprocesses is affectedeven though runoff doesnot occur. The secondcaseis one in which the rainfall rate exceeds saturated the conductivity but is lessthan the infiltration capacity. Curves(I), (2), and (3) of Fig. 4.4 illustratethis condition.It shouldbe observed that the period from the beginningof rainfall to the time of surfacesaturationvaries with the rainfall intensity.The final caseis one in which the rainfall intensityexceeds the infiltration capacity.This condition is illustratedby the infiltration capacitycurve of Fig. 4.5 andthoseportionsof infiltration curves(l), (2), and (3) of Fig. 4.4 that are in their declining stages. Only under this condition can runoff occur. All three to caseshaverelevance hydrologicmodeling,particularly when it is continuousover time.

4.3

MODEL HORTON'SINFILTRATION

57

tsatt

tsal, Time, t

tsat,

Figure 4.4 Inflltration curves for several rainfall intensities.(After Mein and Larson.e)

fo: f"+ Ao-f")"u'

Infiltration capacitY curve

f, 0

,r*"

Figure 4.5 Horton's infiltration curve and hyetograph.

MODEL INFILTRATION .I3 HORTON'S

An The inflltration processwas thoroughly studiedby Horton in the early 1930s.o the following relation for outgrowth of his work, shown graphically in Fig. 4.1, was determininginfiltration capacity: (+.r; fo: f, + ("fr f")e'n'

58

CHAPTER4

INFILTRATION

where fo : k : f" : ,fo :

the infiltration capacity(depth/time) at sometime / inf capacity the a constantrepresenting rate of decrease d final or equilibrium capacity the initial infiltration capacity

It indicatesthat if the rainfall supply exceedsthe infiltration capacity,infiltration in tendsto decrease an exponentialmanner.Although simplein form, difficulties in determininguseful valuesfor/. and fr restrict the useofthis equation.The areaunder the the curve for any time interval represents depth of water infiltrated during that interval. The infiltration rate is usually given in inches per hour and the time r in minutes,althoughother time incrementsare usedand the coefficientk is determined accordingly. By observingthe variation of inflltration with time and developingplots of / / versus asshownin Fig. 4.5,we canestimate fsandft. Two setsof/and / are selected havingtwo unknownsare thus from the curve and enteredin Eq. 4.1. Two equations forfi and k. approximations obtained;they can be solvedby successive Typical infiltration ratesat the end of t hr ( f) areshownin Table4.1. A typical relation betweenf, and the infiltration rate throughout a rainfall period is shown graphically in Fig. 4.6a; Fig.4.6h showsan infiltration capacity curve for normal conditionson turf. The data given in Table4.1 are for aturf areaand must antecedent A be multiplied by a suitablecover factor for other types of cover complexes. range of cover factorsis listed inTable 4.2. Total volumes of inflltration and other abstractionsfrom a given recorded (plot of the streamflow rate versus hydrograph rainfall are obtainablefrom a discharge time) if one is available.Separationof the base flow (dry weather flow) from the for hydrograph resultsin a direct runoff hydrograph(DRH), which accounts discharge Direct surfacerunoff or the direct surfacerunoff. that is. rainfall less abstractions. can readily be precipitation excess inchesuniformly distributedover a watershed in at calculatedby picking valuesof DRH discharge equal time incrementsthrough the hydrograph and applyingthe formulas P": where P" : 4r : A : r?7:

(0.0371e)() q')
Ano

( a) \

(in.) precipitation excess DRH ordinatesat equal time intervals (cfs) area(mi2) drainage hurrber of time intervalsin a 24-hr period

For most cases differencebetweenthe original rainfall and the direct runoff the can be consideredas infiltrated water. Exceptionsmay occur in areasof excessive drainageor tracts of intensiveinterceptionpotential.The calculatedvalue subsurface as of infiltration can then be assumed distributedaccordingto an equationof the form of Eq. 4.1 or it may be uniformly spreadover the stormperiod. Choiceof the method on employeddepends the accuracyrequirementsand size of the watershed. To circumvent some of the problems associatedwith the use of Horton's Fig. 4.5. Notethat where Consider can infiltration model,someadjustments be made.6 the the infiltration capacitycurve is abovethe hyetograph, actual rate of infiltration

MODEL INFILTRATION 4.3 HORTON'S

59

o - 1

1.0 Time (hr)

(a)

3.0 2.8
2.O

-TT
-

2.4 2.2 2.0 1.8 1.6 t.4 7.2 1.0 0.8 0.6 0.4 0.2 0

u l, = U.UUUUJTI . ) 9 ( l

,4 f=0.53+2.4

0691ty

l
e

l
fle
rOa"

9l rras!
o\

r\$4"

-iro$dts"
Infiltration capacity curve (/.

10

70 80

100

t20

140

160

t80

Time (min) from beginning of infiltration capacity cuve' t/ (b)

Figure 4.6 (a) Typical infiltration curve. (b) Infiltration capacity and mass conditionsof turf areas.fAfter A. L. Tholin and curvesfor normal antecedent "The Hydrology of Urban Runoff," Proc. ASCE J. Sanitary Clint J. Kiefer, Ens. Div. S4(SA2),56 (Mar. 1959).1

60

CHAPTER4

INFILTRATION

TABLE 4.1 ryPICAL f, VALUES

Soilgroup
High (sandysoils) Intermediate(loaps, clay, silt) Low (clays,clay loam)

f, (in./hr)

f' (mm/h)

0.50-1.00 0.10-0.50 0.01-0.10

t2.50-25.00 2.so-r2.50 0.25-2.50

Source: After ASCE Manual of Engineering Practice,No.28.

TABLE4,2 COVERFACTORS Cover

Permanentforest and grass Cover fac{or

Good(1 in. humus) Medium(f-1 in. humus; Poor(< j in. humus)

Good Medium Poor Good Medium Poor

Close-growingcrops

Row crops

3.0-7.5 2.0-3.0 1.2-1.4 2.5-3.0 t.6-2.0 1.11.3 1.3- .5 1 1.1-1 3 1.0- .1 1

Source: After ASCE Manual of Engineering Practice, No.2t.

is equal to that of the rainfall intensity, adjustedfor interception,evaporation,and other losses. Consequently, actuafinfiltration is given by the

f(t) : minlfof), i(t)l

(4.3)

where/(r) is the actual infiltration into the soil and i(l) is the rainfall intensity.Thus the infiltration rate at any time is equal to the lesserof the infiltration capacity, f,(t) or the rainfall intensity. Commonly,the typical valuesof foandf" are greaterthan the prevailingrainfall intensitiesduring a storm. Thus, when Eq.4.l is solvedforS as a function of time alone, it shpwsa decrease infiltration capacity even when rainfall intensitiesare in much lessthanfo. Accordingly,a reductionin infiltration capacityis maderegardless of the amount of water that entersthe soil. To adjust for this deficiency,the integratedform of Horton's equationmay be used,

: F(tp)

lr'' ,

o, : f,tp +

(t - a' *,,1

(4.4)

whereF is the cumulativeinfiltration at time to, as shownin Fig. 4.7.rnthe figure, it is assumed that the actual infiltration has been equal to$. As previouslynoted, this is not usually the case,and the tnie cumulativeinfiltration must be determined. This

MODEL INFILTRATION 4.3 HORTON'S

61

0
0

tptpt tl Equivalent time

infiltration. Figure4.7 Cumulative can be done using

F\t) :

,u, l,' o,

(4.s)

usingF,q.4.3. where/(r) is determined Equations4.4 and4.5 may be us6djointly to calculatethe time t, that is, the equivalent time for the actual infiltrated volume to equal the volume under the infiltration capacity curve (Fig. 4.7). The actual accumulatedinfiltration given by Eq. 4.5 is equatedto the area under the Horton curve, F,q. 4.4, and the resulting is expression solvedfor r' This equation,
F:fJrt (l-e-k'n)

(4.6)

It explicitly for to,but an iterativesolutioncan be obtained. should cannotbe solved time r. Thus that the time to is lessthan or equal to the actual elapsed be understood that given as infiltration capacity shownin Fig. 4.7 is equalto or exceeds the available a described,fbecomes functionof the actual by Eq. 4.1. By makingthe adjustments in amount of water infiltrated and not just a variable with time as is assumed the original Horton equation. In selectinga model for use in inflltration calculations,it is important to know shortcomings; a its limitations. In somecases model can be adjustedto accommodate are if in othercases, its assumptions not realisticfor the natureof the useproposed, the model shouldbe discardedin favor of anotherthat better fits the situation. Part One of this book deals with the principal componentsof the hydrologic cycle. In later chapters,the emphasisis on putting these componentstogetherin

62

CHAPTER4

INFILTRATION

for When thesemodelsare designed continuvarioushydrolqgicmodelingprocesses. of is ous simulation,the approach to calculatethe appropriatecomponents the hydroof overtime. A discussion how infiltration could logic equation,Eq. 1.4, continuously be incorporatedinto a simulation model follows. It exemplifiesthe use of Horton's model (SWMM)." equationin a storm water management First, an initial value of /o is determined.Then, consideringthat the value of$ depends the actualamountof infiltration that hasoccurredup to that time, a value on infiltration capacity,fo, availableover the next time stepis calculated of the average using

'' f o : * 1 " = ' *foo ' : l W

-tp

(4.7)

rate of infiltration, /. Equation 4.3 is then usedto find the average

'v - [f, tt

iti >1, iri <f,

(4.8)

rainfall intensity over the time step. where i is the average Following this, infiltration is incrementedusing the expression F(t + Lt) : F(t) * AF : F(t) + f Lt

(4.e)

infiltration (Fig.4.7). whereAF : f Lt is the addedcumulative The next stepis to find a new valueof ro.This is doneusingEq. 4.6.If LF : Fig. 4.7),Eq. 4.6 must f o Lt, tp : to t Ar. But if the new /oris lessthanto + A/ (see using the be solvedby iteration for the new value of lo. This can be accomplished procedure.6 Newton-Raphson horizontal and When the valueof tp > I6f k, the Horton curve is approximately : f". Once this point hasbeenreached, there is no further needfor iteration since fo on f is constantand equal to f" and no longerdependent F.
EXAMPLE 4.I

ft Givenan initial inflltrationcapacityfiof 3.0 in./hr anda time constant of 0.29hr-', if the ultimate infiltration capacity is derive an infiltration capacity vs. time curve 0.55 in./hr. For the first ten hours, estimatethe total volume of water infiltrated in inchesover the watershed. Solution. Using Horton's equation(4.1), valuesof infiltration can be computed for various times. The equationis as follows: f : f" + Lfo + f")e-k, Substitutingthe appropriatevaluesinto the equationyields ,f = 0.55 + (3.0 0.55)s-o'zo' Table4.3, valuesof f atecomputedand Then for the times shownin spreadsheet graphics package,the curve of enteredinto the table. Using the spreadsheet Fig. 4.8 is derived,

MODEL 4,3 HORTON'SINFILTRATION FOR EXAMPLE4.1 TABLE 4.3 CALCULATIONS

63

Time (hr) 0.00 0.10 0.25 0.50 1.00 2.00 3.00 4.00

lnfiltration (in./hr) 3.00 2.93 2.83 2.67 2.38 t.92

1.58

Time (hr) 5.00 6.00 7.00 8.00 9.00 10.00 15.00 20.00

lnfiltration (in./h0 t.l2

0.98

0.87 0.79 0.73

0.68

0.58 0.56

r.32

4.1 can To find the volumeof waterinfiltrated duringthe flrst 10hours,Eq' be integratedover the range of 0-10 v : Y: I J [0.5s + (3.0 0.55)e-o2e' ldt

[0.55/ + (2.45I-0.29)e-o2s\o : 12.47in' V ll

is The volume in inchesover the watershed thus 12'47 in'

3.0

\ ;
l? 6 .:

2.0

\ -\-

15

1.0

"'-0 0.5

L
8

10

12

t4

18

Time (hr) Figure 4.8 GraPh for ExamPle 4'1.

64 4.4

CHAPTER4

INFILTRATION

MODEL GREEN-AMPT
in The Green-Ampt infiltration model,originallyproposed 1911,has had a resurgence inte1gs1.3'6-1t approach based Darcy's law (seeChapter18).In its on This is of of original form, it was intendedfor use where infiltration resultedfrom an excess water at the ground surfaceat all times. In 1973, Mein and Larson presenteda methodologyfor applying the Green-Ampt model to a steadyrainfall input.eThey the alsodeveloped procedurefor determining valueof the capillary suctionparamea the ter usedin the model. In 1978,Chu demonstrated applicability of the model for useunder conditionsof unsteadyrainfall.lo As a result of theseand other efforts, the Green-Ampt model is now employedas an option in such widely used continuous simulation modelsas SWMM.6 that the soil surfacewas The original formulation by Greenand Ampt assumed coveredby ponded water of negligible depth and that the water infiltrated a deep homogenous with a uniform initial watercontent(seeFig. 4.9). Wateris assumed soil to enter the soil so as to define sharply a wetting front separatingthe wetted and unwettedregions as shown in the figure. If the conductivity in the wetted zone is definedas K", applicationof Darcy's law yields the equation Io:

"

K"(r + s)
r

(4.10)

where I is the distancefrom the ground surfaceto the wetting front and S is the capillary suction at the wetting front. Referringto Fig. 4.9, it can be seenthat the cumulativeinfiltration F is equivalentto the product of the depth to the wetting front L and the initial moisture deficit, 0, - 0, : IMD. Making these substitutionsin

Ponded depth considerednegligible

I
Ho

I
I

I I
Figure 4.9 Definition sketch for GreenAmpt model.

MODEL 4.4 GREEN-AMPT

65

we Eq. 4.10 and rearranging, obtain

f- - - , \ r p : K.(-,

t * jt' )
F /

(4.11)

Considering thatfp : dFldt. we can state

#: *,(t.'#)
: 0 at t : 0' we obtain Integratingand substitutingthe conditionsthat F F - S x I M D X l o g " ({ IMD

(4.r2)

+ I MD X {) : K.t
X

(4.r3)

for This form of the Green-Ampt equationis more convenient usein watershed infiltrationto the the cumulative processes than Eq. 4.l}beciuse it relates modeling a of this equation assumes ponded time at which infiltration began.The derivation equal to the infiltration capacityat all surfaceso that the actualrate of infiltration is a the times.Using Eq. 4.13, we can determine cumulativeinfiltration at any time, in All the parameters the equation modeling. featuredesiiablefor continuoussystems The determinaare physicalpropertiesof the soil-water systemand are measurable. s is often difficult, however,particution of suitablevaluesfor the capillary suction be larly for relationssuchas that ihown for a clay-type soil in Fig. 4.10. It can there is a wide variation of capillary observedfrom the figure that for this curve content.3 with soil moisture suction The Mein-Larson formulation using the Green-Ampt model incorporatestwo The stages.3,6 first stagedealswith prediction of the volume of water that infiltrates before the surfacebecomessaturated.The secondstageis one in which infiltration usingthe Green-Ampt equation.In the widely usedstormwater capacityis calculated of the ,nunug"*"nt model, the irodif,ed Green-Ampt model of infiltration is one Computationsare madeusing optiois that can be employedto estimateinfiltration.6

v)

Moisture content,0

Figure 4.10 Capillary suction versus moisture contentcurves.

66

CHAPTER4

INFILTRATION

the following equations:for F ( F,(f : i),

F. :7:

IM?

i/K" - |

rori > K"

\4.r4)

and thereis no calculationof F"for i < K"; for F > F,(f : fi): ' f,: x,(t *\ where f : fo: I : F : { :

t * j") F

(4.11)

acttal infiltration rate (ftlsec) infiltration capacity(ftlsec) rainfall intensity (ftlsec) cumulativeinfiltration volume in the event (ft) cumulative infiltration volume required to causesurfacesaturation (f0 S : average capillary suctionat the wetting front (ft of water) IMD : initial moisture deficit for the event (ftlft) K" : saturatedhydraulic conductivity of soil (ftlsec)

Equation 4.10 showsthat the volume of rainfall neededto saturatethe surface is a function of the rainfall intensity.In the modelingprocess,for eachtime stepfor which I ) K", the value of d is computedand comparedwith the volumeof rainfall infiltrated to that time. If F equalsor exceeds the surface saturates calculations and {, for infiltrationthenproceed usingEq.4.I4. Notethat by substituting/for in Eq.4.l4 i and rearranging, equationtakesthe sameform as Eq. 4.11. the For rainfall intensitieslessthan or equal to K", all the rainfall infiltrates and its amount is used olly to update the initial moisture deficit, IMD.6 The cumulative infiltration volume {" is not altered. After saturation achieved the surface,Bq.4.l1 showsthat the infiltration is at capacityis a function of the infiltrated volume,and thus of the infiltration ratesduring previous time steps.To avoid making numerical errors over long time steps,the integratedform of the Green-Ampt equation(Eq. .B) is used.This equationtakes the following form as it is usedin SWMM: (4.15) &Gr- tr): Fz- Cln (F2+ C) - Fr* Cln(F1+ C) where C : IMD X .t (ft of water) / : times (sec) 1,2 : subscripts indicating the starting and ending of the time steps. Equation4.15 mustbe solvediteratively for F2,the cumulativeinfiltration at the end of the time step.A Newton-Raphsonroutine is used.6 In the SWMM model, infiltration during time step tz - tt is equal to (t, - tr)i if the surfaceis not saturatedand is equal to F, - F, if saturationhas previously occurred and there is a sufficient water supply at the surface.If saturationoccurs during an interval, the infiltrated volumesover each stageof the processwithin the time stepsare computedand summed.When the rainfall endsor becomeslessthan

MODEL 4.5 HUGGINS_MONK

67

theinfiltrationcapacity,anypondedwaterisallowedtoinfiltrateandisaddedtothe cumulative inflltration volume'

MODEL 4.5 HUGGINS-MONKE by problem introducing the have investigators circumvented timedependency Several by proposed equation the following as soilmoisture the depenO"niuutluUfe.2'10-13 and Huggins Monkeis an examPle:2

f : f,*A(
where AandP:

(4.16)

coefficients layer (Q ,,orug" potential of a soil overlying the impeding moisture) minus antecedent F _ the total volume of water that infiltrates impeding stratum T o = the total porosity of soil lying over the The using data from sprinkling infiltrometer studies' The coefficientsare determined process'At the in the iteration for variableF mustbe catculated eachtime increment

s - itr"

exceedsthe inflltration capacity,the rate zone," which determinesthe soil moistu (1) as evaluated follows:10 wherethe moist the field capacity(amountof waterheld in z drained),the drainagerate is considered soil is s to the infiltration rate when the constan| and (3) if the watercontentis be drainagerate is comPutedas

rate drainage : ,"(t 2)'

(4.r7)

pore volume where P, : the unsaturated G:maximumgravitationalwater,thatis,thetotalporosityminusthefield caPacitY Datafromsprinklinginfiltrometerstudiesofvariouswatershedsofinterestare in the usedto estimate coefficients Eq' 4'16''

5.6

EVAPOTRANSPIRATION

101

of a masstransferequationthat has often been employedfor this purpose.However, Linsley and co-workers indicate that there is some question as to the adequate The losses.27 equation is verification of this model to estimate evapotranspiration expressed as ^ E : @ 833x2(e1 er)(V, V,)

(s.2s)

E : K: 1,2 : Vr, Vz : Z =

(in./hr) evaporation (0.4) von K6rm6n's constant (in. Hg) vaporpressures (mph) wind speeds the mean temperature("F) of the layer betweenthe lower level zt and the upper level z2

is in It is assumed Eq. 5.25 that the atmosphere adiabatic and the wind speedand moistureare distributedlogarithmically in a vertical direction. In view of the sm4ll to differencesbetweenwind and vapor pressure be expectedat two levelsso closely evaporation, and sincethesegradientsare directly relatedto the sought-after spaced, highly exactinginstrumentationis required to get reliable results.

PotentialEvapotranspiration
as"the waterlosswhich will occur Thornthwaitedeflnedpotential evapotranspiration if at no time there is a deficiencyof water in the soil for the use of vegetation."In a that potential evapotranhave assumed practical sense, however,most investigators spiration is equal to lake evaporationas determinedfrom National WeatherService the ClassA pan records.This is not theoreticallycorrect because albedo(amountof of incoming radiation reflectedback to the atmosphere) vegetatedareas and soils should be As rangesas high as 45 percent.28 a result, potential evapotranspiration somewhatlessthan free water surfaceevaporation.Errors in estimatingfree water from pan recordsare such,however,asto make an adjustmentfor evapotranspiration value. potential evapotranspiration questionable of by developed the AgriAn equationfor estimatingpotential evapotranspiration Service (ARS) illustratesefforts to include vegetalcharacteristics cultural Research potentialfor any given and soil moisturein sucha calculation.The evapotranspiration day is determinedas follows:2e

E T : G I x ku , " ( # l x
where

(s.26)

potential (in./day) ET : evapotranspiration : growth index of crop in percentage maturity of GI usually 1.0-I.2 for short grasses' K : ratio of G1 to pan evaporation, height,and 1'6-2.0 for forest 1,.2-L6 for cropsup to shoulder (in./day) Eo: pan evaporation ,S: total porosity SA : availableporosity (unfilled by water) AWC : porosity drainableonly by evapotranspiration x : AWC/G (G : moisture freely drainedby gravity)

102

CHAPTER5

EVAPORATION AND TRANSPIRATION

0.2 0.1 ,.} 0.0

ti

>

n ,

0.1

Months (b)

Figure 5.5 Averagedaily consumptionof water: (a) for year 1953 by corn, followed by winter wheat under irrigation; (b) for year 1955, with irrigated first-year freadow of alfalfa, red clover, and timothy. Both measurements taken on lysimeterY 102 C at the Soil and Water ConservationResearchStation. Coshocton.Ohio. (After Holtan et al.2e)

l x F rrllS
N I

]l{

tl rH

'

Weeks

Figure 5.6 Growth index GI = ETfET^,, from lysimeter records, irrigated corn,andhayfor 1955,from Coshocton, Ohio.(AfterHoltanet al.2e)

5.7

ESTIMATINGEVAPOTRANSPIRATION 103

OF CAPACITIES SOIL TEXTURECLASSES TABLE 5.5 HYDROLOGIC

class Texture
Coarsesand Coarsesandy loam Sand Loamy sand Loamy fine sand Sandyloam Fine sandy loam Very fine sandy loam Loam Silt loam Sandy clay loam Clay loam Silty clay loam Sandy clay Silty clay Clay

(v")
24.4 24.5 32.3 37.0 32.6 30.9 36.6 32.7 30.0
JI,J

sa

GD

(Y") t'7.7 15.8 19.0 26.9 27.2 18.6 23.5 21.0 14.4 11.4 13.4 13.0 8.4 rl.6
o 1

AWC" ('/") 6.7 8.7 13.3 10.1 5.4 t2.3 13.1 11.7 15.6 19.9 tr.9 12.7 t4.9 7.8 12.3 11 . 5

x AWC/G 0.38 0.55 0.70 0.38 0.20 0.66 0.56 0.56 1.08 1.74 0.89 0.98 r.77 0.6'l r.34 r.58

25.3 25.7
z5-J

19.4
z l.+

18.8

7.3

aS = total porosity - 15 bar moisture 7o' bG : total porosity - 0.3 bar moisture 7o. "AWC: S - G. "Land Capability: A Hydrologic Response Unit in Source: Adaptedfrom C. B. England, U.SlDepartmentof Agriculture,ARS 41-172' Sept' 1970' Asiicultural'Watersheds," Aiter H. N. Holtan et a1.2e

The GI curveshave been develoPed for evapotranspiration severalcrops (Fig. : (Fig. 5.6).Equation5.26is used daily rate hydrolq USDAHL-74 model of watershed Represental evapotranspiration. late daily Table5.5.

EVAPOTRANSPIRATION 5.7 ESTIMATING

areas, Transpirationis an important componentin the hydrologicbudgetof vegetated on but it is a difficult quantity to measurebecauseof its dependence phytological It is a function of the number and types of plants, soil moisture and soil variables. temperature,and averageannual precipitation. As noted previously, type, qeason, in evaporationand tianspiration are commonly estimated their combinedevapotranspiration form. ' If the precipitation and net runoff for an atea ate known, and estimatesof of groundwateiflow and storagecan be made,rough estimates- ET canbe had using by approachdeveloped hydrologicequatiJn,Eq. 1.1. A more sophisticated ihe basic peaman foilows.t3It iJ representative the methodsmost often used' of

104

CHAPTER EVAPORATION TRANSPIRATION 5 AND

The PenmanMethod
Both the energybudgetand masstransportmethodsfor estimatingevapotranspiration (ET)have limitations dueto the difficulties encountered estimatingparameters in and in making other required assumptions. circumventsomeof theseproblems,penTo man developed method to combinethe masstransport and energybudgettheories. a This widely usedmethodis one of the more reliableapproaches estimatingETrates to 13'rs'23,30 usingclimaticdata. The Penmanequationis of the form of Eq. 5.18; it is theoretically basedand showsthat EZ is directly related to the quantity of radiative energy gained by the exposedsurface.In its simplified form, the Penmanequationisls
t s t : -

LH + 0.27E L + 0.21

(s.27)

where A : the slopeof the saturatedvapor pressurecurve of air at absolute temperature(mm Hg/'F) H : the daily heatbudgetat the surface (estimate net radiation) (mm/day) of E : daily evapoiation(mm) ET : the evapotranspiration consumptive for a given period (mm/day) or use The variablesE and.Fl are calculatedusing the following equations:

( s .28) 0.0098ar) : the saturationvapor pressureat mean ak temperature(mm Hg) where eo e6 : the saturationvapor pressure meandew point (actualvapor pressure at in the air) (mm Hg) u2 : the mean wind speedat 2 m abovethe ground (mi/day)
The equationusedto determinethe daily heat budget at the surface,11,is 11 : R(1 - r)(0.18 + 0.55.t)- 8(0.56 * 0.092e2s)(0.10 0.905) (5.29) + where R : the mean monthly extraterrestrialradiation (mm HrO evaporated per dav) : the estimatedpercentage reflecting surface of B = a temperature-dependent coefficient s : the estimatedratio of actual duration of brisht sunshineto maximum possibleduration of bright sunshine. The empirical reflectivecoefficientr is a function of the time of year,the calmness of the water surface,wind velocity, and water quality. Typical rangesfor r are 0.05 to 0.12.31 valuesof e" andA can be obtained from Figs.5.7 and 5.8, thosefor R and B can be obtainedfrom Tables5.6 and 5.7. The use of Penman's equationrequiresa knowledge vaporpressures, of sunshine duration,net radiation,wind speed, and mean temperature. Unfortunately,regular measurements theseparameters often unof are availableat sites of concern and they must be estimated.Another complication is making a reductionin the valueof EZwhen the calculationsare for vegetated surfaces. While results of experimentsto quantify reduction factors have not completelyresolved the problem, there is evidencethat the annual reduction factor is close to

E : 0.35(e"- e)(l'+

5.7
"C 60 50
I

EVAPOTRANSPIRATION ESTIMATING
.F
111

105

.F t40 't22 104 86 68 50 10 20 30 40 50 60 70 80 90 100

!?

"C 50 40

r04 86
68

313 =
ts

: o
o B t ^
14 1 0

303
ro? z6J | | 0.8 | I I I t273

ct
!d 6 6 F.

+ o
o

o F

t r 1 n

20 10

s0l. zzV |
0.2

0.4 0 6

1.0 1.2 1.4 1.6 1.8

satuated vapor pressure' ea (mm Hg)

Figure 5.7 Relation betweentemperatureand saturatedvapor Pressure.

of Value A (mmHg/'F) Figure 5.8 TemperatureversusA relation for use *i?rt ,n" PenmanLquation. (After Criddle'23)

that using to thereis evidence supportanothervalue'it appears Thus,unless unity.32-34 satisfactoryresults for surfaces a value of 1 for the reduction coefficient may give of free water evaporation having varied vegetal covers.Accordingly, aty estimate by an appropriatereduction could be used ro estimaieEZ, providin[ it is modified coefficient.

EXAMPLE 5.4

ET, giventhe following data: ': using the PenmanMethod, Eqs.5.21 to 5.29, estimate : 30 degrees c' c, 20 degrees temperatureof 1l temperatureat warer r"tru"" month is : (48 mi/day)' the relative humidity : +O p"r""nt, wind-velocity i mph r is given ut 0'07' and S is found to be 0'75' north, Juneat latitude 30 degrees Solution l.Giventhedatafortemperature,thevaluesofeoandeacanbedetermined. UsingFig.5.TorAppendixTableA.2,thesaturatedVaporpressuresare 31'83'andfor foundto be l7.53unO-:t'Sl mm Hg respectively' : ThI'"-: : 31'83 X 0'4 12'73' a relativehumidity of 40 percent,e,t Then,usingEq. 5.28' + .83 E : 0.35(31 - 12.73)(1 0.0098x 48) E:9.83 mm/daY 2.ThevalueofAisfoundusingFig.5.8;forthegivenlatitudeandmonth,R isobtainedfiomTable5.6;andBisgottenfromTable5.'Tforatemperature : I7 '01' found are A : 1'0' R : 16'5' andB of 30"C.fne vatues Then,usingEq.5.29, + H: 16'5(I - 0'07)(0'18 0's5 x 0'75) - 17.01(0.5- 0.092x 12j30)(0.10 + 0'90 x 0'75) 6 = 6.04 mm/daY H

O \ O h O \ 6 \ t o O \ \ O r ) O o r) F. O N + r) \O F-

t--

l--

F-

\O F- :+ f.-' r) O\ O F- * C.l * d \ + \ O O \ i o * r ) \ O \ O \ O \ O n

Ol

o
LL 6

ft

o
IL

trJ

t- * co o o\ * o co cn oo oo n tF-O\dcl\fh!nv1<.cqolO

F-

f U)
J

tr

oN catf,r +s+coNooo\o
: : * * * d d i * -

nv?qnnqqnv.t-nv.tol

z o o
I
tf

F.

O\

cn ci

e.l O

c.i ri + r; ri ri {

cd -.j oi F ri c.i

\n

\O h

O\

I = o
E.

r.)

\O \O \O r.)
* * i

-.: dl .l \

!f
*

oq v? q g oq 9 q cl
c.l r O f\ \A tri d
d

(r o q)
lJ-

r F- tr - \r ) $\O\O**6 \o \o \o \o r) * c7) d o\ F- r) N o

o 6 z
F

9nqq\qqn\\-q*:
$ h n \ o r ) h c n N o \ o $ *
* i i i i * i * d

uJ

z
J T

-j 6i di + ri ri + ri 6i ci *, <j +

F-

O\ oo a]

a-

t1

oo oo \O o

o
4 Z tIO

\ O O \ : c . l o $ n h * o a l ( ) H * * * * r *

e-:q\qoq.l-9qqvln

>tr
u O olu 6 Z ui< f > 1 z
t=P

ds

n q q n r t

q q \ g o q q - \

O r} oO O al cn r) r) \O r) r)

<+ C.l

S IIJ 2O-

cn\oor)66r)ooooo6*\o .j d; \o od o 6i + r; \ci i-- r- r- \c; i : :

ft+
) z
muJ 4 ,Ct) F ] q lo UJ 6

s
z

\Or ) *6al -

- 6l o+r ) \O
E V

E:
a

> 6

SUMMARY

107

TABLE5.7 VALUES TEMPERATURE-DEPENDENT OF B FORUSEINTHEPENMAN COEFFICIENT EQUATION

Tu (K)

B (mmHrO/day)

T" fF)
Jf

B (mm HrO/day)

270 275 280 285 290 295 300 305 310 315 320 325

r0.73
11.51 12.40 13.20 14.26 15.30 16.34 r7.46 18.60 19.85 21.15 22.50

40
A<

50 55 60 65 70 75 80 85 90 95 100

11.48 I 1.96 12.45 12.94 13.45 13.96 14.52 15.10 15.65 16.25 16.85 t'7.46 18.10 18.80

Source: Afler Criddle.23Note that B = oT1 where o is the Boltzmann c o n s t a n t . 2 . 0x l 0 - e m m / d a y . l

3. UsingEq. 5.27, Er : 0.0 x 6.04 + 0.27 x 9.83)/(1+ 0.27) ET : 6.85 mm/day is Thus the estimated evapotranspiration 6.85 mm/day. ll

Simulating Evapotranspiration
The volume of water evaporatedor transpiredfrom a watershedover time can be shouldincorpoAccordingly, continuoushydrologicmodelingprocesses substantial. rate an EZcomponent.The modelsgiven in this chaptertypify suchan approach(see aboundon this subjes1.28'2e'rs':o':z alsothe flow chart of Fig. 1.3).References

Summary
Figure 1.1 and Table5.1 showthe overall importanceof ET in the hydrologicbudget. annual precipitation by a In many regionsof the United States,annual ET exceeds and use must development significantamount.As a result, plans for water resources incorporate estimatesof ET losses.Where irrigated agriculture is practiced, these are estimates especiallyimportant. They generally to A numberof approaches estimatingEThavebeendeveloped. analyttheoretical,basedon the physicsofthe process; fall into the following classes: Equaical, basedon energyor water budgets;and empirical, basedon observations. are tions usedin making ET calculations usually of the type illustrated by Eqs. 5.1, 5.8,5.10,5.19,5.22,and5.26.

1oB

cHAp+R 5 EVApoRATtoN ANDTRANsptRAloN

PROBLEMS
5.1. An 8000-mi2watershed received20 in. of precipitationin a 1-yearperiod. The annual streamflowwas recorded as 5000 cfs. Roughly estimatethe combined amounts of water evaporated and transpired.Qualify your answer. 5.2. Find the daily evaporation from a lake during which the following datawere obtained: air temperature 90oF,water temperature60oF,wind speed 20 mph, and relative humidity 30 percent. 5.3. Find the monthly consumptiveuse of an alfalfa crop when the mean temperatureis 70"F, the average percentageof daytime hours for the year is 10, and the monthly consumptiveuse coefficientis 0.87. 5.4. During a given month a lake having a surfacearea of 350 acreshas an inflow of 20 cfs, an outflow of 18 cfs, and a total seepage of 1 in. The total monthly precipitaloss tion is 1.5 in. and the evaporationloss is 4.0 in. Estimatethe changein storage. 5.5. What are two filethodsthat might be usedto reduceevaporationfrom a small pond? 5.6. Computethe daily evaporationfrom a ClassA pan if the amountsof water required to bring the level to the fixed point are as follows:

Day Rainfall (in.) Water added(in.) Evaporation

1 0 0.29

2 0.65 0.55

3 0.12 0.07

4 0 0.28

5 0.01 0.10

5.7. For Problem 5.6, the pan coefficientis 0.70. What is the lake evaporation(in inches) for the 5-day period for a lake with a 250-aqe surfacearea? 5.8. The pan coefficientfor a ClassA evaporationpan locatednear a lake is 0.7. A total of 0.50 in. of rain fell during a given day. Determinethe depth of evaporationfrom the lake during the sameday if 0.3 in. of water had to be addedto the pan at the end of the day in order to restorethe waterlevel to its original valueat the beginningof the day. 5.9. A 2500 mi2drainagebasinreceives25 in.lyr rainfall. The discharge the river at the of basinoutlet is measured an average 650 cfs. Assumingthat the change storage at of in for the system essentiallyzero, estimatethe EZlossesfor the areain inchesand cm is for the year. Stateyour assumptions. 5.10. Determine the daily evaporationfrom a lake for a day during which the following mean values were obtained: air temperature78'F; water temperature62oF; wind speed,8 mph; and relative humidity, 45 percent. 5.11. Usingthe Meyer and Dunne equations, find the daily evaporation rate for a lake given that the mean value for air temperaturewas 80T, for water temperature60'F, the ziverage wind speedwas 10 mph, and the relative humidity was 25 percent.Refer to Appendix Table A.2 for vapor pressrire values. 5.12. Determinethe seasonal consumptiveuse of truck crops grown in Pennsylvania the if meanmonthly temperatures May, June,July, and August are 62,71,16, and 75'F for respectivelyandthepercentdaylighthoursforthegivenmonths arc10.02,10.1,10.3, and 9.6 as percent of the year respectively.

REFERENCES

109

5.13. Using the PenmanMethod, Eqs.5.27 to 5.29, estimateET, giventhe following data: temperatureat water surface: 20 degreesC, temperatureof air : 32 degrees C, relative humidity : 45 percent,wind velocity : 3 mph, the month is Juneat latitude 30 degrees north, r is given as 0.08, and S is found to be 0.73.

REFERENCES
1 . WaterResources Council,TheNation's WaterResources: 1975-2000,U.S. Govt. Print. Off., Washington, D.C., 1978. 2 . U.S. Department of Commerce, National Oceanic and Atmospheric Administration, National WeatherService,NOAA TechnicalReport 33,EvaporationAtlas of the Contiguous 48 UnitedStates, Washington, D.C., June1982. 3 . U.S. Department of Commerce, National Oceanic and Atmospheric Administration, National WeatherService,NOAA TechnicalReport NWS 34, "Mean, Monthly Seasonal, and Annual Pan Evaporationfor the United States,"Washington, D.C., June 1982. 4. "Water-LossInvestigations,"Vol. 1, Lake Hefner Studies,U.S. GeologicSurveyProfessionalPaperNo. 269 (1954). (Reprintof U.S. Geological SurveyCitc.229, 1952.) 5 . N. N. Gunaji, "EvaporationInvestigations ElephantButte Reservoir,New Mexico," ^lnl. at Assoc.Sci.Hydrol. Pub. 18, 308-325(1968). 6. I. S. Bowen, "The Ratio of Heat Lossesby Conduction and by Evaporationfrom Any Water Surface,"Phys.Rev. 27, 779-7 81(1926). 7 . E. R. Anderson,L. J. Anderson,andJ. J. Marciano, "A Reviewof Evaporation Theory and Development Instrumentation,'o of Lake Mead WaterLoss Investigation;Interim Report, Navy Electronics Lab. Rept.No. 159 (Feb. 1950). 8 . O. G. Sutton, "The Application to Micrometeorologyof the Theory of Turbulent Flow over RoughSurfaces," Meteor,Soc.Q. "I. 75(No. 236),335-350(Oct. 1949). R. 9 . A. F. Meyer, "Evaporationfrom Lakes and Reservoirs,"MinnesotaResources Commission,St. Paul,June1944. 1 0 . T. Dunne and L. B. Leopold, Waterin EnvironmentalPlanning, San Francisco:Freeman and Co., 1978. 1 1 . V. M. Ponce,EngineeringHydrology: Principles and Practices.EnglewoodCliffs. New Jersey: PrenticeHall, 1989. 1 2 . M. A. Kohler, T. J. Nordenson, and W E. Fox. "Evaporationfrom Pansand Lakes," U.S. Department Commerce, of Weather Bureau, Res.PaperNo. 38, Washington, D.C., 1955. 1 3 . H. T. Haan, H. P. Johnson,and D. L. Brakensiek(eds.),Hydrologic Modeling of Small Watersheds, ASAE MonographNo. 5. St. Joseph, MI: American Societyof Agricuitural Engineers,1982. 14. R. K. Linsley,M. A. Kohler, and J. L. H. Paulhus, Hydrologyfor Engineers,3rded. New York: McGraw-Hill, 1982. 1 5 . H. L. Penman,"Natural Evaporationfrom Open Water,Bare Soil, and Grass,"Proc. R. Soc.LondonSer.A 193(1032),120-145(Apr.1948). 16. F. G. Millar, "Evaporationfrom FreeWaterSurfaces,l' CanadaDepartmentof Transport, Division of MeteorologicalServices,Can, Meteor. Mem. vol. l, No. 2, 1937. 1 7 . J. B. Franzini, "Evaporation Suppression Research,"Water and SewageWorks (May 1961). 1 8 . Victor K. La Mer, "The Case Evaporation for Suppression," Eng.(June10, 1963). Chem. 1 9 . D. R. Maidment(ed.),Handbookof Hydrology. New York: McGraw-Hill ,1993. 20. H. F. Blaney,"Water and Our Crops," inWater, the Yearbook Agriculture. Washington, of D.C.: U.S. Department Agriculture,1955. of

110

AND 5 CHAPTER EVAPORATION TRANSPIRATION "Monthly Consumptive Use Requirementsfor Irrigated Crops," Proc. ZI. H. F. Blaney, J. ASCE, Irrigation DrainageDiv. 85(IR1), 1-12(Mar' 1959)' 22. yen Te Chow (ed.), Handbook of Applied Hydrology. New York: McGraw-Hill ' 1964. "Methods of Computing ConsumptiveUse of Water." Proc. ASCE J. 23. W. D; Criddle, Irrigation DrainageDiv. 84(IR1), 1-27(Jan' 1958)' ,;Estimating Evaporation," Trans. Am. Geoplrys. union 37(l), 4324. H. L. penman, "Reductionof Transpiration," Geophy. Res.66(10), 3309-3312(Oct. J. 25. N. J. Roberts, 1961). "Research control of Phreatophytes," Proc. Ninth Annual water conon 26. E. H. Hughes, ference,New Mexico StateUniversity,Las Cruces,Mat. 1964' New York: Hydrology Engineers. for 21. R. K. Linsley,Jr., M. A. Kohler, and J. L. H. Paulhus, McGraw-Hill, 1958. "National WeatherServiceRiver ForecastSysLaboratory, 28. Staff, Hydrologic Research tem ForecastProcedures,"NWS HYDRO 14. U.S. Departmentof Commerce,Washington. D.C..Dec. 1972. "usDAHL-74 Revised 29. H. N. Holtan, G. J. Stiltner,w. H. Henson,and N. C. Lopez, ARS Modelof WatershedHydrology," Tech.Bull. No. 1518.U.S.DepartmentofAgriculD.C., 1975. ture, Washington, "Evapotranspiration-Review of ReH. N. J. Rosenberg, E. Hart, and K. w. Brown, Lincoln, 1968' search,"MP 20, Agricultural ExperimentStation,University of Nebraska, R. H. McCuen, HydrologicAnalysis and Design.Englewoodcliffs, New Jersey:Prentice 31. Hall, 1989. "Correlation of ClimatologicalData with WaterR'equire3 2 . W. O. Pruitt and F. J. Lourence, University of California Water ScienceEng. Paper9001. Davis, June ments of Crops." 1968. "Potential Evaporation:The cornbination concept and Its Experi33. C. H. M. Van Bavel, Res.2' 455-467(1966)' mental Verification," WaterResources "Discussionof Paperby H. L. Penman,'Estimating Evaporation,"' Trans' 34. H. F. Blaney, Union37,46-48(Feb. 1956). Am. Geoplrys. "Digital Simulationin Hydrology:StanfordWater35. N. H. Criwford and R. K. Linsley,Jr., shedModel IV," Tech. Rept. 39, Departmentof Civil Engineering,Stanford University' July 1966. 36. W. C. Huber, J. P. Heaney,S. J' Nix, R. E. Dickinson, and D"J' Polmann' Storm Water ManagementModel (lsei's Manual, Version /11, EPA-60012-84-109a(NTIS PB84198423).Cincinnati, oH: EnvironmentalProtection Agency, Nov. 1981. Model User's R. 37. L.A. Roesner, P. Shubinski,and J. A. Aldrich, StormWaterManagement (NTIS PB84-198431). I, Manual, versionIII: Addendum Extran, EPA-600/2-84-109b Cincinnati, OH: EnvironmentalProtection Agency,Nov' 1981'

5o(1es6).

6 Chapter

Streamflow

r Prologue
The purposeof this chapteris to: . Introduce the conceptof streamflow. . Describethe characteristics a hydrograph. of ' Presentapproaches measuringstreamflow. to The amountof water flowing in surfacewater coursesat any instant of time is small importanceto those in termsof the earth's total waterbudget,but it is of considerable A supply,and management. knowledge development, waterresources with concerned quantity and quality of streamflowsis a requisite for: municipal, industrial, of the flood control; reservoir designand agriculiural, and other water supply endeavors; recreation;navigation;fish hydroelectricpower generation;water-based operation; management;drainage; the managementof natural systemssuch as und .ildlit" treatment. wetlands;and water and wastewater by is generated precipitationduring stormeventsand by groundwaStreamflow ter entering surface channels.During dry periods, streamflowsare sustainedby reservoirsare below streamchannelsWhere groundwater groundwater discharges. to in arid regions-streams cease flow during protractedprecipitationoften the case periods. Relations between precipitation and streamflow are complex, being free influJnced by the factors discussedin the foregoing chapters.As a result, many to approaches relating theseimportant hydrologicvariableshavebeen developed.l-3 in Severalof them are discussed detail in Part Three of the text' of streamflow are based on the use of flow-measuring Field measurements of devicessuchas weirs and flumes and on the measurement channelcross-sections velocities(seeChapter 8). along with streamflow

BASINEFFECTS DRAINAGE
in The quality and quantity of streamflowgenerated a drainagebasin are affectedby Accordingly,it is important and climatic features.a-e physical,vegetative, the bisin's ofthe soils,rocks,plants,topography, havea good understanding hydrologist that the

J12

CHAPTER STREAMFLOW 6 of that land-usepatterns,and otherbasincharacteristics influencethe sequence events precipitation and runoff. It should be pointed out, however,that while separating natural basin featuresare very important elementsin the runoff process,land-use parking lots, agricultural ieatures createdby humans (e.g., housingdevelopments, practicescan patterns)may,in somecases, the dominantones.Land management be suchasin retardingerosion,and they can alsobe detrimentalwhenthey be beneficial, In natural hydrologicprocesses. Chapter10,the principal basin function to accelerate of concernto the hydrologistare discussed. characteristics

6.2 THE HYDROGRAPH

This by at Streamflow, a givenlocationon a watercourse,is represented a hydrograph. graph displays propertiesof streamflowwith respectto time, normally the continuous obtained by nteansof a continuousrecorder that shows stage(depth) versustime hydrographby applica(stagehydrograph),and is then transformedinto a discharge curve. In general, the term hydrograph as used herein means a tion of a rating hydrograph. discharge As was shownin Fig. 1.3, the hydrographproducedin a streamis the result of that occur-duringand after any precipitation event. A various hydrologicprocesses has is of theseprocesses given in Chapter11. A hydrograph more completediscussion or (1) runoff, (2) interflow, (3) groundwater elements: direct surface four component (4) channelprecipitation.The rising portion of a hydrographis known baseflow, and as the concentrationcurve; the region in the vicinity of the peak is calledthe crest depends The andthefalling portion is therecession.lo shapeof a hydrograph segrnent; and precipitationpattem characteristics basinproperties.Figure 6.1 illustratesthe on definitionspresented.

Stormperiod hydrograph

'{1 oa-. . I'a

\%
. Continuous hydrograph

Endof direct runoff

Time Figure 6.1 Hydrograph definition.

6.4

AND RECORDING STREAMFLOW MEASURING

113

FOR 6.3 UNITSOF MEASUREMENT STREAMFLOW

They are units of water flowing in streams. Two types of units are usedin measuring and units of volume.Discharge, rate of flow, is the volume of water that or discharge passes particularreference point in a unit of time. The basicunits usedin connection a of with streamgaugingin the United Statesare the foot and meter for measurements of dimensionand the secondfor measurements time. Commonly usedunits of dischargemeasurement are cubic feet per second(cfs) and cubic meters per second per (m3/sec). in Other units of discharge useare second-foot squaremile (sec-ft/mi2), from a drainagebasinor definedarea,and for expressing average of discharge the rate Units of million gallonsper day (mgd),commonly usedin water supplycalculations. volume used are the cubic foot, cubic meter, liter, gallon, and acre-foot (a volume equivalent to 1 ft of water over an acre, 43,560 ft2, of land). The latter unit is commonly used in irrigation practice in the westernUnited States.Irrespectiveof the whetherEnglish or metric units are usedfor dimensions, standardunit of time for is streamflowobservations the second.

AND RECORDING STREAMFLOW 6.4 MEASURING

usinggaugingdevicessuchas flumes,weirs,and Streamflowratesmay be determined of control sections,or they may be calculated from measurements head (depth), velocity, and cross-sectionalarea.t-t Usually, specific devices such as flumes are

' bI) 6 6
o bo ri

3000
Discharge (sec-ft)

Figure 6.2 Station rating curve for RaquetteRiver at Piercefield,New York. (U.S. GeologicalSurvey.)

114

CHAPTER6

STREAMFLOW

limited to small streamsbecauseof problems of scale.For large stream systems, velocity and using the cross-sectional is discharge normally estimatedby measuring devicesare Where flow-measuring into areato translatethis measurement discharge. rating curve to translatethis into used,it is customaryto observethe head and use a calculations cannot be made, discharge When direct flow measurements discharge. chemical tracers, electrical are often facilitated by use of velocity-area relations, methods,and empirical equationssuchas the Manning formula' In the United States,the primary responsibilityfor gaugingmajor streamslies record of streamflowin terms of with the U.S. GeologicalSurvey,with a systematic mean daily dischargebeing the norm. Usually, a stagerecording is obtained at a by gauging siie and this record is convertedinto discharge one or more of several for this purpose(seeChap-"tttoAr. Rating curves,tables, and formulas are used rating curve' The instrumentsand ter 11).Figure 6.2 rho*r a typical stage-discharge to mustbe carefully adapted the to recordings discharge methodsusedto convertstage gauging site so as to ensurethe natural or artiflcial conditions encounteredat the reliability of conversions.'

AREA AND CROSS.SECTIONAL OF 6.5 MEASUREMENTS DEPTH

devicecanbe installedin a stream,aratity for streams Unlessa direct flow-measuring areawill be needed of of any scale,measurements depth of flow and cross-sectional may be taken using to permit dischargeto be calculated. Depth measurements weighted soundin! [nes, calibrated rods, and ultrasonic soundingdevices.Crosssectionalareasat streamsectionscan be determinedusing ordinary surveyingtechthat ate taken below or niquescombinedwith soundings other depth measurements of measurement depth and the water level at the time of the survey.Although the requirecarefulcalibration determinations simple,accurate al cross-section areaseems conditions. of instrumentsand the ability to deal with submerged

OF 6.6 MEASUREMENT VELOCITY

area,permit calculacombinedwith thoseof cross-sectional Velocity measurements, Point flow velocities can be tion of dischargeat a given stream or river location. devicessuch as the Pitot tube, dynamometer' determinedusing velocity-measuring and current -"i"r. In the United States,the Price current meter has long been a cuppedvanesto the by standardin streamflowgauging.This deviceoperates exposing measuringwind velocity. The used in direction of flow, *o"h lik" the anemometer rotatesin nearproportion to flow velocity and the rate of rotation cup-vaneassembly is convertedto point velocity using a rating table or appropriateequation' , Various chemical and electrical methods are also employed in determining velocities.Commonly usedchemicalmethodsinclude salt velocity, salt dilution' and is the detectionof radioactivetracers. Of these methods,the salt velocity method principle that salt introducedinto the perhapsthe most widely used.It is basedon the of streamwill increaseits electricalconductivity.Electrodesplaceddownstream the

FLOW TO POINT VELOCIry CROSS-SECTIONAL VELOCIry 6.7 RELATING

115

profile Figure6.3 Verticalvelocity

and thesecan be translatedinto velocity by conductivity changes salt injection sense keeping track of time. Electrical methods knowing the spacingof electrodesand oxygen voltagegeneration, electromagnetic includethe useof hot wire anemometers, Streamflowconditionsvary widely and thus no polarography, waves.a and supersonic specificapproachto velocity determinationis universally suitable.The best method of for use at a given site must be determinedon the basis of the characteristics streamflowat that site. Field observationshave shown that the mean velocity in a verticai stream of by sectionis closelyapproximated the average the velocitiesoccurringat depthsof (seeFig.6.3).11'12 percentof the total sectiondepthrespectively 20percentand 80 on the orderof 0.5 feet, singlemeterreadingsat about Where depthsarevery shallow, the 50 percentpoint havebeen shownto yield good results.t' Velocity measurements permit estimatingchannel and geometricdefinitionsof streamchannelcross-sections havebeen made. velocity and depth measurements flows at locationswhere

FLOWVELOCITY TO POINT VELOCITY CROSS-SECTIONAL 6.7 RELATING

are While point velocity measurements important, what is desiredis a method to velocity, flow velocity. This average cross-sectional translatethem into the average area,yields the dischargeat a given stream cross-sectional when multiplied by the at section.One procedureis to take point velocity measurements numerousvertical positions in a crosi section,plot them, and then determinevelocity and horizontal of the contours.By calculatingthe areasbetweenthe contoursand assigning average of the two confining contoursto theseareas,a determinationof the flow velocities is discharge easilycalculated. meanvelocity can be made.Once this is accomplished, make use of the geometric properties of stream channel Other approaches method.To usethis approach, is One cross-sections. suchtechnique the mean-section at to divide the streamchannelcross-section a gauginglocation into a it is necessary

116

CHAPTER6 STREAMFLOW
Water surface

6.1 for cross-section Example Figure 6.4 Channel seriesof geometricshapes(seeFig. 6.4). AL eachvertical location along the crossvelocity of The section,the meanvelocity is estimatedfrom measurements. average of to areabetweentwo verticalsis considered be equalto the average the flow for the betweentwo vertifor eachof the borderingverticals.The discharge meanvelocities velocity for the sectionmultiplied by the areaof the section. cals is thus the average total flow for the are discharges then summedto provide an estimated The individual to at that location. Note that it is important to have enoughmeasurements channel is The the cross-section. procedure illustratedin Example6.1. characterize
EXAMPLE 6.1

at Calculatethe discharge the sectiongiven in Fig. 6.4. Data from field observations are shownin Tables6.1 and6.2.
TABLE 6.1 DATA FOR EXAMPLE6.1

6.1 TABLE6.2 DATAFOREXAMPLE

section # Vertical 0
I

Depth (ft.) 0
4

Avg. vel.

(rps)
0 2.1 2.3 2.7 2.8 2.5

2
3

'7.2
7.4

5
6 7 4.'1

2.2
0

FORDETERMINING DISCHARGE 6.8 THESLOPE-AREA METHOD TABLE6.3 CALCULATIONS FOR EXAMPLE 6.1 Area
A1

,117

Area (sq.ft.) 8.40 14.85 29.28 3't.96 26.83 30.09 13.87 t61,.27

Vel. (fps) 1.05 2.20 2.50 2,75 2.65 2.35 1.10

Flow (cfs) 8.82

A2 A3
A4

32.67 73.20 104.39 7t.09 70.71 15.25 376.t3

A5 A6 A7

Total estimateddischargeis 376.13 cfs

Solution. The first step is to calculatethe individual section areas(A1, A2, etc.). The calculatedareas(usingtriangular or trapezoidalformulas)are shown meanvelocitiesat the verticalsare on spreadsheet Table6.3. Next, the estimated (see multiplied by the section areasto obtain the individual area discharges Table 6.3). Thesedischarges summedto yield an estimated376.13 cfs of are flow being deliveredby the full channelwidth. r I

DISCHARGE METHOD FORDETERMINING 6.8 THE SLOPE.AREA

neededto deter: it In somecases is difflcult to make velocity or other measurements mine discharge. This is often the caseduring large flood events.Under suchcircumof possible estimatethe flow by taking measurements high stances, is sometimes it to areas,and channelslopesand then water lines (after the flood event),cross-sectional the usingthesedatain an equationsuchasManning's to estimate flow. The applicable Manning equationis Q : \I.49ln)APzrzgrrz where Q : n : A : R: S: discharge(cfs) coefficient Manning's roughness cross-sectional area(ft2) the hydraulic radius the headloss per unit length of channel

(6.1)

For streamflows, Manning's n valuesmay rangebetweenabout0.03 and 0.15. When reasonabledeterminationscan be made of n, A, R, and S, Eq. 6.1 can be used to estimate the streamflow that occurred during the high-water period. For a of more completediscussion this and other streamflowdeterminationmethods,the

118

CHAPTER6

STREAMFLOW

4 references the end ofthe chaptershouldbe consulted.References and 10 give an at excellent overview of techniques and include a valuable list of references on streamflow.

r Summary
Streamflowis the result of storm-periodprecipitation, snowmelt,and groundwater discharge.l3 is a primary sourceof water for a host of instreamand offstreamuses. It The graphicalrepresentation streamflowis the hydrograph,a plot of flow versus of time at a prescribedlocation alongthe water courseof interest.As illustratedby Fig. is 1.3, the end product of many hydrologicmodelingprocesses a hydrographwhich is derived from a precipitation input, modified appropriatelyby various abstractions in streamflowin the field were presented suchas infiltration. Methodsfor measuring this chapter.In later sectionsofthe book, a variety oftechniquesfor deriving hydrographs from precipitation and other hydrologic data are covered (seePart Three).

PROBLEMS
6.t. Considerthat you haveobtaineda gaugeheightreadingof4 ft at a gaugingsite on the
to RaquetteRiver (Fig. 6.2). What would you estimatethe discharge be in cfs and in be? m3/sec? the gaugeheighthad been9 ft, what would the discharge Which of the If two estimatesdo you think would be the most reliable?Why? 6.2. SolveProblem 6.1 if the gaugeheight readingswere 5 ft and 7 ft. 6.3. Consulta USGSWaterSupplypaperand plot the streamflowdatafor a drainagebasin ofinterest.Discussthe factorsthat you believeinfluencedthe shapeofthe hydrograph. 6.4. For the major surfacewater course in your locality, discussthe value of making streamflowforecasts. at 6.5. Calculatethe discharge the sectiongiven in Fig. 6.4 if the depth measurements at Give resultsin the verticalswere:0, 3.8, 5.4, 7.1,8.1,7.0,4.5, and 0 ft respectively. cfs and m3/s. Calculatethe discharge the sectiongiven in Fig. 6.4 if the velocitieswere: 0, 2.3, at and 2.6,3.1,2.9,2.7,2.5, and0 fps respective$, the depthsof Problem6.5 applied. Give resultsin cfs and m3/s.

REFERENCES
1 . N. C. Grover and A. W. Harrington, StreamF/ow. New York Wiley, 1943. 2. United StatesDepartmentof the Interior, Bureau of Reclamation,Water Measurement 3 . I. E. Houk, "Calculation of Flow in Open Channels,"Stateof Ohio, The Miami Conservancy District, Tech. Rept. Part IV, Dayton, OH, 1918. Manual. Washington,D.C.: U.S. GovernmentPrinting Office, 1967.

4. Ven Te Chow (ed.), Handbook of Applied Hydrology. New York: McGraw-Hill, 1964. 5 . American Society of Civil Engineers,"Hydrology Handbook," Manuals of Engineering
Practice. No. 28. New York: ASCE. 1957.

REFERENCES 1 19 6 . O. E. Meinzer, Hydrology.New York: Dover, 1942. 7 . A. N. Strahler, "Geology-Part II," Handbook of Applied Hydrology. New York: McGraw-Hill. 1964. 8 . R. E. Horton, "Drainage Basin Characteristics,"Trans.Am. Geoplrys.Union l3r 35O36r(1932\. 9 . W. B. Langbeinet al., "TopographicCharacteristics DrainageBasins,"U.S. Geological of Survey,Water Supply Paper,968-c, 1947. 1 0 . R. K. Linsley, Jr., M. A. Kohler, and J. L. H. Paulhus,Applied Hydrology. New York: McGrawHill, 1949. 1 1 . S. S. Butler, EngineeringHydrology.EnglewoodCliffs, New Jersey:Prentice-Hall, Inc., 1951. 12. C. H. Pierce,"Investigationof Methodsand EquipmentUsedin StreamGauging,"Water Supply Paper 868-A, U.S. GeologicalSurvey,Washington,D.C.: GovernmentPrinting Office,1941. 1 3 . D. R. Maidment (ed.),Handbookof Hydrology.New York: McGraw-Hill , 1993.

PART TWO

ENTS MEASUREM HYDROLOGIC AND MONITORING

Chapter7

DataSources Hydrologic

r Prologue
The purposeof this chaPteris to: . Describethe principal sourcesof data for hydrologicinvestigations. and modeling. forecasting, Data on hydrologicvariablesare fundamentalto analyses, publications of state and federal agencies, Such data *uy 1" found in numerous Severalofthe most significant and otheroganizations. universities, institutes, research in this chapter'1-3 briefly sourcesof hydrologicdata are described

DATA 7 . 1 GENERALCLIMATOLOGICAL
solarradiation,wind, and of The most readily availablesources data on temperature, publishedby the EnvironrnentalData humidity are ilimatological Data bulletins Service of the Nationai Oceanic and Atmospheric Administration (NOAA), and by Monthly Summaryof Solar RadiationData publtshed the National Climatic Data in cooperationwith the World MeteorologCenter.The EnvironmentalData Service, ical Organization(WMO), also publishesMonthly Climatic Data for the World- A 1968 publication of the Environmental SciencesService Administration, entitled humidity, evapowind, temperature, Climalic Atlasof the UnitedStates,summarizes seriesof maps. In addition-to these ration, precipitation, and solar radiation on a and agriculfederal sourcesof data, stateenvironmental,geologic,water resources, universitiesalso publish a variety of shouldbe consulted.Most state tural agencies hydrologicdata through their researchcentersand extensionprograms.

DATA 7.2 PRECIPITATION

There are probably more records of precipitation than of most otherhydrologic The priniipal federalsourceof data on precipitationis NOAA. Climatologvariables. the ical Data, publishedmonthly and annuallyfor eachstateor combinationof states, pacific area,PuertoRico, and the Virgin Islandsby the EnvironmentalData Service,

E # = ,

124

CHAPTER7

DATA SOURCES HYDROLOGIC

departuresfrom normal, and extremes of presentsa table of monthly averages, precipitation and temperatureas well as tables of daily precipitation, temperature, Hourly Precipitawind, and soil temperatarc. snowfall,snowon ground,evaporation, tion Data is issuedmonthly and annuallyfor eachstateor combinationof statesand presentsalphabeticallyby station the hourly and daily precipitation amounts for A with recordingauges. stationlocationmap is alsoincluded.This stationsequipped publication is availablefrom the EnvironmentalData Service.Another publication, periods.Data are regionsfor 10-year by is Records, issued geographic World Weather listed by country or areaname, station name,latitude and longitude,and elevation. pressure, and tempersea-level Monthly and annualmeanvaluesof stationpressure, order.Aside aiure, and monthly and annualtotal precipitationare given in sequential and universitiespublish precipitation from NOAA, other federal and state agencies dataat varying intervals,often in a storm o1 site-specificcontext.In addition, many utilities also collect and maintain precipitamunicipalitiesand water and wastewater tion and other related data. Computeized precipitation data are availablefrom the National Climatic Data Centerin Asheville,North Carolina.

DATA 7.3 STREAMFLOW

of The principal sources streamflowdatafor the United Statesare the U.S. Geological and (SCS), U.S.ForestService, U.S. Service U.S. Soil Conservation Survey(USGS), Service(ARS). In addition,the U.S. Army Corpsof Engineers Agricultural Research (COE), the Tennessee Valley Authority (TVA), and the U.S. Bureauof Reclamation and (USBR) make somestreamflowmeasurements tabulatestreamflowdata relative also organizations and variousresearch universities, Stateagencies, to their missions. compile and publish a variety of streamflowdata. The USGS Water Supply Papers(WSf; are the benchmark for referencing data are also availablefrom the USGS. streamflowdata. Furthermore,computerized Publications of the Geological Survey, publishedevery 5 years and supplemented annually, arean excellentsourceof information on that agency'sreports. The SCS historically published data on streamflow from small watershedsand plots in its Hydrologic Bulletin series,but much of the data have been republishedby ARS. "pilot watersheds" publishedin cooperationwith the USGS. are Recordsfrom SCS U.S. Forest Service streamflow data are publishedat irregular intervals in various papers. technicalbulletins and professional

DATA AND.TRANSPIRATION 7.4 EVAPORATION

of Monthly and annualissues ClimatologicalData, publishedby NOAA, includepan and water utilities are evaporationand related data. The ARS, agricultural colleges, are often other sourcesof information. In particular, data on evapotranspiration working through their Agricultural Experiment obtained by university researchers Stations.

REFERENCES 125

r summary
Numerous in Climatic and other data are keystones hydrologicmodelingprocesses. to sources of data exist and may be accessed support model developmentand verification, statisticalanalyses, and specialstudies.

PROBLEM
7.1 Developa list ofdata sources your stateorlocality by visiting the library or through in other channels.

REFERENCES
1 . Soil ConservationService,U,S. Departmentof Agriculture, SCSNational Engineering Handbook, "Hydrology", Sec. 4. Washington,D.C.: U.S. GovernmentPrinting Offlce, Au,g.1972. J. F. Miller, "Annotated Bibliography of NOAA Publications of Hydrometeorological Interest," NOAA Tech. Mem. NWS HYDRO-22, National Oceanic and Atmospheric Administration, Washington, D.C., May 1975. , D. R. Maidment (ed.), Handbookof fudrology. New York: McGraw-Hill, 1993.

ChapterB

lnstrumentation

Prologue
The purposeof this chapteris to: ' Describeinstrumentsusedin measuring hydrologicvariables. . Indicate waysin which data are recordedand transmitted. . Presentlimitations on measurements. must be obtainedin sufficientquanThe data neededto supporthydrologicanalyses frequency,and in an appropriateform if they are to be of value. tity, with adequate are A variety of instruments usedto obtain and transmitthe data.They are the subject of this chapter.

8.1 INTRODUCTION
problemanalyses, research, supportsarealinvestigations, Hydrologicinstrumentation are A planning,and environmentalpolicymakingand analysis. hostof measurements neededto support efforts in water resourcesplanning, management,design, and analysis,solid wastemanageconstructionrelatedto suchsubjectsas aquifer systems water supplyavailability,water quality management, ment, flood hazardassessment, groundwaterrecharge,protection of fish and wildlife, and navigation. Historically,instrumentswere often usedto obtain cumulativerather than conFqrtinuousinformation abouthydrologicvariablessuchasrainfall and evaporation. thermore,there was often no attemptto correlatewater quality constituentloadings, many historic data havelimited for example,with ratesof water flow. Consequently, but instrumentation, ratherfrom using of utility, not somuchbecause lack of adequate availableinstrumentsto measurethe wrong thing or in too limiting a fashion.Today it is widely recognizedthat it is important not only to selectappropriateinstruments of but to selectthem in the contextof datanetworksthat meetthe needs moderntimes. More will be said about this in Chapter9, In Section8.2, instrumentsfor measuring hydrologicvariablesand waysin which they can be usedjointly to createa complete of representation a functioning hydrologicsystemare discussed.

'

INSTRUMENTS 127 8.2 HYDROLOGIC

INSTRUMENTS 8.2 HYDROLOGIC

Good sourcesof information abouthydrologicinstrumentsare the National Weather U.S. Army Corpsof Survey,U.S. Bureauof Reclamation, U.S. Geological Service, Theseagencies and instrumentmanufacturers, Service, Soil Conservation Engineers, of in the business measuringhydrologicvariablesand and industrieshavelong been Someofthe of theycanprovidedetaileddescriptions state-of-the-artmeasuringdevices. is here,but the coverage far from are described instruments major typesof measuring should consult the appropriatereferences.l-3 exhaustiveand the interestedreader

Precipitation
for Gauges measuringrainfall and snowfall may be recording or nonrecording.The 8-in. gauge. Servicestandard gauge the U.S. Weather is mostcommonnonrecording interval but often this is daily. The gaugeis The gaugemay be read at any desirable calibrated so that a measuringstick, when inserted, showsthe equivalentrainfall depth. Such gaugesare useful when only periodic volumes are required, but they cannotbe usedto indicate the time distribution of rainfall' the continuouslysense ratb of rainfall and its time of occurRecordinggauges type or the tipping are rence.Thesegauges usually either ofthe weighing-recording gauges run for a period of 1 week, at which time usually buckettype. Weighing-type with Problem2.5 is typical of the The figure associated their chartsmustbe changed. depth versustime is the product, and this recordedoutput. A masscurve of rainfall curve can be translatedinto an intensity-time graph by calculating the ratios of rainfall to time for whatevertime stepis desired.Tipping bucketgauges, accumulated rainfall accumulationwhen it reachesa on the other hand, senseeach consecutive 1 mm of rain, A small calibratedbucket is prescribedamount, usually 0.01 in. or located below the rainfall entry port of the gauge. When it fills to the 0.01-in. bucketinto position.Thesetwo small buckets incrementit tips over,bringing a second are placed on a swivel and the bucketstip back and forth as they fiIl. Each time a an bucketspills it produces indication on a strip chart or other recordingform. In this way a record of rainfall depth versustime (intensity)is the outcome.For rain gauges to record snow accumulations,some modifications must be made. Usually these involveproviding a melting agentso that the snow can be convertedinto measurable water. Figure 8.la is the diagram of a self-reportingrain gaugingstation.The tipping a bucketmechanismgenerates digital input signal whenever1 mm of rainfall drains The signalfrom the gaugeis automaticallytransmitted through the funnel assembly. to a receiving station where it records the station ID number and an accumulated was amountof rainfall. The receiving stationrecordsthe time at which the message periods can be calculated accordingly' Figreceived and rainfall rates for desired ure 8.1b showsa similar gaugeequippedto measuresnow.In this case,a glycometh solution is usedto melt the snow.The melt water overflowsthrough a temperaturethe by and is measured the tipping bucket,which operates mechanism compensating in Fig. 8.1 can easilybe incorporated of station's transmitter.Gauges the type shown that can be used in a variety of forecastingand into real-time monitoring systems operatingmodes.

"

128

CHAPTER INSTRUMENTATION 8

Antenna

Antenna mast

Antenna

Antenna mast

Temperatffe compensation overflow mechanism Glycometh collecting section Tipping bucket Drain holes (4)

Funnel assembly

Tipping bucket

Signalcable Lifting rope Antennacable

Signalcable Lifting rope Antennacablb

(a)

(b)

Figure 8.1 Self-reporting (a) rain and (b) snow stations. (Courtesy of SierraMisco, Inc., EnvironmentalProducts,Berkeley,CA.)

Evaporation and Transpiration

Evaporationpans have been widely used for estimatingthe amount of evaporation Devicessuchasthat depictedin Fig. 8.2 areeasyto use,but from free water surfaces. relating measurements taken from them to actual field conditionsis difficult and the -data_lhey-produce often of questionable value for making areal estimates.A are

..

INSTRUMENTS 8.2 HYDROLOGIC

129

tr'igure 8.2 U.S. Weather Bureau Class A pan.

but variety of pan types havebeen developed the U.S. WeatherBureauClassA pan havebeenusedto Pan is the standardin the United States.4 evaporationobservations from well-watered and evapotranspiration both free water(lake) evaporation estimate vegetation.Field experimentshave shown a high degreeof correlation of pan data from surroundingvegetationwhen there is full cover and with evapotranspiration pan gauges, datacan be recorded goodwatersupply.a in the caseof precipitation As and transmittedcontinuouslyto a central receiving station. are measurements often made using lysimeters.ThesedeEvapotranspiration vices are containersplaced in the field and filled with soil, on which sometype of growth is maintained.The object is to study soil-water-plant relationsin vegetative a natural surrounding.The main featureof a weighinglysimeteris a block of undisin turbed soil, usually weighing about 50 tons, encased a steel shell that is 10 ft by 10 ft by 8 ft. The lysimeteris buried so that only a plasticborder marksthe top of the containedsoil. The entire block of soil and the steelcasingare placedon an underground scale sensitiveenough to record even the movement of a rabbit over its surface.The soil is weighed at intervals, often every 30 min around the clock, to most of the in changes soil water level. The scalesare set to counterbalance measure deadweight of the soil and measureonly the active changein weight of water in the The scalescan weigh accuratelyabout 400 g (slightly under 1 1b), which is soi1.5 equivalent to 0.002 in. of water. The weight loss from the soil in the lysimeter Added water coverplus any soil evaporation. waterusedby the vegetative represents is also weighedand thus an accountingof water contentcan be kept. Crops or cover are plantedon the areasurroundingthe lysimeterto provide uniformity of conditions surroundingthe instrument.Continuousrecordsat the set weighingintervalsprovide monitoring of conditions.The data obtainedcan be transmittedto almostcontinuous can produce location for analysisand/or other use.Weighinglysimeters any desirable over short periodsof time. But they are expenvaluesof evapotranspiration accurate which are lesscostly,havealsobeenused,but sive.Nonweighingtypes of lysirneters, unless the soil moisture content can be measuredreliably by some independent method,the data obtainedfrom them cannotbe relied on exceptfor long-term meaevents.s precipitation suchas between surements

and Wind,Temperature, HumiditY

to and humidity are needed supportmany types of Measurements wind, temperature, a using an anemometer, device is commonly measured Wind of hydrologicanalyses. elementsuchas a cup (Fig. 8.2) or propellerwhosespeed that has a wind-propelled is calibratedto reflect wind velocity. Wind direction is obtainedusing a vane,which orients itself with the direction of the wind.

130

CHAPTER INSTRUMENTATIoN 8

Temperaturemeasurements madeusing standardthermometers various are of types,while hurnidity is measured using a psychrometer. psychrometer A consistsof two thermometers, one called a wet bulb, the other a dry bulb. Upon ventilation the thermometers measuredifferently, and this differenceis called the wet-bulb depression.By usingappropriate tables,dewpoint, vaporpressure, relativehumidity can and , be determined.6 Figure 8.3 depictsa completeweatherstation incorporatingmeasurements of precipitation, wind, temperature,barometricpressure,and humidity. Sucha station - can automatically report weatherdata from remote siteson either an event and/or

Funnel assembly Solar panel Tipping bucket

Lifting rope Antennacable

Main housing

Signal cable

Ground level

Transmitter

Figure 8.3 Self-reporting weather station. (Courtesy of Sierra-Misco, Inc., Environmental Products, Berkeley, CA.)

131 INSTRUMENTS 8.2 HYDROLOGIC timed basisto a central site. A station suchas this can be used for marine weather of forecasting,quantitativedeterminations oncoming storms,determinationof wind effect on tidal areas,and establishinga data basefor irrigation.

OpenChannelFlow
of Measurements open channel(natural and created)flow are made using standard measuringdevicessuchas flumes and weirs, and they are also madeby calibrating of special control sectionsalong rivers and streamssuchthat measurements depth devicesare designedso Flow-measuring (Jtage)of flow can be related to discharge. that sensingsomeparametersuchas depth automaticallytranslatesthe observation of When a control sectionis used,observations crossinto units of flow (discharge). flow velocitiesmust and average sectionalareafor variousdepthsmust be obtained, In so for be ascertained variousstages that a sectionrating curve can be established. U.S. Bureau of Reclamation,the the United States,the U.S. GeologicalSurvey,the havedoneextensive Service,andthe U.S. Army Corpsof Engineers Soil Conservation instrumentsand proceduresfor flow measuringand havebeen active in developing rates of flow.2'6 ascertaining devices.When they are properly inWeirs Weirs are common water-measuring simple and accuratemeansfor gauging stalled and maintainedthey can be a very The most often usedweir types are the rectangularweir and the V-notch discharge. weir (Flg. 8.4). To be effective,weirs usually require a fall of about0.5 ft or more in the channelin which they are placed.Basically,a weir is an overflow structureplaced acrossan open channel.For a weir of specificsize and shapewith free-flow steadystateconditionsand a proper weir-to-pool relation, only one depth of watercan exist by Flow rate is determined measuring in the upstreampool for a particular discharge. part of the weir to the watersurface from the crestof the overflow the vertical distance this recordeddepth in the upstreampool. The weir's calibrationcurvethen translates into rate of flow at the device. flow section openchannel shaped flumeis a specially Parshall Flumes A Parshall The 8.5 depictsone of thesedevices. that can be installedin a channelsection.Figure (1) it can operatewith a relatively small head flume has severalmajor advantages: to loss;(2) it is fairly insensitive the approachvelocity; (3) it can be usedevenundet conditions;and (4) its flow velocity is usually sufficientto precludesedisubmerged by The Parshallflume was developed the late Ralph ment depositsin the structure.2 flume. The constrictedthroat of the L. Parshill and it is a particular form of venturi Thus,asin the case a flume produces differential headthat canbe relatedto discharge. the (head)is all that is requiredto determine rate ofdepth .ofthe weir, an observation are generally best suitedto gauging of flow at the control point. Weirs and flumes weirs can be installed althoughlargebroad-crested and openchannels, small streams For major rivers, other measuringapat dam sites as part of overflow structures. proachessuchas developingfield ratings at a specifiedcontrol sectionmust be relied on.

132

CHAPTER8

INSTRUMENTATION

Figure 8.4 Field installation of weirs: (a) rectangular and (b) Vnotch. USDA CooperativeExtensionService,Mountain StatesArea.

Golttrol Sections where the installation of a weir, flume, or some other flowmeasuring deviceis impractical, it is sometimes possibleto developa rating curve at somelocation alonga streamby taking measurements depth,cross-sectional of area, and velocity and calculatingthe rate of flow for a particular stageat the location.By doing this for a range of depths of flow, a station rating curve can be developed. Instrumentsrequired to developsuch a curve are depth-sensing devices,surveying instruments,and velocity meters.The velocity meter is similar to an anemometer. It is placedat variouspositionsin the channeland a velocity is recorded.By doing this at a numberoflocations, a velocity profile for a given depth can be developed. From this an average flow velocity can bJcomputed, ind uy uiing that determinationand

INSTRUMENTS 133 8.2 HYDROLOGIC

Diverging sectlon Throat section
a
J

Altemate45" wing wall PLAN

t x t xf,tngle

SECTIONZ-f, Service.) flume.(U.S.Soil Conservation Figure 8.5 Parshall

can area,discharge be calculatedas the product of meanvelocity the cross-sectional can area.If observations be madefor a rangeof depths,a rating and cross-sectional of developed the control sectionsothat only measurements depthwill for curvecanbe to estimaterate of flow at somelater time. Additional information on this be needed proceduremay be found in Refs.2 and 6.

Valve shut-offkeys Connecting band Valves

Figure 8.6 Recorderhouseand stilling well for a streamgaugingstation. (U.S. Bureau of Reclamation.)

134

CHAPTER INSTRUMENTATIoN 8

Depth (stage) Measurements Most depth measurements made using a float are and cable arrangement a stilling well or a bubbler gauge.In the first insiance,a in stilling well connectedto the channel(Fig. 3.6) is used to housea float devicethat activatesa recorder as it movesup and down. Figure 8.7 illustratesa self-reporting stilling well liquid-level station. Data from this station can be transmitted to un! central location for analysisand/orother use.A bubbler-typeinstallation makesusl of dry air or nitrogenas a fluid for bubblingthrough an oriflce into a channelbed. As the depth of flow changes,the changein head above the bubbler orifice causesa

Antenna mast Cover with special accessscrews

Transmitter
Hydraulic damping device Level sensor Counter weight Mounting brackets

lnlet tubes

Sidecleanout port Bottomcleanout port

Figure 8.7 Self-reportingstilling well liquid-levelstation. (Courtesy of Sierra-Misco,Inc., Environmental products. Berkeley,CA.)

8.4 REMOTE SENSING 135 corresponding pressure change. This resultsin a fluid-levelchangein the manometer connected the gassupplyand this in turn is usedto reflect stagevariationover time. to The foregoingdescriptionsare of a few of the instrumentsused in hydrologic work. Both the limitations associated with their use and their reliability must be understoodif they are to be used correctly and their outputs are to be considered credible.

8.3 TELEMETRY SYSTEMS

Historically,many gauges were read periodically by an individual making the rounds of installations.This servedwell whenthe purposeof the data wasto establish base a record of somevariablesuchasrainfall. But today,undermany circumstances, has it becomenecessary make continuous to recordingsof rainfalls, streamflows, evapand oration rates and to have thesedata availablefor the real-time operation of water management systems and for forecastinghydrologicevents.Someexamples activof ities requiring real-time hydrologicdataarc managingreservoirs, issuingflood warnings, allocating water for various usessuchas irrigation, monitoring streamflows to ensure that treatiesand pactsare honored,and monitoringthe quality and quantity of waterfor regulatoryand environmental purposes. Accordingly,gaugingstationscapable of electronicallytransmitting their data to a central location for immediateuse havenow becomecommon.The advantages suchstationsincludeproviding inforof mation to usersin a time frame that meetsmanagement needs,reducingthe costsof collecting data, and providing a continuousand synchronous record of hydrologic events.Figure 8.8 showsa streamgaugereporting station using radio transmission. Figure 8.9 illustratesa satellitedata collection and transmitting operation.T-|2

8.4 REMOTE SENSING

Sincethe 1960s,remote sensing becomea commonhydrologictool. Examplesof has aircraft and satellite datacollectionand transmission abound.13-16 Figure8.10 illustratesthe useof aircraft and satellites a snowsurveysystem. in Other typesof surveys suchasthoserelatedto determining imperviousareas, classifyingland usesfor assessing basin'wide runoff indexes, determining lake evaporation, and groundwater prospectingcan be depictedin similar fashion.Table 8.1, which summarizes operational uses of satellite data in hydrology circa 1981, showsthe great diversity of remote sensingand data transmission options that can be exercised.l6 The principal value of remote sensing its ability to provide regionalcoverage is and at the sametime providepoint deflnition.Furthermore,satellitecommunications can be digitized and are thus compatiblewith the transferof computerized information. Following the evolution of linkages between computer and communications technology, new softwaresystems incorporatingpowerful data management systems havebeen developed. Thesesystems facilitate the storage,compaction,and random access large data banks of information. one data management of option, geographic (GIS), allows the overlayingof many setsof data (particularly information systems satellite-deriveddata) for convenientanalysis.Versatilecolor pictorial and graphic display systems also becomingattractiveas their costshavedecreased.ra are

136

CHAPTER INSTRUMENTATIoN 8

I
L. I

Utah
Anzona

q)

B .H

Boulder City VIIF Radio Station

Mt. Hualpai Repeater Station N

20

40

60

"&*" rq*
s
Taylor'sFerry gauge

,&

Scale miles of
Lake Havasu

EXPLANATION Parker Dam

Black Point Repeater Station

fl Water level gauge

X

il VIIF radio antenna tower H @ Transmitter-ieceiver radio

Cibola gauge

Colorado nver

VHF Very high frequency radio

lmperial Dam Repeater Station

Dam
-'-:4

az*<3e -{/co '-Figure 8.8 Streamgaugereporting system Water stage using radio transmission. information is requested from the gaugingstationsby VHF radio signal.In turn, this water stage information is obtained from the stream gaugesand automatically encodedand transmitted to the Boulder City receiving station. All downstream releases from Hoover Dam are determinedand integrated with this streamflow information in controlling the flow of the lower ColoradoRiver. (U.S. Bureau of Reclamation.)

8.4

REMOTESENSING

137

Figure 8.9 Hydrologic data collection by satellite.(U.S. GeologicalSurvey.)

Low-altitude gamma-ray light.

Water 4gency forecastcenter

Figure 8.10. Satellite snow survey system.(AftenCalabreseand Thome'rs)

jo (d

E o E,

32
o

o o o o

c
k

9 o
IY

(u (E o

89'
d r i

o z

F-> 'o ri
X a ! c

s
o
IJJ

Fl

t J
J

tr|
F

a
lJ-

U) uJ a f
J

eE;
* E E
o
(E o CL

o
F

(r tu o
d ut 6

' E c c

SEE
! q o H o 9r

5 E K
.=

E - E s> ,, Y q F .t

hiEE. g b " E6
F

E R ! ' i

F
o o a

;+

o
E

t T

c o

E E a a

e o
o

E
!o

Ec,

o x

o E

o r A 0

boo

o o U' o (o

6 6

l 6 f i E

, .

bo o o (B ,o o. e z a

s
o
dt uI

Ch

(E
L . =

!t e qt a)

3 -

E
o .
O

o E a

e Ci

E O :.i c >r-!!

o g o 4 =
O F , Y o E

I E t

Ee
L

+e
d

8-:

at)

142

CHAPTER INSTRUMENTATIoN 8 With the advancement satellite technology,the use of satellitesas remote of platformshasspread.Currently available sensor can sensors operatein a multitude of electromagnetic radiation wavelengths and the information content of their signals pollutants,and other types can include Surface temperatures, radiation, atmospheric of meteorological data. As remote sensors improved to permit the attainmentof are greater radiometric and geographicresolution, and as computer image-enhancing techniques becomemore sophisticated, is certain that this powerful watermanageit ment tool will seeeven sreaterand more diversifieduse.

r summary
Hydrologic data are important componentsof model design and testing and of a variety of statistical analyses. The quality of data obtained relate to attributes of measuring instrumentsand to the featuresof gaugingsites.It is important to understandthe pros and consof variousinstruments and to know how they canbestbe used.

REFERENCES
1. "Irrigation WaterMeasurement," Mountain StatesRegionalPubl. 1, revisionof Extension Circ. 132,Irrigation Water Measurement, University of Wyoming, Laramie, June 1964. 2. U.S. Bureau of Reclamation,Water Measurement Manual, 2nd ed. Washington,D.C.: U.S. GovernmentPrinting Office, 1967. Data Book, 4th ed. Beaverton,OR: 3. Leupold & Stevens,Inc. StevensWater Resources Leupoldand Stevens, Jan. 1987. 4. W. Brutsaert,Evaporationinto the Atmosphere. London: D. Reidel Publishing, 1982. 5. "TexasLab InstallsWeighingLysimeters," Irrigation J. 37(3),8-12(May/June 1987). 6. R. K. Linsley, Jr., M. A. Kohler, and J. L. H. Paulhus,Applied Hydrology. NeW York: McGraw-Hill, 1982. 7. R. J. C. Burnashand T. M. Twedt, "Event-Reporting Instrumentationfor Real:TimeFlash Flood Warning," American Meteorological Society, Preprints, Conference on Flash Floods:Hydro-meteorological Aspects, May 1978. 8. D. E. Colton and R. J. C. Burnash,"A Flash-FloodWarningSystem,"American MeteoAspects, rologicalSociety,Preprints,Conference Flash Floods:Hydro-meteorological on May 1978. "AutomatedPrecipitation Aug. 1980. 9. R. J. C. Burnash, Measurements," 10. R. J. C. Burnashand R. L. Ferral, "A Systems Approach to Real Time Runoff Analysis with a Deterministic Rainfall-Runoff Model," International Symposium on RainfallRunoff Modeling,Universityof Mississippi, May 18-21, 1981. 11. Hydrologic ServicesDivision, National WeatherService,WesternRegion, "Automated Local Evaluationin Real Time: A CooperativeFlood Warning Systemfor Your Community," Feb. 1981. Involvedin 12. R. J. C. Burnashand R. L. Ferral, "Examplesof Benefitsand the Technology Optimizing HydrosystemOperation Through Real:Time Forecasting," Conferenceon Real:TimeOperation of Hydrosystems, Waterloo,Ontario, June24-26, 1981. 13. M. Deutsch,D. R. Wiesnet, and A. Rango (eds.),Satellite Hydrology. Bethesda,MD: AmericanWaterResources Association. 1981.

REFERENCES 143

t4. J. F. Bartholic,

"Agricultural Meteorology: SystemsApproacli to Weather and Climate Thirty-Second in Needsfor,tgricuttrire, Forestry,and Natiiral iesources,o' Proceedings, Institute, Agiicultural R'searcih'Institute. Bethesda,MD: Agricultural Research tuteeting, 1 9 8 3p p . 7 5 - 8 5 . , *NASA Water Resources/HydrologyRemote Sensing 1 5 . M. A. Calabreseand p. G. Thome, Programinttrelg80's,"insatettiteHydrology(M'Deutsch'D'R'Wiesnet''andA'Rango' Association' 1981' pp' 9-15' MD: American Water Resources eds). Bethesda, 16. G.K.Moore,..AnlntroductiontosatelliteHydrology,''hsate.tlite-Hydrology(M. MD: American WaterResqurces Deutsch,D. R. Wiesnet,and A. Rango,eds.).Bethesda, t7-4L. 1981,PP. Association,

ChapterI

Networks Monitoring

Prologue
The purposeof this chapteris to: . Describeelementsof systems monitoring hydrologicvariables. for . Indicate the importanceof real-time and continuousrecording of hydrologic events. for Information (data)is the requisitefoundationfor designing schemes manipulating (managing) hydrologicsystems, evaluatingthe efficacyof actionstaken to correct for problem situations,and for identifying trouble spots deservingattention. But to be useful, the data must be of the right type, in the right form, and appropriatelyrepresentativeof critical spaceand time dimensions. Modeling hydrologic systems requires an understanding how these syster-ns of requirestracking the effectsof actually function; cleaningup a toxic wastedischarge remedial actions;enforcing environmentalregulationsrequires knowledgeof what has happenedsincethe rules were implemented;and regulatingreservoirreleases to meet specifiedtargetsrequiresa continuous understanding the stateof the system of lies beingoperated. The key to meetingsuchrequirements in the productsof carefully designedand managedmonitoring networks.Developingsuchnetworksis no small task, however,as the numberof variablesthat must be observedmay be very large, and the instruments measure to them costlyto install and operate,andthe datastorage managementrequirementsextensive.Accordingly, a monitoring network's design must beginwith a thoroughunderstanding its purposeso that the degreeof resoluof for but tion providedby its observations adequate, not excessive, the task at hand. is of within the constraints what A goodrule is to keepthe networkassimpleaspossible, must be accomplished.

9.1 THEPURPOSE MONITORING OF

The purpose of monitoring is to gather information in a continuum such that the dynamics of the systemcan be ascertained.According to Dressing,objectives of of monitoring for nonpoint source pollution control include development baseline

9.2 SPECIAL CoNSIDERATIoNS 145 information, generatingdata for trend analysis,developing and/or verifying models, and investigating singleincidentsor events.z Theseobjectivesare alsovalid for hydrologic monitoring in general,but they shouldbe supplemented the following obje"by tives:planning,real-time system operating,enforcingregulatoryprograms,and environmental policymaking. In the flnal analysis, ultimate purposeof monitoring is the to enhancedecisionmaking, whetherit be for development, management, regulation, or research aims.

9.2 SPECIAL CONSIDERATIONS

Beforean acceptable monitoringplan can be devised, theremustbe a full understanding of the hydrologic systemto be monitored and of the objectivesto be met by monitoring. The costs of monitoring can be very high and thus it is essentialthat monitoring networksbe efflcient and cost effective.r-5

Timeand SpaceVariability
In general, monitoring networks are designedto have both spatial and temporal dimensions. Although monitoringa specificpoint locationmay be all that is necessary undersomecircumstances, is more commonthat what is happening a regional it in settingis of importance.The temporal aspectis similar. While a snapshotat some point in time may suffice for some purposes, time varianceof conditions to be the tracked is usually critical for effective analysesandlor decisionmaking.Both the short-termand long-term variabilities of many targefsof monitoring must be ascertained. For example,water quality in a streamcan changerapidly with time, while changes lake levels,suchas thoseexperienced the GreatLakes in the 1980s,are in in the result of long-termhydrologicvariability. Spatial variability must also be represented a monitoring network: for examin ple, infiltration ratesmay vary considerably within a region, rainfall intensitiesmay be quite different within even short distances, and water quality in a river might be different in upstreamand downstreamlocations.Topography,soils, vegetalcovers, and many other factors affecting the performanceof a hydrologic systemare also distributed differently in space,and these differencesmust be recognizedin the monitoring plan. The trick is to developa monitoring system that can (1) provide the neededdata,(2) recognizeregionaland temporal variabilities,and (3) keep installation, operation,and maintenance coststo a minimum. To do this requiresa comprehensiveknowledgeof the system be monitored,an understanding what the data to of obtainedby the systemwill be used for, and a knowledgeof the level of detail in collecting the data that must be exercised spaceand time. in

DataRequirements
The amountand type of datato be generated a monitoringsystem by mustbe carefully consideredin its design. Selectingappropriateinstruments,determining sampling frequency,and settingdata formats are elements that must be considered. Questions such as how much do we need to know and when do we need to know it must be answered. The form and extensiveness data must be tightly relatedto monitoring of

146

CHAPTER9

MONITORING NETWORKS

Flood warnlng Water managemeff

---:
Telephone

Lift station monitoring control

Hardwire
+ + , . .

lnfrared + + xgnr
Water quality monitoring

(Courtesy Sierra-Misco, Environof Inc., Figure9.1 A telemetry monitoring system. mental Products. Berkeley. CA. insteadof the to objectives.Furthermore,it might be necessary monitor surrogates condition to be tracked.' For example,if lake eutrophicationis the issue,phosphorus If is flre&sureS; this approach taken, andchlorophyllconcentrations might be surrogate is selectionof appropriatesurrogates very important and the foregoingcomments aboutdata formats and so on are also applicable.Hydrologic,waterquality, land use and treatment, topographic,soils, vegetativecover, meteorologic,and many other in typesof datamaybe requiredin combinationor separately a monitoringplan. It is easyto seethat the amount of data required for a monitoring program can be enorgreat care must be taken to seethat the data collection effort is mous.Consequently, not in excessof the objectivesof the monitoring program. Figure 9.1, depicting a gives an indication of the variety of data that might be telemetrymonitoring system, collectedin a monitoring program.

QualityControland QualityAssurance
that the data The costsof monitoring are usually substantialand thus it is essential generatedbe of consistentlyhigh quality. Accordingly, most monitoring systems (QA/QC) elements. include quality control and quality assurance Quality control is to a plannedsystem activitiesdesigned producea quality product(datain this case) of is that meetsthe needsof the user. Quality assurance a plannedsystemof activities designed guarantee to that the quality control programis being carried out properly. A quality managementplan should be part of the overall monitoring program to and should be preparedwhen the monitoring program is being developed ensure that the data collectedwill be of a satisfactorynature for the monitoring program's objectives.3

9,4

NETWORKS HYDROLOGICAL-METEOROLOGICAL

147

To useror

Figure 9.2 Microcomputeruse in streamgauging.

IN 9.3 USEOF COMPUTERS MONITORING

miof With the rapid technologicaldevelopment computers,especiallyinexpensive crocomputers, oppoitunities foiautomated collection of all types of hydrologic the and water quality data have increased substantially.Microcomputers, used with analog-to-digital converters,pressureor liquid-level sensors,and the appropriate software can, for example, be used in hydrologic monitoring systemsas flow (Fig.9.2).4Furthermore, are suchsystems highly acquisition systems metering/data versatile and they are rblatively inexpensive.Computer systemscan be customapplicationand they are often lesscostly than for designed almost any dataacquisition for designed the samepurpose.Comhardwaresystems other commerciallyavailable puterscan convert raw datainto other more usefulforms, storedata for later use,and As with other computerterminalsif necessary. such,they are a powercommunicate Figure 9.3 ful and important componentof modern hydrologicmonitoring systems. illustratesthe use of computersin a real-time telemetry system.

NETWORKS ETEOROLOGICAL e.4 HYDROLOGICAL-M

Most modern hydrologic-meteorologicnetworks are designedto provide real-time releasingflows for irrigainformation for purposessuchas hydropowerscheduling, models,regulatingreservoirdischarges, and tion, developing testinghydrologicsystem allocating water from multiple sources,streamflow forecasting,tracking pollutant transport, and enforcing environmental regulations. Hydrological-meteorological

148

CHAPTER9

MONITORING NETWORKS

Water level

Weather station

Computer

,/: :
Modem Data Collection

(Courtesy Sierra-Misco, Figure9.3 Computer in a real-time use telemetry system. of Inc., Environmental Products, Berkeley, CA.)

networks may be designedto monitor physiographic, climatic, hydrologic,biologic, and chemicalfeatures, combinations these,in a region or river basin.They must or of have gaugedensitiesand distributionsthat are sufficientto permit interpolationbetweengaugesitesin a mannerpermitting valid conclusions be drawn for the entire to areacoveredby the network. Typically, measureinents madeof suchvariablesas are precipitation, solar radiation, temperature,relative humidity, barometric pressure, snow depth, soil moisture,wind speed,streamflow,and water quality. In any event, special basin or regional climatic factors must be given due consideration.Each hydrological-meteorological network is different in its purposeand setting and thus its designmust reflectboth the spatialand temporally varying featuresat the locality to be monitored along with the objectivesof the monitoring program.s

REFERENCES 149

r Summary
Monitoring of hydrologicsystems essential better understanding system is to interof actions and to the designand testing of hydrologicmodels.It is also the meansby which a determinationcan be made of the effectiveness measures of taken to alter performance. watershed

REFERENCES
t . S. J. Nix and P. E. Black (eds.),Proceedings the Symposiury. Monitoring, Modeling, of on
and Mediating Water Quality. Bethesda, MD: American Water Resources Association, 1987. 2. S. A. Dressing, "Nonpoint Source Monitoring and Evaluation Guide," in Ref. 1, pp. 69-78. J. J. Lawrenceand A. S. Y. Chau, "Quality Assurancefor EnvironmentalMonitoring," in Ref. I, pp. 165-176. 4. H. E. Postand T, J. Grizzard, "The Monitoring of StreamHydrology and Quality Using Microcomputers," Ref. 1, pp. 199-208. in 5 . P. J. Gabrielsen A. J. Carmeli, "Operation of a Hydrologic-MeteorologicMonitoring and Networkin a Severe Winter Environment," Ref, l, pp. lI3-122. in

PARTTHREE .

WATER SURFACE HYDROLOGY

C h a p t e r1 0

Runoffand the Catchment

r Prologue
The purposeof this chaPteris to: . ' . . . the Expandon definitionsof termsfrequentlyusedin describing runoff process. Presentconceptsof surfacerunoff and drainagebasin discharge. Introduce elementsof drainagebasin geomorphology' characteristics. of Describequantitativemeasures watershed Familiarize the readerwith elementsof frequencyanalysis'

Surfacewater hydrologydealswith the movementof water alongthe earth's surface as a result of precipitation and snow melt. Detailed analysisof surfacewater flow is highly important to suchfieldsasmunicipaland industrialwater supply,flood control, water qualreservoirdesign,navigation,irrigation, drainage, stieamflowforecasting, and flsh and wildlife management. recreation, ity control, water-based precipitationandrunoffis influencedby variousstormand The relationbetween of Because thesecomplexitiesand the frequentpaucity of adebasin characteristics. to quaterunoff data, many approximateformulashavebeendeveloped relaterainfall the whereas were usuallycrudeempirical statements, andrunoff. The earliesfofihese descriptiveequationsbasedon physicalprocesses. trend now is to develop

BASINS AND DRAINAGE WATERSHEDS 10.1 CATCHMENTS,

the Runoff occurswhen precipitationor snowmeltmovesacross land surface-some natural or artificial streamsand lakes'The land areaover of which eventuailyreaches which rain falls is called the catchmentand the land area that contributessurface runoff to any point of interestis called a watershed.This can be a few acresin size can or thousandiof ,quut" miles.A large watershed contain many smaller subwatersheds. awayfrom high water and groundwater Streamsand rivers conveyboth surface preventingsurfaceflooding and rising groundwaterproblems'The tract water areas,

154

CHAPTER RUNOFF THE 1O AND CATCHMENT of land (both surfaceand subsurface) drainedby a river and its tributariesis called a drainage basin. A watershedsuppliessurfacerunoff to a river or stream,whereasa drainage basinfor a given streamis the tract of land drainedof both surface runoff and groundwaterdischarge. Rain falling on a watershed quantitiesexceeding soil or vegetation in the uptake becomes surface runoff. Waterinfiltrating the soil may eventuallyreturn to a stream and combine with surfacerunoff in forming the total drainage from the basin. The network of overlandflow courses and defineddrainagechannelscomprisethe watershed.Surfacerunoff from tracts of land beginsits journey as overlandflory, often calledsheetflow,beforeit reaches definedswaleor channel,usually beforeflowing a more than a few hundredfeet. The lines separating land surfaceinto watersheds the are called divides.Thesenormally follow ridges and moundsand can be delineated using contourmaps,field surveys, stereograph pairs of aerial photographs idenor to tify gradient directions.

Area Contributing
In the majority of hydrologicanalyses, magnitudeof total surface the areacontributing direct runoff to somepoint of interestis needed. Because variationsin topography, of the true surface areacannotbe easilymeasured. horizontalprojectionofland area The is easilyobtainedand normally adoptedin hydrologiccalculations.This resultsin an error in actual watershedarea whereverthe projected area is less than the actual. Somesurfaceareain watersheds may not contributeto surfacerunoff, so the error in using the projectedwatershed areais somewhatoffset.

PartialArea Hydrology
For light storms,or for someflat areas,portions of the catchmentdo not contribute to runoff. Precipitationfalling on or flowing into depressed blockedareascan exit or only by seepage evaporation,or by transpirationif vegetated. sufficientrainfall or If occurs, such areasmay overflow and contribute to runoff. Thus the total area contributing to runoff varies with the intensity and duration of the storm. Methods for incorporating this phenomenonin hydrologic studies are calegorized under proceduresfor partial area lrydrology. In partial areahydrology,watershed methods areasare dividedby one of several into contributing (active) and noncontributing (passive)subareas. For infrequent (severe) storms, largerpercentages the watershed may contributeto the peak surface of flow and volume of runoff, which are the primary variables of interest to design engineers. For more frequent storms,significantly smaller portions of some waterpartial areahydrologyis seldomincorposhedsmay contribute.As a consequence, ratedin hydraulic structuredesign,and is of greaterinterestin watersupplyand water quality studies. will be shownlater (Chapter12) unit hydrograph As theory andrunoff curve numbermethodsare basedon linearity of rainfall and runoff, and assume that the full watershed contributesto runoff in all stormsand in proportional amountsat different times in the samestorm. Application of thesemethodsto watersheds that havesignificantnoncontributingzonescould, and do, introduceerror ifthe zonesare nof first delineated and the distributedeffectsproperly modeled.

10.2 BASINCHARACTERISTICS AFFECTING RUNOFF

155

Subdivisionof contributing from noncontributingareashas traditionally been subjectively accomplished from siteinspection, topographic and soilsmaps,and aerial photos.Soils having good drainageclassifications, dark tones or colors on aerial or photos, can often be consideredas passiveareas.Other signs of noncontributing areaswould includepresehce wetlands,grassed of areas,rooftops (unlessconnected to the drainage), terraces, erosioncontrol structures, stockwateringponds,and flood control dams. Boughtontdeveloped quantitativemethod of determiningthe proportion of a a watershed that contributessurfacerunoff in different storms,and at different times duringthe samestorm.by analyzingrainfall andrunoff records.His logic is asfollows: 1. Watersheds be idealizedas a group of "surfacestorage can capacity" cells, eachrepresentinga fraction of the watershedarea and eachhaving some capacityto abstractrainfall into storage, infiltration, or evapotranspiration. 2. Runoff from eachcell occurs when rain fills the surfacestoragecapacity. 3. Runoff occursfrom the cell with the smallestcapacitybeforeflowing from the cell with the next largestcapacity(this is an assumption Boughton by that has not been fully verified). 4. Using theseprinciples,storm data for the watershedare evaluatedfirst to find thosein which runoff occurs only from the area of smallestcapacity. This is done using a graphicalmethod outlined in the article that looks at slopechanges the rainfall-runoff graph.Both the capacityof the cell and in its area as a percentage the watershed estimated. of are 5. After subtracting contributionto runoff from the smallestcapacitycell, the the capacityand contributingareafor the secondsmallestcapacitycell are determinedby the sameprocedure. 6. The process is repeateduntil all the runoff is accountedfor, or until 100 percent of the watershed contributing,whicheveroccurs first. is When Boughton applied the procedureto a test watershed,r was found that it runoff occurredfrom the entire watershed only 3 of 30 eventsin the l5-year study on period.In abouttwo-thirds of the runoff events, discharge occurredonly from the cell with the smallestsurfacestoragecapacity.

10.2 BASINCHARACTERISTICS AFFECTING RUNOFF

The natureof streamflowin a regionis a function of the hydrologicinput to that region and the physical,vegetative, and climatic characteristics. indicatedby the hydroAs logic equation,all the waterthat occursin an areaas a result ofprecipitation doesnot appearas streamflow. Fractionsof the grossprecipitationare divertedinto pathsthat do not terminatein the regional surfacetransport system. Precipitation striking the groundcango into storage the surface in the soil andinto groundwater on or reservoirs beneaththe surface.The characterof the soil and rocks determines a large extent to the storagesystem into which precipitatedwaterwill enter.Opportunity for evaporation and transpirationwill also be affectedby the geologicand topographic natureof the area.

156

CHAPTER1O RUNOFFAND THE CATCHMENT

StreamPatterns
Wind, ice, and water act on land surfaces createseveraltypes of drainagepatterns to seen in nature. The particular design that results is a function of several factors including slope,underlying soil and rock properties,and the historiesof hydraulic action, freeze-thaw activity, and sedimenttransport.

(a) Dendritic system

(b) Trellis system

(c) Radial system

(e) Meandering stream

(f) Braided stream

(g) Anabranching stream

(h) Reticulate stream

and individual streamshapes). Fi-gxre 10.1- -Streampatterns(combinedsystems

10.2 BASINCHARACTERISTICS AFFECTINGRUNOFF

157

Typical streampatterns shownin Fig. 10.1.The most common,dendritic, are is characterizedby numeroussmall tributariesjoining at right angles into higher-order streams,eventually forming the major rivers in the region. The smaller tributaries often occur in sufficientquantitiesthat little land surfaceareais left unintercepted by a definedchannelof someform, The maximum overlanddistancebetweenchannels in theseareasseldomexceeds few hundredfeet. a Trellis patterns characterizedby are long main streams intercepted numerous by shorterright-angletributaries.They are commonin the Appalachians (EasternUnited States)and are also seenin the Rocky Mountains along the foothills, Multi-basin patterns,also called derangedsystems, occur in low gradient swampy areaswith numeroussurfacedepressions normally haveonly a few tributaries.Theseoccur and in glaciated,windblown,and permafrostareas,and are common in plains and mountain valley regionsof the United States. Radial patternsare typically found in foothill areasor mountain areaswith more advancedsoil development. Individual streamsfavor one or more of the four patternsshown on the lower portion of Fig. 10.1.Streams rarely straightexcepton steepslopes homogeare in neous materials. Braided streams are characterizedby numerous interconnected channels flowing aroundand overislandsandbars,inundatingmostduringhigh flows. Braided streamsare generallytransportinglarge amountsof sediment,but often less than the amount supplied. They have been called incipient forms of meandering streams dueto the fact that many revert to meandering other forms when sediment or suppliesor other factorschange. Meanderingin an otherwisestraightchanneloccurs as a result of transverse currents.Thesecurrents are consideredto be the result of forcesactingon the streamparticles,includingbed and bank shearforcesand coriolis effects. In evaluatingthe effects of changesin streamflows,a relationship known as Lane's Law is often applied.It states that the productof bed slopeand waterdischarge is proportional to the product of sedimentsize and transportrate. Changingany one of thesefour terms results in the likelihood of a shift in one or more of the others. Constructinga reservoit,for example, reduces sediment the transported into the reach just downstream.By Lane's Law, either the slope or dischargemust decrease or sedimentparticle sizemust increaseto offsetthe changein sedimenttransport.Most often, the slope decreases when the sediment-hungry flows deepenthe bed in the reach.Degradation(downwardcutting) of the bed of the Missouri River, for example, has occurredbelow someof its upstreamreservoirs, isolatingboat marinasand water intake structuresin somelocations.

Geomorphology Drainage of Basins

The principal geologicfactorsthat affect surface watersare classified lithologic and as structural. Lithologic effects are associated with the composition,texture, and sequenceof the rocks, whereas structuraleffectsrelatemainly to discontinuities suchas faults and folds. A fault is a fracture that resultsin the relative displacement ofrock that was previouslycontinuous.Folds are geologicstrata that are contortedor bent. Variationsin the erodibility of the different strata can easily lead to the creation of distinctive forms of drainagesystems. Both large-scale local effectson the storage movement surface and and of waters existbecause geologicactivity and structure.For example,drainagepatternsare of

1 58

CHAPTER10

RUNOFFAND THE CATCHMENT

determinedto a large extentby the nature of land forms. On the other hand, flowing surfacewatersalso affect the surfacegeometrythrough the processof erosion.Thus significant land forms resulting from volcanic activity, folding, and faulting affect drainage, whereas patterns,havingbeengenerated, alsomodify the land drainage can forms by creatingvalleys,deltas,and other geomorphicfeatures. Streamsare classified beingyoung,mature,or old on the basisof their ability as to erodechannelmaterials.Youngstreamsare highly active and usually flow rapidly so that they are continually cutting their channels.The sedimentload imposed on thesestreamsby their tributariesis transportedwithout deposition.Mature streams are those in which the channel slope has been reduced to the point where flow velocitiesarejust able to transportincoming sedimentand where the channeldepth is no longerbeingmodifiedby erosion.A streamis classified old whenthe channels as in its system havebecomeaggraded. The flow velocitiesof old streamsare low dueto gentleslopes that prevail.Wide meander belts,broad flood plains,anddeltaformation are alsocharacteristic old streams. of The lower reaches the Mississippi, of Rhine, and Nile are examples. Flows in young river basinsare often o'flashy,"*h"t"u, sluggish flows are common to older streams. The description of a drainagebasin in quantitative terms was an imponant forward stepin hydrologyand can be tracedback in largepart to the efforts of Robert E. Horton.2Strahler,Langbein,and othershaveexpanded Horton's original work.3-a To quantify the geometry of a basin, the fundamental dimensionsof length, time, and mass are used. Many drainagebasin featuresthat are important to the hydrologistcan be quantified in terms of length, length squared,orlength cubed. Examplesare elevation,streamlength, basin perimeter,drainagearea, and volume. The conceptof geometricsimilarity can be appliedto drainagebasinsjust as it is to many other systems.3 Most readerswill be awareof model-prototypestudiesof aircraft, dams,and turbomachinery. Suchstudiesinvolveconsiderations geometricas of well as dynamic similarity. In the samemannerthat inferencesas to the operationof a prototype can sometimes drawn from a geometricallysimilar model, inferences be canalsobe drawnaboutthe operationof one drainage areaon the basisof information obtained from a similar one. Perfect similarity will never be realized if natural drainagesystems compared,but striking similarities have been observedwhich are can often be put to practical use.

Measuresof DrainageBasin Characteristics

Important measures drainagebasin characteristics of include overlandflow lengths and streamlengths.The conceptof streamorder is often associated with the dimension of streamlength.

AFFECTINGRUNOFF 1O.2 BASIN CHARACTERISTICS

159

'/, "81 {,'v '\'rr

;)

Figure 10.2 Sketch indicating definition of streamorder.

The order to the streamorder, provided that a large enoughsampleis investigated. number permits comparisonsof drainage systemsthat are quite different in size quantity.Suchcomparisons shouldbe madeat the because numberis a dimensionless locations in the two systemsthat have a similar geometry; that is, second-order streams, third-order streams,and so forth. of Streamlengthsare determinedby the measurement theh projectionsonto a If horizontal plane. Topographicmaps are useful for obtaining suchmeasurements. possible Z, the meanlenglh of a streamsegment of order a is definedas L,, then it is to determine Z, using2
tr-

-2!i, L,,
N,

(10.1)

of whereN, is the number of streamsegments streamorder u. Another measurerelated to stream length is the distanceL"o from a point of the intereston the main streamto a point on the primary channelthat is nearest center of gravity of the drainage arca (center of gravity of the plane atea of the drainage basin).Studiesof basinlag (time betweenthe centersof massof effectivestorminput and the resulting runoff ) havemadeuse of this dimension. of development a drainagebasin in Of particular significance the physiographic is the overlandflow length Ls. This is the distancefrom the ridge line or drainage alongthe path of surfaceflow which is not confinedin apy defined divide, measured flow channel.If a to the intersectionof this flow path with an established channel, then basin of the first order is the basicelementof a larger drainagesystem, drainage overlandflow length can be determinedfor thesefirst-orderbasins. a representative One apprbachis to measurea numberof possibleflow paths from a map of the area these.In somecases(for example,with the rational method,Chapter and-toaverage

160

CHAPTER1O RUNOFFAND THE CATCHMENT

15), the use of the longestoverland flow length is prescribed,measuredfrom the upstreamend of the first-order stream to the most remote point of flow that will terminateat this point.

Areal Measurements
Just as linear measuresrelate to many factors of hydrologic interest, so do areal measures. example, quantity of discharge For the from any drainage basinis obviously a function of the areal extentof that basin. Correlationshavebeenobserved between average the area, of basinsof order Au, u, andthe average length of streamsegments, Thesevariablesare often relatedby 2,. an exponentialfunction. For example,studiesof sevenstreamsin the Maryland-Virginia areaby Hack haveproducedthe relationship6
L : I.4Ao6

(10.2)

where L = the streamlength measured miles to the drainagedivide in A : the drainage area(miz) Hack's observations indicatethat as the drainagebasinincreases size,it becomes in longer and narrower; thus precisegeometricsimilarity is not preserved. Drainageareahas long beenusedas a parameterin precipitation*runoff equations or in simpleequationsindexingstreamflowto area or other parameters. Many early empirical equationsare of the form3 Q: cA: a measureof flow suchas mean annual runoff where Q A : the size of the contributine drainaeeatea

(10.3)

Valuesof c and m are determined regression by analysis(seeChapter26); Fig. 10.3 illustratesa relation of this form.

r u o
8 ( ( ) 6
= a

./

10

20

30 40 5060 80 100 Area tmi2)

200

Figure 10.3 Runoff-drainage arca correlation for five Maryland streams (1933 storm

RUNOFF AFFECTING 1O.2 BASINCHARACTERISTICS

161

includedefinitionsof the basinshapeandthe densityof the Other arealmeasures drainagenetwork or drainage density, definedas the ratio of total channelsegment lengthi cumulatedfor all streamorderswithin a basin to the basin area.The stream in segments a drainagebasin (total frequency is defined as the summation of all of number of segments all orders)divided by the drainagearea.

Channeland BasinGradients
havea very strongeffecton the surface basinand its channels of The slopes a drainage streamchannelprofiles exhibit the characteristic runoff pro""r, of thairegion. Most direction.Figure 10.4illustratesthis slopeproceedingin a downstream of decreasing particular.trait. Also illustrated in drop dividedby the channt elevation suchthat the areasbetweenthe ave that is, A, : Azin the figure. The g to asparameters describedrainaget describe make this clear. Some mathemalicalfunctions that are used to more fully streamprofilesarelinear, exponenti ical value to representthe Primal This factor,known as th Schwartz.T a uniform channelthat is equivalen sametraveltime. This factorhasber maximum to from the centerof massof rainfall excess the peak rate of runoff ) and discharge. In additionto the slopeofthe streamchannel,the overall land slopeofthe basin factor.A quantitativerelation betweenvalley wall slopes is an important topographic method A has and streamcnannetstopes been derivedby Strahler.3 commonly used has been presentedby Horton'SThe method of determiningthe slopesof a basin map of the drainagearea grid over a topographic a involvessuperimposing transparent b"t*""n its intersectionswith the drainage in question.nacfr grlJnne is measu."d needed' of divide; the numberof intersections eachgrid line with a contourline is also then be made using A determinationof the land slopecan nsec0, S:--t (10.4)

Distance from head of stream

Figure 10.4 Typical streamprofile'

162

oHAPTER RUNoFFAND oATCHMENT 10 THE

Drainage boundary Horizontal grid line

uContour'

Contour interval = 50 ft scaie

Figure 10.5 Determinationof meanland slope:numberof vertical intersections: 72; tumber of horizontal intersections= 120; total : length of vertical grid segments 103,900ft; total length of : horizontalgrid segments 101,200 ft. 72x50 120 x 50 : 0.035 : 0.059 ftlft Ss : ftlft " 10 t , 2 0 0 103;900 .s + (.. 0.035 0.059 + : 0.047 Mean slope : ftlft 2

,s:

where n : the total number of contour intersections the horizontal and bv vertical grid lines t (horizontal and vertical) the total length of grid line segments h - the contour interval 0 * the anglemeasured betweena normal to the contoursand the grid line
L _

Because0 is very difficult to measureit is often neglected,and separatevaluesof average slopein the horizontal and vertical are computedand then averaged obtain to an estimate the meanland slope.This procedure illustratedin Fig. 10.5. of is

Area-ElevationRelation
How the areawithin a drainagebasinis distributedbetweencontours(Fig. 10.6)is of interestfor comparingdrainagebasinsand gaining insight into the storageand flow characteristics the basin.For suchstudies.an areadistribution curve suchas that of shown in Fig. 10.7 is used.The curve can be obtained by planimeteringthe areas

RUNOFF AFFECTING CHARACTERISTICS 10.2 BASIN

163

. ' , \: \ I . i/ z ' r '

\

l l \ ' / t r 1 l l \ / / / / r r \ f t /

,ffi_-_.w;i.i
.Figure 10.6 Topographic of WendyRun drainage area map lines. 20-,40-,and60-ftcontour showing betweenadjacentcontoursor by using a grid as in Fig. 10.5 and forming the ratio of the number of squaresbetweencontoursto the total number of squarescontained within the drainageboundaries.The mean elevationis determinedas the weighted betweenadjacentcontours.The medianelevationcan be deteraverage elevations of from the area-elevationcurvesas the elevationat 50 percent, rnined

300

E 2so
o

I o
r! '

2oo l)u
100 Median elevation

10

20

30

40

50

60

70

80

90

100

of Percentage area

Figure 10.7

An area-elevation distribution curve.

1 64

CHAPTER10

RUNOFFAND THE CATCHMENT

Drainage Basin Dynamics

Geomorphology' likg hydrology,was largely qualitative in nature in its formative years'With the passing time and the greatlr needfor of reliablequantitativeinformation, the science progressed the point whererational relations has to between variables are being developed. Theserelationsire usually intendedto quantify theinieractions betweenthe factorsthat modify the land forrnand the land iorm ifsef. In addition, equations relatingthegeomorphic propertiesto hydrologic,climatologic, vegetative or factorsare beingsought.some of the iunctional ielations of particulir significance to the hydrologistwill be discussed the following chapters. in

10.3 RUDIM ENTARY ECI PR PITATION-RUNOFF RELATIONSHIPS

ritation and runoff has beento plot annual rend line, and estimatethe percentageof rtities determinedthis way, however, are :e ofreliability is higherfor drainageareas rnal or other types of variation, that is, an e procedure.The resulting equationtakes

the form

o:($)rr-Pr)
1902.
I

(10.5)

1898r

l .
a 903

^ 4 0
190r l
- J 6 .E

I t28 a Slope

8 3 6
. ) J +

--l-I --i-1920 1l : -1 9 1 8 i q)5-oL f--a 7921o 193 0 . 1897 90 1924 1932

,77 X 23
1 926
a

0.57

trg

29_

32

o1!

1 0 1 1 t 2 t 3 ! 4

Runoff (in.)

Figure 10.8 Annual precipitation and annual runoff in the NeoshoRiver basin above Iola, Kansas.(U.S. GeologicalSurvey Data.)

165 10,3 RUDIMENTARYPRECIPITATION-RUNOFFRELATIONSHIPS

S : the slopeof the line (LPILQ) Pa : & baseprecipitation value below which Q is zero From Fig. 10.8 the relation for the examplewould be Q: 0'57(P 24) where Q and P are the annual runoff and precipitation, respectively,in inches' relation indicatesthat scatterof severaldata points from the assumed Considerable be used with care. For rough approximationsin this type of computation should preliminary planning studies,suchmethodsare frequently helpful, however.Equaprecipsuchasantecedent Eq. 10.5areimprovedif otherparameters tionsresembiing are characteristics included. Suchrelations and storm itation, soil moisture,season, or techniques graphicalmethods'Linsley using multiple regression can be described correlatreatmentof methodsfor developing and co-workerspresenta very complete tions involving severalvariables.e Soil moisturerelationsnormally havea soil moistureindex as the independent soil of variable,sincedirect measurements actualantecedent moistureare not generat insertedare groundwater,flow the beginning ally practical. Indexesthat havebeen Groundwatervalues precipitation, and basin evaporation.lo of the storm, antecedent shouldbe weightedto reflect the effectsof precipitation occurring within a few days from previousrains may affect results. of the storm becausesoil moisture changes employedto estimatesoil moistureamounts, can measurements be Pan-evaporation since eviporation is related to soil moisture depletion.tl Antecedentprecipitation precipitationis readily indexes(API) haveprobably receivedthe widestusebecause deficiencyof the basin. measuredand relatesdirectly to moisture precipitation index is A typical antecedent (10.6) Po: aP6'l bP, -f cP., where

p,." po, :

P.:

index (in.) precipitation the antecedent

fi"HTlJ:;fJil:rr:Hifl3*ff"-e

and presenr for vear

Coefficientsa, b, andc are found This index links annualrainfall and runoff values.r2 to by trial and error or other fitting techniques producethe best correlationbetween precipitation index. The sum of the coefficientsmust the runoff and the antecedent be 1. haveproposedthe following API for use with individual Kohler and Linsley13 storms:
+ P o: b r P l + b 2 P 2 ' ' ' + btPt

(10.7)

on wherethe r subscript P refersto precipitationwhich occurredthat many daysprior to b to the given storm,and the constants (lessthan unity) are assumed be a function In by of t. Valuesfor the coefficientscan be determined correlationtechniques' daily to evaluationof the index, b, is considered be related to / by b,: K' (10.8) normallyrepoftedin the range0.85-0.98.The initial constant whereK is a recession

166

cHAprER1o RUNoFFANDTHE cAToHMENT

value of the API (P"s)is coupled to the API / days later (P",) by
Por: PagK'

(10.e)

To evaluatethe index for a particular day basedon that of the precedingone, Eq. 10.9becomes
Po, : KPox

(10.10)

because : I. t Various empirical relations for API have been proposed.Most are based on correlating two or three variablesand at best yield only rough approximations.In many casesthesewere developed physical principlesor dimenwithout considering sional homogeneity. addedshortcomingis that many formulas fit only a specific An watershedand have little generalutility. Empirical equationsdemandgreat caution and an understanding their origin. of EXAMPLE 10.1 PrecipitationdepthsP, for a l4-day period are listed in Table 10.1. The API on April 1 is 0.00. Use K : 0.9 and determinethe,API for each successive day. Solution. Equation 10.9 reducesto API,: K(API,-1)+ P,

which was applied in developingthe successive valuesof API, in Table 10.1. TABLE 10.1
(A Precipitation April 1 2
J

API,

'6 7 8 9 10
ll

t2 l3 t4

0.0 0.0 0.5 0.7 0.2 0.1 0.0 0.1 0.3 0.0 0.0 0.6 0.0 0.0

0.00 0.00 0.50 1.5 l t,24 t.22 1.10 1.09 1.28 t.l5 1.04 1.54 1,39 |.25

10.4 STREAMFLOW FREOUENCY ANALYSIS

Hydrologists estimatestreamflows. Two approaches employed. are The first is a physical processes approachin which runoff is computed on the basis of observedor precipitation.The secondis foundedon statisticalanalyses expected ofrunoffrecords .- --vvi{}out resott-to precipitation data. Such investigationsusually include frequency

ANALYSIS FREQUENCY 10.4 STREAMFLOW

167

C a o a ao

12 months

1.0 Example 1.0 in drought with R.I. = 17 yr

x.

0.1 10
RecurrenceInteryal (Yr) 100

for dataconsolidated FiveRivfrequency Figure 10.9 Low-flow (AfterWhipple.r5; ers. studies(Chapter 27) to indicatethe likelihood of certain runoff eventstaking place. A knowledgeof the frequencyof runoff eventsis helpful in determiningrisks associare analyses Frequency or atedwith proposeddesigns anticipatedoperatingschemes. of usuallydiiectedtowardstudies miximum (flood) andminimum (drought)flows.ra'1s Figure 10.9 illustrates a typical drought frequency analysis.Unfortunately,many existingrunoff records are short-term; as a result they limit utility for reliable frequency analyses.Few adequaterecords are availableearlier than about 1900. In some cases,sequentialgeneratingtechniques(Chapter 22) can be used to develop syntheticrecords. Time-series analysesare particularly pertinent to the problem of estimating trends, cycles, and fluctuationsin hydrologic data. They also permit derivation of by generatingprocesses which syntheticrecordsof runoff can be developed.

Intervaland Frequency Recurrence

interval over a long period of The recurrenceinterval (R.I.) is definedas the average magnitudeof somehydrologicvariableis at least yearsduring which a corresponding met. This parameteris also called the return period, and sometimes,though less appropriately,thefrequency of the event.For the examplein Fig. 10'9, droughtsless in than 1 in. occurred 8 ofthe 136yearsofrecords.The 1.0in. droughthasan average one recurrenceinterval of about 17 years.Statedanotherway,on the average, year of a drought of at most 1.0 inch. is every l7-year sequence expectedto experience droughtis 8/136 = 0'059, or about Similarly,eachyearthe probabilityof a 1.O-in. probability or frequency, and is the 6 percent. This is defined as the exceedence re-iprocal ofthe return period.It shouldbe obviousthat the 1.0-in.droughtcould

168

CHAPTER RUNoFFAND oAToHMENT 10 THE occur in any year,or in severalconsecutive years.This type ofanalysiscannottell the investigatorwhat will happenthis year or next, and allows only an estimateof the average recurrenceinterval and the probability of occurrencein any given year. This subjectis fully developed Chapter27. in

10.5 STREAMFLOW FORECASTING

Surfacewaterhydrologyis basicto the designof many engineering works and important in waterquality management schemes. addition, the ability to providereliable In forecasts flowsfor shortperiodsinto the future is of greatvaluein operatingstorage of and other works and in planning proper actions during times of flood.e'16 good A exampleis the operationof a reservoirwith an uncontrolledinflow but with a means of regulatingthe outflow. If information on the natureof the inflow is determinable in advance, then the reservoircan be operated somedecisionrule to minimize downby streamflood damage.Suchoperationscan be computerizedto continually improve estimates basedon incoming dataandthus offer direction on the natureof the releases to be made.For river forecaststo be reliable, adequate, dependable data on various watershedand meteorologicconditions are neededon a continuing basis.Modern monitoring stationscapableof telemetering data to computercontrol centersprovide an important supportfunction for forecasting. The methodsusedto forecastflows are basically the same ones empfoyedin design: precipitation-runoff equations,unit hydrographs, watershed mldels, and flow-routing techniques.

r Summary
Runoff is probablythe most complexyet most important hydrologicprocess underto stand.It has attractedthe attention and focus of engineersand scientistsand comprisesthe greatestpercentage far of most hydrology textbooksand publications. by The conceptsintroducedhere will be more fully developed the next six chapters, in as well as in significantportions of Parts Five and Six.

PROBLEMS

10'1'ffi,',T'#:fl '#':iJf :Jfr Hfi ;J3;:1,!;";ffi *:l;J:ff :?'fi#.n.}il:H:i*

what purposesmight you use this? 10.2. Selecta rain gaugerecord of interest.Use the annual valuesas data to calculatethc coefficientsof an antecedent precipitation index of the form of Eq. 10.8. 10.3. Determinethe drainagedensityof the basin shown.Area : 6400 acres.Lengths are in miles.

REFERENCES 169

Lengths of channel segments in mi between points

Figure for Problem 10.3 from one or two other hydrologytexts, or L0.4. Using any dictionary,plus indexes glossaries find and compare definitions of the following terms: runoff, direct runoff, direct surface runoff, surface ruryoff,surface water, overlandflow, streamflow,drainage, watershed,catchruen&inage basin, subbasin,dr ainage divide.

"systematic Procedure for Evaluating Partial Areas of Watershed 1. Boughton, W. C., Runoff," Proc. ASCE J. Irrigation and Drainage Engineering116, 1 (February 1990). "Drainage Basin Characteristics,"Trans.Am. Geophys.Union 13(1932). 2. R. E. Horton, "Geology-Part II," in Handhok of Applied Hydrology. New York: 3. A. N. Strahler, McGraw-Hill, 1964. "TopographicCharacteristics DrainageBasins,"U.S. Geological of 4. W. B. L4ngbeinet al., Survey,Water Supply Paper,968-c(I941 ). "Erosional Developmentof Streamsand Their Drainage Basins:Hydro5. R. E. Horton, physicalApproach to Qualitative Morphology,"Bull. Geol. Soc.Am.56(1945). "studies of Longitudinal StreamProflles of Small Watersheds," Tech. Rept. 6. J. T. Hack, 18, Columbia University,Departmentof Geology,New York, 1959. 7. A. B. Taylor and H. E. Schwartz,"IJnitHydrographT ag and PeakFlow Relatedto Basin s. s," Characteristic Trans.Am. Geoplty Union 33(1952). "Discussionof Paper,Flood Flow characteristicsby c. S. Jarvis," ?ans. R. E. Horton, 8. ASCE 89(1926)'. 9. R. K. Linsley, Jr., M. A. Kohler, and J. L. H. Paulhus,AppliedHydrology,2nd Ed' New York: McGraw-Hill, 1975. 10. ven Te chow (ed.), Handbook of Applied Hydrology. New York: McGraw-Hill,1964. 11. R.K.Linsley,Jr.,andWC.Ackerman,"MethodofPredictingtheRunofffromRainfall," Trans.ASCE 107(1942). 12. S. S. Butler, EngineeringHydrology. EnglewoodCliffs, NJ: Prentice-Hall, 1957.

170

oHAPTER 10 RUNoFF ANDTHEoAToHMEI 13. M. A. Kohler and R. K. Linslev. , / WeatherBureau,Res.Paper34, l95L T4, Leo R. Beard, "statistical Methodsin Hydrology," U.S. Army EngineerDistrict, Sacrf mento,CA,1962. 1 5 . William W. Whipple, Jr., "RegionalDrought FiequencyAnalysis," Proc. ASCEJ. Irrigd

16.

1'1 Chapter

Hydrographs

r Prologue
The purposeof this chapteris to: . Characterizea hydrographas a time plot of the dischargeof surfacerunoff arld groundwaterfrom drainagebasins. ' Introduce the componentsof hydrographs that the reader can rel9le them so of to the quantltative assessments runoff presentedin subsequdnf-chapters. . Describethe time relationshipsmost commonly usedin hydrographanalysis. Hydrographanalysisis the most widely usedmethod of analyzingsurfacerunoff. Its presentationin most textbooks is normally confined to one chapter. Becauseof to three chaptersare dedicated the in numerousdevelopments hydrographanalyses, the and expands conceptsintrosubjectin this text. Chapter 11 defineshydrographs duced in Chapter 6. The conceptsreferred to as unit hydrographtechniquesare packaged togetherin Chapter 12. Individual streamflowhydrographshapesvary as called hyflow travels downstream,and the conceptsfor analyzingthese changes, ' in drographrouting methods,are presented Chapter 13.

HYDROGRAPHS 11.1 STREAMFLOW

the hydrographprovides rateof flow at all pointsin time duringand after A streamflow or a storm or snowmeltevent.l Hydrologistsdependon measured computed(synthesized) hydrographsto provide peak flow rates so that hydraulic structurescan be designedto accommodatethe flow safely.Becausea hydrographplots volumetric flow any two rates againsttime, integration of the area beneath a hydrographbetWeen the points in time givesthe total volume of water passing point of interestduring the allow analysisof sizesof time interval. Thus, in addition to peak flows,hydrographs storage tanks, detentionponds,and other facilities that deal with volumes reservoirs, of runoff. A knowledge of the magnitude and time distribution of streamflow is and of to essential many of theseaspects watermanagement environmentalplanning.

172

CHAPTER11

HYDROGRAPHS
Storm period hydrograph
o u

q b0

Y41 oar.

Hydrograph in period of no direct runoff and where no reservoir regulation exists rcflects discharge from groundwatet Storm period hydrograph End of Beginning of

Inflection point

\?

direct runoff

direct runoff

,l

Continuous hydrograph

Figure 11.1 HydrographO"o;;.

SHAPE AFFECTING HYDROGRAPH 11.2 FACTORS

A hydrographhas four componentelements:(1) direct surfacerunoff, (2) interflow, The rising pott(rn of a (3) groundwater baseflow, and (4) channelpregipitation.2 or curye;the regionin the vicinity of the\eak hydrograph known astheconcentration is The shapeof a is called the crest segment;and the falling portion is the recession.3 precipitation pattern characteristicsand basin properties. hydrographdependson presented. Figure 11.1illustrates definitions the

Processes Precipitation-Streamflow
to Duriirg a given rainfall, water is continually being abstracted saturatethe upper levelsof the soil surface;however,this saturationor infiltration is only one of many Rainfall is also interceptedby trees, plants, 4nd roof continuous abstractions.4-6 surfaces,and at the same time is evaporated.Once rain falls and fulfills initial collect falling rain to form small requirementsof infiltration, natural depressions puddles,creatingdepression storage.In addition, numerouspools of water forming within the waterand impermeablesurfaces detentionstoragebuild up on permeable shed.This storedwater gathersin small rivulets, which carry the wateroriginatingas then into larger channels,and flnally as channel overlandflow into small channels, to the watershed outlet. Figure 11.2aillustratesthe distribution of a prolonged flow uniform rainfall. Although such an event is not the norm, the conceptis useful for storagewould be distributed. showingthe mannerin which detentionand depression possesses certain amountof baseflow a In general,the channelof a watershed during most of the year. This flow comesfrom groundwateror spring contributions as and may be considered the normal day-to-dayflow. Dischargefrom precipitation excess-that is, after abstractions are deducted from the original rainfallthe constitutes direct runoff hydrograph(DRH). Arrival of direct ninoff at the outlet enoughtime continues, for accounts an initial rise in the DRH. As precipitationexcess the qlapqgs progressively for distant areasto add to the outlet flow. Consequently,

11.2 FACTORSAFFECTINGHYDROGRAPTI SHNPC

179

Depressionstorage
fi

Time (b)

Figure 11.2 (a) Distribution of a uniform storm rainfall for condition of is no interceptionloss.Note that all water storedin depressions ultimately is to evaporated infiltrated while somedetentionstorage also subjected or (b) theselosses. Equilibrium discharge hydrograph.

duration of rainfall dictatesthe proportionatearea of the watershedamplifying the peak, and the intensity of rainfall during this period of time determines resulting the greatestdischarge.

Shapes Hydrograph
If the rainfall maintainsa constantintensityfor a long enoughperiod of time, a state of equilibrium dischargeis reached, as depicted by curve A in Fig. 11.2b. The inflection point on curveA often indicatesthe time at which the entire drainagearea contributes to the flow. At this time maximum storase of the watershedis only

174

CHAPTER11

HYDROGRAPHS

and partially complete.As rainfall continues,maximum storagecapacityis attained (runoff)] is reached.The condition of iqrliUri"t" finflow (rainfall) equals outflow Extended maximum storageand equiiibrium is seldom if ever attained in nature' its duration negateany rainfall lnuy o"Iu., but viriations in intensity throughout possibility of u IRH of the theoretical shapefor constantrainfall intensity' Anormalsingle.peakDRHgenerallypossessestheshapeshownbycurveBin peak magnitudeof this in Fig. 11.2b rather tian iy the "ur',oJ Fig. 11.2a.The time to ofthe rainfall' and the size,slope, onih" intensity and duration hyirograph depends for once peak flow has beenreached a capacityof the watershed. shape,and storage its source of supply coming given isolatedra'instorm,itt" ORg beginsto descend, suchas detentionand channel within the watershed largely from water accumulated storage. by processes involvedin forming the DRH can be better understood visualizing theprecipitationexcessaspartiallydisposedofimmediate$bysurfacerunoffwhile later from boundariesand is released a portion remainsheld within the watershed integratedeffectsof the duration Thus the shapeand timing of the DRH are storage. factorsaswell asthe effect of intensityofrainfall and other hydrometeorological and storagecapacity' upon the factorsof the watershed the physiogiaphic

COMPONENTS 11.3 HYDROGRAPH

how the hydrographcan be subdividedinto its com\It is important to understand ofprecipitation andwatershed shape and nenrpar15 to look at the effecton hydrograph purpose' Figures11.3and Il'4 areusedfor this features. graphshowingthe rate of streamflowwith respect A hydiographis a continuous strip recorderthat indicatesstage to time, ,rtr*ityoutained by -"anr of a continuous hydrograph which is then transformedto a discharge vefsustime (stale hydrograph), lrydrographis generallytaken to by applicationof a rating curve.Hereafter,the term hYdrograPh' indicate a discharge a pgriod Figure t t.gi lttustrai"* ih" hydrographof a permanentstreamduring lrydrographbecausegroundbetween precipitation events,known i" i bot" flow of the base water sustainsthe flow. Four general conditions cause modification HortonTusing the following sets of flow hydrographshape.They aL describedby inequalities: Setl i<f F(S, Set2 i<f F)S, set3 i>f F(S, Set4 i>f F)Sr.

11.3 HYDROGRAPH COMPONENTS 175

(e)

(0

fo)

(h)

Figure 11.3 Effectsof stormandbasincharacteristics hydrograph on shape. or addedgroundwatercomponents develops. The entire effect of the storm would be to slightly reducethe soil moisturedeficiencysr. The field capacityis the amountof water held in the soil after excess gravitationalwater has drained. The conditions describedby Set 2 still do not produce direct surfacerunoff, although the comlponentsof interflow and groundwater flow are added to channel precipitation.The initial hydrograph would be modified,sincethe field capacityof the soil is exceeded. Figure 11.3b illustrates this condition. Note that deviation of the hydrographfrom the original baseflow curve is likely to be very small under these conditions. Figure 11.3c illustrates a casewhere surfacerunoff becomesa componentof flow because > f. In this example,interflow and groundwater i flow are zero,assoil

176

CHAPTER11 HYDROGRAPHS

End ofrainfall

9C

Groundwater flow

Time

of Figure 11.4 Components thehydrograph.

moisture deficiency still exists,although at a reduced level. Channel precipitation likewise constitutesa component. In the final set, Fig. 11.3d, all four componentsexist with rainfall intensity This casewould inflltration rate andthe field capacityofthe soil is reached. exceeding be typical of a large storm event. Figures 1l.je-h illustrate how hydrograph shape can be modified by areal Minor variations in rainfall and rainfall intensity and by watershedconfiguration.8 are linked to variationsin stormintensity.In fluctuationsshownin thesehydrographs Fig. 11.3eonly the delayingeffectspertinent to a storm over the upstreamsectionof Figures11'3g of the Figure 11.3fshows reverse this condition. the areaareindicated. effects of basin geometry. and h depict the comparative interflow and channelprecipitation are grouped In most hydrographanalyses, Channelprecipitation begins treated independently. with surfacerunoff rather than endswith the storm.Its distributionwith respectto time with inceptionof rainfall and is highly iorrelated with the stormpattern.The relative volume contributiontendsto increasesomewhatas the storm proceeds,since stream levels rise and the water and areaoccupiedby streams The fraction of watershed areatendsto increase. surface on the order of 5 percent or less,so the percentage lakes is generallysmall, usually of rrnoff relatedto channelprecipitation is usually minor during important storms. rate up to by Distribution of interflow is commonly characterized a slowly increasing period, followed by a gradual recessionthat terminatesat the the end of the storm intersectionofthe surfaceflow hydrographand base flow hydrograph.Figure 11.4 illustrates the approximatenature of the componentsof channelprecipitation and interflow. The baseflow componentis composedof the water that percolatesdownward as reservoirand then flowsto surfacestreams groundthe until it reaches groundwater The groundwaterhydrographmay or may not show an increase water discharge.

11.4 BASE FLOW SEPARATION

177

'

during the actual storm period. Groundwateraccretion resulting from a particular storm is normally releasedover an extendedperiod, measuredin days for small watersheds and often in months or yearsfor large drainageareas. The surfacerunoff component consistsof water that flows overland until a streamchannelis reached.During large stormsit is the most significanthydrograph component. Figure 11.4illustratesthe surface runoff and groundwater components of a hydrograph. pointed out in Fig. 11.3,the relative magnitudeof eachcomponent As for a given storm is determinedby a combinationof many factors.Hydrographsare analyzed provideknowledgeof the way precipitationand watershed to characteristics interact to form them. The degreeof hydrographseparation requireddepends the on objective of the study. For most practical work, surfacerunoff and groundwater components only are required.Research projectsor more sophisticated analyses may dictate considerationof all components.When multiple storms occur within short periods,it is sometimesnecessary separate overlappingparts of consecutive to the surfacerunoff hydrographs.

11.4 BASEFLOWSEPARATION
Severaltechniques usedto separate hydrograph's are a surface and groundwater flows. Most are basedon analyses groundwater recession depletioncurves.If there is or of no addedinflow to the groundwater reservoir,and if all groundwater discharge from the upstreamareais interceptedat the stream-gauging point of interest,then groundwater discharge can be described either e'to by er: eoK' or er: eo-K' where eo : q, : K: e: a specifiedinitial discharge the discharge any time / after flow 46 at arecessionconstant baseof natural logarithms

( 11 . ) 1

Time units frequently used are days for large watersheds and hours or minutes for small basins.A plot of either yields a straight line on semilogarithmicpaper by plotting / on the linear scale. groundwaterdepletioncharacteristics approximately For most watersheds, are stable,sincethey closelyfit watershed geology.Nevertheless, recession the constant varieswith seasonal effectssuchas evaporation and freezingcyclesand other factors. Becauseq, dt is equivalentto -dS, where S is the quantity of water obtainedfrom storage, integration ofEq. 11.1produces
4 s - Q t - Ko log"

(rr.2)

This equation determinesthe quantity of water releasedfrom groundwaterstorage betweenthe times of occurrenceof the two discharges interest,or it can be usedto of calculatethe volume of water still in storageat a time some chosenvalue of flow occurs.To get the latter, 4, is setequalto zero and qsbecomes reference the discharge. Figure 11.5ais a plot of Eqs. 11.1and II.2 andprovides additionaldefinition.

178

CHAPTER11

HYDROGRAPHS

qo

.. ,?

K
\r3
t (a.)
?r

qo discharge at rcginningof arr interval discharge at :nd of an inten,a\ Qt

I
I

I
q d

I
/

I I

Interval = unit period in / whicht is expreissed .

qo
Qt

Time (b) flow model' Figure 11.5 Base to

t1

by Groundwaterdepletioncurvescan be analyzed variousmethodsto evaluate Data from a stream-gaugi.ng constantK. One of thesewill be described. the recession and shouldreflectrainlessperiodswith no upstreamregulastationare a prerequisite tion,'suchas a fesefvoir,to affect flow at the gaugingpoint. Otherwisean adjustment with its own enors is introduced. From the streamflow data, plot a portion of the recession hydrograph (Fig. 11.5b) to find values of dischargeat the beginning and end of selectedtime to intervals.Flows at the beginningof eachinterval are analogous 4e,whereasthose to at the end are analogous q1.Next, selectseveraltime intervalsand plot correspondvalues consecutive q,'s shownin Fig. 11.6.The time periodbetween ing qe'sversus curvesof of 4 shouldbe identical for eachdatum set. Figurestaken from recession times that still reflect surfacerunoff will usually fall below and to the right of a 45o with larger numbersfor line drawn on the plot. Thesevalueswill also be associated Points taken from true groundwater recessionperiods should approximately 4.

11,4 BASEFLOW SEPARATION 179

r.-,x. ^are points plotted from discharges for different groundwater discharge periods eo nd qt are taken for equal time

,/
o

I
A

,(
.5r"2'

/(^lI
- s l o p e o 0 A= f
inEq11.1

9/ .,

fi:

/#

/
Figure 11.6 Graphical rnethod for determiningrecessionconstant K. (U.S. Departmentof Agriculture, Soil ConservationService.)

describe straightline. a Because : eo : 0 when4o: 0, a straightline canbefitted h graphicallyto the datapoints.The slopeof this line is qrfqo : K. Usingthis value,the depletioncurve plots as a straight line on semilogarithmic paper (r is the linear scale variable) as a curveon arithmeticpaper,Fig. 11.5a. or

Techniques Separation
r Severalmethodsfor baseflow separationare used when the actual amount of base flow is unknown. During large storms,the maximum rate of discharge only slightly is affectedby baseflow, and inaccuracies separationmay not be important. in The simplest baseflow separation technique to draw a horizontalline from the is point at which surfacerunoff begins,PointA in Fig. 11.7, to an intersectionwith the hydrograph recession wherethe baseflow rate is the sameasat the beginningof direct runoff as indicatedby Point B. A secondmethodprojects the initial recession curve downwardfromA to C, which lies directly below the peak rate of flow. Then point D on the hydrograph,representing daysafter the peak, is connectedto point C by a N straightline definingthe groundwater component.One estimateof N is basedon the formula3
N - Ao'2

(11.3)

where N : the time in days A = the drainageareain squaremiles A third procedureis to developa baseflow recession curve using Eq. 11.1 for data from the segmentFG, and then back-calculateall base flow to the left of Point fl

180

cHAPTER HYDRocRApHS 11

9p F

Time

Flgure 11.7 Illustration some of hydrograph separation techniques. wherethe computedcurve beginsto deviatefrom the actualhydrograph, marking the end of direct runoff. The curve is projected backward arbitrarily to some Point E below the inflection point and its shapefrom C to E is arbitrarily assigned. fourth A widely usedmethod is to draw a line betweenA and F, and a fifth common method is to project the line AC alongthe slopeto the left of A , and then connectPointsC and

o d d

12N 6P

t21|'{{ 6A

12N 6P

tz]0|{ 6,{

I2N

6P

IzM

Figure 11.8 Illustration of base flow separation:hydrograph for the Uharie River near Trinity, North Carolina, February 25, 1939. (U.S. Departmentof Agriculture, Soil ConservationService).

TIME RELATIONSHIPS 181 11.5 HYDROGRAPH

is of B. All thesemethodsare approximatesincethe separation hydrographs partly a subjectiveprocedure. techniques determinesurface to Figure 11.8illustratestwo graphicalseparation the . runoff and groundwaterflow components Line AD represents simpleprocedure of connectingthe point of the beginningof direct runoff with the flrst point on the because overthe horizontal line technique groundwater recession curve (an advantage from the the time baseof direct runoff is much shorter).Ctrve ABCD is constructed curve. extensionof the baseflow recession

TIMERELATIONSHIPS 11.5 HYDROGRAPH

Wavetravel time is definedas the time required for direct runoff originating at the most remotepoint in the channelto reachthe outlet. The last drop of direct runoff to and reaches outlet at the the passthe outlet conceptuallytravelsoverthe watersurface velocity of equal to the average speedof a small surfacewave,rather than at a speed velocity and varieswith channel flow. The wavetravel time is fasterthan the average 5/3 (see shapeand other factors.For a rectangularchannel,the ratio is approximately (Fig. 11.4) of a hydrographis Section 13.1 for other wavevelocities).The time base to considered be the time from which the concentrationcurve beginsuntil the directzero. An equationfor time basemay take the form runoff componentreaches T6:t"*t, where To : the time baseof the direct runoff hydrograph /" : the duration of runoff-producingrain rainfall releasetime t- : the excess Watershed time, illustrated in Fig. 11.4, is definedas the time from the penterof lag tt massof effectiverainfall to the centerof massof direct runoff. Other definitionsand are characteristics providedin S_ecseveralequationsrelating lag time to watershed chapters. tion 11.7 and subsequent theory, the excess-rainfallrelease Because ofits importancein unit hydrograph time is introduced.This is definedasthe time requiredfor the last, most remotedrop of to ofexcessrain that fell on the watershed passthe outlet, signallingthe cessation direct runoff. It is easilydeterminedas the time interval betweenthe end of rain and as the end of direct runoff. Only that part of the outflow which classifles direct runoff outflow nor(excess rain) is consideredin dqterminingthe releasetime. Watershed of after cessation direct runoff, in the form of interflow andbaseflow. mally continues Release time is very similar by definitionto wavetraveltime and time of concentration (Section 11.6). A foundational assumptionof unit hydrographtheoryl2 is that the watershed of releasetime is a constant,regardless the storm duration, and is related to excess release time is also The characteristics. excess basinfactorsratherthan meteorological unit conceptuallyidenticalwith the time baseof an instantaneous hydrograph(IUH). rain applieduniformly over the This is the runoff hydrographfrom 1.0 in. of excess in watershed an instant of time (seeChapter 12). Both wavetravel time and excesswith time of concentration. rainfall releasetime are often used synonymously

(rr.4)

182

CHAPTER11

HYDROGRAPHS

11.6TIMEOF CONCENTRATION
The most common definition of time of concentrationoriginatesfrom consideration the of overlandflow. If a uniform rain is appliedto atract, the portions nearest outlet contribute runoff at the outlet almost immediately.As rain continues,the depth of growsand discharge ratesincrease throughout.Runoff contribuexcess the surface on to tions from variouspoints upstreamarrive at later times, addingthemselves continuing runoff from nearerpoints, until flow eventually arrives from all points on the watershed,"concentrating" at the outlet. Thus, concentrationtime is the time required,with uniform rain, for 100percentof a tract of land to contributeto the direct runoff at the outlet.e As a secondpopular definition, the concentrationtime is often equatedwith the either the excess-rainfall releasetime or the wave travel time because time for is runoff to arrive at the outlet from the most remotepoint after rain ceases assumed to be indicative of the time required for 100 percent contribution from all points during any uniform storm having sufficient duration. The latter definition is often preferred becausefew storm durations exceedthe time of concentration,making rain recession. determinationof /. possibleonly by examiningexcess Because time of concentrationis conceptuallythe time requiredfor 100percent of the watershedto contribute, it is also often defined as the time from the end of excessrainfall to the inflection point on the hydrographrecessionlimb (e.g., see at Fig. 12.2).The reasoning usedin this definitionis that direct runoff ceases the point of inflection. For a small tract of land experiencing uniform rain, the entire areacontributes an at approximately sametime that the runoff reaches equilibrium.This givesrise the the to yet anotherdefinition of time of concentration.If rain abruptly ceased, direct releasetime t,. On the basis runoff would continueonly as long as the excess-rainfall release time and time of concentrationcan be considof the seconddeflnition.excess ered equivalent. Numerous equationsrelating time of concentrationto watershedparameters havebeen developed. severalpopular versions.Other variaTable 11.1 summarizes in tions are presented Chapters12, 15,16, and25.

11.7 BASIN LAG TIME The relative timing of rainfall and runoff must be known if drainageareashaving subbasins to be modeledor if continuoussimulation is desired.A basic measure are positionrelativeto the causative lag, which locatesthe hydrograph's of timing is basi'n stormpattern.It is most often definedas the differencein time betweenthe centerof massof effectiverainfall and the centerof massof runoff produced.Other definitions are also used.Two of theseare ( 1) the time interval from the maximum rainfall rate to the peakrate of runoff and(2) the time from the centerof massof effectiverainfall by to the peak rate of flow. Time lag is characterized the ratio of a flow length to a mean velocity of flow and is thus a property that is influencedby the shapeof the and geometry,and drainagearea,the slopeof the main channel,channelroughness thg'storm pattern.

11.7 BASIN LAGTIME TABLE11.1 SUMMARY TIMEOF CONCENTRATION OF FORMULAS

Method and date Kirpich (1940)
tc : L : S :

183

Formula t" (min) for

0.00782077S-o38s length of channel/ditch from headwater to outlet, ft average watershed slope, ftlft

Remarks
Developedfrom SCS data for sevenrural basins in Tennessee with well-definedchanneland steepslopes(37o to 1O7o); overlandflow on for concreteor asphaltsurfaces multiply t;by 0.4; for concretechannelsmultiply by 0.2; no adjustments for overland flow on bare soil or flow in roadsideditches. Essentiallythe Kirpich formula; developed from small mountainousbasinsin California (U.S. Bureauof Reclamation, 1973,pp. 67-7I).t4

USBR Design of Small Dams

t" : L : Il :

(r973)

60(lI.9L31H)o38s length of longest watercourse, ml elevation difference between divide and outlet, ft 41.025(0.0007, t c)Lozz

lzzatd (7946)ts
-c

S0.333i0.66?

I : c: Z : ,S: FederalAviation Administration ( 1970)r6

rainfall intensity,in/h retardancecoefficient length of flow path, ft slopeof flow path, ftlft

Developedin laboratory experiments Bureau of by Public Roadsfor overland flow on roadwayand turf surfaces; valuesof the retardance coefficientrange from 0.0070 for very smooth pavementto 0.012 for concretepavementto 0.06 for denseturf; solution requiresiteration; product i times Z shouldbe = 500. Developedfrom air field drainagedata assembled by the Corps of Engineers;method is intended for use on airfield drainageproblems,but has been used frequently for overlandflow in urban basins. Overland flow equation developedfrom kinematic wave analysis of surface runoff from developed method requiresiteration sinceboth i surfaces; (rainfall intensity) arrdt, are unknown; superposition of intensity-duration-frequency curve gives direct graphical solution for t". Equation developed SCS from agricultural waby tersheddata; it has been adaptedto small urban basinsunder 2000 acres;found generally good where areais completelypaved;for mixod areas it tendsto overestimate; adjustmentfactorsare applied to correct for channelimprovementand imperviousarea; the equationassumes that t" : 1.67 X basinlag. Overland flow charts in Ref. 20 provide average velocity as function of watercourse slope and surfacecover.

r" = 1.8(1.1 C)Losofso333 C : rational method runoff coefficient l, : length of overland flow, ft S : surface slope, Va O.94Lo6no6
(io 45o 3)

Kinematic WaveFormulas Morgali and Linsley (1965)'? Aron and Erborge (1973)18 SCS Lag Equation (1972)te

l, n I S

: : : :

. , ._- @

length of overland flow, ft Manning roughnesscoefficient rainfall intensity in/h averageoverland slope ftlft t.67 Lo8[(tooo/cN)- 9]oi

L : hydraulic length of watershed (Iongestflow path), ft CN : SCS runoffcurve number S : average watershedslope,Vo

SCS AverageVelocity Charts(1975, 1986),0

',' : l v L 60- v Z : length of flow path, ft V : ureragevelocity in feet per secondfrom Fig. 3- 1 of TR 55 for various surfaces

Sarrce.' After Ref. 13.

184

CHAPTER11

HYDROGRAPHS

Various studieshave been conductedfor the purpose of developingrelations descriptiveof time lag. Most prominent of thesewas the work by Snyderon large natural watersheds.2l original equation has been widely used and modified in His variouswaysby other investigators. Eagleson proposedan equationfor lag time has on sewered drainageareashavinga minimum sizeof 147 acres.2z early investigaAn tion (1936) on small drainageareas(2-4 acres)was conducted Horner in his by classical work on urbandrainage St. Louis,Missouri.23 in Horner'swork wasinconclusive in that it did not yield a deflned procedure,but he did conclude that the comparativelywide rangein the lag time at eachlocation led to the inferencethat the lag was a variable,its valuebeingdetermined more by rainfall characteristics than by characteristics the drainagearea. of Snyder'sstudybasedon data from the AppalachianMountain regionproduced the following equationfor lag time:z1
t,1: c,(L""L)o3

(1 .s) 1

where

/1 : the lag time (hr) betweenthe centerof massof the rainfall excess for a specifiedtype of storm and the peak rate of flow I'"o : the distancealongthe main streamfrom the baseto a point nearest the centerof gravity of the basin (mi) I : length of the main stream channel (mi) from the base outlet to the upstreamend of the streamand including the additional distanceto the watershed divide C, : & coefficientrepresenting variationsof types and locationsof str'eams

For the areastudied,the constantC, was found to vary from I.8 to 2.2, with somewhatlower valuesfor basinswith steeperslopes.The constantis considered to includethe effectsof slopeand storage. The value of 4 is assumed be constantfor to a given drainagearea,but allowanceis madefor the useof different valuesof lag for different types of storms. The relation is consideredapplicable to drainage areas rangingin sizefrom 10 to 10,000mi'. In a studyof sewered areasrangingin size from 0.22 to 7.51 mi2,Eagleson22 developed equation the
t " ', : -

L
(l.5 /n)R2/3 t/2 s

(11.6)

where

tr : L :

lag time, the center of mass of rainfall excess to the peak discharge (sec) the mean travel distance (ft), which is equal to the length of that portion

of the sewerwhich flows full n : the weightedManning's coefficientfor the main sewer : the weightedhydraulic radius of the main sewerflowing full { S : the weightedphysicalslopeof the main sewer Eagleson'sequationdirectly includesthe effects of channelgeometry and slope,as well as basin shape,and thus represents refinementof the Snyderapproach. also a It indirectly includesthe important effect of the storm pattern.

PROBLEMS185 of Linsley and Ackerman give examples applicationof the following modified equation.24 form of Snyder's

t,:C,+!
t t =K +

(rr.1)

where s is a weighted slope of the channel and the other variablesare as defined previously. time lag by equationsof the form haverepresented Other investigators ( 11 . 8 ) Vs

Numerous other derivations of relations for watershedlag times can be found in standardhydrologictexts and periodical literature. Others are includedwith someof in the syntheticunit hydrographdiscussions Chapter 12.

r Summary
is the Understanding structureof hydrographS important to many designand water the hydrographrepresents portion of the hydrologiccycle supply applications.The for ratesof flow in streams setting needin orderto determine mostoften that engineers and establishing designingflood protection measures, bridge lengthsand elevations, areal extent of flooding. Similarly, the volume of drainageinto a reservoiror past a water supply diversionis determinedfrom the areaunder the hydrograph.Accurate pipelines,and of estimates thesevolumesare important to designof dams,reservoirs, ' numerousother structures. including the time components, of After graspingthe fundamentals hydrograph presentedin this chapter,the reader should be well preparedfor the relationships presented throughquantitativedevelopments hydrographtheory and applications of 12 through 16 and in Part Five. out Chapters

PROBLEMS
to and usetwo different techniques separate 11.1. Referto Fig. 11.1.Replotthis hydrograph the baseflow. for 11.2. Obtain streamflowdata for a water courseof interest.Plot the hydrograph a major the event and separate baseflow. runoff 11.3. For the event of Problem I1.2, tabtlate the precipitation causingthe surfacerunoff the and determine duration of runoff-producing.rain. Estimatethe time of concentraComparethis with tion and useEq. I 1.4 to estimatethe time baseof the hydrograph. the time basecomputedfrom the hydrograph. rates at a cross sectionof a stream.The 11.4. Tabulatedbelow are total hourly discharge drainagearea abovethe sectionis 1.0 acre. a. Plot the hydrographon rectangular coordinate paper and label the rising limb and the recessionlimb. (concentrationcurve), the crest segment,

186

CHAPTERll HYDROGRAPHS plot of Q versus of b. Determinethe hour of cessation the direct runoff usinga semilog time. c. Use the base flow portion of your semilog plot to determine the groundwater constantK, recession d. Carefully constructand label baseflow separationcurveson the graph of Part a, using two different methods.

Time(hr) 0 1 2 3
I

Q (cfs)

Time (hr)

@ (cfs)

5 6 7

102 100 98 220 512 630 460 330

8 9 l0 11 t2
IJ

t4 15

2lo 150 105 75 60 54 48.5 43.5

11.5. On a neatsketchof a typical total runoff hydrograph,showor dimensionthe (a) storm

(b) hyetograph, beginningofdirectrunoff, (c) cessationtimeofdirectrunoff' (d) base fllw separationassumingthat additional contributions to base flow are negligible of during ihe period of rise, and (e) crest segment the hydrograph' of by your instructor,obtain measures the watershed assigned 11.6. For an urban watershed of length, and slope,and compareestimates the time of concentrationusing the area, in Kirpich, USBR, FAA, and SCSLag equations Table 11.1'

REFERENCES
1. American Society of Civil Engineers,Hydrology Handbook, Manuals of Engineering No. 28. New York: ASCE, 1957. Practice, "A 2. Donn G. DeCoursey, Runoff HydrographEquation," U.S. Departmentof Agricul.ture, Feb. 1966,pp.4I-116' Service, AgriculturalResearch 3. R. K' Linsley, M. A. Kohler, and J' L. H' Paulhus,Applied Hydrology' New York: McGraw-Hill, 1949. "Erosional Developmentof Streamsand Their DrainageBasins:HydroA R. E. Horton, physicalApproach to QuantitativeMorphology,"Bull. Geol' Soc'Am' 56(1945)' "An Approach Toward a PhysicalInterpretationof Infiltration Capacity," 5 . if.-n. gorton, Soil Sci.Soc.Am. 5,399-417(1940). Proc. "An Inflltration Equation with Physical Significance,"Soil Sci.77(1954). 6 . J. R. Philip, 7 . R. E: Horion";Surface Runffi Phenomena.Ann Arbor, MI: EdwardsBros., 1935. New York: Wiley' 1965' 8 . R. J. M. DeWiest,Geohydrology. ,.Hydrology,,'in EngineeringHandbook, Sec. 4, U.s. Department of Agriculture, Soil 9. Service,1972. Conservation "Discussionof Analysis of Runoff characteristicsby o. H. Meyet," Trans. 1 0 . B. S. Barnes, ASCEl0s(1940). "Unit HydrographLag and PeakFlow Relatedto Basin 1 1 . A. B. Taylorand H. E. Schwartz, s. s," Characteristic Trans,Am. Geophy Union 33(1952).

REFERENCES 187 "Streamflow 12. L, K. Sherman, from Rainfall by the Unit-GraphMethod," Eng.News-Rec. 108(1932). 1 3 . D. F. Kilber, "Desk-top methods for urban stormwatercalculation," Ch. 4 in Urban Stormwater Hydrology, Water ResourcesMonograph No. 7, American Geophysical Union, Washington, C., 1982. D. 14. U.S. Bureauof Reclamation, Designof SmallDams,2nd ed., Washington,D.C., 1973. 1 5 . C.F.Izzafi,, "Hydraulicsof RunofffromDeveloped Surfaces," Proceedings,26th Annual Meetingof the HighwayResearch Boad,26, pp. 129-146, December1946. 16. FederalAviation Administration,"Circular on Airport Drainage,"ReportA/C 050-532058, Washington, D.C., 1970. n. J. R. Morgali, andR. K. Linsley,"ComputerAnalysisof OverlandFlow,"./. Hyd.Div., Am. Soc.Civ.Eng.,9l, no. HY3, May 1965. 1 8 . G. Aron, and C. E. Egborge,"A PracticalFeasibility Study of Flood PeakAbatementin Urban Areas,"U.S. Army Corpsof Engineers, Sacramento, Calif., March 1973. 1 9 . Soil ConservationService,"National EngineeringHandbook, Sec. 4, Hydrology," U.S. Dept. of Agriculture,U.S. GPO, Washington, D.C., 1972. 20. Soil ConservationService,"Urban Hydrology for Small Watersheds," TechnicalRelease 55, Washington, D.C., 1975(updated, 1986). 2 t . F. F. Snyder,"Synthetic Unit Graphs," Trans.Am. Geophys.Union 19, 447-454(1938). 22. Peter S. Eagleson,"CharacteristicsofUnit Hydrographsfor SeweredAreas," paperpresented beforethe ASCE, Los Angeles,-CA, 1959,unpublished. 23. W. W. Horner, and F. L. Flynt, "RelationBetweenRainfall and Runoff from Small Urban Areas," Trans.ASCE 62(101), 140-205(Oct 1956). "Methodof PredictingtheRunoff aA R. K. Linsley,Jr., andW. C. Ackerman, fromRainfall," Trans.ASCE 107(1942\.

C h a p t e r1 2

Unit Hydrographs

r Prologue
The purposeof this chapteris to: . Define unit hydrographsand show their utility in hydrologicstudiesand design. . Developfully the current methodsof obtaining, analyzing,and synthesizing unit hydrographs. . Presentmethodsfor converting unit hydrographsfor one storm duration to other storm durations. Waysto predict flood peak discharges discharge from rainfall events and hydrographs havebeenstudiedintensivelysincethe early 1930s.One approach receivingconsiderable use is called the unit lrydrographmethod. 12.1 UNIT HYDROGRAPH DEFINITION The concept of a unit hydrographwas first introduced by Shermant'z 1932. He in defineda unit graph as follows:2 produces1-in.depth given one-day drainage area, Ifa given rainfall a ofrunoffoverthe thehydrograph showing rates whichtherunoffoccurred be consideredunit at can a the graphfor that watershed. baseflow) Thus, a unit hydrographis the hydrographof direct runoff (excludes net rain (the total runoff after abstracfor any stormthat produces exactly 1.0inch of tions).Sucha stormwould not be expected occur,but Sherman'sassumption that is to the ordinatesof a unit hydrographare t.O/P times the ordinatesof the direct runoff hydrographfor an equal-duration storm with P inchesof net rain. The term "unit" hasto do with the net rain amountof 1.0inch and doesnot mean to imply that the duration of rain that producedthe hydrographis one unit, whether an hour, day,or any other measure time. The storm duration,X, that producedthe of has unithydrographmustbe specified because watershed a differentunit hydrograph a

DEFINITION 12.1 UNIT HYDROGRAPH

189

is for eachpossiblestorm duration. An X-hour unit lrydrograp,h defined as a direct runoff hydrograph havinga 1.0-in. volumeand resultingfrom anX-hour stormhaving would havea 1.0-in. volume a net rain rate of 1,/Xin.lhr. Az-hr unit hydrograph producedby a 2-hqstorm,and a 1-dayunit hydrograph would be producedby a storm having 1.0 in. of eicessrain uniformly producedduring a 24-hr period. The valueX is often a fraction. Figure 12.1illustratesa2-ltr,l2-hr, and24-hrwthydrograph for a given watershed.

t (b)

24hr
T (c)

)z nl

Figure 12.1 Illustration of 2-br, I2-hr, and 24-ht unit hydrographsfor the same watershed(Note: a : b : c : 1' X A).

190

CHAPTER UNIT 12 HYDROGRAPHS rainfall By Sherman'sassumption, applicationof an X-hourunit graphto design excess amountsother than 1 in. is accomplished simply by multiplying the rainfall excessamount by the unit graph ordinates,since the runoff ordinatesfor a given duration are assumedto be directly proportional to rainfall excess.A 3-hr storm producing2.0 in, of net rain would haverunoff rates2 times the valuesof the 3-hr unit hydrograph. One-half inch in 3 hr would produceflowshalf the magnitudeof the 3-hr in unit hydrograph. This principle of proportional flows is expanded Section 12.3 and appliesonly to equal duration storms. Implicit in deriving the unit hydrographis the assumptionthat rainfall is distributed in the sametemporal and spatialpattern for all storms.This is generallynot true; consequently, variationsin ordinatesfor different stormsof equal duration can be expected. first, then presentmethods This chapteris organizedto defineunit hydrographs of deriving unit hydrographs from actual rainfall and runoff records (Section 12.2). Section I2.3 preAfter familiarizing the readerwith the origin of unit hydrographs, for sentsmethodsof applyingunit hydrographs generate to direct runoff hydrographs any storm with durationsthat are multiple integersof the U.H. duration. The constructionof unit hydrographs stormswith other than integermultifor ples of the derived duration is facilitated by a method known as the S-lrydrograph The procedure,as explainedin Section 12.4, developed Morgan and Hulinghorst.3 by employs a unit hydrographto form an S-hydrographresulting from a continuous appliedrainfall. The need to alter duration of a unit hydrographled to studiesof the unit rainfall. The concept of shortestpossiblestorm duration-the instantaneous instantaneous unit hydrograph(IIJH) is tracedto Clark6and can also be used(Secfor tion 12.5) is constructingunit hydrographs other than the derived duration. The previousdiscussion assumes the analysthasrunoff and rainfall datafor that deriving a unit hydrographfor the subject watershed.The application of unit hyand drographtheory to ungauged receivedearly attentionby Snydera also watersheds of by Taylor and Schwartz,5 who tried to relate aspects the unit hydrographto watershed characteristics. a result, a full set of synthetic unit-hydrographmethods As emerged. numberof theseare presented Section12.6. A in

12.2 DERIVATION UNITHYDROGRAPHS FROM OF DATA STREAMFLOW

can Data collection preparatoryto deriving a unit hydrographfor a gaugedwatershed recordsof haveavailable many watersheds be extremelytime consuming. Fortunately, with office records of the streamflowand rainfall, and thesecan be supplemented Rainfall records pay be Water Resources Division of the U.S. GeologicalSurvey.T by from ClimatologicalDatas publishedfor eachstatein the United States the secured National Oceanicand AtmosphericAdministration (NOAA). Hourly rainfall records for for recordingrainfall stationsare publishedas a Summaryof Hourly Observations the location. Summariesare listed for approximately300 first-order situationsin the United States. it to To developa unit hydrograph, is desirable acquireas rnanyrainfall records aspossiblewithin the study areato ensurethat the amountand distributionof rainfall

12.2 DERIVATION,OF UNIT HYDROGRAPHS FROM STREAMFLOW DATA

191

over the watershedis accurate$ known. Preliminary selectionof storms to use in deriving a unit hydrographfor a watershedshouldbe restrictedto the following: 1. Stormsoccurring individually, that is, simple storm structure. 2. Storms having uniform distribution of rainfall throughout the period of rainfall excess. 3. Stormshavinguniform spatial distribution over the entire watershed. Theserestrictionsplace both upper and lower limits on size of the watershed be to employed.An upper limit of watershedsize of appro5imately1000 mi2 is overcautious, althoughgeneralstormsover suchareasare not unrealisticand somestudiesof areasup to 2000 mi2 have used the unit-hydrographtechnique.The lower limit of watershed extentdepends numerousother factorsand cannotbe preciselydefined. on A generalrule of thumb is to assume about 1000acres.Fortunately,other hydrologic techniques help resolveunit hydrographs watersheds for outsidethis range. The preliminary screeningof suitable storms for unit-hydrographformation shouldmeet more restrictive criteria before further analysis: 1. Duration of rainfall event should be approximately10-30 percent of the drainagearea lag time. 2. Direct runoff for the selected storm shouldrangefrom 0.5 to 1.75 in. 3. A suitablenumberof stormswith the sameduration shouldbe analyzedto obtain an average of the ordinates (approximately five events). Modificationsmay be madeto adjustdifferent unit hydrographs a single to duration by meansof S-hydrographs IUH procedures. or 4. Direct runoff ordinatesfor eachhydrographshouldbe reducedso that each eventrepresents1 in. of direct runoff. 5. The final unit hydrograph a specificdurationfor the watershed obtained of is by averaging ordinatesof selected eventsand adjustingthe result to obtain 1 in. of direct runoff. Constructingthe unit hydrographin this way producesthe integratedeffect of runoff resultingfrom a representative of equal duration storms.Extremerainfall set intensityis not reflectedin the determination. intensestormsare needed, study of If a recordsshouldbe madeto ascertain their influenceupon the discharge hydrograph by comparingpeaks obtainedutilizing the derived unit hydrographand actual hydrographsfrom intensestorms. Essentialstepsin developing unit hydrographfor an isolatedstorm are: a l. Analyzethe streamflowhydrographto permit separationof surfacerunoff from groundwaterflow, accomplished the methodsdeveloped Secby in tion l 1.4. 2. Measurethe total volume of surfacerunoff (direct runoff ) from the storm producingthe original hydrograph.This is the area under the hydrograph after groundwaterbaseflow has been removed. 3. Divide the ordinatesof the direct runoff hydrographby total direct runoff volume in inches,and plot theseresultsversustime as a unit graph for the basin. \-

192

12 HYDROGRAPHS GHAPTER UNIT 4. Finally, the effective duration of the runoff-producing rain for this unit (time history of rainfall intensity) graphmustbe found from the hyetograph of the storm eventused. Procedures other than thoselisted are requiredfor complex stormsor in developing synthetic unit graphs when data are limited. Unit hydrographscan also be An transposed from one basinto anotherundercertain circumstances. exampleillustratesthe derivationof a unit hydrograph.

EXAMPLE I2.1 Using the total direct runoff hydrographgiven in Fig. I2.2, derive a unit hydrograph for the l7I5 ac drainage area. Solution 1. Separatethe base or groundwaterflow to get the total direct runoff hydrograph.A commonmethodis to draw a straightline AC that beginswhen

2-hr rainfall duration

I
I

*zTotal

precipitaion = 4.2 in.

o Time (hr)

1l

t2

Directrunoff. /T\ ordinate Yl

*zTotal directrunotf of l.4l5in.on1715ac 2-hr unit hydrograph of l.u rn. on I / I) ac

500
o FA

400 300 200 .100 0

Basef-tow / / \,

/^
3 4 5 6

t'-# -j
7 8 Time(hr) Dfuect runoffduration

Baseflow separatlon

Figure 12.2 Illustration of the derivation of a unit hydrograph from an isolatedstorm.

12,2 DERIVATION UNIT HYDROGRAPHS FROM STREAMFLOW OF DATA

193

the hydrograph rise curve startsan appreciable and endswherethe recession intersects baseflow curve.The importantpoint hereis to be consistent the in methodologyfrom storm to storm. 2. The depth of direct runoff over the watershedis calculatedusing > (DR x Ar) _ 2447 cfs-hr : 1 4" - ''* area l7l5 ac

(r2.r)

runoff ordinateduring a chosen whereDR is the average heightof the dfuect time period Ar (in this caseA/ : 1.0 hr) . The valuesof DR determined from Fig. I2.2 are listedin Table 12.1. 3. Computeordinatesof the unit hydrographby using

8 "_ Q , 1 v,

(r2.2)

where Q, : the magnitudeof a hydrographordinate of direct runoff having a volume equal to % (in.) at someinstant of time after start of runoff Q, : the ordinateof the unit hydrographhavinga volume of 1 in. at someinstant of time In this examplethe valuesare obtainedby dividing the direct runoff ordinates 1.415.Table12.1outlines computation the unit-hydrograph of by the ordinates. 4. Determinethe duration of effectiverainfall (rainfall that actuallyproduces surfacerunoff). As statedpreviously,the unit hydrographstorm duration

TABLE 12,1 DETERMINATION2.HR HYDROGRAPH OFA UNIT

FROM AN ISOLATEDSTORM (1) Time (hr) (2) Runoff (cfs) (3) Base flow (cfs) (4) Direct runoff, (2)-(3) (cfs) (5) 2-hr unit hydrograph (4) ordinate, + 1.415 (cfs) 0 8.5 84.8
i-l I

I 2 3 4 4.7 5 6 7 8 9 10 10.5 lt t2

110 t22 230 578 666 645 434 293 202 160 1t7 105 90 80

t10 t10 t10 110 110 110 110 110 110 110 1r0 105 90 80

0 t2 120 468 556 535 324 183 92 50 7 0 0 0

393 379 229 129 65.0 35.3 4.9 0 0 0

194

CHAPTER12

UNIT HYDROGRAPHS

should not exceedabout 25 percent of the drainage atea lag time' but violatesthis rule for the example.From Fig. 12.2, the rain duration is 2 hr. 5. Using the values from Table I2.1, plot the unit hydrograph shown in Fig. 12.2. r I

METHODS BY APPLICATIONS LAGGING 12.3 UNITHYDROGRAPH

has Once an X-hr unit hydrograph beenderivedfrom streamflowdata (or synthesized parameters,Section 12.6) it can be used to estimatethe direct runoff from basin hydrographshapeand duration for virtually any rain event.Applications of the X-hr usedfor stormshavingdurations UH to other stormsbeginswithlagging procedures, duration. Applications to stormswith fracthat are integermultiples of the derived in are tional multiples of X, known as S-hydrographandIUH procedures, discussed Sections12.4 and 12.5. are Because unit hydrographs applicableto effective (net) rain, the processof lossesfrom applyrngUH theory to a storm beginsby first abstractingthe watershed in an effective rain hyetograph'Any of the the precipitation hyetograph,resulting proceduresdetailedin Chapter 4 can be applied. The remainder of this discussion lossesfrom the storm. watershed that the analysthas already abstracted assumes is an integermultiple of X, the storm is'treated If the duration of anotherstorm from eachX increment as a seriesof end-to-endX-hourstorms.First, the hydrographs The ordinatesare then added X-hourunit hydrograph. from the ofrain are determined times to determinethe total hydrograph. at corresponding EXAMPLE 12.2 rates for the 2-hr unit hydrographshownin Fig' I2.3 are'. Discharge

Time (hr) O (cfs)

0 0

1 100

2 250

3 200

4 100

5 50

6 0

Develophourly ordinatesof the total hydrographresultingfrom a 4-hr designstorm amounts: havingthe following excess

Hour (in.) Excess

1 Q.5

0.5

1.0

4 1.0

Solution. The 4-hr duration of the designstorm is an integermultiple of the can duration.Thus,the total hydrograph be foundby addingthe unit hydrograph rain, asshowninFig. l2'3c. of of contributions two 2-hr increments end-to-end a has The first 2-hr stormsegment 1.0 in. of net rain and thus reproduces unit 2.0in. of netrain (in 2 hr);thus 2-hr stormsegmenthas The hydrograph. second its ordinatesare twice those of a 2-hr unit hydrograph.The total hydrograph,

12.3 F

BY APPLICATIONS LAGGINGMETHODS UNIT HYDROGRAPH

195

1.0

r>
h

u.) 0
2 3 4 . 5 6 (a) 2-hourunit storn excess

300 200

i90 zoo
1oo '6 0

1.0
ir

0.5 0 2 3 4 5 6 (c)Design stom excess 600 500

'6 400
:i

H 300 i5 200 100 0 r 2 3 4 5 6 7 (d) Contribution each of 2-hourstorm 600 500

a

t +oo ; ff:oo '$ zoo

-100 0
0 1 2 3 4 5 6 1 8 (e) Total design hydrograph

Figure 12.3 Example 12.2 deivation of total runoff hydrographusing a 2-hr unit hydrograph.

196

cHAPTER i2

uNtr HyDRocRApHS

Fig.12.3e, is found by summingthe fwo contributionsat corresponding times. Note in Fig, 12.3d that runoff from the secondstorm beginswhen the second rain begins,not at the beginningof the first storm. r I This methodof "lagging" is basedon the assumption that linear response the of watershed not influencedby previousstorms-that is, one can superimpose is hydrographs offset in time and the flows will be directly additive. The simplestway to developcompositedirect runoff hydrographs multiple-hourstormsis in a spreadfor sheet.Care must be taken, however,in visually confirming, as in ExampleI2.2, that the start and end points of runoff from eachcontributingX-hr incrementof rain are properly selected. commonerror is to lag eachadditional contributinghydrograph A by Ar, the time interval betweenreadings, rather than X, the associated duration with the given unit hydrograph. Also, the multiplier for the UH ordinatesmust be the net rain occurring in X hours, not the rain occurring in the time increment A/. Example I2.3 illustrates thesepoints. EXAMPLE 12.3 Using the derived 2-hr lunit hydrograph Table 12.1, determinethe direct runoff in hydrographfor a 4-hr. storm havingthe following excess rain amounts:

Hour Excessrain. in.

2 0.7 t.2 1.2

o.7

Solution 1. Tabulatethe unit hydrograph intervalsof the selected at time interval, A/, as shown inTable 12.2.
TABLE 12.2 UNIT HYDROGRAPH APPLICATION EXAMPLE12.3 OF Effective rainfall (in.)

Time (h0 0
I

Unit hydrograph (cfs) 0 8.5 84.8

JJI

Contrib. of first 2-hrrain U HX 1 . 4 ' 0 I1.9 l19 463 531 321 181 91 49.4 6.9 0

Contrib. of second 2-hr rain

uH x 2.4'

Total outflow hydrograph (cfs)

2
J

4 5 6 7 8 9 10 1l t2

0.7 0.7 1 . 2" 1.2

379 229 129 65 35.3 4.9 0

0 20.4 203 794 910 550 310 156 84.7 11.8 0

0 11.9 119 483 734 11 1 5 1091 641 359 163 84.7 I 1.8 0

12.gUN|THYDRoGRAPHAPPL|CAT|oNSBYLAGG|NGMETHoDS197

X 2. Determine the correct UH multiplier for eachX-hr interval. Because is storm producea total net 2 hrs for this example,the first two hours of the rain of 1.4 inches. Similar$, the last two hours of the storm produce 2.4 inchesof net rain. 3. Determinethe correct start and end times for eachof the two hydrographs and tabulatethe contributionof the l.4-inch and 2.8-inch rains at the X-hr storm startedat / : 3 hrs, the appropriatelag times. Because second : 3 hrs as shown inTable I2'2. runoff for this-stormcannotbegin until / the 4. Add the contributionsat eachtime to determine total runoff hydrographs for the 4-hr storm. and 5. Checkthe tabular solutionby plotting eachof the two hydrographs sum rr at the ordinates each/, as showninFig.I2'4' In addition to using a given X-hr UH for determiningthe runoff hydrographfor a given storm, tagging of ttt" X-hr UH can be used to developother duration unit is hyirographs.The pioCedure the sameas applyingthe X-hr UH to 1.0 in. of net rain in f nouri. As earlier, Y must be an integermultiple of X. For example,if a 1-hr unit resultingfrom a 2-hr a for is hydrograph available a given watershed, unit hydrograph siorm-is tbtained by plotting two L-hr unit hydrographs,with the secondunit hyin drographlagged t hr, adding ordinates,and dividing by 2' This is demonstrated unit hydrograph'Thus nigl ti.S, riliere the dashedline rept"sentsthe resulting2-ht the t in. of rainfall containedin the original 1-hr duration has been distributedover a Z-hr period.

Time units

hydrographfor'Example 12'3 Figure 12.4 Synthesized derivedby the unit hydrographmethod'

\_

198

CHAPTER UNITHYDRoGRAPHS 12

l-hr unit hydrogaph 2-hIunithy&ogaph

Time

Figure12.5 Unithydrographlaggingprocedurero develop another hydrograph. unit Modificationsof the original unit-hydrograph duration can be madeso that two 1-hr unit hydrographs usedto form a 2-hr unit hydrograph;two 2-hr unit hydroare graphsresult in a 4-fu diagram,and so on. Care must be taken not to mix durations in the lagging procedure, since errors are introduced; a l-hr and a}-hr unit hydrograph do not representa 3-hr unit hydrograph.Lagging procedureis therefoie restrictedto multiples of the original duration accordingto the expression D1 : nD (I2.3\ where Dl : the possibledurationsof the unit hydrographby lagging methods D : the original duration of any given unit hydrograph fl: I,2,3,..

12.4 S.HYDROGRAPH METHOD

The S-hydrograph methodovercomes restrictionsimposedby the laggingmethodand allowsconstructionof any durationunit hydrograph. observingthe lagging system By just described, is apparentthat for a l-hr unit hydrograph,the l-in. rainfall excess it hasan intensityof 1 in./hr, whereas 2-hr unit hydrograph producedby a rainfall the is intensity of 0.5 in./hr. Continuouslagging of either one of theseunit hydrographs is comparable a continuouslyappliedrainfall at either 0.5 in./hr or 1 in./hr intensity, to dependingon wfuch unit hydrographis chosen. As an example, usingthe 1-hr unit hydrograph, continouslaggingrepresents the direct runoff from a constantrainfall of 1 in./hr as shownin Fig. r2.6a. The cumulative addition of the initial unit hydrograph ordinatesat time intervalsequalto the unit storm duration resultsin an S-hydrograph (seeFig. r2.7). &aphically, construction of an s-hydrographis readily accomplished with a pair of dividers. The maximum

12.4 S-HYDROGRAPHMETHOD

199

*l

Dhr

D-hr S-hydrograph laggedt hr

Time
(a)

*l

l-

Figure 12.6 S-hydrograph method.

discharge the S-hydrograph of occursat a time equalto D hourslessthan the time base of the initial unit hydrographas shown inFig. 12.6a. To constructa pictorial 2-hr unit hydrograph,simply lag the first S-hydrograph by a secondS-hydrograph time interval equalto the desiredduration.The difference a in S-hydrograph ordinatesmust then be dividedby 2. Any duration r unit hydrograph may be obtained in the samemanner once another duration D unit hydrqgraphis known. Simply form a D-hr S-hydrograph; this S-hydrograph hr, andmultiply the lag f difference in S-hydrographordinatesby D/t. Accuracy of the graphical procbdure dependson the scaleschosento plot the hydrographs.Tabular solution of the Shydrograph methodis also employed, hydrograph but tabulationsmustbe at intervals of the original unit.hydrographduration. EXAMPLE 12.4 Given the following 2-hr unit hydrograph,use S-hydrograph procedures construct to a 3-hr unit hydrograph.
Time (hr) 0 (cfs)

o 0

| 100

2 250

3 200

4 100

5 50

6 0

Solution. The 2-hr unit hydrograph the runoff from a 2-hr stormof 0.5 is in./hr. The S-hydrograph formedfrom a net rain rate of 0.5 in./hr lasting is indefinitely shown Fig. 12.6a. ordinates foundby adding 2-hr as in the Its are (UH) runoff ratesfrom eachcontributing unit-hydrograph 2-hr block of rain:

200

CHAPTER12

UNIT HYDROGRAPHS

/,/

90

-'{.3
u0 Figure 12.7 S-hydrograph. Time (min.)

Time (h0 0 I
z

1st2-hr 0 100 250 200 100

5n

2nd Z-hr

3rd2-hr

S-hydrograph 0 100 250 300 350 350 350 350 350

3
4

5 6 7 8

0 100 250 200 100 50 0

' 0 100 250 200 100

0 100 250

as To find a 3.-hrhydrograph, S-curveis laggedby 3 hr and subtracted shown the in Fig. 12.6b.This results in a hydrographfrom a 3-hr storm of 0.5 in./hr, or 1.5 in. total. Thus the ordinatesneed to be divided by 1.5 to producethe 3-hr unit hydrograph:

12.5 THE INSTANTANEOUS UNIT HYDROGRAPH

201

Time (hr)
0 1 2 3 4 5 6 7

S-hydrograph

Lagged S-hydrograph

Difference

3-hrUnit hydrograph 0 67 167 200 167 . 6 7 33 0

0 100 250 300 350 350 350 350

0 100 250 300 350

0 100 250 300 250 r00 '50 0

12.5 THE INSTANTANEOUS HYDROGRAPH UNIT

The unit-hydrograph methodof estimatinga runoff hydrograph be usedfor storms can of extremelyshortduration.For example, the durationof a stormis 1 min and a unit if volume of surfacerunoff occurs, the resulting hydrograph is the 1-min unit hydrograph.The hydrographofrunofffor any 1-min storm ofconstant intensitycan be computedfrom the l-min unit hydrographby multiplying the ordinatesof the 1-min unit hydrographby the appropriaterain depth.A storm lasting for many minutescan be described a sequence 1-min storrns(Fig. 12.8).The runoffhydrographfrom as of each l-min storm in this sequence be obtainedas in the precedingexample.By can superimposing the runoff hydrograph from each of the l-r.nin storms, the runoff hydrographfor the completestorm can be obtained. From the unit hydrographfor any duration ofuniform rain, the unit hydrograph for any other durationcanbe obtained.As the durationbecomes shorter,the resulting unit hydrograph approaches instantaneous hydrograph. The instant4neous unit an unit hydrograph(IUH) is the hydrographof runoff thaf would result if 1 in. of waterwere spreaduniformly over an areain an instant and then allowed to run off.e To develop an IUH, any I in.lhr S-hydrographmust first be obtained. The resulting S-curve is laggedby the interval Ar to developa Ar-hour unit hydrograph. The resulting At-hour unit graph becomesan IUH when Ar is set to 0.0 in the limit. If a continuing 1 in./hr excessstorm produces the original and lagged Sof hydrographs Fig. 12.6b,the Ar-hour unit hydrographis the differencebetweenthe two curves,divided by the amount of excess rain depth in A/ hours,or
Qo- Q" Q,(Lt-hr UH) : ILt

(r2.4)

The Qo - Q" dtfferences dividedby I Lt to convertfrom a stormwith 1Al inches are in Al hours to one with 1.0 in. in At hours,which is the definition of a Ar-hour unit graph. As Ar approaches zpro, Eq. 12.4 becomes

: a 0,(ruH) ! d ts I

(12.s)

202

CHAPTER12 UNIT HYDROGRAPHS

Figure 12.8 Unit-hy&ographdescriptionof the runoff process. Unit hydrograph; a sequence l-min (a) (b) of (c) storms; superposition runoffhydrographs each for of of (After Schaake.e) the l-min storms. which showsthat the flow at time I is proportional to the slopeof the S-hydrograph at time r. In applications, slopeis approximatedbyLQ/A,I, and the IUH ordinates the can be estimatedfrom pairs of closelyspacedpoints of the S-hydrograph. Ifan IUH is supplied,the aboveprocesscan be reversed,and any X-hour unit graph can be found by averaging IUH florvsat X-hr intervals, or

UH) : 1(IUH,+ IUH,_X) Q,(X-tu

{r2.6)

Use of this approximate equation is allowed for small X values and permits direct calculationof a unit graph from an IUH, bypassing normal S-hydrograph the procedure.

UNIT 12.5 THEINSTANTANEOUS HYDROGRAPH

203

EXAMPLE 12.5 determinethe IUH, and then use it to Given the following 1.0 in./hr S-hydrograph, a estimate 1-hr UH.
3.5 750 4.0 800

Time (hr) S-curve (cfs)

0 0

0.5 5 0

1.0 200

1.5 450

2,0 500

2.5 650

3.0 700

Solution. The IUH is foundfrom Eq. l2.5.The slopeat time r is approximated bY (Q,*o., Q,-o)lLt

IUH = AQlAf

0 0.5 1 1.5 2 2.5 3 3.5

+

0
JU

200 450 500 650 700 750 800 800 800

0 200 400 300 200 200 100 100 50 0 0

at from F,q.12.6,usingreadings 1-hr intervals: The 1-hr uH is obtained

1-hr UH 0 200 300 150 75 25 n "

luHf

luHr_1

0 1 2 3
A

5 6

0 400 200 100 50 0 0

0 0 400 200 100 50 0

The readershouldverify that the 1-hr UH obtainedthrough use of the IUH is t t[e approximately sameas that obtainedby lagging the S-hydrograph hr, subtractthe differenceto a 1-hr UH. ing, and converting rainfall The ordinatesof the IUH representthe relative effect of antecedent the runoff rate at any instant of time. By plotting the IUH with time intensitieson increasingto the left rather than to the right (seeFig. I2.9), andthen superimposing this plot over the rainfall hyetograph(plotted with time increasingto the right as in pig.iZ.g),the'relative weight given to antecedent rainfall intensities(asa function of the past) is easily observed.In other words, the runoff rate at any time is time into

204

12 CHAPTER UNITHYDROGRAPHS

Time-reversed image of the instantaneous unit hydrograph

Time into the past Antecedent iainfall intensities

oo=[;f@xi(t-r)dr

unit Figure 12.9 Calculationof runoff rates with the instantaneous hyof average the antecedent is a weighted The runoff rateat arrytime drograph. unit imageof the instantaneous hyThe time-reversed rainfall intensities. (After Schaake'v) furtction. the represents weighting drograph of computedas a weighted average the previousrainfall intensities.Therefore,the of.the rainfall pattern computedrunoff hydrographis the weighted,moving average image of the unit hydrograph'e and itre weighting irtt"iiott is the time-reversed the Statedmathematically, runoff rate at any time is given by the convolution integral

Q(A:

- r)dr ,<,tu, l,'

(r2.7)
ll:

12.6 SYNTHETICUNIT HYDROGRAPHS

205

where

Q(t) : the surfacerunoff rate at time t f(r) : the ordinate of the IUH at time r i(t - r): the rainfall inten'sity (after abstraction of the appropriate infiltration losses) time t - r at

The variable 7 represents time into the past so that time r - r occursbefore time r. The limits on the integral allow r to vary betweena pastand presenttime (i.e., r : 0, t - r : 0)' The integral givesa continuous weightingof prbviousrainfall intensities by the ordinatesof the IUH.

12.6 SYNTHETIC UNITHYDROGRAPHS

As previously noted, the linear characteristics exhibited by unit hydrographs a rf6i watershedare a distinct advantagein constructingmore complex rto.- Oir.fruig. hydrographs. Generally,however,basicstreamflowand rainfall data arcnot available to allow construction of a unit hydrographexcept for relatively few watersheds; therefore, techniqueshave evolved that allow generationof syithetic unit lrydrographs.

GammaDistribution
The shapesof hydrographsoften closely match a two-parametergamma function, given by

f(x) :

xoe-*/B

B"+lf(d + 1)

(12.8)

where0 ( r ( m. The parametera is a dimensionless shapefactor (mustbe greater tltul - 1), and B is a positive scalefactor havingthe sameunits as x and contiolling the baselength. The product of a and B givesthe value-r corresponding the apexl to or maximum value ofl(x). For a ) 1, the distribution has a single upe* und'plot, similar to hydrograph shapes,as shown in Fig. 12.10. The dislribution mein is is F@ + 1), and variance p2(a + I). Many of the syntheticunit hydrographproceduresresult in only three to five . points on the hydrograph,through which a smooth curve must be fitted. In addition to the requirementthat the curve passes through all the points, the area under the hydrographmust equal the runoff volume from one unit of rainfall excess over the watershed. This latter requirementis often left unchecked and can result in considerablg errors in performing calculationsthrough the use of ordinatesof a hydrograph that do not representa "unit" of runoff. The mostusefulfeatureof the gammadistributionfunction (explained greater in detail later) is thit it guarantees unit areaunder the curve. It can convenientlybe a used to synthesize entire hydrographif the calculatedpeak flow rate an eo and its associated time to are known. This usesa proceduredeveloped Aron und'White.ro by If time r is substituted x in Eq. 12.g, the time to peak tois aB. At this point, for the function/(r) equalsthe peak flow rate e,, or

o,:ffi=ffrr*1

(r2.e)

206

CHAPTER12

UNIT HYDROGRAPHS

1.0

B--1

0
x

Figure 12.10 Gamma function shapesfor various shape and scale parameter values. factor A. whereC, A is the unit volumeof runoff from a basinwith arca The conversion = 1.008 is selected make @(a)dimensionless. to C, The function f(a) is shownby Aron and white to be relatedto a by1l (12.10) + a : 0.045+ 0.5d + 5.6Q2 0303 well in the range 1 reasonably Collins showsthat this can be approximated
byt'

( a ( 8

q.:05Q+5.902 Combiningthis with F,q. 12.9 *tu"t , o : o s tf f i . t n ( H ) '

(r2.Lr) (r2.r2)

To fit a unit graph usingEqs. I2.9 and 12.12,the peak flow rate and time must allow this. Next. @(a)is subsequ-ntly Severalof the methodsdescribed be estimated. can foundfrom Eq.12.9, and a from Eq.12.10 ot l2.Il. The unit hydrograph now to. Substitutingalo by be constructed calculatingQ at any convenientmultiple, a, of for x in Eq. 12.8 gives the flow at t : atp as
Qoto: QraoeQ-o)"

(r2.r3)

which can be solvedfor all the flow rates of the hydrograph. EXAMPLE 12.6 is watershed 1720cfs and of The peak flow rate for the unit hydrograph a 36,000-acre the the initiation of runoff. Use Eq. 12.8 to synthesize rest of o".oi, 12 hr following the hydrograph.

12.6 SYNTHETICUNIT HYDROGRAPHS

207

Solution. From Fq. 12.9,

r720(12) : 6@) : 1.008(36,000)0.57

From Eq. 12.10(and t2.It), q.:2.2 The hydrographis then found from Eq. L2.I3: a) e2'2(rQ*, : 1720a2'2 Solving for a few points, we obtain the followine values:
10.0 0 0.5 1.0 2.0 5.0 10.0
t = afo(hr)

O(ctu)
0 tt25 1720 876 9 0

0 6 l2
)4

60 120

sufficient intermediatepoints shouldbe generated define the entire shapeof to the hydrograph. tl

Snyder's Method
one- techniqueemployedby the corps of Engineersl3 many othersis and basedon methods developedby Snyderaand expandedby Taylor and 3chwartz.sIt allows computation of lag time, time base,unit_hydrograph duration, peak discharge,, and hydrograph time widths at 50 and75 perceniof peukflor. ny uring ttteseseven points, a sketchof the unit hydrograph obtained,rig.lz.rt,and cleckel to seeif it contains is 1 in. of direct runoff.

Alternate recessions I to produce 1.0 in. ofrunoff

Time,r F'igure 12.11 Snyder'ssyntheticunit hydrograph.

208

CHAPTER12

UNIT HYDROGRAPHS

that a Time to Peak Snyder'smethodof synthesizing unit hydrographassumes the from Eq. 11.5.Its locationis lag, estimated peak flow rate occursat the watershed rate are both as established shownon Fig. 12.11.The lag time and peak discharge For characteristics. the lag time, the watershed correlatedwith variousphysiographic and C, is variablesL and L"o for Eq. 11.5 are estimatedfrom map measurements, TableL2.3summaor for developed the locale,usingSnyder'sestimates other sources. rizes a variety of C, valuesfor variousregions. that lag time is a constantfor a particular watershed-that'is, It is assumed uninfluencedby variations in rainfall intensitiesor similar factors. The use of L"o shape,andC, takescare of wide variationsin topography, for accounts the watershed from plains to mountainousregions. lower valuesof C,, with extremesof 0.4 nqtec Steeperslopestend to generate in SouthernCalifornia and 8.0 alongthe Gulf of Mexico and Rocky Mountains.W{ren valuesof C' will be betweenone influencepeak discharge, snowpackaccumulations sixth to one third of Snyder'svalues. Time Base The time base of a syntheticunit hydrograph(seeFig. 12.11) by Snyder'smethod is
t , : ?
T ;

, t l 6

(r2.r4)

where t6 : the basetime of the syntheticunit hydrograph(days) t1: the lag time (hr)

FOR U.S. LOCALITIES TABLE 12.3 TYPICALSNYDER'SCOEFF|CIENTS Location

Appalachian Highlands West Iowa Southern California Ohio Eastern Gulf of Mexico Central Texas North and Mid-Atlantic states Sewered urban areas Mountainous watersheds Foothills areas Valley areas Easlern Nebraska Corps of Engineers training course Great Plains Rocky Mountains SW desert NW coast and Cascades 21 urban basins Storm sewered areas "Channel slope S

Range of C, t.8-2.2 0.2-0.6 0.6-0.8 0.4-2.3 0.2-0.5

Average Cr 2.0 o.4 0.4 0.7 8.0 1.1

Range of Co 0.4-0.8 0.7-1.0 0.6-0.7 0.3-1.2 0.1-0.6

Average Co

0.6 0.8 0.9 0.6 0.6 0.8 0.3

0.6/\4" 0.3
1.2 o;7 0.4 0.8 0.3-0.9 1.3
1,4

0.4-1.0 0.4-8.0 0.8-2.0 1.5-8.8 0;t -1.9 2.0-4.4 0.3-0.9 0.2-0.3

0.5-1.0

0.8

3.1 0.6 0.2

12.6 SYNTHETIC HYDROGRAPHS UNIT 209 Equation L2.14 gives reasonableeslimatesfor large watersheds but will produce excessively large valuesfor smaller areas.A generalrule of thumb for small areasis to use three to five times the time to peak as a base value when sketchinga unit hydrograph.In any event, the time base shouldbe adjustedas shown in rig. tz.lt until the areaunder the unit hydrographis 1.0". Duration The duration of rainfall excessfor Snyder's synthetic unit-hydrograph development a function of lag time is
l -

'

tt

5.5

(r2.1s)

where /, : duration of the unit rainfall excess (hr) t1 : the lag time from the certtroidof unit rainfall excess the peak of thb to unit hydrograph This synthetictechnique alwaysresultsin an initial unit-hydrograph duration equalto in fi/5.5.However, sincechanges lag time occur with changes duration of tlie unit in hydrograph,the following equation was developed allow lag time and peak disto chargeadjustments other unit-hydrograph for durations. tm:h+0.25(t*-t) where t1p: tt : to: t, the adjustedlag time (hr) the original lag time (hr) the desiredunit-hydrograph duration (hr) the original unirhydrograph duration : t,/5.5 (hr)

(r2.16)

Peak Discharge If one assumes that a given duration rainfall produces1 in. of direct runoff, the outflow volume is some relatively constantpercentage inflow of volume.A simplified approximationof outflow volume is r, X er, andthe equation for peak discharge can be written
Vr: 640CPA -Ltn

(12.r7)

where Qp : peak discharge(cfs) C" : the coefficientaccountingfor flood waveand storageconditions;it is a function of lag time, durationof runoff ptoducingrain, effectivearea contributing to peak flow, and drainagearea A : watershedsize (mi2) tp : the lag time (hr) Thuspeakdischargetanbe calculated givenlag time andcoefficientofpeak discharge C.. Valuesfor Cp range from 0.4 to 0.8 and gengrally indicate retention or storage capacityof the watershed. Larger valuesof C" aregenerallyassociated with smaller valuesof C,. Typical valuesare tabulated inTable 12.3. Hydrograph Construction From Eqs: 1I.5, lZ,I4, I2.I5, and,12.1,7 plot three points for the unit hydrographand sketcha syntheticunit hydrograph,remembering

210

CHAPTER12

UNIT HYDROGRAPI.IS

that total direct runoff amountsto 1 in. An analysisby the Corps of Engineers(see in Fig. 12.12) gives additional assistance plotting time widths for points on the hyAs drographlocatedat 50 and 75 percentofpeak discharge.l3 a generalrule ofthumb, proportionedeachsideof the pegk ttr" ti-l width at l/so and IV^ ordinatesshouldbe in a ratio of I:2 with the short time sideon the left of the syntheticunit-hydrograph Eq. peak. As noted earlier,for smallerwatersheds, I2.I4 givesunrealisticvaluesfor by ihe bur" time. If this occurs,a value can be estimated multiplying total time to the modified basedon the amount and peak by a value of from 3 to 5. This ratio can be boundaries. ii." tut" of depletionof storagewater within the watershed by The envelopecurvesin Fig. I2.I2 ate defined Wso: 8301(Qo/A)" (Qo/A)" W15: 470f

(12.18)

(t2"]a)

The sevenpoints formed through the use of theseequationscan be plotted and a smooth curve drawn. To assurea unit hydrograph,the curve shape and ordinates shouldbe adjusteduntil the areabeneaththe curve is equivalentto one unit of direct runoff depth over the watershedarea.This can be doneby hand-fittingand planimetering or by curve-fitting. to techniques fit a Pearson regression Hudlow and Clarkla used least-squares III (gamma)probability densityfunction (referto Chapter26) throughthe seven type points. This function has an areaof 1.0 and a shapesimilar Snyaeruoit-nyAtograph
1000 800 400
N

,2

9 200
oo

1oo -? 8 0 E 6 0 9" 40

) \

4-l

'.,

s
*

t *tS
08,

F\

\, \ \ $ \
40 60 80100

10 8 6
0.2

I 0.4 0.60.8

810

Width of unit hYdrograPh (hr)

Figure L2.12 Unit hydrographwidth at 50 and 75 percent of peak flow' o, a, observedvalue of Wso. observedvalue of lfi5.

c uNrr HYDR.GRAPHS1" 21 to thator natural hydrographs. ,h;:l rh"

Q,:

,lll::i
e-G-tp')/b

nr(;

(r2.20)

where a and b are shapeand scaleparameters.Hudlow and Clark presenta trial-anderror solution to the least-squares normal equations,using Newton's method, to developestirnatesof a and b. The application of Snyder's syntheticunit-hydrographmethod to areasother than the original study area should be precededby a reevaluationof coefficientsC, and C, in Eqs; 11.5 and I2.I7 . TIns analysiscan be accomplished the use of unit by hydrographs the region under study which havethe proper lag time-rainfall durain tion ratio; that is, t, : ttf 5.5. If anotherrainfall durationis selected, variations C of and C, can be expected.

SCS Method
A methoddeveloped the Soil Conservation by Servicefor constructingsyntheticunit hydrographs basedon a dimensionless is hydrograph(Fig. 12.13). This dimensionless graph is the result of an analysis of a large number of natural unit hydrographs from a wide range in size and geographii locations.The method requires only the

0.6

cSlo
b 0.5 0.4 0.3 0.2 0.1

als

t tp

Figure 12.13 Dimensionless unit hydrographand masscurve. (After Mockus.1s)

212

CHAPTER UNIT 12 HYDRoGRAPHS determination of the time to peak and the peak discharge as follovi,s:

tp:

D t+

tt

(r2.2r)

(hr) where to : the time from the beginningof rainfall to peak discharge D : the duration ofrainfall (hr) (hr) t1 : the lag time from the centroid of rainfall to peak discharge The ratios corresponding Fig. 12.13 are listed in Table 12.4. The peak flow to for the hydrographis developed approximatingthe unit hydrographas a triangular by shapewith basetime of ! t, and unit area.The readershouldverify that this produces
U r :

4844
t

(r2.22)

(cfs) where Qp: peak discharge A : drainagearea (mi2) tp : the time to peak (hr) The time base of ! ro is based on empirical values for averagerural experimental peak flow) for steepconditions watersheds shouldbe reduced(causing and increased or increased (causing decreasedpeak flow) for flat conditions. The resulting coefficientinBq.12.22 rangesfrom nearly 600 for steepmountainousconditionsto 300 for flat swampyconditions. A relation of /, to size of watershedcan be used to estimatelag time. Typical relations from two geographic regionsare
tt : t,: TABLE 12.4 COORDINATES SCS OF DIMENSIONLESS UNIT HYDROGRAPH OF FIGURE 2.13 1
Q/Q, Q/Qp

l.44A0'6 Texas 0.54A0'6 Ohio

(r2.23a) (r2.23b)

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 t.2 1.3

0 0.015 0.075 0.16 0.28 0.43 0.60 0.77 0.89 o.9'7 1.00 0.98 0.92 0.84

1.4 1.5 1.6 1.8 2.0 2.2 2.4 2.6 2.8 3.0 3.5 4.0 5.0

0.75 0.66 0.56 0.42 0.32 0.24 0.18 0.13 0.098 0.075 0.036 0.018 0.009 0.004

12.6 SYNTHETIC HYDROGRAPHS UNIT 213. The average is 0.6/",where/" is the time of concentration, lag definedby SCS aseither (calledthe the time for runoff to travel from the furthermostpoint in the watershed rain to the inflection of the unit upland method) or the time from the end of excess hydrograph.For the first case,
t":L1tp-D

(r2.24)

The dimensionless unit hydrograph,Fig. 1.2.1.3,has point of inflection at approxia mately 1..7t,.If the lag time of 0.6t" is assumed, Eqs. 12.2I and 12.24 give D - 0.2t0 D = 0.I33t"

0).zs)
{r2,26)

A small variation in D is permissible, it shouldnot exceed0.25t, or 0.I7t". pnce but for the 0.133r"-hourunit hydrographis developed, unit hydrographs other durdtions can be developed using S-hydrograph IUH procedures. or By finding a value of t,, a syntheticunit hydrographof chosenduration D is obtained from Fig. 12.13. Atother equationusedby the SCS is . t "'- : where fi : I : Y : S: /0.'(s * 1;o.z

lgooyo's

(r2.27)

the lag time (hr) length to divide in feet average watershedslopein percent the potentialmaximumretention(in.) : (1000/CN) - 10, where.CN is a curve number described Chapter4 in

or The lag from Eq. 12.27is adjustedfor imperviousness improved watercourses, or both, if the watershedis in an urban area.The multiple to be applied to thq lag is M : 1.- P(-6.8 X 10-3 + 3.4 x 10-4CN- 4.3 x 10-?CN, -2.2 x 10-8CN3) (12.28) where CN is the curve number for urbanized conditions, and P can be either the percentage that imperviousor the percentage the main watercourse is hydraulically of improved from natural conditions.If part of the area is imperviousand portions of the channelare improved,two valuesof M are determined,and both are multiplied by the lag. EXAMPLE 12.7

.\ For a drainagearea of 70 r4i2 having a lag time of 8 | hr, derive a unit hydrograph of unit hydrograph. duration 2 hr. Use the SCS dimensionless Solution 1. Using Eq.l2.21we obtain

t,:?*8*=9*t'tr

214

CHAPTER12

UNIT HYDROGRAPHS

2. From Eq.12.22 nP _ 4 8 4 x 7 0 Y
9.5

att Qo:3J60 cfsoccurring :9|hr 3. Using Fig.12.13, we find the following: a: The peak flow occursar tfto: 1 or at t : 9+hr. b. The time baseof the hydrograph: 5toor 47.5 hr. c. The hydrographordinatesare: c l . A t t / t o : 0 . 5 ,Q / Q r : 0 . 4 3 ; t h u sa t t : 4 . 7 5 h r ,Q : 1 5 3 1 f s . 2. At tlto : 2, Q/Q" - 0.32; thus at t : 19 ht, Q : 1139cfs. 3. At tft, : 3, Q/Q, : 0.07; thus at t : 28.5 hr, Q : 249 cfs. 4. Check D/to:0.21; OK. rl

Gray'sMethod
Another method of generatingsynthetic unit hydrographshas been developedby An size for applicationof this method Gray.16 approximateupper limit of watershed to the geographic areasof central Iowa, Missouri, Illinois, and Wisconsinis 94 mi2. The method is basedon dimensionalizingthe incomplete gamma distribution and graph of the form resultsin a dimensionless
O,t" : 25.0(y')n

r(q)

t 1o_7,t/pR\( \s-1

\Pol

(r2.2e)

where

Q,lPo: q and y : f : e: Pp : I :

percentflow in 0.25 PRat any given r/P^ value respectively shapeand scaleparameters, the gammafunction of q, equalto (4 - 1)!* the baseof natural logarithms the period of rise (min) time (min)

The relationfor 7' is definedzs yt : yPpandq : l.+ y'. (Fig. 12.14)allowscomputation This form of the dimensionless hydrograph unit ofthe discharge ordinatesfor the unit hydrographat times equal to I intervals of the period of rise P*, that is, the time from the beginningof rainfall to the time of peak discharge the unit hydrograph. of can be develof characteristics the watershed Correlationswith physiographic opedto get the valuesofboth Poand y'. parameAs 4qexample, storage the factorP^fy' hasbeenlinked with watershed ters Lf\/S", where L is the length of the main channel of the watershedin miles measured from the outlet to the uppermostpart of the watershed(Fig. 12.15); S. is definedas an average slopein percentobtainedby plotting the main channelprofile rrf4isnotaninreserf(q) + Z) : (N - I + z)(N- 2 + z)... (t + Z)/f(r + Z), : f(N
where N equalsthe integer.""-rl q, the function is approximated by

: qe.-al';l. ;uFr(q) & h.

- #fu.

12.6 SYNTHETICUNIT HYDROGRAPHS

215

Muckey Creek near Mapleton, lowa

F< a-

K 1 6
F re o ' E o
6

1.0

l(Atloi -=:, penoo or nse (mrn.)

1.5 time (min.)

2.0
t PR

Figure 12.14 Dimensionlessgraph and fitted two-parameter gamma distribution for Watershed (After Gray.r6) 5.

='(ft)'
957o confidence belts

a-l> ": lH l0
e ttr x t9!
o l H

tlr

for Ppll'

Ttta
b

IL, IA,MO,WI OH NE andW. IA

(After Gray.16)

9.27 0.562 11.4

A A

al

0.531 0.498

(r = 0.92)

tgilrgljggtq. Rudo,
lchannel

-L
VS"

slope(Vo)

r.ir '

Figure 12.15 Relationof storage parameter, factor,Ppfy', and watershed (After LV 5",for watershedsNebraska, in Iowa, Missouri, Illinois,andWisconsin. Gray.r6) and drawing a straight line through the outlet elevationsuch that the positive and negativeareasbetweenthe streamprofile and the straightline are equal.The storage factorPofy' can also be correlatedwith the period of rise P^ as shownin Fig. 12.16. These two cerrelations allow solution of Eq. 12.29 and produce a synthetic unit hydrographof duration P^f4 for an ungauged area.

216

12 CHAPTER UNITHYDROGRAPHS

\to
. F h

ql-

"!'

0.0139 2.6761Po+

.Fl6 o t * A
'l

x t!9

40

80

120

160

200

240

280

320

360

Period of rise, Pp (min.)

factor y' andperiodof risePn.(After P*f of Figure12.16 Relation storage Gray.r6; The solutionproceedsas follows: watershed' 1. DetermineL, 5", and A for the ungauged P*, Determineparameters T', and q. 2. Ppfy'. a. With Lf\/5", useFig. I2.I5 to select With P*/y', use Fig. 12.16 to obtain Pa. Compute y' as the ratio b. Pol@*ly'). Substitute7' obtained in Step 2b into the equation Q : | + y', and c. solvefor 4. graph using Eq. 12'29' Com3. Computethe ordinatesfor the dimensionless pute the percent flow in 0.25PRfor values of tf P*: 0.125, 0.375, incrementof tf P* : 0'250 until the sum 0.625,. . . , and every succeeding of the percent flows approximates100 percent. Also compute the peak tfPR: 1. percentage substituting by 4. Computethe unit hydrograph. factorto convertthe volume of the direct runoff a. Computethe necessary over the graph to 1 in. of precipitation excess under the dimensionless entire watershed. 1. The volume of the unit hydrograph: V

V:1in.xAnizx640Tf nu'
I

n'rnln

x 43'560 ;

.- --^ ft2

graph: Vo of 2. Thevolume the dimensionless

fiun 3. Solve fot 2 q, by equatingV and Vo, sincethey must be equal'

x V o : 2 q , x 0 . 2 5 x P 'R 6 0 -

12.6 SYNTHETIC UNIT HYDROGRAPHS

Rdinfall duration = 6 mln = Pnl 4

217

s 400 300

60 Time(min)

Figure 12.17 Derived hydrographof Example12.8usingGray'smethod. Notethat Gray's method results in a unit hydrograph for a P^/4-hr storm.

b. Convert the dimensionless graph ordinatesto the unit-hydrograph ordinates n _ percntflow in 0.25PR\ u,-T.aqi c. Translate time base of dimensionless graph to absolutetime units by multiplying t/PR x P^ for each computedpoint. Rememberthat runoff doesnot commenceuntil the centroid of rainfall, or at a time P^/8. An example problemdemonstrates solutionof Gray'smethodfor a Missouri the watershed. plot representing derivedhydrograph shownin Fig. l2.r7.Dkect A the is runoff commeirces the centroid of rainfall. Thus it is necessary add D/2 or Po/B at to to column 2,Table r2,.5,to obtain the proper times of the unit trydrograph ordinates, showsin Column 6.
EXAMPLE 12.8

For the given data, use Gray's method to constructa unit hydrographfor the Green Acrewatershed, wheredrainage area: 0.62mi2,length: 0.98mi, andS" : l.45%a. Procedure 1. a. b. c. 2. a. Figure12.15; L/t/S: 0.813 ni: Pnfy' : 8.25min. Figure 12.161, P^/^y' : 8.25 min; P*: 24.9 min. q : I I Y' : 4.02; Y' : 3.02. Tabulate percentflow in 0,25PRfortfP*: 0.250:

218

CHAPTER12

UNIT HYDROGRAPHS

TABLE 12.5 TABULATIONFOR EXAMPLE12.8

(1)
r/P

(2) Time, (min) 0 3.1 9.3 15.6 2r.8 24.9 28.0 34.2 40.4 46.7 52.9 59.3 65.5 71.7 78.0 84.2

(3) Percent flowin 0.25Pn 0 0.45 5.80 t2.70 16.35 16.85 16.25 14.20 11 . 1 0 7.97 5.55 3.56 2.28
l.4l

(4) Cumulated flow 0 0.45 6.2 18.9 35.3 5r . 5 65.7 76.8 84.8 90.4 93.9 96.2 97.6 98.5 99.0

(5) UH (cfs) 0
lt.J

(6) time Actual (min) 0 6.1 12.3 18.6 24.8 27.9 31.0 37.2 43.4 49.7 " 55.9 62.3 68.5 '74.7 81.0 87.2

0.000 0.125 0.375 0.625 0.875 1.000 r.t25 t.375 t.625 1.875 2.125 2.375 2.625 2.875 3.125 3.375

490 631 651 628 548 428 308 214 138 88.0 54.4
JJ.J

0.86 0.50

19.3

3. a. 1. V : I x 0.62 x 640 x 43,5601t2: 14.4X 10sft3' x 2.Vo:0.25 x 24.9 60 X > q,:373.52 q' 3.2 q,: 3860. b. Column 5 is tabulated by multiplying 3860 times values in Column 3 divided by 100. c. Column 2 is obtainedby multiplying24.9 times valuesin Column 1. d. Column 3 comesfrom solutionof Eq. l2'2J' rr

SyntheticUnit Hydrograph Espey10-Minute

by was of A regionalanalysis 19 urbanwatersheds conducted EspeyandAltmanlTand equations.thatprovide sevenpoints of a 10-min hyset of regression resulted in a by drograph.The entire hydrographis developed fitting a smooth curve through the eye-fitting or curve-fitting procedures.In either case, a unit area is points using necessary. (cfs),time base(minThe equationsfor time to peak (minutes),peak discharge and 75 percentof the peak flow rate are utes),and width at 50
* fo'1801 57 S-o'2s To : 3.ILo'23 Qp : 3L62 X I03Ao'e6T-t'o7 Ta: 125.89 x I}3AQ;o'es W5o: 16.22x I03Ao'e3Q-oe2 7eQ-o78 w 1 5: 3 . 2 4 x 1 0 3 4 0

(r2.30) (r2.3r) (r2.32) (12.33) (r2.34)

where

L : total distance (ft) along the main channel from the point being boundary to considered the upstreamwatershed

12.6 SYNTHET|C HYDROGRAPHS UNIT 219 : main channel slope (ftlft) defined by H/0.8L, where 11 is the S differencein elevationbetweenthe point on the channelbottom at a distanceof 0.2L downstream from the upstreamwatershed boundary and a point on the channelbottom at the downstreampoint being considered 1 : imperviousarea within the watershed(7o) watershed conveyance factor Q : a dimensionless A : watersheddrainagearea (mi2) T, : time of rise of the unit hydrograph (min) Q, = peak flow of the unit hydrograph (cfs) I, :,time baseof the unit hydrograph(min) ITso: width of the hydrograph at 507o of Q, @in) = I4zrr width of the unit hydrographat75%oof p. (min) The coefficients of determination (explained in Chapter 26) for the five equationg ranged from 80 to 94 percent. The watershedconveyancefactor is found from Fig. 12.18.The Wso andWrt widths are normally drawnwith one-thirdof the calculated width placedto the left of the peak and two-thirds to the right.

Clark's IUH Time-AreaMethod

A syntheticunit hydrograph that utilizes an instantaneous hydrograph(IUH) was unit developed 1945 by clark.6It has been widely used,is often called the time-area in method,and hasappeared severalcomputerprogramsfor hydrographanalysis(see in Chapters24 and25).

80

b 7 0 . 6 0

F s o
;o 4 0
?
d

6 3 0

0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.10 0.11 0.12 0..130.14 0.15 0.16 0.17
Main channel Manning z value

Figure 12.18 Watershedconveyancefactor Q as a function of percent watershed impervious cover l and weighted main channel Manning n value, for Espey method.

220

CHAPTER12

UNIT HYDROGRAPHS

at recognizes that the discharge any point in time is a function of The technique of the translation and storagecharacteristics the watershed.The translation is obtained by estimatingthe overland and channeltravel time of runoff, which is then effectsof a watershed. by combinedwith an estimateof the delaycaused the storage mouth rainfall from its point of falling to the watershed The translationof excess This is a histogramof is accomplished using the time-area curve for the watershed. lines . as incrementalrunoff versustime, constructed shownin Fig. 12.19 The dashed in Fig. I2.I9a subdividethe basin into severalareas.Each line identifiesthe locus of "times" are pointshavingequaltraveltimesto the outlpt. The isochrones drawnequal to apart, and sufficient zonesare selected fully definethe time-area relation. The time-area graph of Fig. l2.I9b is a form of unit hydrograph.The area to beneaththe curve integrates 1.0 unit of rain depth over the total areaA, and it has \ are a translationhydrographshapeif sufflcient subareas delineated. at If one unit of net rain is placedon the watershed t : 0, the runoff from At rate of At units of runoff would passthe outlet during the first At period at an average would be At units of areatimes one unit of pei unit of time. The volume discharged rain. After all areascontribute,one unit of rainfall over the entire area would have passed outlet. the

Lt

I
I

J
^1

.l
A2

J
^3

Lt
A

As
Time interval

ft) of Figure L2.19 Development time-area histogramfor use with Clark's method: (a) isochronesspacedLt apart (shown as dashedlines) and (b) time-area histogram.

12.6 SYNTHETICUNIT HYDROGRAPHS

221

The impact of watershed storageon the translationhydrographis incorporated by routing the time-area histogramthrough a hypotheticallinear reservoirlocatedat the watershedoutlet, having a retardance coefficient K equivalentto that of the watershed. the simplest For form of reservoir,the storage at time f is linearly related S, to the outflow Q, at time /, or S,': KQ,

(r2.3s)

whereKis a constantof proportionality calledthe storage coefficient.It hastime units and is often approximated the lag time of the watershed. by From continuity,the inflow, storage, and.oq,{flow the reservoirare relatedby for

I,-Q,:#:U#
Q'

(r2.36)

If the differentialis discretized LQ/LI, andif Q, andQ, are the flowsat t andt - l, to thenEq. 12.36becomes Io,- Ao,: YQ'Lt

(12.37)

BecauseQ : (Q, + Qr)/2, the flow at the end of any Al is Qz: CoI * CtQ,

(12.38) (r2.39) (r2.40)

where
and

cr: #+T;
- Lt 2K " - ,t -- z K + L t

The IUH is found from Eq. 12.38by solvingfor Q, at the end of eachsuccessive time interval. EXAMPLE 12.9 Given the following 15-min time-area curve, find the IUH for the 1000-acrewatershed.Then determinethe 15-min syntheticunit hydrograph.The storagecoefficient K is 30 min.

Timeinterval (min) 0 - 1 5" 15-30 30-45 45-60

Areabetween isochrones (acres) 100 300 500 100

Solution. From Eqs. 12.39and 12.40,Cs : 0.4 and Ct : 0.6.Routing most is easily accomplished a tableauas follows: in

222

CHAPTER12

UNIT HYDROGRAPHS

Time (hr) 0

I (acre-in./Af )

I (cfs)

col + c1Q1

IUH (cfs)

100 0.25 300 0.50 500 0.75 100 1.00 0 1.25 0 1.50 0

400 1200 2000 400 0 0 0

160+0 480 + 96 800 + 346 160+ 688 0+509 0+305

0 80 368 861 99'7 6't9 l

due to the magnitudeof K for long recession The IUH has a characteristically hr, the flow becomes0.6 times the previous this example.Note that after 1.25 in to flow andcontinues decayat this rateindefinite$.As discussed Section11.5, time, which is release the excess-runoff the time baseof the IUH shouldequal Clark's method often producesproone definition of time of concentration. of longedrunoff because this shortcoming. The 15-min unit hydrographis found using Eq. 12.6, ot

UH) : lGUH, + IUH,-15) O,(l5-min

This resultsin:
Time IUH 1S-min UH (cfs)

(h0
0 0.25 0.50 0.75 1.00 t.25 1.50 :

(efs)
0 160 576 1146 848 509 :

0 80 368 861 997 679 : tl

If the waterri"O ,uU time is not available,the K value can also be estimatedby recognizingthat Q, = KdQldt when the inflow is zero in Eq. 12.36.This occurs at of approximatelythe inflection point on the recession Fig. t2.2, when inflow to the of the estimate the Kvalue is the ratio If dataare available, channelceases. hydrograph at this particular point on the hyof the flow rate to the slope of the hydrograph drograph.

12.6 SYNTHETIC UNITHYDROGRAPHS 223

Nash'sSyntheticIUH
One of the earliest formulations of the IUH was developedby Nash.l8Instead of characteizingrunoff as translationfollowedby storage a singlelinear reservoiras in Clark did, Nash viewed the watershedas a seriesof n identical linear storagereservoirS,eachhavingthe samestoragecoefficientK. The first instantly (r : 0) receives a volume equal to a full inch of net rain from the entire watershed. This water then passes through reservoirsI,2,3, . . . , fl, with eachproviding an additional diffusion effecton the original I -in. rain. The number of reservoirs, n, is uniquely related to the reservoir storage coefficientK andthe watershed time. Once the IUH is developed; can be used lag it to synthesizeany other hydrograph by application of the convolution integral, Eq. I2.7, or from the approximate methods discussed Section12.3. in The derivation of Nash's equationfor IUH beginsfrom continuity at the first reservoir:

I,- Qu: #1,

wherc Qt, is the outflow from reservoir 1 at time t. SubstitutingS,: r ) 0 (for an IUH, I, is zero after t : 0), we obtain

(r2.4r)
KQr, at time

_Qr,: u#1,,, #1,,,:

which can be written dQ"

(r2.42)

eu:
Integrationfrom / : 0* to time t gives

' - kI o '

(r2.43)

l n Q r , - l n Q r , l , - o :- which reducesby taking antilogarithmsto

Qt,

(r2.44)

o,l;:
BecauseQr, : S,/K and S,=o: 1 in., then

e ''-ttK

(r2.4s)

Qu:f,"-'r*

(r2.46)

wherc Qy has units of depthper unit of tirne. Equation 12.46is an exponentialdecay function having an initial value of I/K at t : 0. This monotonically decreasing outflow from reservoir 1 becomesinflow to the secondreservoir. The secondreservoiris initially empty. The continuity relation
Qu * Qz,:

K+
dt

(r2.47)

224

CHAPTER12 UNIT HYDROGRAPHS

is solved,giving | ,.. (r2.48) 9r, : V;te-"^ This equationhas a full hydrographshape,beginningwith zero flow at time zero, peaking at the maximum of the function, and eventuallyrecedingto zeto. Similarly derived,the hydrographflowing from ruthreservoirhas the form
Qn, : |n-r n-t/K

, (12.49)

which is the two-parametergammafunction,

e,,:

(12.50)

#1n-1r-t/K

rain falling by Because outflow from the nth reservoirwas caused 1 in. of excess the an Eq. instantaneously, 12.50describes IUH.

for EStimatiOn of K and n Valuesof K and n arc needed applicationof Nash's IUH. By integration,the centroid of the distribution (Eq. 12.50) occurs at t : nK. - 1). The From classicalcalculusmaximization,the peak flow occurs at t : K"(n : 0 is n(n + l)Kz. Trial combinationsof n and secondmoment of the IUH about / K can be usedto developthe IUH from Eq. 12.50, and the momentsof the plotted distribution can be estimatedto verify the productsnK and n(n * I)Kz ' If the IUH by is discretizedinto m At increments,the momentsare approximated First moment = > ttQi Lt
m

(r2.sr) (r2.s2)

and

moment 2 t?Q, Lt Second

Another less tediousapproachis to use the definition of lag time as the time from For an IUH, this is the sameas the centroid of rain to the centroid of the hydrograph. time can be determinedfrom equationssuchas centroidal distance.Thus, if the lag reducing the number of those in Section 11.6, the product nK can be established, trials. to haveattemptedto relateNash'sK andn parameters basin Someinvestigators relations developed Rao et al.1e techniques. usingregression and storm characteristics for urban areasgreaterthan 5 mi2:

K :
and
Lt -

po.zzz g.57 54o.zto (l + 1)o'azzPo:oa 0.83140.458D0.37r (I + 1yt'ezzPozat

(r2.s3) (r2.s4)

is wherenK : tlandA is areain squaremiles,D is net rain duration in hours,Pn", the net (effective)rain depth in inches,and / is the ratio of imperviousto total area.

12.6 SYNTHETICUNIT HYDROGRAPHS

225

ColoradoUrbanHydrograph (CUHP) Procedure

Synthetic unit hydrographscan be tailored for regional use. As an example,the Coloradourban hydrographprocedure20 provides 5-min syntheticunit hydrographs for use in the Denver metropolitan area. lt is based on Snyder's method and is considered applicableto watersheds the size rangefrom 90 acresto 10 mi2, with in "regular" shapes(length 4 times width). It was developed the early 1980sand in modifiedin 1984to reflectrefinements from earlyapplications. its pre-1984'form, In the Snyder/CUHPC, and C, valuesfor use in Eqs. 11.5 and 12.17 are
C, : 7.8I1P2378 and Co: 0.89C?'46

(r2.ss) (r2.s6)

where P" is the percent impervious.Theseregression equationswere developed for P" > 30 percent, using data for 96 storms over 19 urban watersheds. The given equations apply to normal watershed conditionsandneedto be adjusted steep, for flat, or sewered basins. an urbanareais fully sewered, calculated from Eq. 12.55 If the C, is decreased percent. If sparselysewered,a 10 percent increaseis made. If the 10 averageslope S of the lower 80 percent of the main water courseis flat (less than 0.01 ftlf| the C, value becomes C,:3.12/Po78so2 and for steepareas(,S> 0.025 ftlf$, C, : 3.75/Po78s'02

( r 2.57) ( 12.s 8)

The C, coefficient is determinedfrom Eq. 12.56 and adjustedto 10 percentup or down for fully or sparselysewered conditions,respectively. For the CUHP applications, Eqs. 12.18 and 12.19become ryro: SO}A/Qo w75:260A/Qo

(r2.se)
(12.60)

whereA andQohaveunits of squaremiles and cubic feet per second. plotting ft., For the smallerof 35 percentor 0.670is placedleft of the peak.For W15,45 percentis placedto the left, or O.424To 0.67pwas usedfor W'r6. is the time from beginning if I of runoff of the unit rainfall to the peak time. Severalinvestigatorshave suggested that Snyder's C, and slope S are correlated.a'S original CUHP procedurewas alteredto recognizethis relation, making The the adjustments C, unnecessary. the modifiedversion,Eqs. 12.57and 12.58are in For bypassed, thetime to peakratherthanlag time is usedin Snyder's 11.5,where and Eq. tp: c,(LL,o/\8100' (12.6r)

The revisedtime coefficientC, is obtainedfrom Fig. 12.20a,and the peak coefficient is found from
Co : PC,Ao'1s

(r2.62)

where

S : weightedaverage slopeof basinalongthe streamto the upstream basin boundary (ftlft)

CHAPTER12

UNIT HYDROGRAPHS
0.18

Equations curve: of Ct= aI?+ blo+ c

0.16 tro. I 00
b c -0.00371 0.163 0.146 0.000023 -0M224 ] 3 x 1 o r -8.01 x l0- 0 12(

Basic equations I LL--\0.48 'r: "\fi) 4p : 640c, --7-

0.14
r i

\r

tp = time to peak (br) L = watershed length (mi) Ic, = distance to centroid (mi) So = waterway slope (ff/ft)

E q
o

U 0.10

0.08 3q. 0.10

Ec,.2l
20

Eq.3 80
100

40 60 Percentimpervious,l, (a)

12
Equations of curve: P = aI?+ blo+ c

10

Eq.

\t<

2 | -o.t --T-T-l )0091 +0.228|

b -0 012

+216

06

z-r ,ql.: v t Y

\ o 6

Eq. 1 0 20

84.2
40 60 Fercent impervious, l, (b) 80 100

Figure 12.20 Snyder's C, and C, coefficientsfor urban areas,for use with CUHP: (a) relation between C, and imperviousness (b) relationbetweenpeakingparameand ter and imperviousness.

PROBLEMS

227

from Fig. I2.20b P : coefflcient,dependingon imperviousness, A : drainagearea (mi2) L : length of the main stream channel (mi) from the outlet to the divide L"o : length along the main channel (mi) from the outlet to a channelpoint centroid nearestthe watershed usingEqs. 12.17 can The peakrate andtime of the unit hydrograph be developed 12.6t. After four additional points definedby Eqs. 12.59 and t2.60 areplotted, and the rest of the 5-min hydrographcan be fitted to provide a total area representing 1.0 in. of direct runoff. A hand fit is applied, or the mathematical curve-fitting techniquesdescribedearly in Section 12.5 can be used if a gamma (or any other) appropriate. distribution is considered (lessthan 90 acres),the time to peak is watersheds For small
Lp

: o.zg(P=l- * 0- . 0 7(')^, ".^?3,6=P: 1 +: Pz - 0.49P" 0 .1 4

6P,)

02.63)

where I is the time of concentrationin minutes, and P^ is the percentimpervious.

r summary
Unit hydrographmethodsallow the hydrologistto estimaterunoff volumesand rates for virtually any storm. By far, the greatestnumber of problems in practice are Most of the current computermodelsuse procedures. usingunit hydrograph evaluated rs in procedures described Chapte 23,24, and25. Thesemodelsare as unit hydrograph and simply computerprogramsthat perform the unit hydrographsyntheses convoluthe in tion stepsdescribed this chapter.Any softwareuser shouldunderstand origin, hydrographmethod for applicability,and parameterestimationprocedures eachunit usesof the computermodelswill resultfrom a thorough The seiected. most successful in described this chapter. familiarity with the processes

PROBLEMS
a 12.1. Given the following storm pattern and assuming triangular unit hydrographfor one determinethe compositehydrograph. time unit,

Stormpattern
Time unit Rainfall

1 1

2 1

3 4

4 2

Unit hydrographbaselength : 6 time units; time of rise : 2 time units; and maximum ordinate : I rainfall unit height. 12.2. Given a rainfall duration of 1 time unit, an effectiveprecipitation of 1.5 in., and the following hydrograph,determine(a) the unit hydrographand (b) the compositehydrographfor the given storm sequence.

228

CHAPTERl2

UNIT HYDROGRAPHS Hydrograph 1.S in. net rain in .t time unit for Time units 1 2 3 Flow (cfs) r00 98 220 4 4.5

srz 620 585 460

7 8 330 2to

9 r 0 150 105

l 1 l2 13 75 60 54

Storm sequence
Time units Precipitation (in.)
I

0.4

z l.l

2.0

1.5

12.3. Solve Problem lZ.2 if the storm sequence as follows: is

Storm sequence
Time units Precipitation (in.)

r 0.3

2 r.4

0.9

Using U.S. GeologicalSurveyrecords,or other data, selecta streamflow hydrograph for a large, preferably single-peaked runoffevent. Separate baseflow and deterthe mine a unit hydrograph for the area. 12.5. For the unit hydrographof problem 12.1, constructan S_hydrograph. 12.6. For the unit hydrqgraphcomputedin problem 12.2, construct an S_hydrograph. 12.7. use the S-hydrograph problem 12.6to find a 3 time-unit of unit hydrograph. 12.8. Given a watershedof 100 mi2, assumethat C, = 1.g, the length of main stream channelis I 8 mi, and the length to a point nearestthe centroid is I O mi. Use Snyder,s methodto find (a) the time rag, (b) the duration of the syntheticunit hydrograpl, and (c) the peak discharge the unit hydrograph. of l2'9' Apply Snyder's method to the determinationof a synthetic unit hydrographfor a drainage area of your choice. 12'10' Use Fig. 12.13to determine a2-hr pnrthydrograph the if drainageareais 60 mi2 and Eq. 12.23ais applicable. 12.11. SolveProblem12.10using Eq.12.23b. 12.12. Assuminga Nebraskalocation, use Gray's method determine te a unit hydrograph: drainagearea = 1.0 mi2, length : 0.6 mi, S" : 1.3 percent. 12'13' A drainageareain Nebraskacontains30 mi2.The tenjtn ortne main channelis 10 mi andthe'representative watershed slopeis 2.5 percent.iJseGray's methodto determine a unit hydrograph. 12'14' Dischargerates for a flood hydrographpassingthe point of concentration for a 600acredrainagebasinare given in the tabl; belo;. The flood wasprotluced by a uniform rainfall rate of 2.15 in.rru, which started,at9 A.M., abruptry ended at il A.M: and resulted in 5.00 in. of direct surfacerunoff. The base flow (derived tiom influent seepage) prior to, during, and after the storm was 100 cfs.

l2'4'

PROBLEMS

229
6

8 a.u. Time Measured discharge 100

10

11

12

I P.M

100 300 500 700

800

600 400 300 200 100 100

cease? a. At what times did direct runoff begin and Determinethe @index (in'/hr) for the basin' b. ordinates(cfs)for eachtime.listed' c. Derive the 2-hr'unit-nydrograpir time) for the basin. release concJnffation(excess d. Estimatethe tirne of e.Atwhattimewoulddirectsurfacerunoffceaseiftherainfallof2.T5in./hrhad than 2? and begunat 9 .q.'l\a. had lastedfor 8 hr rather .ur" (cfs) and the direct runoff (in.) for a uniform urg" f. Determined;;;i;t rainfall of 2j5 it'lhr and a duration of 8 hr' rates Lz.lS. Measuredtotal hourly discharge table' The hydrr in the accompanying uniform intensity of 2'60 in'/hr startir baseflow from 8 A.M' to 3 P'v' was a ' determined as the area under the dir

Time Measured discharge

8 a.v 100

10 100 300

11 600

12 400

I P.M

2 100 100

200

a. At what time did the direct runoff begin? to the volume of the direct surface b. betermine the net rain (in') corresponding runoff of 1000 cfs-hr' the c. Determine { indexfor the basin' basin by tabulating time in hours and d. Derive " , nt ltii ttyJrog'api' for the in discharge cfs. time of the basin? release e. What is the excess the direct tt" the derived2-hr unit hydrograptr-to-determine f. For the ,"*"'t"'it, beganat 1 p.u. and ralnfall runoff rate (.fr) ;;; ;.*, on a day when excess-(netj 't"i at 5 P'M' iti"tt ity of 2 in'/hr for 4 hr' ceasingabruptly "t " continued |2.|6.A5-hrunithydrographfora425}-acre.basinisshownintheaccompanyingsketch. Thegivenhyd.og.-uphactuallyappearedasadirectrunoffhydrographfromthebasin' of 5 hr,beginning

by caused rainf;ffi;;;;ii"'"iriv

in./hrfor a duration "i 0.30

att = 0. time of the basin' release a. Determinethe excess index for the basin' b. Determinethe @ contributingto direct runoff 4 hr after c. what n".""*"J" "'i,ti" JJr"g" t"r* was rain began(r = 4)? as shownin the sketch'Do not scale to d. Use your response part c to determineQp' Qp from the drawing' t = 3 andr : 5' Why did the hydrograph e. Note that rain continuedto fall between : 3 andt = 5, rathet than continue to rise during those form a ptateaut"i*""n t 2 hours?

230

CHAPTER12

UNIT HYDROGRAPHS
QO

,i: 700 = ouu I 500

e 40o

! :oo $ zoo
100

Time, t

f. Usethe given 5-hr unit hydrograph determine direct runoff rate (cfs)at 7 p.r'r. to the on a day when rain fell at an intensity of 0.60 in./hr from I p.rrr. 11 p.vr. to 12.17. The 2-hr unit hydrographfor a basin is given by the following tabie:

Time (hr) O@fs)

0 0

1 60

2 200

3 300

4 200

5 120

6 6 0

7 8 3 0 1

9 0

a. Determinethe hourly discharge values(cfs)from the basinfor a net rain of 5 in./hr and a rainfall durationof 2hr. b. Determinethe direct runoff (in.) for the storm of part a. What is the direct runoff for a net rain of 0.5 in./hr and a duration of Z Iv? c. Rain falls on the basinat arateof 4.5 in.lhr for a2-hr period and abruptly increases to a rate of 6.5 in./hr for a second2-hr period. convert theseactual intensitiesto netrain intensities usinga {index of 0.5 in./hr. Constructa tablethatproper$ lags and amplifiesthe 2-hr unit hydrograph, and determine hourly ordinates(cfs)of the direct runoff for the storm.The deriveddirect runoff hydrographshouldbegin and end with zero discharge values. 12.18. Given the following 2-hr unit hydrograph for a drainage basin, determine hourly ordinatesof the 4-hr unit hydrograph:

Time (hr) 0@f9

0 0

1 50

2 300

3 400

4 200

5 s0

6 0

12.19. Usethe following 4-hr unit hydrograph a basinto determine peak discharge for the rate (cfs)resulting from a net rain of 3.0 in./hr for a 4-hr duration foliowed immediately by 2.0 in.lhr for a 4-hr duration.

Time (hr) 0(cfO

0 0

2 200

4 300

6 100

8 50

10 0

PROBLEMS 231 12.20. Compare the time from the peak to the end of runoff for the SCS triangular unit hydrographwith the time of concentration,/". Discuss. hydrograph 12.21. Prove that the areaunderthe rising limb of the SCSbasic dimensionless equalsthat of the triangular unit hydrograph,that is, 37.5 percefi of the total. 12.15occurswhenx:aB, 12.22. Bycalculus,showthatthemaximumvalueof/(x)inEq. by for a > 1. Also solvefor the centroidaldistance taking the flrst moment aboutthe y axis. 12.23. Accordingto the rational method(seeChapter15) of estimatingpeakflow from small areasn peak rate for a storm with uniform continuingintensity is equal to the net the For what conditions,if any, would rain rate and occurs at the time of concentration. ' of Eqs. 12.63 and 12.17result in agreement the peak magnitudeand time, estimated by CUHP, with those of the rational method?Discuss. L2.24. Describetwo methodsthat could be usedto constructa 2-hr unit hydrographusing a l-hr unit hydrographfor a basin. rates (cfs)from a 2.48-n12drainagebasin are tabu12,25. Measuredtotal hourly discharge lated below.The hydrographwas producedby a rainstormhavinga uniform intensity of 2.60 in./hr starting at 9 A.M.and abruptly ending at 11 l.rvl. The baseflow from 8 ,q,.lt. 3 p.lrl.was a constant 100 cfs. to

Time Discharge(cfs)

8 e.u. 100

9 100

10 300

11 450

12 300

1 p.vr. 2 150 100

100

a. At what time did direct runoff begin? b. Determinethe grossand net rain depths(inches). c. Derive a 2-hr unit hydrograph for the basin by tabulating time in hours and in discharge cubic feet per second. d. Derive a 4-hr unit hydrographfor the basin. e. Derive a l-hr unit hydrographfor the basin. 12,26. Given below is a 3-hr unit hydrographfor a watershed.The {-index is 1.5 in./hr. 3-hr rainfall ratesof 2.5, Desiredis the DRH for an 18-hr stormhavingsix successive 3 . 5 , 1 . 5 ,4 . 0 , 6 . 5 , 2 . 5i n . l h r .

Time (hr)

10 11 12 13 14

o(ruH) 0 10 40 60 80 100 90 70 60 50 40 30 2 0 1 0 0
12.27. Use the following 2-hr unit hydrographto determinethe peak direct-runoffdischarge rate (cfs)resulting from a net rain of 2.0 in./hr for 5 hr'

Time (hr) 0(cf9

0 0

1 50

2 200

3 300

4 200

5 1s0

6 100

7 0

12.28, The ordinatefor a 5-hr unit hydrographis 300 cfs at a time 4 hr after the beginning of net rainfall. A storm with a uniform intensity of 3 in,/hr and a duration of 5 hr occursover the basin.What is the runoff rate after 4 hr if the @index is 0.5 in./hr?

232

CHAPTER UNITHYDROGRAPHS 12 12.29. Given below is an IUH for a watershed. Use the IUH to find hourly DRH rates for a net rain of 4 in. in a 2-hr oeriod.

Time (hr)

o(ruH)

0 0

1 10

2 40

3 50

4 60 80

6 100

7 80 20

8 r0

9 10 0

12.30. A 2-fu unit hydrographfor a basin is shownin the sketch. (cfs)for a net rain of 5.00 in./hr and a duration of a. Determinethe peak discharge 2 hr. b. What is the total direct surface runoff (in inches)for the storm described part a? in c. A different storm with a net rain of 0.50 in./hr lastsfor 4 hr. What is the discharse at 8 p.vr.if the rainfall startedat 4 p.tvt.?
I I I

{, +oo
po
I I

--t

.A 2oo

T I I I

--i

I I I I I
I

I I -T I
I I

Time (hr)

12.31,. Recordedflow rates for a net rain of 1.92 inchesin 12 hours are shownin the table. ' If the baseflow is 375 cfs throughoutthe storm,determinethe 12-hrunit hydrograph, and convert it to a 6-hr unit hydrograph.Then apply the 6-hr unit hydrographto determinethe total hydrograph(including 400 cfs baseflow) for a24-hr storm having four 6-hr blocksof net rain at ratesof 0.7, 3.8, 10.8,and 1.8 in. per hour.

Time in hours

Observedflow (cfs)

0 6 12 18
aA JU

36
A'

48 54 60 66 72 78

375 825 2200 36s0 3900 3200 2375 1,725 1250 900 650 490 410 375

12.32. Starting with a triangular-shaped unit hydrographwith a baselength of 2.67toand a height of qo,deriveEq. 12.22,qo : 484A/tp. Statethe units of eachterm usedin the .- --- derivatign"

REFERENCES 233 rate 12.33. The SCS syntheticunit hydrographis derivedby computingthe peak discharge (cubicfeet per second)from qo : 484A/tp.In the derivation,it was actually assumed thatqrin.lhr:0.7sv/tp,whereVisthevolumeofdirectrunoff(inches),roisthetime to peak flow (hours),and A is the basin area(squaremiles).Derive the first equation from the second. for a 12.34. Which of the techniques synthesizing unit hydrographrequiresthe leastcompuWhich probably requiresthe the tationai effort in developing entire unit hydrograph? most?

REFERENCES
Division, 1948. 1. W. D. Mitchell, "Unit Hydrographsin Illinois," illinois Waterways "Stream-Flowfrom Rainfall by the Unit-GraphMethod," Eng.News-Rec. 2. L.K. Sherman, 108,501-505(Apr. 1932). "Unit Hydrographsfor Gaugedand Ungauged 3. Rand Morgan and D. W. Hulinghorst, Watersheds," U.S. EngineersOffice, Binghamton,NY, July 1939. 4. F. F. Snyder,"synthetic Unit Graphs," Trans.Am. Geophys.Union 19,447 -454(1938). 5. A. B. Taylorand H. E. Schwartz,"lJnit HydrographLag and PeakFlow Relatedto Basin s. Characteristic Trans. Am. Geoplry Union 33' 235-246(19 52). s," ASCE Trans.ll0,1419-1446(1945). 6. C. O. Clark, "storageand the Unit Hydrograph," Division. Washington, 7. Water SupplyPapers, U.S. GeologicalSurvey,Water Resources D.C.: U.S. Government PrintingOffice, 1966-1970. 8. Hourly PrecipitationData, National Oceanicand AtmosphericAdministration.WashingPrinting Office, L971. ton, D.C.: U.S. Government Jr., "Synthesisof the Inlet Hydrograph,"Tech.Rept.No. 3, Department 9. JohnC. Schaake, The of Sanitary Engineeringand WaterResources, JohnsHopkins University,Baltimore, MD, 1965. 10. G. Aron andE. White, "Fitting a GammaDistribution overa SyntheticUnit Hydrograph," Resources Bull.18(l) (Feb.1982). Water Bull. l9(2) (Apr. 1983). WaterResources 11. G. Aron and E. White, "Replyto Discussion," 12. M. Collins, "Discussion-Fitting a Gamma Distribution over a Synthetic Unit HyBull. 19(2) (Apr. 1983). drograph,"WaterResources Engin' 13. "Flood-HydrographAnalysisand Computations,"U.S. Army Corpsof Engtneets, D.C.: U.S. Government neering and Design Manuals,Emlll0-2-1405. Washington, PrintingOffice,Aug. 1959. "Hydrograph Synthesisby Digital Computer," Proc. 14. M. D. Hudlow and R. A. Clark, ASCEJ. Hyd. Div. (May 1969). in Characteristics SyntheticHydrographAnal15. V. Mockus, "Use of Storm and Watershed Service,1957. ysisandApplication,"U.S.Department Agriculture,Soil Conservation of for 16. D. M. Gray,"synthetic Unit Hydrographs Small DrainageAreas," Proc' ASCEJ. Hyd. Div. 87(HY4) (July 1961). for for 17. W. H. EspeyandD. G. Altman, "Nomographs Ten-minuteUnit Hydrographs Small Urban Witersheds,"EnvironmentalProtectionAgency,Rept.EPA-600/9-78-035,Washington, D.C., 1978. Unit Hydrograph," IASH Publ. No. 45' Vol. 18. J. E. Nash, "The Form of the Instantaneous ' 3. 1951. "ConceptualHydrologicModels for Urbanizing 19. R. A. Rao, J. W. Delleur, and B. Sarma, Basins,"Proc. ASCE J. Hyd. Div. (HY7) (July 1972). "First Short Courseon Urban Storm WaterModeling 20. University of Coloradoat Denver, Using ColoradoUrban HydrographProcedures,"Departmentof Civil Engineering,June 1985.

C h a p t e r1 3

Routing Hydrograph

r Prologue
The purposeof this chapteris to: . Presenttechniquesfor determiningthe effect of streamsand reservoirson hydrographshapes the hydrographs movedownstream through the systems. as . Distinguishbetweenthe two major classifications hydrograph of routing techniques. . Familiarizethe readerwith procedures determiningwhen to apply eachof for the variousrouting methods. water Flood forecasting, reservoirdesign,watershedsimulation,and comprehensive planning generallyutilize someform of routing technique.Routingis used resources a to predict the temporal and spatial variationsof a flood wave as it traverses river reach or reservoir. Routing techniquesmay be classified into two categorieshydrologicrouting and hydraulic routing. Hydrologic routing employsthe equationof continuity with either a linear or curvilinear relation betweenstorageand dischargewithin a river or reservoir. Hydraulic routing, on the other hand, usesboth the equation of continuity and the equationof motion, customarilythe momentumequation.This particular form utilizes the partial differential equationsfor unsteadyflow in open channels.It more adequately describes dynamicsofflow than doesthe hydrologicrouting technique. the Applications of hydrologicrouting techniques problemsof flood prediction. to of evaluations flood control measures, assessments the effectsof urbanization of and are numerous.Most flood warning systems instituted by NOAA and the Corps of Engineersincorporatethis techniqueto predict flood stagesin advanceof a severe storm.It is the methodmost frequentlyusedto sizespillwaysfor small, intermediate. and large dams. Hydrologic river and reservoirrouting and hydraulic river routing techniques presented separate are in sectionsof this chapter.

RIVER ROUTING 13.1 HYDROLOGIC

235

RIVERROUTING 13.1 HYDROLOGIC

The first referenceto routing a flood hydrographfrom one river station to another was by Graeff in 1883.1 The technique wasbasedon the useof wavevelocity and a rating curve of stageversus discharge.Hydrologic river routing techniquesare all founded upon the equationof continuity dS (1 3 . 1 ) I - O : dt where 1 : the inflow rate to the reach O : the outflow rate from the reach dS/dt : the rate of changeof storagewithin the reach

Three of the most popular hydrologic river routing techniques described subseare in quent paragraphs,

Method Muskingum
Storagein a stableriver reachcan be expected dependprimarily on the discharge to of into and out of a reach and on hydraulic characteristics the channelsection.The as2 storagewithin the reach at a given time can be expressed

s:2;y7^n+(1 - X)O-n1
a

(r3.2)

of at Constants andn reflectthe stagedischarge a characteristics control sections each characteristics the section. of end of the reach,and b and m mirror the sta$e-volume The factor X defines the relative weights given to inflow and outflow for the reach. The Muskingum method assumesthat mfn - L and lets b/a : K, resulting in

S: KIXI + (1 - X)o]
where K : the storagetime constantfor the reach x : a weighting factor that variesbetween0 and 0.5.

(13.3)

closeto the wave Application of this equationhas shownthat K is usually reasonably trarel time through the reach andX averages about 0.2. Behavior of the flood wave due to changesin the value of the weighting factor X is readily apparentfrom examinationof Fig. 13.1. The resulting downstreamflood waveis commonlydescribed the amountof translation-that is, the time lag-and by As by the amount of attenuationor reduction in peak discharge. can be noted from Fig. 13.1,'thevalueX : 0.5 resultsin a pure translationof the flood wave. procedure Application of Eqs. 13.1 and 13.3 to a river reachis a straightforward if Kand X are known. The routing procedurebeginsby dividing time into a number Eq. 13.1 in finite difference form, using of equal increments,A/, and expressing subscriptsI and2 to denotethe beginningand ending times for Ar. This gives I t + 1 2_ O 1 + 0 2 _ S ' - S ' 2 2 A t

(13.4)

236

CHAPTER13

ROUTING HYDROGRAPH

99 E

Time

Figure 13.1 Effect of weighting factor. The routing time interval A/ is normally assigned any convenient value between the limits of K/3 and K. i in change the river reachduringthe routing interval from Eq. 1 3 . 3s The storage

s, - s' : Klx(I' - 1')+ (r - x)(o2- o')l

O2: CsI2+ CII. + CzOl

(13.s)
( 13.6)

and substitutingthis into Eq. 13.4 resultsin the Muskingum routing equation

in which

co

-KX + 0.5Ar K-KX+0.54/ KX + 0,5 Lt ct: K-KX+0.54t K - K X- 0 . 5 A r

cz:

(r3.7)
(13.8)

K-KX+0.54t

(13.e)

Note that K and Ar must havethe sametime units and aiso that the three coefficients sum to 1.0. if Theoretical stability of the nunrericalmethod is accomplished Al falls be- X). The theoreticalvalue of K is the time required tweenthe limits 2KX and2K(I for an elemental(kinematic) waveto traversethe reach.It is approximatelythe titne interval betweeninflow and outflow peaks, if data are available.If not, the wave volocity as for velocity can be estimated variouschannelshapes a function of average uniform flow canbe estiinated for steady flow rate Q. Velocity Vfor any representative by either thi Manning or Chzyequation: The approximatewavevelocities for differare ent channel'shapes given in Table 13.1. FOR WAVE 13.1 KINEMATIC VELOCITIES VARIOUS TABLE '
CHANNELSHAPES Channelshape Wide rectangular Triangular Wide parabolic Manningequation
t-V !v
lJrl 9 '

Ch6zyequation
3-tl 2 l

tv Zrl
6 '

RIVER 13.1 HYDROLOGIC ROUTING 237 by Since1, and Irareknown for every time increment,routing is accomplished incrementsusing each02as Ol for the next time time solvingEq. 13.6for successive incremeni. Example 13.1 illustratesthis row-by-row computation'

EXAMPLE 13.1 : 2 days.The Perform the flood routing for a reach of river given X : 0.2 and K : 1 day is shownin Table 13.2,column 1. Assumeequal inflow hydrographwith Ar inflow and outflow rates on the 16th. Solution. If Ar : l dayandX = 0.2 andK :2days, thenEqs.13.7toI3.9 is give Co : 0.0477,C1 : 0.428, andC2 : 0'524' Row-by-rowcomputation ll given in TabIe13.2. routing of Determination of Muskingum K andX Values K andX for Muskingum average usingK equal to the travel time in the reachand an are commonly estimated value of X : 0.2.If inflow and outflow hydrographrecords are availablefor one or more floods,the routing processis easilyreversedto provide better valuesof K and values of S versus X for the reach. To illustrate the latter method, instantaneous 13.2 TABLE
(1)
(2\

(3)
Crl't

(4) C,Q,
) )7)

Date 3-t6 17 18 19 20 21, 22

z3

lnflow

voI2

(5) Computed outflow 4,260 4,419 6,rr9 8,783 12,791 16,941 1 9 ,1 0 23,578 34903 46,705 51,469 49,109 41,514 32,67' | 34,120 39,559 43,729 42,199 37,569 29,166 22,128 t6,932 13,222 10,576 8,497

24 25 26 27 28 29 30
J I

4-l 2 4 5 6 7 8 9

4,260 7,646 11,167 16,730 21,590 20,950 26,570 46,000 59960 57,'t40 47,890 34,460 21,660 34,680 45,180 49,r40 41,290 33,830 20,5t0 t4,720 11,436
o )04

364 532 798 1,029 999 L,267 2,194 2,860

)'754 ) )24

t,823 3,2'72
4 114

7,160 9,240 8,966 tl,37l 19,688 25,662

)a1l)

2,315 3,206 4,602 6,702 8,877 10,013 12,355 18,289

)4 417

1,643 1,033 1,654 t r55 r,969 t,613 9'78 702 443

3 t J

20,496 14,748 9,270 14,843

1q??7

26,9'70
)\ 117 )'t 171 t7 1))

7,831 6,228 6,083

2t,031, 17,672 14,479 8,778 6,300 4,894 3,977

? ?{1 ) 66\

l7,879 20,729 22,914

)t 11)

29'| 290

19,686 t5,283 11,595 8,872 6,928

238

CHAPTER13

HYDROGRAPH ROUTING

XI + (l -- X)O are flrst graphed severalselected for valuesof Xas shownin Example 13.2.Because andXI + (1 - X)O are assumed be linearlyrelatedvia Eq. 13.3, S to the accepted value of X is that which gives the best linear plot (the narrowestloop). After plotting, the valuefor K is determined the reciprocalof the slopethrough the as narrowest loop, sincefrom Eq. 13.3

K :

n+0-x)o

(1 .10) 3

Instantaneous valuesof S for the graphsin Example I3.2 were determinedby solvingfor S, in Eq. 13.4for successive time increinents.A valueof S, : 0 was [sed for the initial increment,but the value is arbitrary since only the slope and not the intercept of E_q.13.3 is desired.The 52 valuesare plotted against.average weighted discharges, + (l - X)O in Table 13.3. A preferablemethod would be to plot S, XI (rather than average) valuesagainstcorresponding valuesof instantaneous valuesof XIz + (l - X)Or, using recordedvaluesof inflow and outflow (not provided).

TABLE 13,3
l.ll"

Date 3- 16 t7 18 t9 20 21 22 23
z+

2 (cfs) 5,870 9,310 12,900 20,500 21,000 23,400 32,500 s5,400 62,700 52,600 43,200 25,200 22,800 41,200 s0,400 45,300 38,800 2?,000 16,200 12,400 10,200 8,080 6,010 5,050

u =

Oj+02

2 (cfs) 4,180 6,970 7,560 14,200 18,300 18,500 21,300 29,300 39,700 48,700 53,300 48,700 37,r00 35,800 35,800 35,800 42,700 44,lOO 35,400 25,200 t6,400 11,500 9,380 7,860

sz
(1O3cfs-days)

Weighted discharge (cfs) X+(1 -X)O

X: 0.1 4,350b 7,200 8,090 14,800 18,600 19,000 22,400 31,900 42,000 49,100 52,300 46,400 35,700 36,300 37,300 36.800 42,300 42,400 33,500 23,900 15,800 1r,200 | 9;040 7,300

X : 0.2 4,520 't,440 8,630 15,500 18,800 19,500 23,500 34,500 44,300 49,500 51,300 44,000 34,200 36,900 38,700 37,700 41,900 40,800 31,600 22,600 15,200 10,800 8,710 7,300

X: 0.3 4,690 7,670 9,160 16,100 19,100 20,000 24,700 37,100 46,600 50,000 50,300 41,700 32,800 37,400 40,200 38,600 41,500 39,000 29,600 2r,400 14,500 10,500 8,370 7,020

1.7 4.0 9.4 15.7 18.4

z5-J

25 3-26 27 28 29 30 3l 4-r 2
J

34.5 60.6 83.6 97.5 87.4 73.9

Jv.o

4 5 6 7 8

65.0 79.6 r 89. 85.2 68.0 48.9 36.1 29.9 26.5 23.1 20.3

"Note: ,S2 ,Sr* f Ar - d Al = fseeEq. 13.4]. ' E x a m p l e :4 3 5 0 : 0 . 1 ( 5 8 7 0 )+ ( 1 - 0 . 1 X 4 1 8 0 ) .

13.1 HYDROLOGIC RIVERROUTING

239.,

EXAMPLE 13.2 Given inflow and outflow hydrographson the Muckwamp River, determine K and X for the river reach.(SeeTable 13.3.)

o po

E e .E b0 o

ro X
I

g t\ X

,a o x 5 A

80.000 cfs-davs Storage, S

Solution. Selecting the narrowest loop gives X : 0.3; K : 80,000 cfsdays/40,000cfs : 2.0 days. These values could now be used to route other floods through the reachas in Example 13.1. Inherent in this procedureis the postulatethat the water surfacein the profile betweenupstreamand downstream reachis a uniform unbrokensurface ends of the section.Additionally, it is presupposed that K andX are constant throughoutthe range of flows. If significantdeparturesfrom theserestrictions are present,it may be necessary work with shorterreaches the river or to to of employ a more sophisticated approach. I I it Muskingum Crest Segment Routing Sometirnes is desirableto solve for a single outflow rate or route only a portion of an inflow hydrographby the Muskingum method(e.g.,the crest segment when only the peak outflow is desired).This is easily accomplishedby successively numbering the inflow rates as 11, 12, 13,. . . , In, In+t,. .. , and rewritingEq. 13.6as
O,: CsI, + Ctln-t + C2O^-l

( 1 3 .1 ) 1

where Onis the outflow rate at any time n. The outflow O,-1 is next eliminated from Eq. 13.11by makingthe substitution
On-t : Coln-t + ClIn-2 * C2O,-2

(r3.r2)

By repeatedsubstitutions the right-sideoutflow tetm On-2,On-2,. . . can eachbe for eliminated and On can be expressed a function only of the flrst n inflow rates or, as finally,
On : where Kr: Kz= K3: K,: K l I n + K 2 I n - 1 + K 3 I n - 2+ ' .' + K,I,

(13.13)

Co CoC2+ C1 K2C2 K,-rC.rfori>.2

240

CHAPTER13

HYDROGRAPH ROUTING

Using data from Example 13.1 to find the outflow rate on 3-26, we obtain Kr : C6: 0.0477 Kz: CoC2 Cr : 0.0477(0.524) 0.428:0.4530 + + : Kz : KzCz: 0.4530(0.524) 0.2374 Krt: KrcG : 0.0013

Thus the outflow on 3-26 is calculated O11: 0.0477 X (47,g90)+ 0.4530 X as : (57,740)+ . . . + 0.0013(4260) 51,469cfs.

SCS ConvexMethod
The U.S. Soil ConservationService (SCS) developeda coefficient channelrouting technique,similar to the Muskingum method, in their National EngineeringHanibook.3It has had widespreadapplication in planning and,designand can be used successfully evenwhenlimited storage datafor the reachare available. Until 1983,the procedurewas usedfor all streamflowhydrographrouting in TR-20, the SCS storm event simulation computer program describedin Chapter 24. Newer versions of TR-20 usethe att-kin methoddescribed Section13.3. in Analysis of Fig. 13.2 producesthe working equation for the convex routing method.Because areasunderboth curvesare equal,and because peak outflow the the

Figure 13.2 Geometricrelationsusedin the scs convexrouting method.(After U.S. Soil Conservation Service.3)

RIVERROUTING 13.1 HYDROLOGIC

241

is lessthan (and occurslater than) the peak inflow, the curvescrossat somepoint A, resultingin the fact that the value Orwill alwaysfall betweenIl andOt. At any time, the vertical distanceof 02 aboveOr (or below 01 on the right of A) is a fraction C, of the differenceIt - 01 as shown in the inset of Fig. 13.2. By proportionatevertical distances
Oz: Ot + C,(11 O)

(r3.r4)

This could be usedto route the entire inflow hydrographif C, could be established. FromEq. 13.14,

-'-

or- o, 1,.-o,

(13.1s)

Al Because is one limb of the triangle in the inset to Fig. 13.2, Lt _O"- O, Ir-O, K (13.16)

where the constantK is the horizontal time from O, to the interseptionof the line passingthrough Ol and 02. Thus C, is a function of both A/ and K, or
n w r_- L t v

(r3.r7)

routing method. is Determination of K and C, Proof that K from Fig. 13.2is a constant left to the by reader.It is a storageparameterwith time units and can be approximated the K twice the MuskingumX. from the Muskingummethod.Similarly, C, is approximately from Table 13.1) will provide The reach length divided by wavevelocity (estimated of another estimateof K, or actual measurements reach travel time can be used. by Equation 13.l'l canthenbe solvedfor C,. The valuerecommended the SCS,in the is of absence other estimates, C,:

(13.18)

steadydischarge,andV * 1.7 approxiwhere Vis the velocity for a representative (speed)of a kinematic wavetravelingthrough the reach.The units matesthe celerity (fps). of V in Eq. 13.18are feet per second The after the C, valueis estimated. accomplished Routingby Eq. 13.14is easily lessof the time to peak of the inflow hydrographto interval Ar shouldbe one-fifth or assurea sufficient number of calculatedoutflow rates to deflne the hydrograph.As so with all routing methods,the time interval shouldbe selected that one point falls peak and other locationsof rapid change. at or near the

242

13 CHAPTER

ROUTING HYDROGRAPH

the Unlike otherrouting methods, convexmethodequationfor O, is independent be used to forecast outflow from a reach without of 1r. Thus the proceduie can tnowing the concurrent inflow. This provides a method for early calculation and to warninf for floods.Flow recorderscanbe linked through microprocessors warning flood potentials at least one full routing-time systemsthat calculate downstream interval aheadof the flood. to The procedurecan be reversed find the inflow hydrographfor a given outflow cumulativemasscurve of inflow to the reachinsteadof hydrograph,or it can route a itself. the hydrograPh

Method Muskingum-Cunge
been the Severalattemptsto overcome limitations of the Muskingummethodhavelot complexity or difficulties in physically totally ,u"""riful becauseof computational The Muskingum parametersare best derived interpretingthe routing parameteri.a's are not easilyrelatedto channelcharacteristics. and fiom streamflowmeasurements blendedthe accuracyof the diffusion wave method (seeSection 13.3) Cunge6 with the simplicity of the Muskingum method, resulting in one of the most recomgives as for mendedtechniques generaluse.It is classified a hydrologicmethod,yet it methods' resultscomparablewith hydraulic Cungeshowedthat ihe finite-differenceform of the Muskingum equationbefor comesthe diffusion waveequationif the parameters both methodsare appropriI3'3, the Muskingumequationis atelyrelated'From Eqs. 13.1 and

K+ln+(l-x)ol:t-o
dt-

( 1 1e) 3.

obtain SubstitutingQ,for I andQi*rfot O, andrewritingin finite-differenceform' we

+ (1 fir"a':' - x)Q"+l xQI- (1: - x)Qi*,]Oi+i+ Qi- Q',*') t(Q',*'

form of If K is set equalto Lxfc,Eq. 13.20is alsothe finite-difference

(13'20)
(t3.2r)

u #' H : o *

which is calledthe kinematicwave equation(seeEq. 13.59)and can be derivedby The variable Ax combining the continuity and -oln"ntu- (or friction) equations.T denotesunincrerrr"ntofdistancealongthestreamaxisandcisthewavespeed' The equationto be usedfor routing is obtainedfrom Eq . 13.20by solvingfor the unknown flow rate,

+ OiIi: coQ',*'ctQi* crQi*,

where

(r3.22)

co

Lt/K - 2X 2(r-x)+Lt/K

(r3.23)

RIVERROUTING 13.1 HYDROLOGIC

243

cr:
4 v2 -

LtlK + 2x
2(r-x)+Lt/K 2(I-x)-cLt/Lx
" '

(r3.24) (r3.2s)

2(I-x)+Lt/K
./r .-\ L Aal?

the K Because : A,xfc,it represents time for a waveto travelthe routing reachlength at velocity t. Congeshowsthat the velocity c is the celerity of a kinematic Ax, moving (Table 13'1). described wavepreviously translation X : 0.5 and c Lt/A,x: 1.0, the routing equationproduces When When Ax : 0 (zeroreachlength),no translationor attenuation without attenuation. occurs. by the If previousflood data are available, routing parameterc canbe extracted can also be obtained calculations.Estimatesof the parameters reversingthe routing from flow and channelmeasurements. The value of X for use in Cunge'sformulation is

x:t('-#t)
c: mV

(r3.26)

where So : channelbottom slope (dimensionless) for the peak rate per 4o : discharge unit width (cfs/fO,normally determined velocity V by The value of celerity c can be estimatedas a function of the average Q3.27) The velocity QfA, andm is aboutI for wide natural channels. where v is the average m comesfrom the uniform flow equation coefficient Q: bA-

(13.28)

by which reduces, taking partial derivatives,to

o Q:
A

*g:
A

*y

(r3.2e)

Substitutingthis into the continuity equation

Q*4:o
6x At

(13.30)

givesEq. I3.2t if c : mv.If dischargedata ate available,m canbe estimatedfrom are nq. tZig. Valuesfor commonshapechannels given in Table 13'1' The rout{rg can now be done using either constantm and c parameters(i.e., using a single ai'eragevelocity) or variableparameters(usingeachnew velocity v). equition I-3.2i rs solvedfor c, the valueX is derivedfrom Eq. 13.26,andE6' 13.23 to 13.25are solvedusing K : Lx/c. to When using this rnethod,the valuesof Ax and Ar shouldbe selected assure time to peak of inflow detailsare proper$ routed.Nominally' the that the flood wa-ve is broken into 5 or 10 time incrementsAt. To give both temporal and spatialresolution, the total reachlength l, can be dividedinto severalincrementsof Ax length, and outflow from eachis treated as inflow to the next.

244

CHAPTER13

HYDROGRAPH ROUTING

EXAMPLE 13.3 Usethe Muskingum-Cungemethodto route the hydrographfrom Example13.2.Use areaat Q : 59,960is 5996 ft2, So: 0.0001,Lx : 545 mi, flow cross-sectional : 59,960is 60 ft, and Ar : 1.0 day (asin Example13.2). width at Q Solution. From the inflow, the peak rate of 59,960cfs gives

So:T

o^

59.960

,, r-: T - 59,960: V Qo f c : ZVo: 16.7ps From Eq. 13.26

l0fps

":;1,
L* K :

1000 :0.4 (16.7)s4s L (s280) 0.0001

I I

't

7:

545(5280) -ff: 172,800 sec C": -0.1765 Ct: 0'7647 Cz: 0'294I

and

The routing for a portion of the hydrographis as follows:

Date, t

CoQ'*L*

Ct Qti*ro*

Cz Qlutno*

Qli*.*
r9r0 (3-r7) 4440 6900 11,010 16,050 16,050 16p20 29,580 44,360 48,750 44,880 35,730 (3-29) 20,9s0

3-16 17 18 t9 20 2l 22
L3

24 25 26 27 3-28

- 1350 -r970 -2950 -3810 -3700 -4690 -8120 - 10,580 10,190 -8450 -6080 -3820 -6120

3260 5850 8540 12,790 16,510 16,020 20,320 35,180 45,850 44,150 36,620, 26,3sO 16,560

0 560 1310 2030 3240 4720 4720 4980 8700 13,050 14,340 13,200 10,510

Note that the peak outflow of 48,750 cfs on March 26 occurson the samedate as in Example I3.2 but has experiencedslightly greaterattenuationfrom the Muskingum-Cunge example. The value C, is alwayspositive,and negativevaluesof C, arenot particuin this condition should larly troublesome. Although C0is negative this example, valuesof Coare avoided be avoided practice.As seenfrom Eq. 13.23, negative in

ROUTING 245 RESERVOIR 13.2 HYDROLOGIC

when

^|,
Other Methods

,* rr

(13.31)

includingthe working havebeendeveloped, river routing procedures Other hydrolSgic method. method,Tatummethod,and multiple storage R&D method,straddle-stagger They all appear as options in HEC-I, the U.S. Army Corps of Engineer's event in simulation and routing model described Chapter24.

ROUTING RESERVOIR 13.2 HYDROLOGIC

The storageindication method of routing a hydrographthrough a reservoiris also reservoiris through a storage A calledthe modified Pulsmethod.8 flood wavepassing both delayedand attenuatedas it enters and spreadsover the pool surface.Water storedin the reservoiris gradually releasedas pipe flow through turbines or outlet spillway. or works, calledprincipal sprllways, in extremefloods, over an emergency from Flow over an ungatedemergencyspillway weir sectioncan be described by energy,momentum,and continuity considerations the form O: CYH'

(r3.32)

where O : the outflow rate (cfs) Y : the length of the spillway crest (ft)
H : deepest reservoirdepth abovethe spillway crest (ft) coefficientfor the weir or section,theoretically 3'0 C : the discharge x : exponent, theoretically J pipe is similarly described Eq. 13.32 by Flow through a free outlet discharge where' I H C .r : : : : pipe (ft') the cross-sectionalareaof the discharge head above the free outlet elevation (ft) coefficient, theoretically\/29 the pipe discharge exponent,theoreticallyj

Flow equations for other outlet conditions are availablein hydraulics textbooks. from in Storagevaluesfor variouspool elevations a reservoirare readily determined from topopool areasmeasured of computations volumesconfinedbetweenvarious graphic maps. Since storageand outflow both dependonly on pool elevation,the relation (Eq. 13.32) can curve and the outflow-elevation resulting storage-elivation graph. Storagein a reservoirdepends easily be combinedto form a storage-outflow on to only on the outflow, contrasted the dependence the inflow and outflow in river routing(Eq. 13.3). "surchargestorage"or the storage ,S For convenience, is often defined as the is spillwaycrest.Normally the overflowrate is zero when,S zero. abovethe emergency and ifthe slopeofthe line relation is found to be linear, Ifthe graphedstorage-outflow

246

CHAPTER13

ROUTING HYDROGRAPH

is definedas K, then S : KO (13.33) is a and the reservoiris called a linear reservoir. Routing through a linear reservoir 13.l-usingr: 0'0inEq. 13.3' showninFig. Muskingumriverrouting specialcaseof only while the inflow exceeds rate Note alsothat the,outfl-ow in Fig. tg.t is increasing that assumptions the inflow immewith the is the outflow.This observation conJistent only andthat the outflow depends overthe entirepool surface diatelygoesinto storage on this storage. time by Routing through a linear reservoiris easily accomplished first dividing s2 : Ko2into Eq. I3'4 and into a numberof equal incrementsand then substituting increment' solving for oz, wtrictris the only remainingunknown for eachtime reservoir,the storage-outflow through anonlinear Ro'oa To route an emergency the outflow relation and the continiity equation,Eq. L3'4,are combinedto determine I3'4 canbe rewritten as and storageat the end of'eachtime inciement A/. Equation

( r , + r n +.1 * - o , )

:'+

t on+t

(r3.34)

side'Pairs in which the only unknown for any time incrementis the tgt_*:n the right that satisfy Eq. 13.34 and checked b" generated of trial valuesof S"*, andO,*1"ould "uru" fot confirmation.Rather than resort to this trial procein the storage-outflow outflow curve are replottedas and dure, a valrieof At is selected points on the storage indication" curve shownin Fig. 13.3' This graph allowsa direct determithe ';storage has been a nation of the outflow O,11 o11ce value of the ordinate Zl,*rfLt I On+r from Eq. f :.:4' ftre secondunknown' S,*t, can be read from the S-O calculated 13.3)orfoundfromEq'13'34' curve(whichcouldilsobeplottedonthegraphinFig.
t200 I

/ / {,
: .il< 600

0 4 0 Outflow (cfs)

O' Figure 13.3 Curveof zsl\t + O versus

RESERVOIRROUTING 19.2 HYDROLOGIC

247

using of integration Eq. 13.34with Fig. 13.3is illustrated This row-by-rownumerical : t hr in ExampleI3.4. Ar
EXAMPLE 13.4

inflow hydrographand the 2S/Lt + O curve ofFig. 13.3 Giventhe tria4gular-shaped find the outflow hydrograph for the reservoir assumingit to be completely full at the of beginning the storm.(SeeTable 13.4.)

o 90

o Time (hr)

In selecting routing period Ar, generallyat leastfive points on the rising limb a number of of the inflow hydrographare employedin the calculations.An increased the accuracy,sinceas A/ - 0 points on the rising limb, that is, a small A/, improves the the numerical integrationapproaches true limit of the function being integrated, in this casedSfdt. Column 3 in Table 13.4 comesfrom the given inflow hydrograph,column 4 is simply the addition of I, + In*r, andColumns5 andT are initially zero, sincein this of problem the reservoir is assumedfull at the commencement inflow. Therefore, there is no availablestorage.
TABLE 13.4 ROUTINGTABLE (1) Time (hr) 0 I z 3
+

(2)

(3)
ln

(4) I, + In+l (cfs)

-tt,

'n-o.
(5)

(6)
2Sn+1 , a -;fT vn+1

(7)
an+t

(8)
Sn+t

(cfs)

(cfs) 0 20 74 160 284 450 664 853 948 953 870 746 630

(cfs)
JU

(cfs)

(cfs-hr)

5 6 '7 8 9 t0 ll t2

1 2 3 4 5 6 7 8 9 10 ll 1 t

0 0 0 0 120, 150 180 135 9 0 45 0 2 0 3 0 3 6 9

90 150 2lo 270 330 315 225 135 0 0 0

110 224 370

))4

5 18
3Z AA

52
)6

780 979

r078 1085 998 870 746 630

63 65 65 64 62 58 54

12.5 46 96 164 250 361 458 506 510 467 404 344 288

248

ROUTING 13 CHAPTER HYDROGRAPH the sum of columns4 The starting value for n : 1 in column 6 is computedas and 5 from F,q.13.34 (L + 12)- ( * - o ' )

25" - + o2 Lt 25. 3 0 * 0 : - - + o2

Lt

column 6 gives a valuefor Enteringthe ordinateof Fig. 13.3 with the value 30 from end-of-time-interval o, of 5 cfs, which is recordedin column 7 , The corresponding to columns6 and7 and recordedin column 8. Moving s,, storage, is calculatedfrom n : 2 usings' be found fof the secondfow, a value of the term in column 5 can now and O, from columns7 and 8' ThestepwiseprocedureusedtogetoutflowfiguresforallncanbeSumma. rized as l.Entriesincolumnsland3areknownfromthegiveninflow.hydrograph. column 3' - 2. Entries in column 4 arc the additionsof I' t I'*' in 3.Theinitialvalueofthetermincolumn5iszero,thoughitcouldalsobe 4 and 5 are based on any arbitrary starting storagevalue' and columns addedto producethe value in column 6' 4.The2SlA,t+oversusoplotisenteredwithknownvaluesotzSlLt+o to find valuesof O for column 7' 5.Columns6andlaresolvedforS,+r,whichisrecordedincolumn8. for column 5 usingthe 6. Advanceto the next row and calculatethe next value 7 and 8' valuesin the precedingrow for O and S {om columns sum in column 4 and enterthe 7. Add the value in cotuin 5 to the advanced result in column 6 for the new period under consideration' 8.ThenewoutflowforcolumnTisagainfoundfromtherelationof ZSlLt + O as in Fig. 13.3. g.Thecorrespondingnewstorageincolumn8isfoundbysolvingfrom columns6 andT' l.0.Steps6throughgarerepeateduntiltheentireoutflowhydrographisgenerated. rl

ROUTING RIVER 13.3 HYDRAULIC

ploys both the equationof continuity and utions to the complete hydrarllic routing

ili"'nTil::ll;lffi ffi::: ffffii.;fi

tions. Hydraulicfoutingtechniquesarehelpfulinsolvingriverroutingproblems,oversolution the simultaneous land flow, or sheetnow. uyaraulic routing proceedsfrom 1'negeneralforms of the combinationfor of of expression, continoiti una-o*"ntuir, fl'ow equations' rivers are called the spatically varied unsteady

13.3 HYDRAULIC ROUTING 249 RIVER Theseequationsalso apply to sheetflow or overland flow and include terms for laterally incoming rainfall. They can be simplified and used to resolveriver routing problems.e For completeneis presentation, generalform of the spatially varied of a unsteady flow equations will be presented first.

Equationof Continuity
The equation of continuity statesthat inflow minus outflow equals the changein storage. relatethis conceptto a river sectionundera condition ofrainfall or lateral To inflow, consideran elementof length Ax and unit width into the page as shown in Fig. 13.4. The total inflow is

,,*. 35) ,(, - X+)(, - * *) * * , [.' l,'*o' r)dtdx (13

The total outflow is

+ o(vt*+)(,.#?) o'
The changein storageis

(13.36)

* ofrl,"
l(x, t)

(r3.37)

,(,#+)(,l

o ( v * o t) ( r . * + ) ?
f
A x L (a) x

2--*l-- 2-

i(x, t)

ft) Figure 13.4 Continuity and momentum elements(where p : the density of water, V : the averagevelocity. 1l : the depth. i : the lateral inflow per elemental Ax, and S = the slopeof the river bottom).

250

CHAPTER13

HYDROGRAPH ROUTING

Consequently, continuity gives

-o(r{*tx + vfia') a, + pl A,xL, -

oX L,xL,t: o

(13.38)

where i is the average lateral inflow resulting from rainfall over Ax and Ar. The continuity equation of unsteadyflow with lateral inflow is obtained by simplifyingEq. 13.38 6V .,0y , 0y v-r " 6 x| v - -6 x At :

( 13 .3e)

For otherthan a unit width, Eq. 13.39takesthe form

e d{x + v d x * *d tY
whereA is the width times the depth,y.

l : 0

(13.40)

Momentum Equation
In accordance of with Newton's second of motion, the change momentumper unit law of time on a body is equal to the resultantof all externalforces acting on the body. The following derivationof the momentumequationof spatiallyvariedunsteadyflow (1) is presentedsubjectto the following assumptions: the flow is unidirectional and is velocity uniform acrossthe flow section;(2) the pressure hydrostatic;(3) the slope of the river bottom is relatively small; (4) the Manning formula may be used to evaluate friction lossdueto shearat the channelwall; (5) lateral inflow entersthe the stream with no velocity componentin the direction of flow; and (6) the value of I represents spatial and time variationsof lateral inflow. the Forcesactingon an elementof lengthAx andunit width are shownin Fig. i 3.4b. The forcesF1andF, represent as hydrostaticforceson the elementand are expressed

f-, - atvA) F,: ylyA -;zl axl - - .f-^ a(la)Axl rz: lL,!^A- Z l

(13.41)

(r3.42)

wherey is the distancefrom the watersurface the centroidofthe area.The resultant to hydrostaticforce is F, - Fr, or
\ 1

r ,6 :

-ra(!A) *
x

(13.43)

By assuminga small slope for the river bottom, the gravitational force is given by Fr : yASA.x (13.44)

The frictional force alongthe bottom is equalto the friction slope$multiplied by the weisht of water in an elementAx. \: vASyLx

( 13 .4s )

13.3 HYDRAULICRIVER ROUTING

251

as The rate of changeof momentumin the length Ax may be expressed d(mv)

dt
: iii____ T

dV
dt

,.dm
V____

(r3.46)

dt

in which m is the massof fluid. that the incoming lateral inflow entersthe moving fluid with no If it is assumed velocity componentin the direction of flow and I representsthe spatial and time variationsof the lateral inflow, the rate of changeof momentumfor the elementcan as be expressed

dv dtuv) T:pALxi+pvi
wherc dV/dt represents

Lx

(r3.47)

dv av ,,av
dt at
T v -

Ex

(13.48)

The rate of changeof momentumis therefore

oe*(Ya,. "#) + pviL,x

(r3.4e)

Equating the rate of changeof momentumto all externalforces acting on the elementresults in

{*r{+gaFA)
A t O x A O

+L:s(s-s)
x A

(13.50)

Now for a unit width element,the relation simplifiesto

{u, r{u** tX+Yri : s{s s) : o *

( 1 3 . s) 1

that expressions can be Equations13.50and 13.51form a set of simultaneouS solvedfor V and y subjectto the appropriateboundary conditions.

and Diffusionn DynamicWaves Kinematic,

For the casewith zerclaterclinflow, l, Eq. 13.51can be solvedfor the friction slope
c _ c _ :

Ev

VAV

IAV

(r3.s2)

Friction slope

Bed slope

Kinematic wave Diffusion wave

Water surface slope

Convective acceleration

Temporai accelelation

dynamic wave Fu11

252

ROUTING CHAPTER13 HYDROGRAPH

are flow routing in open channels classifiedas of The three typesof analysis unsteady depending kinematic,diffusion (also called noninertia), anddynamicwaveanalyses, differ not only by on which terms in F;q. 13.52 are retained. The three techniques regarding flow including different terms of Eq. 13.52 but also in the assumptions 13.5 showstheseassumpconditions for satisfyingthe momentum equation.Table tions.
HYDRAULIC USED 13.5 ASSUMPTIONS IN VARIOUS TABLE METHODS ROUTING
Method Kinematic Diffusion Full dynamic Commonflow condition
Steady Steady Unsteady

profile Watersurface
Uniform Nonuniform Nonuniform

Steadyflow is definedas flow that doesnot changewith time, and uniform flow is flow with a water surfaceparalleling the bed slope.For steadyuniform flow, the loops. Steady curve)is a singlecurve without hysteresis rating curve (stage-discharge but nonuniform flow has constantdischarge varying water surfaceslopesuchas that found at the entranceto a reservoiror at the approachof a waterfall. One way of selectingthe applicablemethod is to examinethe rating curve and whether it is the same for rising and falling stages.The choice of routing assess equation dependson whether the difference is small (kinematic), relatively large (dynamic),or somewhere in-between(diffusion). that the inertia terms of Eq. 13.52 ate The kinematic wave method assumes negligibleand that the friction slopeequalsthe bed slopeS. Momentumconservation by by is approxi(nated assumingsteadyuniform flow, and routing is accomplished combiningthe continuity equationwith any form of friction lossequation.Typically, either the Manning equationor Ch6zyequationis used.The Chdzyequationis
V : C\/RS

( 13.s 3)

and the Manning equationis

V : l'486 n O2/3St/2

( 13.s 4)

where C andn arefriction coefficients,S is the friction slope,and R is the hydraulic if radius (areadividedby wettedperimeter).Both give velocity in feet per second area and wetted qerimeterare input using squarefeet and feet units. *of these equations can be substitutedinto the kinematic portion of Either Eq. 13.52,equatingslopeof energygradeline with bed slopeto accountfor momentum. The continuity equation,Eq. 13.39,for this casereducesto dQ *64:o At Ex The Manning or Ch6zyequationhas the form Q: b'4^

( 13.5s )

(13.56)

ROUTING 253 RIVER 13.3 HYDRAULIC

is which, after taking derivatives, 6Q = 6*4*-t AA Multiplying Eq. 13.55by dQ/dA gives
: m--a : M V A

(r3.57)

, ,a Q mv -dx

+ 9t : o a

( 13.s 8)

or. if c - mV,the kinematic routing equationis

. dQ * 9 : o - x
At

(13.se)

":dQ:ldQ dA BdY

(13.60)

to the slope of the where B is the top width of the channel.Thus celerity is related that c is constantand assume Most applications rating curve,which varieswith stage.

9 * , 9 : dd{x9 dx At

(13.61)

accounts for the The left-hand side is the kinematic wave equation and the right diffusion d is diffusion effect of nonuniform water surfaceproflles. The hydraulic given by " 2 5
d : -

(r3.62)

This term reveals where4 is the flow per unit width of channel,and s is the bed slope. (resultingin small d) is why kinematiowaveanalysis valid whenbed slopesare steep the channells extremelywide (resultingin small 4)' For flat bed slopes, or when the hydraulic diffusion coefflcientis particularly important' Numerical solutionsof Eq. 13.6I ate presentedby severalinvestigators'12-ra Thesenormally involve substitutionof the relation

ao
A--

- B3 v +
dt

(13.63)

254

CHAPTER13

HYDROGRAPH ROUTING

into Eq. 13.61,whereB is the channeltop width. The Manning or Ch6zyequationis usedfor the friction slope,where

tr:#

(r3.64)

K in which the conveyance is Q/\/Srfromeither equation.Diffusion wavesapply to a wider range of problems than kinematic formulations, but their use may not be it warrantedbecause requiresabout the sameeffort as dynamic routing. for accounts all terms inBq.13.52, includingthe The third type of waveanalysis "dynamic" or It nonuniform,unsteady, and inertia components. is referredto as the "full dynamic" formulation. Dynamic wave solutionsare far more complicatedbut for are often necessary analysisof flow along very flat slopes,flow into large reservoirs, highly unsteadydam-breakflood waves,or reversing(e.g.,tidal) flows. These on conditions are often encountered coastalplains. As a generalrule, full dynamic when wave analysisbecomesnecessary

s>
where

S - bed slope T o : time to peak (sec) of the inflow hydrograph D : averageflow depth (ft) o : eravitationalacceleration
d

1 5 F 4\;

( 13 .6s )

SCSAtt-Kin TR-20Method
method(Section13.1)with themodifiedatt-kin the In 1983,the SCSreplaced convex (attematron-kinematic)methodas the agency'spreferredchannelrouting method.ls The 1.964SCS TR-20 (Chapter 24) single-eventsimulation model used the convex modified to route by the att-kin method. method but was subsequently indication and kinematicwavemethods. The procedureis a blend of the storage Figure 13.5 showsthe two-step processof simulating attenuationfirst by meansof

o PP

Flgure 13.5 Routing principlesusedin the SCS att-kin method,

RIVER r3.3 HYDRAULIC ROUTING 255 routing and then translatingthe wavein time by the kinematic wavemethod storage in to accountfor the fact that routed flow ratesnot only decrease magnitudebut also require time to traversethe length of the routing reach.The storagerouting portion providesattenuationwith instantaneous translation,and the kinematic waverouting providestranslationand distortionbut doesnot attenuate peak. Both are needed the to produce the desired effect. The previously rhentioned full dynamic equations simultaneously accountfor both effectsbut are difficult to solve. The samevolume V1of waterflowing Figure 13.5helpsto visualizethe process. reservoirduring into the reachduring time /, would flow out of a hypotheticalstorage by interval 12.This samevolume would translateand distort downstream kinematic actibn, flowing out of the reachduring interval /r. and selectionof routing coefficients,the Through its theoretical development att-kin methodequationssatisfythe physicalpropagationand timing of the peak flow (areasunder the three hydrographs of rate first. Conservation massis also assured of Fig. 13.5are equal). thentranslates through storage, routesthe inflow hydrograph The actualprocess the peak flow rate, without attenuation,to its final time location in the outflow equal to that correhydrograph. The location in time of the peak outflow is assumed of spondingto the maximum storagein the reachduring passage the flood. Becausecelerity changeswith storage,the other flows of the storage-routed hydrographare translated,pachby a different celerity,to their respectivefinal times and values. by The storageindication routing is accomplished substitutionof the relation Q: KS^

(13.66)

into the continuity equation, Eq. 13.30, where S is the storageand K and m are coefficients.Kinematic routing solvesthe unsteadyflow equation(13.59) with Q: bA*

(r3.61)

area.If Z is the length whereb andm are input coefficients,and A is cross-sectional areathroughoutL is relativelyconstant,the ofrouiing reach,andifthe cross-sectional storageis given by S: LA (13.68)

Theseequatiohsare combined in an iterative fashion to assurethat the peak flow routing, resultingfrom the kinematic routing equalsthe peak resulting from storage and simultaneously ensuringthat the time of the peak outflow occurs at the time of maximum storagein the reach,or Qo: KS;

(13.69)

of Input to the methodrequiresselectionof a reachlength and estimates b andm for 13.1and Eq. 13.27,m canbe shownto be for usein Eq. 13.59.As discussed Table velocity (under bankfull conditions)with wave celerity, or a factor relating average

(13.70)

256

CHAPTER13

ROUTING HYDROGRAPH

the The larger ru becomes, shorterthe travel time, A value of m I 1.0 would incorby flow velocity.Studies SCSresulted rectly makethe celerity slowerthan the average Signif,canterrors resultedfor m values in a recommendationof ! for general use. greater than2.0. Equation 13.67 is appropriatefor cross sectionshaving a single but ihannel with regulaishape.Complexcrosssectionsare more diffcult to evaluate, table for the stream'rs la valuescan be developid from a rating attenuationof the peak flow increasesdue to As the coefficient b decreases, by in storage the reach.The value b canbe estimated reducedvelocity and increased fitting the linear form of Eq. 13.67 (referlo plotting Q andA on log-logpaper and faatei6.+1. The slope*oUa be m andtheinterceptut 4_: 1 would be b. The SCS for nomographs estimatingb and m'" has also developed As a generalguideline,the reach length L shouldbe increasedto a value that time increment,or than the selected resultsin a kinematic wavetravel time c greater
Lo> cA,t
Z*io ft LR ft 18,000

mV = wave celerity
l*ln LR

9,000 7,500 15,000

6,000 5,000

= minimum acceptablereach length = minimum recommended reach length MainTime Increment Hours

12,000 10,000 7,000 5,000 4,000 3,000 2,000 1,500 1.000

mV ft/sec

0.5 0.7

3,500 r <nn ? nnn

2
I

1,500 1,000 750

0.5
n?5

0.3
0.2'

3 /
5 7 10 T2 15

0.1 0.05

350 700 250 500

Example: mV = 4ftlsec Main time increment = 0.2 hr L*n= 1,450ft La = 2,900 ft

t00 reach length for Figure 13.6 SCS nomographfor determining atl-kin method of routing. (After U.S. Soil ConservationService, "ComputerProgramfor Project Formulation," TechnicalRelease AppendixG, 1983.) 20, Revised,

PROBLEMS 257 The length' c : mV' and Ar is the time increment' where Lo is the recommended minimumrecommendedLoisthatgivingawavetraveltimeequaltoabouthalfthe lengthy iray iesult in analytical difficulty when rime incremenr. This 1j[$;"1* inflowhydrographsorsteepStreams.areencountered.Italsoresultsinthepeak were time incrementAr. If severalreaches outflow time being ,ourJ"J"rp io the full a reachlength between accumulate'Thus routed, this incrementattime^errorwould but a length greater thal c. A/ is recommended' c Lt andc Lt/2 is acceptable' Figure13.6providesth"rangeofminimumacc"eptableandminimumrecommended routing reachlengths.

ff "u.n ff Ji,ffiffi ;'il;;'n"'uu"'i11,, ;t'1 ":".1-':^:"::::T,it?ii: ffi :3:[";ffi;i;;;ffi;"'"-*:i"o1j.u""]'mJ1't^:-:T:1T':1"::i"%i"$1 presented , i. wru,iil

iir"T-l"ii,'.i,"n" ;;lrt':?
HX"J"H;il;il#il;;;;;,;;i;#

hvdraulicroutingtec|1ioue11p.p"11'i^'"Tt*::?:Yi:: ff ffi ff;;iicationsof

ll :^11ti:: understand an 1"*Tl student iil::1,:: that interesred can

modeling' of the structuringprocesses hydrologic

Summary
ComputersoftwareforhydrographsynthesiS^androutingisavailablefromnumerous federal agelcy routines are detailed in public and private vendors. wid;ry used includesone or amflow processes Chapter24. Virtually every ( m o r e h y d r o g r a p h r o a s H E C - I ( C h a hydroloeig a1d hydraulic d e p t e r 2 4 ) , p r o v i lplied severalchoices.In I the models can be found in the routing procedures' suggestions g]]:i,:Ytditg .T" At 1iterarure.,6,17 sel ln The readerwho will engage rlver or when eachof the methodsshouldbe applied' reservoirroutingisencouragedtodwelopthecomparisonrequestedinProblem13.25 before leavingthis chaPter'

PROBLEMS
techniques' betweenhydrologic and hydraulic routing 13.1. Discussthe main diff-erences 13.2.

equation' river TheMuskingum routing ?:,:,':'?,*

t+}i ;q":: ;;:..ii+l q- 191 "'*ifi:'i I :il# 1iT .'SX,ffff ffi?#t r" :, Iit:t?;;i,Y -Yllli ilY?iBi'"".U'1-'"d;i"?';:fli!11iy":;T"':'.."t"jfl i:;#"J I"a'^#i'.h;r::;;'i;A--)-r",t'i'i""1'"'lT1,1l.o:'^"iui:.'* i:il3iil;
v\v p:.'i"l' the or time *u,'::9,":1i: 4-_ of the Siif#ffi"#ru:=#;"i"e-i*inetlme oerro.; i'||w 12' v2' .xg 9z lT"jll outflow, and storage at the beginning ;;:E;: n"t*#:J:'::3*4":::i3";ffi';

u,,t'l uutu", corresponding G :fd;#?;##ir*"p"'i"a'lndAsistl"*"1c-",'istorage'Perrorm "and ffi:Tt"'ff for Cs' C1' and C2'
A",inution the described verify ttre equations

l"3.3.IftheMuslongumKvalueis12hrforareachofariver,andiftheXvalueis0.2'what value of Ar for routing purposes? would be a reasonable

258

ROUTING, 13 CHAPTER HYDROGRAPH 13.4. A river reachhas a storagerelation given by S; : ali + boi. Derive a routing equation for O2 analogousto the Muskingum equation (13.6). Give equationsfor the coefficientsof 11,01, and 12. 13.5. List the steps(starting with a measuredinflow and outflow hydrographfor a river reach)necessaryto determinethe Muskingum K and X values. If the inflow and outflow are recordedin cubic feet per second,statethe units that would result for K and X if your list of stepsis followed. 13.6. Given the following inflow hydrograph:

'

lnflow (cfs) 6 e.v. Noon 6 p.tu. Midnight 6 A.M. Noon 6 p.Iu. Midnight

Outflow (cfs) 100

100 300 680 500 400 310 230 100

Assumethat the outflow hydrographat a section 3-mi downstreamis desired. : a. compute the outflow hydrographby the Muskingum methodusing valuesof K : 0.13. 11 hr andX b. Plot the inflow and outflow hydrographs on a single graph' c. Repeatsteps(a) and (b) usingX : 0.00. at discharges both endsof.a30-mi river reach: I3,7. Giventhe following valuesof measured a. Determinethe Muskingum K and X valuesfor this reach. b. Holding K constant (at your determined value), use the given inflow hydrograph to determineand plot three outflow hydrographsfor valuesofX equal to the computed value.0.5. and 0.0. Plot the actualoutflow and nurnericallycomparethe root mean squareof residualswhen each of the three calculatedhydrographsis compared outflow. with the measured

Time
6 e.Ira. Noon 6 p.v, Midftight 6 l.rra. Noon 6 p.tvt. Midnight 6 n.n. Noon 6 p.v. Midnight 6 A.M.

lnflow (cfs)

Outflow (cfs) 10 12.9 26.5 43.1 44.9 41.3 35.3 27.7 t9.4 1 5I. 12.7 11 . 5 10.8

l0 30 68 50 40 3l
ZJ

10 10 10 10 t0 10

PROBLEMS 259 regionthat hasrunoffrecords' 13.8. Selecta streamin your geographic X' method of routing to find K and Usethe Muskingum

l3.g.PrecipitationbeganatnoononJune14andcausedafloodhydrographinastredrn.As thehydrographpassed,thefollowingmeasuredstreamflowdataatcrosssectionsA and B were obtained:

Inflow, SectionA (cfs)

Time June14-17
6 e.v. Noon 6 p.tvt. Midnight 6 e.u. Noon 6 p.u. Midnight 6 n.u. Noon 6 p.tvt. Midnight 6 e.u. Noon

Outflow, B Section (cfs) 10 10 t3 26 43 4l 35 28 19

IJ
I J

10 10 30 10 50 40 30 20 10 l0 l0 10 10 10

1l 10

for the a. Determine the Muskingum K and X values b. DeterminethehydrographatSectionBifadifferentstormproducedthefollowing hydrographat SectionA:
lnflow (cfs)

river reach'

Time 6 e.u. Noon

6 p.tt. Midnight 6 l.rra.

Time
Noon 6 p.tvt. Midnlght 6 e.u. Noon

lnflow (cfs) 400 300 200 100 100

100 100 200 500 600

emergencyspillway of a certain 13.10. The outflow rate (cts) and storage(cfs-hr) for an units of hours'Use

: Sl3,wherethe number3 has reservoirare linearly t"fui"a Uy d = I tt + sr - or Ltl2to determine the continuity equationss2 + 02 Ltlz this and the following inflow event: itr" p""f. outflow rad fr;m the reservoir for Time (h0 0
L

I (cfs)

o
(cfs)

(cfs-hr)

4 6 8

0 400 600 200 0

260

CHAPTER13 HYDROGRAPH ROUTING ' 13.11. A simple reservoir has a linear storage-indication curve defined by the equation

" : 2* .
whereAr is equalto 10 hr. If s at 8 e.u. is 0 cfs-hr,usethe continuity equationto route the following hydrograph through the reservoir:
Time I (cf9
8 a.lra.

9 a.u. 200

10 e.Ira. 400

1l A.M.

200

Noon 0

I P.M.

13.12. For a vertical-walled reservoir with a surface area A show how the two routing equations(73.32 and 13.34) could be written to contain only o2, sz, and known, values(computedfrom or, s,, and so on). Eliminate .FIfrom all the equations.How could thesetwo equationsbe solvedfor the two unknowns? 13.13. Given: Vertical-walled reservoir, surface area: 1000 acres; emergency.spillway width : 97.l ft (ideal spillway);H : watersurfaceelevation(ft) abovethe spillway crest; and initial inflow and outflow are both 100 cfs. a. In acre-ft and cfs-days, determinethe valuesfor reservoirstorage corresponding S to the followingvaluesof I1..0, 0.5, l, 1.5,2,3, 4 ft. b. Determine the values of the emergencyspillway Q correspondingto the depths namedin part a. c. Carefully plot and label the discharge-stcirage curve (cfs versuscfs-days)and the storage-indication curve (cfsversuscfs,Fig. 13.3) on rectangularcoordinategraph paper. d. Determinethe outflow ratesoverthe spillwayat the endsof successive dayscorrespondingto the following inflow rates(instantaneous ratesat the endsof successive days):100, 400, 1200,1500, 1100, 700, 400, 300, 200, 100, 100, 100. Use a routing table similar to the one used in Example 13.4 and continue the rotating procedureuntil the outflow drops below 10 cfs. e. Plot the inflow and outflow hydrographson a single graph. Where should these curvescross? 13.14. Routethe given inflow hydrograph through the reservoir by assumingfhe initial water level is at the emergencyspillway level (1160 ft) and that the principal spillway is plugged with debris. The reservoir has a 500-ft-wide ideal emergency spillway (C = 3.0) locatedat the 1160-ft elevation.Storage-area-elevation data are Elevation (ft) 1l{0 1120 1140 I 158 I 160 1162 1164 I 166 I 168 1180
Area of pool (ft2x 106)

Storage (ft3x 106) 0 4.25 50.25 172.15 r92.75 2r5.35 239.95 266.5s 295.20 528.55

0 0.85 3.75 9.8 10.8 I 1.8 12.8 13.8 14.85 25,0

PROBLEMS dataare The inflow hYdrograPh

I (cfs)

261

Time(h0 0.0 0.5 1.0 1.5 2.0 2.5 3.0

J.)

0 3,630 10,920 to,'720

5 010

4.0
A <

1,600 460 100 10 0

table. a.Findthel5-minunithydrographbySnyder'smethod.

the 30 and 1 of for nrst min 063 l i"li:l! ff;ru#iHfllSiii.?om 8in. rain
.. il"i:i.,i[T;iitf;,ersus

the of period 15minand a using routing o.cuwe

curvesprovloeo' outflow and storage to the bottom it the reservoirassuming is full e. Route,n" uJ#f,iu.;;;n";;"dh elevation980' of the sPillwaY height of water^inthe reservor' Indicate maxi-mum f. plu'"d to obtain 5 ft of freeboard? "f th;;;;;; g. At what "tJ;;;;iJ'n"'op Total storage
104(ft3)

Elevation (ft)

lncremental storage 1oo(ft") 40

960 970 980 990 1000

0 40 250 840 2080

zto
590 1240

the reservoirinitially empty' 13.16. RepeatProblem 13'15 with l3,IT.AfloodhydrographistoberoutedtytheMuskingummethodthroughalO-mireach ro-ml fvl,r;ach be divided in with K = Zhr.tnto how *uny ,ubr"u"h"*-**rirt" Kl3 < Lt < Kl ;;J stitt satisry ttre ,tuuitity criterid order to or" lr ]'6.i;

262

CHAPTER HYDROGRAPH 13 ROUTING 13.18. RepeatProblem 13.6aby dividing the 3-mi reach into two subreaches with equal K valuesof 5.5 hr. Comparethe results. 13.19. Discussthe problemsassociated with the useof a reservoirrouting techniquesuchas the storage-indication method in routing a flood through a river reach. 13.20. VerifyEq. 13.51. t3.21. Precipitationbeganat noon on June14 and caused flood hydrograph a stream.As a in the storm passed,the following streamflow data at cross sectionsA and B were obtained:

Time June14-17
6 .q,.N{. Noon 6 p.u. Midnight 6 .r.lr. Noon 6 p.lr. Midnight 6 a.Ira. Noon 6 p.rra. Midnight

lnflow SectionA (cfs)

Outflow B Section (cfs) 10 l0 13 26 43

+l

6 e.u.
Noon

l0 10 30 70 50 40 30 20 10 10 l0 t0 l0 10

35 28 t9 l5 l3
ll

10

a. Determinethe Muskingum K and X valuesfor the river reach. b. Determine the hydrograph at Section B if a different storm produced the following hydrograph SectionA (continuecomputations at until outflow falls below 101 cfs):

Time 6 a.u.
Noon 6 p.Ira. Midnight 6 a.Ira.

Inflow (cfs)

Time (cont.)
Noon 6 p.u. Midnight 6 a.Ira. Noon

lnflow (cfs)

100 100 200 500 600

400 300 200 100 100

13.22. If the MuskingumK valueis 12 hr for a reachof a river, and if the X valueis 0.2, what would be a reasonable value of Ar for routing purposes?

PROBLEMS 263 at discharges both endsof a 30-mi river reach: 13.23. Giventhe following valuesof measured

Time 6 n.u. Noon 6 p.rra. Midnight 61.v. Noon 6 p.Ira. Midnight 6 n.rra. Noon 6 p.tu. Midnight 6 e.lr.

lnflow (cfs)

Outflow (cfs)

10 30 68 50 40 3l 23 10 10 10 10 10 10

10 12.9 26.5 43.r 44.9 4t.3 35.3 27.7 t9.4 15.1 12.7 I 1.5 10.8

a. Determinethe Muskingum K and X valuesfor this reach' b.HoldingKconstant(atyourdeterminedvalue),usethegiveninflowhydrographto for andptot tt ree outflow hydrographs valuesofX equalto the computed determine value,0.5, and 0.0. 13.24. Given the following inflow hydrograph:

lnflow (cfs) 6 e.v. Noon 6 p.v. Midnight 6 e.l"l. Noon 6 p.v. Midnight

Outflow (cfs) 10

10 30 68 50 40 3l
ZJ

10

Assumethat the outflow hydrographat a section3-mi downstreamis desired' using values of a. Compute the outflow hydrograptrby the Muskingum method K - 1 1 h r a n dX : 0 . 1 3 ' on b. Plot the inflow and outflow hydrographs a singlegraph' : 0'00' c. RepeatSteps(a) and (b) usingX textbookfor all refert3,25. Carefully review the chapterand consult one other hydrology presented' Compile the results toipplicability ofeach ofthe routing procedures ences retained and consultedfrequently' into a list or table, ihis taUleshouldbe

264

ROUTING 13 CHAPTER HYDROGRAPH

REFERENCES
1. Graeff, "Trait6 d'hydraulique." Paris,1883,pp. 438-443. 2. Y. T. Chow, Open ChannelHydraulics. New York: McGraw-Hill , 1959. 3. U. S. Soil ConservationServiie, National EngineeringHandbook, Notice NEH 4-102. PrintingOffice,August 1972. D.C.: U.S. Government Washington, PrevenDisaster Bulletin 1. Kyoto,Japan: 4. S. Hayami, On the Propagationof FloodWaves, t i o n I n s t i t u t e1 9 5 1 . , "Multiple Linearization Flow Routing Model," Proc. 5. T. N. Keefer and R. S. McQuivey, ASCEL Hyd. Div.100(HY7) (Iuly 1974). "On the Subjectof a Flood Propagation Method," J. Hyd. Res'IAHRT(2), 6. J. A. Cunge, 20s-230(1967). "KinematicFlood Routing,"Trans.ASCE L0(3) (1967). 7 . D. L. Brakensiek, "Water Studies,"Bureauof ReclamationManual, Yol. 8. U. S. Departmentof the Interior, D.C.: U. S. GovernmentPrinting Offrce, 1941. IV Sec. 6.10. Washington, "Numerical Techniques Spatially VariedUnsteadyFlow," University for 9. T. E. Harbaugh, Center,Rept. No. 3, 1967. of Missouri Water Resources 10. R. K. Price, "Comparisonof Four Numerical Methods for Flood Routing," Proc. ASCE J. Hyd. Div. 100(HY7) (July 1974). "NonlinearKinematicWaveApproximation and M. A. Stevens, 11. R. M. Li, D. B. Simons, Res. for WaterRouting,"WaterResources AGU II(2) (Apr. 1975). "WavePropagation Rivers," HEL Ser. 8' No. 1. in 12. J. A. Harder and L. V. Armacost, Hydraulic Engineering Laboratory, College of Engineering,University of California, Berkeley,June 1966. "Numerical Solution of UnsteadyFlows in Open 13. D. J. Gunaratnamand F. E. Perkins, MA, July 1970. Rept. 127,MIT, Cambridge, Hydrodynamics Channels" DiSilvio, "Flood Wave Modiflcation Along Channels," Proc. ASCE J. Hyd. Div. 14. G. 9s(HY7) (1e69). "Simplifled Dam-Breach 15. U.S. Department of Agriculture, Soil ConservationService, 66,Mar. 1979. Tech.Release RoutingProcedure," "ComparativeAnalysis of Flood Routing Methods," Research Doc. 16. Streldoff, T., et al., 24,Hydrologic EngineeringCenter,U. S. Army Corps of Engineers'Davis, CA, 1980. "Guidelinesfor Calculatingand Routinga Dam-Break l7 . D . L. Gunlachand W. A. Thomas, Center,U. S. Army Corpsof Engineers, Note 5, HydrologicEngineering Flood," Research 1977. Davis.CA.

C h a p t e r1 4

Snow Hydrology

Prologue
The purposeof this chaPteris to: in . Indicatethe importanceof snowmeltto water supplyand management cold regions. . Describemethodsfor measuringsnowfall and describingits water-producing caPabilities. ' Describethe physicsof snowmelt. . presentmodeisfor estimatingsnowmeltunder variousconditionsof temperaground cover,and snowpack. topography, ture, relativehumidity, wind speed,

14.1 INTRODUCTION
In many regions,snowis the dominantsourceof water supply.Mountainousareasin Goodellhasindicatedthat about90 percentof the year$ the Weit areprime examples. of suppliin the high Llevations the ColoradoRockiesis derivedfrom snowfall.' water high proportions are also likely in the Sierrasof California and numerous Equally t"giotrr m tne Nortfrwest.A significantbut lessershareof the annualwateryield in the Northeastand Lake statesalso originatesas snow.It is important that the hydrologist involvedin the the understand natureand distributionof snowfallandthe mechanisms snowmeltprocess. Snowmelt usually beginsin the spring. The runoff derived is normally out of such phase with the periodsof gieatestwaterneed;therefore,variouscontrol schemes to reseivoirshavJ been developed minimize this problem. An additional as storage is point of significanc.e that $omeof the greatestfloods result from combinedlargeon rainstormsand snowmelt.Streamflowforecastingis highly dependent adeicale of knowledgeof the extentand characteristics snowfleldswithin the watershed. quate fhe water yield from snowfall can be increasedby minimizing the vaporization of snow and melt water. Timing the yield can be managedwithin limits by controlling the the rate of snowmelt.Early resultscan be be obtainedby speeding melt process' by or the snowmeltperiod can be extended delayed retarding it. The annual whereas snowfall distribution in the United Statesis shownin Fig. 14'1'

&

-v /

S:.-

I O

) \
,4

00 co

I @

()

!1 a !? o

E
q

{)^ > j

5E
p P
.=.Y

o
t

tu5

AND RUNOFF 14,2 SNOW ACCUMULATION

267

factorsis as much a prerequisite of understanding meteorological An adequate The in consideringthe snowmeltpro""rs as it is in dealingwith evapotranspiration. of water vapor on suppliesmoisturefor both snowfall and condensation atmosphere of the snowpack,regulatesthe exchange energy within a watershed,and is a controlling factor in snowmeltrates. and geologic, topographic, vegetaAs in the rainfall-runoffprocess, geographic, runoff process. tive factors also are operativein the snow accumulation-snowmelt For rainfall-runoff relations, point rainfall measuresare used in estimating areal and time distributions over the basin. A similar approachis taken in snow Mathematical hydrologyalthoughthe point-arealrelationsare usuallymore cornplex. of snowmeltat a given equationscan be usedto determinethe various components of location. Adequatemeasures depth and other snowpackproperties can also be in obtainedat specificlocations.The use of thesemeasurements estimatingamount , and distribution in area and time of snow over large watershedareasis a much less conditions related to particular areal subdivirigorous procedure.Usually, average sionsover time are usedas the foundationfor basin-widehydrologicestimates'Such proceduresare often in the categoryof index methods(Section 14.6). Snowmelt routines have been incorporated in numeroushydrologic models, some of which also include water quality dimensions.A good accountingof the fundamentalsof the snowmelt process and of contemporary snowmelt modeling may be found in Refs.2-15 listed at the end of this chapter. approaches

AND 14.2 SNOWACCUMULATION RUNOFF

in Under the usual conditions encountered regionswith heavy snowfall, the runoff from the snowpackis the last occurrencein a seriesof eventsbeginningwhen the the snowfallreaches ground.The time interval from the start to the end of the process might vary from aslittle asa day or lessto severalmonthsor more. Newly fallen snow of hasa densityof about 10 percent(the percentage snowvolumeits waterequivalent the settling and compactionincrease would occupy),but as the snow depth enlarges, density.ls The temperaturein a deep layer of accumulatedsnow is often well below after prolongedcold periods.When milder weather setsin, melting occurs fueezing flrst atlhe rno*pu"k surface.This initial meltwater moves only slightly below the surfaceand again freezesthrough contact with colder underlying snow.During the from meltwaterraisesthe snowpack rcfreezingpro""rr, the heat of fusion released t"*p"rutur". Heat is also transferredto the snowpackfrom overlying air and the ground. During persistent warm periods, the temperatureof th9 entire snowpack Iontinually rises'and finally reaches32'F. With continued melting, water begins flowing down through the pack. The initial melt component is retained on snow crystals in capillary films. Once the liquid water-holdingcapacity of the snow is the reached, snow is said tobe ripe. Throughoutthe foregoingprocess'pack density due increases to the refreezingof meltwaterand buildup of capillary films. After the the water-holdingcapacityis reached, densityremainsrelatively constantwith continthe ued rnelt. Meltwater that exceeds water-holdingcapacity will continueto move

268

14 CHAPTER SNOW HYDRoLoGY down through the snowpackuntil the ground is finally reached.At this point runoff can occu.r.Three situationsthat may exist at the ground interfacewhen meltwater reaches are described Horton.t6 it by First, considerthe casewherethe melt rate is lessthan the infiltration capacity of the soil. In addition, downward capillary pull of the soil coupled with gravity exceeds samepull of the snow lessgravity. The meltwaterdirectly entersthe soil the and a slushlayer is not formed. The secondcaseoccurs when a soil's infiltration capacity is greaterthan the melt rate, but the net capillary pull of the snowpackexceeds that of the soil aidedby gravity. Capillary water builds up in the overlying snow until equilibrium is reached at which upward and downwardforces balance.A slushlayer forms and providesa supply of water that infiltrates the soil as rapidly as it entersthe slushlayer. The final situationis one in which the melt rate exceeds infiltration capacity. the A slush layer forms and water infiltrates the soil at the infiltration capacity rate. Excesswater acts in a manner analogous surfacerunoff but at a much decreased to overlandflow rate. As warm weathercontinues, melt process maintainedand accelerated the is until the snowcover dissipated. is

14.3 SNOWMEASUREMENTS SURVEYS AND

Snow measurements obtained through the use of standardand recording rain are gauges,seasonal storageprecipitation gauges,snow boards,and snow stakes.Rain gauges usually equippedwith shieldsto reducethe effect of wind.3Snow boards are are about 16 in. square,laid on the snow so that new snowfall which accumulates periodswill be found abovethem. Care must be taken to assure betweenobservation that adverse wind effectsor other conditionsdo not producean erroneoussampleat the gauginglocation.Snow stakes calibratedwoodenpostsdriven into the ground are for periodic observation the snowdepthor insertedinto the snowpack determine to of its depth. Direct measurements snow depth at a single station are generally not very of useful in making estimatesof the distributon over large areas,since the measured depth may be highly unrepresentative because drifting or blowing. To circumvent of this problem, snow-surveying procedures provide havebeendeveloped. Suchsurveys information on the snowdepth,waterequivalent, density,and quality at variouspoints along a snow course.All thesemeasures of direct use to a hydrologist. are The water equivalentis the depth of water that would weigh the sameamount as that of the sample.In this way snow can be described terms of inchesof water. in Density is the percentageof snow volume that would be occupied by its water equivalent.The quality ofthe snow relatesto the ice contentofthe snowpackand is expressed a decimal fraction. It is the ratio of the weight of the ice contentto the as total weight. Snow quality is usually about0.95 exceptduring periodsof rapid melt, when it may drop to 0.70-0.80 or less.The thermal quality of snow,Q,, is the ratio of heatrequiredto producea particular amountof waterfrom the snow,to the quantity

269 CHABACTERISTICS SNOW AND 14.4 POINT AREAL of heatneededto producethe sameamountof melt from pure ice at 32"F.Valuesof The densityof dry snow is temperatures. Q,may exceed100 percent at subfreezing With variability betweensamples. approximately10 percentbut thereis considerable to aging,the densityof snow increases valueson the order of 50 percentor grPater. - A snowcourseincludesa seriesof samplinglocations,normally not fewer than about50- 100ft apartin a geometric 10 in number.ttThe variousstationsare spaced in pattern designed advance.Points are permanentlymarked so that the samelocations will be surveyedeachyear-very important if snow coursememorandaare to be correlatedwith areal snowcoverand depth, expectedrunoff potential, or other significantfactors.Survey dataareobtaineddirectly by forestersand others,by aerial and by automatic recording stations that telemeter photographs and observations, location. information to a central processing In the westernUnited Statesthe Soil ConservationServicecoordinatesmany are and private enterprises also ensnow surveys.Various states,federal agencies, in of gagedin this type of activity. Sources snowsurvey dataaresummarized Ref. 14.

ANDAREALSNOWCHARACTERISTICS 14.4 POINT

The estimationof areal snow depth and water equivalentfrom point measurement data is highly important in hydrologicforecasting.

Estimatesof Areal Distributionof Snowfall

or Normally, taking arithmetic averages using Thiessenpolygonsdoesnot provide g results for estimatin areal snow distribution from point gaugings'This is reliable orographicand topographiceffectsare often pronounced,and gaugingnetbecause frequently are not denseenoughto permit the straightforwarduse of normal works However, regional orographiceffects are relatively constant averagingtechniques. yearto yearand stormto stormfor tractsthat are small whencomparedwith the from areal extent of general storms occurring in the region.2This circumstancepermits in many useful approaches estimating the areal snow distribution once the basic patternhas beenfound for a region' asOne method used to estimatebasin precipitation from point observations the ratio of station precipitation to basin-precipitationis approximately sumesthat constantfor a storm or storms.This can be statedas'
P o _ Nb Po
No PoNu No

(14.1) (r4.2)

or where
D t b --

Pu:

the basin precipitation

P o : the observedprecipitation at a point or group of stations N t : the annual precipitation for the basin N o : the normal annualprecipitation for the control station or stations

27O

CHAPTER14 SNOW HYDROLOGY

from a map (carefullypreparedif it is The normal annualprecipitationis determined displayingthe meanannualisohyetsfor the region.The precipito be representative) tation is determinedby planimeteringareasbetweenthe isohyets.If the number of depict the basin,Eq. I4.2 canprovide stationsusedand their distribution adequately not uniformly distributed,weighting coefficients a good approximation.For stations used of basedon the percentage the basin areaportrayedby a gaugeare sometimes group. in determiningN, for the method.In Another system usedin estimatingareal snowfallis the isopercental as precipitation is expressed a percentage this approach,the storm or annual station on of the normal annualtotal. Isopercentallines are drawn and can be superimposed the (NAP) to producenew isohyetsrepresenting a normal annualprecipitation map storm of interest.A NAP map indicatesthe generalnature of the basin'stopographic from this pattern.The advanthe map shows deviations effects,while the isopercental map directly is that relatively consistageof this methodover preparingan isohyetal tent storm pattern featuresof the NAP can be taken into considerationas well as observedindividual storm variations.

Water Equivalent Estimatesof Basin-Wide

A hydrologistmust be concernednot only with the amount and areal distribution of snowfall, but also with estimatingthe water equivalentof this snowpackover the runoff. Basin water basin, sincein the final analysisit is this factor that determines equivalentmay be given as an index or reported in a quantitative manner such as inchesdepth for the watershed. The customaryprocedurefor determiningthe basin water equivalentis to take data from snowcoursestationsand to provide an index ofbasin conditions. observed to and other approaches weighted averages, Various proceduresemploy averages, of this.2The important point to rememberis that the usefulness any index accomplish the is basedon how well it represents overall basinconditions,not on how favorably it describes particular point value. Indexesdo not actually provide a quantitative a in evaluationofthe property they cover.Instead,they give relativechanges the factor. By introducingadditional data, however,an index can be usedin a prediction equation. For example,if the basin water equivalentcan be estimatedby subtractingthe from the precipitationinput, the index can be correlated runoff and losscomponents with actual basin water equivalentin a quantitativemanner.

Areal Snowcover
Estimatesof the areal distribution of snowfall are very helpful in making hydrologic on A forecasts. knowledgeof actualarealextentof snowcover the groundat any given and in making seasonalvolumetric time is also*applied in hydrograph synthesis are of forecasts ofthe runoff. Observations snowcover generallyobtainedby ground photography. Between snowcoversurveys,approximaand and air reconnaissance data. hydrometeorological are tions of the extentof the snowcover basedon available patternswithin a given basin are normally relatively uniform Snowcoverdepletion from data gathered indexes often be developed can from yearto year; thus snowcover stations. at a few representative

14.5 THE SNOWMELTPROCESS

27'I

PROCESS 14.5 THE SNOWMELT

The snowmeltprocessconvertsice content into water within the snowpack.Rates later. Thesediverfactorsto be discussed differ widely due to variationsin causative drainagefrom the snowpack, gencies not as strikingly apparentwhen considering are however,since the pack itself tends to filter out thesenon-uniformities so that the rate. drainageexhibits a more consistent

for EnergySources Snowmelt

radiation, to The heatnecessary inducesnowmeltis derivedfrom short-andlong-wave ofvapor, convection,air and groundconduction,and rainfall. The most condensation and are important ofthese sources convection,vapor condensation, radiation. Rainfall ranks aboutfourth in importancewhile conductionis usually a negligiblesource.

EnergyBudgetConsiderations
as If snowmeltis considered a heat transferprocess,an energybudgetequationcan of the be written to determine heatequivalent the snowmelt.Suchan equationis of the form2 H^: H,r + H," + H" + H" + Hs + He + Hq (14.3)

where H*: the heat equivalentof snowmelt H4 : rrlt long-wave radiation exchange between the snowpack and surroundings H,, : the absorbedsolar radiation H" : the heat transferredfrom the air by convection H": the latent heat of vaporizationderived from condensation Hr : the heat conductionfrom the ground Ho : the rainfall heat content Hn: the internal energychangein the snowpack In this equationH,", Hs, andHo are all positive;I1,1is usually negativein the open; H" and Hnmay take on positive or negativevalues;andH" is normally positive.The actual amountof melt from a snowpackfor a given total heatenergyis a function of thermal quality. The heatenergyrequiredto producea centimeterof the snowpack's waterfrom pure ice at32"F is 80 langleys(g-cal/cm2).Therefore,203.2langleysarc neededto get 1 in. of runoff from a snowpackof 100 percentthermal quality. If the the term H. represents combinedtotal heat input in langleys,an equationfor snowmelt M in inches.is2
MH-

203.2Q,

(r4.4)

where Q, is the thermal quality of the snowpack.For sirbfreezingsnowpacks,Q,, havingsomewatercontent,Q,is lessthan one.A typical 1; exceeds for ripe snowpacks value for these conditions is reported to be 0.97.6Figure 14.2 gives a graphical solution to this equationfor severalvaluesof Q,.

272

CHAPTER14

SNOW HYDROLOGY

w
0 100 200 300 400 500 600
Net heat flux to snow pack, f(langleys)

Figure 14.2 Snowmelt resulting from thermal energy' (After U.S. Army CoJpsof Engineers.2)

EXAMPLE 14.1 with thermal quality of 0.90, determinethe snowmeltin inchesif Given a snowpack the total input is 137 langleys. Solution. UseEq. 14.4 M: H^/203.2Q, : r37/203.2 x 0.90 M M : 0.75in. rl

Turbulent Exchange is The quantity of heat transferredto a snowpackby convectionand condensation has Suchan approach been equations. from turbulentexchange commonlydetermined mustbe madein and vapor pressure of widely used,sincemeasurements temperature and wind velocity gradithe turbulent zonewherevertical watervapor, temperature, several ents are controlled by the action of eddies.In the following two subsections and practical equationsfor estimatingcondensation convectionmelt are given. Here to a combinedtheoreticalequationis presented acquaintthe readerwith the theory of turbulent exchange. equationcan be written' The basic turbulent exchange

4 o : Adz
where

( 14.s )

property of the air suchaswater Q : thetime rate of flow of a specified horizontal area vapor through a unit dqldT = the vertical gradient of the property

PROCESS 273 14.5 THESNOWMELT

e : the property z : lhe elevation A: an exchange coefficient Property 4 must be unaffectedby the vertical transport.Propertiespertinent to this discussion the air temperature, are water vapor, and wind velocity.Theoretically,the potential temperatureshouldbe used,but air temperatures measured normal disat of tancesabovethe snowpackdo not causeseriouserrors.The potential temperature dry air is that which the air would take if broughtadiabaticallyfrom its actualpressure pressure. to a standard Gradients of the various properties of importance here follow a power law distribution where conditions of atmosphericstability exist.rTThis qualification is Logarithmic profiles are characteristicof the atmosphere'sstate over snowfields.2 more nearly representative neutral or unstableatmospheric conditions.The power of law providesthat the ratio of valuesof a property determinedat two levelsabovethe snow is equivalentto the ratio of the levelsraised to somepower. Thus,
Qz_ Qt

(:31
\Zr,/

/ . \r/n

(r4.6)

where

the value of the property the elevation(with subscripts denotingthe level) power law exponent the

to If 3, is made equal to I, q, assumed be the property value at this height, and the subscript droppedfor the second level,Eq. 14.6becomes
q : qtz\/"

(r4.7)

at The magnitu of q is takenasthe differencein valuesof 4 measured the level de is For example,if T : 38'F at height z, and temperature the z and the snow surface. property of interest,then q: (38 - 32) : 6"p. The gradient of the property dqldz can be obtainedby differentiatingEq. 14.7:

ff: (et),,,_",,"
o : AIL)zo-d/"

(14.8)

equation(14.5),the If this expression substituted the basicturbulentexchange is in followins relation is obtained:

(r4.e)

from Thus,eddyexchange elevation is determined ofthe property at a specified e coefficientis alsorelatedto observations ofthe ptoperty at unity level.The exchange of elevationz. For equilibrium conditions up to the usual levels of measurement gradientsof thesevariablesare suchthat the eddy transfer moistureand temperature, coefficientA mustbe of moistureand heatis constantwith heisht. Then the exchange

274

HYDROLOGY 14 CHAPTER SNOW

related the prope"t r.:O*r,i to inversely - )n,Orr, dqld, h result: the in Substitution Eq. 14.8for dq/dz gives following
A - Ar4"-ttr"

(i4.10)

(14.11)

in since(dq/dz)t : et/nfor I : 1. Now, if the valueof A from Eq. 14.11is inserted Eq.14.9,

n: o,(T)

(r4.r2)

level has been shownto be directly The exchange coefficientat an observation Therefore,it may be proportional to the wind velocity measuredat that elevation.ls written that A': ko1

(14.13)

where o1 : the wind velocity at Level one k : a constantof proportionality for Substituting 41 in Eq. I4.I2 gives

a: \;)a'"'
Using the power law equation(14.7), we find that q, : q;1/"
and D, : 117-1/n

/k\

(r4'r4)
(14.15)
(14.16)

level of o inBq.14.14 and denotingthe observation After making thesesubstitutions thesemay be different),Eq. 14.14becomes as Tu,andthat of q as za(since

n: (!")rr,zu)-'/nq.ou

(r4.r7)

equation. Considerationis now given to turbulent exchange This is a generalized and convectionmelt. specifictheoreticalequationsfor condensation developing melt. The property to be tlansported consider caseof the condensation the First, the coefficientexpresses transfer in this caseis water vapor, and sincethe exchange it to air maqs, is necessary determinethe moisturecontentof the air mass.This of an by can be accomplished using the specifichumidity, which gives the weight of the containedin a unit weight of moist air. Equation 14.18 can be usedto water vapor calculatespecifichumidity: q : 0.622e (14.18)

where e : the vapor pressure of . Po: the total pressure the moist air

PROCESS 275 14.5 THESNOWMELT Inserting this expression in Eq. 14.17 for q"yields2

"u, )-' (9J4:), n" : (I)u"z /^

(r4.re)

," : t s(l)u.,,)-'^(99),"u,

(r4.20)

The proportionality constantk is a complexfunction relatedto the air density and other iactors.Sinceihe densityof air is a function of elevation,k also varieswith of height.The constantmay be madeindependent density,and thereforeof elevation, "introducing a factor to compensatedirectly for the density-elevationrelation' by this, and the equationcan be adjustedby to serves accomplish Atmosphericpressure muttipiying Uy tne ratio pf po, wherepis the pressureat the snowfieldelevationand tevetpr"rrur". Introducingthis ^ratioinBq. 14.20and a new constantk1, po is the se-a gives' which is related to sealevel pressure,

m : a.s(\){r"r,r',' (99),,u, "

{r4.2r)

importanceinEq. 14.17is airtemperature. Forconvectionmelt,thepropertyof into ttreimat units, the specific heat of air crmust b-e To convert air temperature H" introduced. Putting these values in Eq. I4.l7,heat transfer by eddy exchange convertsto2

I/n o, : (I) uoz6)- coror),

(14.22)

cgs Sincethe latent heatof fusion is 80 callg, convectivesnowmeltM" in gramsin the is system givenby H"l8O, or / t \/lc\. (r4.23) ,. : (*J ())(,"20)-'h, or"uu Introducing the elevationdensitycorrectionp/po,we obtain

(l)r,",,r',(o1),,r", * : (#)
M"" =

(14.24)

Equations I4.2I and 14.24 can be^ combined into a single convectionmelt M"" equationof the form' condensation

Lk"ru)-,,"(#fir,r"

(r4.2s)

276

CHAPTER14

SNOW HYDROLOGY

theoreticalequationfor snowmeltthat resultsfrom the turbuleni This is a generalized that the exchange tfansfer of water vapor and heat to the snowpack.It is assumed by coefficientsfor heat and water vapor are equal. Their evaluationis accomplished experimentation.r A combined physical equation of the general nature of Eq. 14.25 has been melt by developed Light.le Widely used,its individual convectionand condensation in componentsare discussed following sections.The combined form of the Light equationis

D :
where

pk',

80 rn(alz) ln(b/zo)

+ ulc,r ( e - 6.IDryf P J

(r4.26)

D - the effective snowmelt(cm/sec) air density L _ von K6rm6n's coefficient= 0.38 ^ 0 parameter: 0,25 the roughness and vaporpressure levelsat which the wind velocity,temperature, r b : the resPectivelY measured, are U : the wind velocity co : the sPecific heat of air 7 :;the air temperature e : the vapor pressureof the air p : the atmosPheric Pressure

Convection
by to Heat for snowmeltis transferredfrom the atmosphere the snowpack convection. processis relatedto temperatureand wind velocity. The amountof snowmeltby this The following equationcan be usedto estimatethe 6-hr depth of snowmeltin inches by convection:2O (r4.27) D: KV(T - 32) where V : the mean wind velocity (mph) Z - the air temperature('F) , a On the basisof the theory of air turbulenceand heat transfer (turbulent exchange), been K coefficient of 0.00184 X 10-0'0000156'has valuefor the exchange theoretical in given by Light.le In this relation h, the elevation feet, is usedto reflect the change is in barometric pressuredue to the difference in altitude. The expression said to of representconditions for an open, level snowfieldwhere measurements wind and abovethe snow.Values are temperature madeat heightsof 50 and 10 ft, respectively, vary of the exprtssion10-00000156' from 1.0 at sea level to 0.70 at 10,000ft of elevation.The actual valuesof K are normally lessthan the theoreticalflgure due to such factors as forest cover. Empirical 6-hr K values have been reported in the literature.20 due for Anderson and Crawfordls give an expression the houily sno.wmelt to convectionas M _

cv(T"_ 32) Q,

04.28)

14,5 THE SNOWMELT PROCESS 277 where M : V : To : Q, : c: the hourly melt (in.) the wind velocity (mi/hr) the surfaceair temperature('F) the snowquality a turbulent exchange coefficientdeterminedempirically

Temperature measurements at 4 ft, with wind gaugedat 15 ft. The corresponding are valueof c is reportedas 0.0002.

Condensation

snow.A total yield of around 8.5 in. of snowmeltincluding the condensate thus is derived. A water vapor supply at the snow surfaceis formed by the turbulent exchange process; consquently, masstransferequationsimilar to thosepresented evapoa for ration studiesfits the melt process. equationfor hourly snowmeltfrom condensaAn tion takesthe formrs

u : W n { r "6 . 1 1 )
where

(14.2e)

b : an empirical constant eo : the vapor pressureof the air (mb), the numerical value 6.11 : the saturationvapor pressure(mb) over ice at 32'F (e, must exceed 6 . 1) l

Also, M, Q,, and v are as previouslydefined.The constantb hasa valueof 0.001 for temperatureand wind measurements 4 and 15 ft, respectively.ls at A similar expression for 6-hr snowmelt(D) is given as but
D:KrV(eo-6.11)

(14.30)

where the theoretical value of K, is said by Light to equal 0.00579 if wind and

kV(e" - e,) ' " Q,

(14.3r)

where E : the hourly evaporationin inches es : the saturationvapor pressure over the snow k : an empirical constant

278

HYDROLOGY 14 CHAPTER SNOW Also, V, eo,andQ,areasdefinedbefore.lsIntheexpressionk : 0.0001,temperature and wind measurements are taken as for Eq. 14.30, and the temperature of the air is assumed equal to that of the snow surface for temperatures below 32oF.

Melt Radiation
radiation receivedby a snowpackcan be a The net amount of short- and long-wave very important source of heat energy for snowmelt. Under clear skies, the most in significantvariables radiationmelt are insolation,reflectivity or albedoof the snow, and air temperature.Humidity effects, while existent, are usually not important. in When cloud cover exists,striking changes the amount of radiation from an open snowfield are in evidence.The general nature of these effects is illustrated in as radiation exchange a function of cloud .2 Fig. 14.3 Combinedshort- and long-wave Radiationmelt is shownto be more significantin the height and coveris represented. spring than in the winter. It should also be noted that winter radiation melt tendsto cloudheightasa resultof the more dominant with cloudcoveranddecreasing increase radiation during that period. role playedby long-wave in Forest canopiesalso exhibit important characteristics regulating radiative from thoseexhibitedby the cloudcover, Theseeffectsdiffer somewhat heatexchange. Cloudsand treesboth limit insoradiation is concerned. especiallywhereshort-wave insolation lation, but cloudsarevery reflective,while a largeamountof the intercepted the by is absorbed the forest.Consequently, forestis warmedand part of the incident radiation; an additional energydirectly transferredto snow irl the form of long-wave fraction is transferredindirectly by air also heatedby the forest. Figure 14.4illustratessomeeffectsof forestcanopyon radiation snowmelt.The In conditions for a coniferouscover in the middle latitudes.2 figure typifies average with completeforest cover, and in winter, the maximum radiation melt is associated shouldnot be spring the greatestradiation melt occursin the open. Generalizations effects of forest cover on drawn from thesecurves,which indicaterelative seasonal Another factor affectingradiation melt is radiation melt for the conditionsdescribed. the land slopeand its aspect(orientation).Radiationreceivedby north-facingslopes for inclinesin the northern hemisphere, example. is lessthan that for south-exposure Solar energy provides an important sourceof heat for snowmelt. Above the the earth's atmosphere, thermal equivalentof solarradiation normal to the radiation 3.97 x 10-3 Btu/cm2).The path is 1.97 lingleys/min (1 langleyis approximately actual amount of radiation reachingthe snowpackis modified by many factorssuch and ofcloudiness,topography, vegetalcover.The importanceofvegetal asthe degree has cover in influencingsnowmelt,long recognized, promptedmany forest managesnowmelt.t'to'"''o to ment schemes regulate Two basic'lawsare applicableto radiation.Planck'slaw statesthat the temperature of a blackbodyis relatedto the spectraldistribution of energythat it radiates. producesStefan'slaw, Integrationof Planck's law for all wavelengths Ro: cT' where R, : the total radiation constant o : Stefan's [0.813 x 10-10langley/(min-K-')] T : the temperature(K)

(r4.32)

14.5 THE SNOWMELTPROCESS

crurar,.ier,iiTth

M,, = 2.0o[1_ (0.82- 0.0244N] Mt=4.4r[r-(1 -}.\UAM

Amount ofclouds, N (a)

M,, = 0.50U- (0.S2 O.O'U4I\tl Mt=4.84[1 * (1 _ 0.0242)t{l

J"*qv;
o u.u 4.2 0.4 0.6 0.8 l.t Amount ofclouds, N (b)

Daily radiation melr in the open with cloudy {igure.fa3 skies: (a) during spring, May 20; and (bf during wintei, February 15. (After U.S. Army Corps of Engineeis.r;

280

CHAPTER14 SNOW HYDROLOGY

t t I I

r.)
^ r _ t "n"n t : t lr I

> 10

\
.= rt \

M,= M^

Mrt

-4
\M,

(r =3.821F .7s7 'F))- 3.30 +

0.6 1.0

-o.5
0.0 0.2 0.4 .'
(a)

F cover, canopy Forest

0.5

(l) ,*^=l,r*
I

H o.o
-0.5

'--1

Mf

Mr, + Mr1

1-1.0
0.0

Mrt

? 'qTF r

-n 7s7Q. - 3.30
0.6 0.8

0.4
(b)

F cover, canopy Forest

melt in the forestwith clear - Figure 14.4 Daily radiation skies:(a) duringspring,May 20; and (b) duringwinter, 15. February (AfterU.S.Army Corpsof Engineers.2) Becausesnow radiates as a blackbody, the amount of radiation is related to its temperature(Planck's law), and total energyradiated is accordingto Stefan'slaw. Long-waveradiation by a snowpackis determinedin a complexfashionthrough the interactionsof temperature,forest cover, and cloud conditions'

SNOWMELT PROCESS 281' 14.5 THE Direct solar short-wave radiation receivedat the snow surfaceis not all transferredto sensible heat.Part of the radiationis reflectedand thus lost for melt purposes. Short-wavereflection is known as albedo and ranges from about 40 percent for melting snow late in the season approximately80 percent for newly fallen snow. to This Valuesas high as 90 percenthavealsobeenrepofted in severalcases.22 property of the snowpackto reflect large fractions of the insolation explainswhy the covers persistand air temperatures remain low during clear, sunny,winter periods. That portion of short-wave radiation not reflectedand availablefor snowmelt may becomelong-wave radiation or be conductedwithin the snowpack.Some heat may also be absorbed the ground with no resultantmelt if the ground is frozen. by An expression hourly short-wave for radiation snowmeltis given as2
MH^

203.2Q,

(r4.33)

radiation (langleys) H^ : the net absorbed where ' 203.2 : a conversionfactor for changinglangleysto inchesof water betweenthe snowcover radiation is exchanged When the snowquality is 1, long-wave and its surroundings.Snowmelt from net positive long-wave radiation follows Eq. 14.33.If the net long-waveradiation is negative (back radiation), there is art equivalentheat loss from the snowpack. An approximate methodof estimating12-hr snowmeltDn (periodsmidnight to The noon, noon to midnight) from direct solarradiation has been given by Wilson.2o relation is ofthe form (14.34) Dp = Do(l - 0.75m) where Do : the snowmeltoccurring in a half-day in clear weather (0 m : thedegree cloudiness for clearweather,1.0 for completeovercast) of Suggested valuesfor Do are 0.35 in. (March), 0.42 in. (April), 0.48 in. (May), and 0.53 in. (June)within latitudes 40-48".2o

Rainfall
Heat derivedfrom rainfall is generallysmall, sinceduring thoseperiodswhenrainfall of occurson a snowpack, temperature the rain is probablyquite low. Nevertheless, the at highertemperatures, rainfall may constitutea significantheat source;it affectsthe aging processof the snow and ffequently is very important in this respect. An equationfor hourly snowmeltfrom rainfall isrs

(14.3s)
P : the rainfall (in.) to T- : the web-bulb temperatureassumed be that of the rain This equationis basedon the relation betweenheatrequiredto melt ice (I44 Btu per pound of ice) and the amountof heatgiven up by a pound of water when its temperature is decreased one degree. by where

282

14 HYDROLOGY CHAPTER SNOW

are Daily snowmeltby rainfall estimates given by Ma: 0.007Pd(T" 32)

(r4.36)

where Md : the daily snowmelt(in.) , P d : the daily rainfall (in.) air To : the meandaily air temperature('F) of saturated taken at the 10-ft

level23

Conduction
are of Major sources heatenergyto the snowpack radiation, convection,and condensation. Under usual conditions, the reliable determinationof hourly or daily melt plus rainfall ifit occurs.An additional quantitiescanbe foundedon theseheatsources sourceof heat,negligiblein daily melt computationsbut perhapssignificantover an is entire melt season, ground conduction. Ground conductionmelt is the result of upward transferof heatfrom ground to snowpackdue to thermal energythat was storedin the ground during the preceding summerand earlyfall. This heatsourcecanproducemeltwaterduringwinter and eady springperiods when snowmeltat the surfacedoesnot normally occur. Heat transfer by by ground conductioncan be expressed the relation2 dT Hn: K-where K : the thermal conductivity "r'J"n dTlda : the temperaturegradient perpendicularto soil surface (14.37)

small. Wilson The snowmeltfrom ground conductionis generallyexceedingly notes that after about 30 days of continuoussnowcover,heat transferredfrom the The ground to the snow is insignificant.20 amountof snowmeltfrom ground conduchas been estimatedat approxirnately0.02 in.lday'23 tlon during a snowmelt season Groundconductiondoesact to providemoistureto the soil; thus,whenotherfavorable of conditionsfor snowmeltoccur, a more rapid development runoff can be expected. physicsof snowmelt.The mannerin which heat This sectionhasemphasizedthe -I4.3I and 14.27 Equations was to canbe provided initiatethemelt process discussed. L4.33-I4.37 inclusivecan be usedto estimatethe melt at a given point. The task of in computingrunoff from snowmeltin a basin cannot be approached sucha simple fashion,sincethere are many complexfactorsoperative.The remainderof this chapFigter is devotedto the general subject of runoff from snowmelt investigations. ure 14.5 illustrateshourly variation in the principal heat fluxes to a snowpackfor a cloudy day. EXAMPLE 14.2 existedfor During a completelycloudy April period of l2hr, the following averages ft above sea lel'el at a latitude of 44" N: air a ripe snowpacklocated at 10,000 temperature50' F; mean wind velocity, 10 mph; relative humidity, 65Vo;avenge reading,48oF. Estimate rainfall intensity,0.03 in./hr for I2hr; wet bulb psychrometer radiation, and warm rain for convection,condensation, the snowmeltin in. of water for the 12 hr period.

April 23 100 80 9 6 0
Eo
<2

Apnl24

Aprii 25

1600 2000 2400 0400 0800 1200 1600 2000 2400 0400 0800 1200
--- Short-waveradiation Long-wave radiation ffi
nm

lt ii
l,Incident short-wave radiation (t)

Incident short-wave radiation

Reflected shon-wave raoanon

G *
40

Absorbedshort-wave radiation

Downwad

long-wave radiation (R2)

H . n E z v

l -

-20

=
o

_40

naiatipn tn) J.bwaa]one-wa1e

Reflected shofr wave radiation (1,)

E E o.o4 On?

.,
6

E 0.02
(Each bs reDresents mem value

E o.ot
0
1700-1800hr 2400*0100hr' 0900-1000Itr
no condensation

0.01 0.25 0.20 0.15

E 0.10
o n

no condensation -

no convgctiotr or -

meltllcondensationmeltmelt t t t t l E Radiation melt (Mr) Convrction condensation 6 mett (Mce) --- Radiationmelt total (Hourlyconvection-cotrde$ation is melt added of subtacbdftom totalhouly to radiation to arnveattotalhouly melt
computed melt) I Totalcomputedmelt

Total Snow Melt (M) ud Runoff (O)

Computed melt (net for period0900- 1800 = hr l22ir-022in = I 00 in.)

_o*
0 -0.05 Apil23-----l-Net nighttime loss = 0.22 in

1600 2000 2400 0400 0800 1200 1600 2000 2400 0400 0800 1200 Apfl24
April 25

Figure L4.5 Hourly variation in principal heat fluxesto a snowpackfor a cloudyday.(After U.S.Army Corpsof Enginers.2l

284

CHAPTER14

SNOW HYDROLOGY

Solution a. Convectionmelt, 6 hr

D:KV(r-32) x D : 2 x 0 . 7x 0 . 0 0 1 8 4 1 0x ( 5 0 - 3 2 ) : 0 . 5 0 i n . melt, b. Condensation 6 hr

D:KrV(e"-6.11) x x D : 2 x 0 . 0 0 5 7 8 1 0 x ( 1 2 . 1 9 0 . 6 5- 6 . 1 1 ) : 0.21in. . Radiation melt, 12 hr Dn: d. Rainfall melt. hourly M : P(r-'- 32)lt44Q, 14 : 10.03x 12 x (48 - 32)l/(144 x 0.97) : s.64 Thus,total melt is 0.86 in. rr D o ( 1- 0 ' 7 5m ) D r 2: 0 ' 4 2 x ( 1 - 0 . 7 5x 1 ) : 0 ' 1 1

(14.27)

(14.30)

(r4.34)

( 14.3s )

RUNOFF DETERMINATIONS 14.6 SNOWMELT

from snowmelthavebeenfollowed.They to Variousapproaches runoff determination that completelyignore the physical rangefrom relatively simplecorrelation analyses Most methodsusing physicalequations. snowmeltprocessto relatively sophisticated ofrecescorrelations,analyses as can techniques be considered basedon degree-day sion curves, correlation analyses,physical equations,or various indexes.Each is in discussed turn.

Purposesof SnowmeltRunoff Estimates

Snowmeltrunoff estimatesare extremelyimportant for many regionsof the United water yields for a diversity of Statesand other countriesin (1) forecastingseasonal water supply purposes,(2) regulating rivers and storageworks, (3) implementing flood control programs, and (4) selectingdesign floods for particular watersheds. Maximum floods in many areas are often due to a combination of rainfall and of snowmeltrunoTf.In effect, the determination snowmeltrunoff hasthe sameutility as the calculationof runoff from rainfall. In someareasit will, in fact, be the more important of the two.

Condition Snowpack
The mannerin which runoff from eitherrainfall or snowmeltis affectedby conditions Variousviewson prevalentwithin the snowpack of primary interestto a hydrologist. is of storagecharacteristics a snowpackhave been advanced.These range from the

14.6 SNOWMELTRUNOFFDETERMINATIONS

285

conceptthat a snowpackcan retain large amountsof liquid water to the hypothesis that snowpackstoragi is negligible.Thereis no universallyapplicablerelation, and it on important to baseany runoff considerations a knowledgeof the character becomes condition, at of a snowpack the time of study.Winter runoff is relatedto a snowpack's in whereas the spring,onceactivemelt begins,little or no delayin the transportof melt or rainfall through the snowpackoccurs. For drainagebasins in mountainousareas,snowpackstorageeffects may be into relatively uniform areas.Normally, by approximated subdividingthe watershed by this will be accomplished usingelevationzones.Snowpackat the lowestlevelsmay a be conditionedto transmit readily rain or meltwater,whereasin higher elevations be very the liquid water deficit may prevail. At uppermostelevations, snowpackmay dry and cold and thus in a condition for the optimxm storageof water. The storage of measurements the zonesmustbe basedon representative potentialof the watershed water equivalent,moisturecontent,and snowpack snow depth,density,temperature, relatesto the physicalstructureofthe pack' UnforThe snowpackiharacter character. tunately, adequate1;r"usrrr"rof all thesefactors are not always availableor easily obtained. Estimatesof changesbetween sampling periods are usually indexed to variables. readily observedmeteorologic can be The formulation of snowpackstorageand time delay characteristics is pack.In this case,storage relateddirectly to a by fashioned assuming homogeneous liquid water defici1and cold content of the pack. Time delay is a function of the the inflow rate. It is consideredthat the snowpackstoragepotential must be entfuely permits but In beforerunoff begins. reality this is not the case, the assumption satisfied snowpack an analysisto be made. As melt proceeds,the storagepotential of any diminishes. to is beforerunoff commences considered be the sum of of Storage a snowpack to of waterrequilementto raisethe temperature the snowpack 0'C (cold the equivalent If capacityof the snowpack. the cold content contentlV")and the HqJidwater-holding to temperature 0oC,it may be is given in inchesof water neededto bring a snowpack by2 represented

w"T" *r=ffi
where

(14.38)

7] : the mean snowpacktemperaturebelow OoC lyo : the initial watei equivalentof the snowpackin inchesfor an assumed heatof ice of 0'5 specific is to temperature OoC thus given by to The time L in hoursneeded raisethe snowpack , " -- r 6 o 1 i + * 7 WoT,

(14.3e)

whereI is the rainfall intensity(in./hr) andm is the rate of melt (in'/hr)' Storage requiredto meet the liquid water deficit of the snowpackis given by

tr:#(%+w.)
( where ".1- : the amount of water stored(in.) T _ the percent deficiencyin liquid water of the snowpack JP

(14.40)

286

CHAPTER14

SNOW HYDROLOGY

The time in hours r, neededto fill the storageSyis given by

+ t': fi\wg+ W") 100(i m)

(r4.4r)

It has been specifiedthat the total storagepotential Soto be met prior to the runoff is given as (r4.42) Sr:W'"*Sy to sinceit is not available the runoff until This is also known as "permanent" storage, componenttransitory storage has the snowpack finally melted.An additional storage while movingthroughit to becomerunoff. Until S,is that waterstoredin the snowpack as initiation of runoff, the transitory storagein inchescan be expressed

^ __T__ ",: D ( i + m )
where D : the depth of the snowpack (ft) through the snowpack(ftlhr) V : the rate of transmission The delay time of water in passingthrough the snowpackt' is thus
a t

(r4.43)

D
v

(r4.44)

is for /, in hours.Assumingthat I4z" very small comparedwith Wo,the depth of the snowpackis given by

O:%
P'

(r4.4s)

Then of with p" the density the snowpack. wo (r4.46) nv in the Beforethe runoff commences, total waterS storedin the snowpack, inches, givenby is t'-

S:W"+++S,
which can also be written

(r4.47)

t = * ( 1 f t + f i d i. + m ,\ r ) "r
before runoff is producedis thus2 The total time in hours that passes t:t"+tr+tt

/ r

(14.48)

(r4,49) (14.50)

z: f!,{1 t = w'1160(i r, -, Too1,-,r -+ p,v) f +

term the runofffromthesnowpack, onlysignificant the After establishing active with the overallbasinlag and is in Eq. 14.49 /,, andthis is usuallysmallcompared

RUNOFFDETERMINATIONS 14.6 SNOWMELT

287

33
End rain

31 -:
'*2
= )

30
Q

6 - ' E
.Z
d o

te

- 2 6

24 23 0 6 12 18 24 30 36 42 48 54 60 66 72 78

(br) Time duringrainfall' in balance a snowpack Figure 14.6 Water can be neglected.With increasedsnowmelt and runoff, additional increments of water prevlouslywithheld by snowblockageto drainageoutlets and other factorsare A at Adequatequantifi;ation of this cannotbe accomplished present'24 deep released. -5oC, could storeabout 4 in. of ft, havinga meantemperatureof snowpack,say 15 fiquid waterbeforethe onsetof runoff. Figure 14.6illustratesthe nature of the water balancein a snowpackduring a rainstorm. EXAMPLE 14.3 A core sampleof a snowpackproducesthe following information: air temperature, 10 ft; 68"F; relative humidity, 2b percent; snowpackdensity,0'2; snowpackdepth' snowpacktemperature,22"F. a. b. c. d. What is the vapor pressureof the air? on Will condensation the snowpackoccur, basedon the vapor pressure? is the cold content of one sq ft of surfacearea of the snowpack? what Is the snowpackripe? -

Solution

a. From AppendixTableA.2, eo: 233 mb for a relative humidity of 20 percenr,o = 4'66 mb will not occur sincethe air is unsaturated" b. Condensation c. Cold content

160 W": WsT"f

288

CHAPTER14 SNOW HYDROLOGY

: Temperature 22oF: -5.6oC in. W": 0.20x l2Ox 5.6/160: 0.84 is its is d. The snowpack not ripe since temperature belowfreezing. I I

Indexes
Hydrologic indexesare made up of hydrologicor meteorologicvariablesto describe their functioning. The index variable is more easily measuredor handier than the When mean fixed relations are known to exist betweenpoint elementit represents. and measurements watershedvalues,indexescan be used to record both areal and temporal aspectsof basin values. Indexes serve to permit (1) readily obtainable cannotbe which themselves to observations depict hydrologicvariablesor processes and (2) simplificationof computationalmethodsby allowing indieasily measured, valuesin time and to or vidual observations groupsof observations replacewatershed of space.The adequacy an index is basedon (1) the ability ofthe index to describe it the , adequately physicalprocess represents(2) the randomvariability of the obseris (3) the degreeto which the point observation typical of actual conditions, vation, and basin means.2 (4) the nature of variability beiweenthe point measurement and Indexesmay be equationsor simplecoefficients,and variable or constant. thermal budgetstudiesare The types of data requiredto make comprehensive ones.As or in normally unavailable wh^ole part for wut"tJh"dt otherthan experimental a result, a hydrologistmust make the best use of information at hand. The most humidity, commonly availabledata aredaily maximum and minimum temperatures, of measurements thesedata, andfew and wind velocity.Lessprevalentare continuous datacan Hourly cloudiness stationsrecord solarradiation or the durationof sunshine. be sometimes obtainedfrom local airport weatherstations. A completelygeneralindex for reliably describingsnowmelt-runoff relations valid only for includecoefficients Most indexes for all basinshasnot beenestablished. conditionsand aretherehydrologic,and seasonal meteorologic, specifictopographic, Table 14.1 shows some types of fore limitedlnipplicability to other watersheds. in indexesthat havebeen used successfully snowmeltinvestigations. VARIABLES BUDGET THERMAL USED INDEXES TODESCRIBE 14,1 SOME TABLE
component Thermal budget
radiatron Absorbed short-wave Long-waveradiation' Convectiveheat exchange Heat of condensation Index Duration of sunshinedata Diurnal temperaturerange Air temperaturefor heavy forestedareas radiation shouldbe estimated For open areaslong-wave (7" - T)V, where f is air temperature, T6 the snow surface temperatureor basetemperature,and Vthe wind speed of (e" - e")V,whereeoand e, are vapor pressures air and snow surfaceot a basevalue, and V is wind speed

" Figure 14.7 iilustrates an approximatelinear relation betweenmelt and long-waveradiation usedby the U.S, Army Corps of Engineers for index purposes.

14,6 SNOWMELT RUNOFFDETERMINATIONS

289

The snowmeltrunoff equationstatedin terms of thermal budgetindexesis

Y:a+>b$t
where I : a : b, : 4 : the snowmeltrunoff aregression constant the regression coefficients individual indexes

( 14.s1)

Variousindexes usableto represent termsof Eq. 14.51are selected a standard the and regression performedto determine andb,.It shouldbe notedthat everyterm analysis a in the heatbudgetequationis not alwayssignificantfor a particular analysis, and thus the numberof Xr will vary for different basinsand conditions.A final melt equation
1.6
Long-wave radiation melt in forest or with low overcast sky in open Ma= 1Z-L n:{-r'1 ,, toTf,- oT!) {referto Kelvin scale)

-3.355 = 604x lA-1274

E 'c o

n -v.+ /

Long-wave radiation melt in open

Ma= zoittlpl rcrf,-orlt = 604x IO-12x- 3.355 T = o.o29 _ 32) (Ta

(refer to Fahrenheit scale)

1L O

r l t l M, = long-waveradiationmelt, in./day

-r.o

P = ftee water content of snow (taken as 0.03 in this case) rA = absolute air temperature, K ?s = absolute snow surface temp = 273K

-2.0

o = stefm's constant = 0.813 x 10. lolmgley/min-K-a Io = mean daily air temperatue, F

1 A

10

20

30 40 60 50 Meanair temperature, Zo("F) 270 2'75 280 285 Absolutetemperature, (K) Z1 290 300

260

265

Figure 14.7 Long-wave radiation melt, with linear approximation. (After U.S. Army Corps of Engineers.2)

290

CHAPTER14 SNOW HYDROLOGY

t.6
dataof Equationderivedfrom combined 1954 1955 and (T G RO = 0.00238 + 0.0245 ^ -'77) D=0.90 r=0.95 = 0.36 in. S"- . = 0.1| in'

t.4

'

1 n
z}

ts E U.6
a

"r_"n
.C .l

,/
Mean ru noff 0.59

Z
\C "y-x

ri

5 0.6
0.4

,/,

aa,

'24.
0.2

/ a o

0.2 a .

0.8 0.4 0.6 runoff(in.) Estimated

1.0

1.2

Figure 14.8 Observedversusestimatedrunoff for (X) snowmelt 1954 and (O) 1955. RO = the daily generated runoff (in.) depth over a snow-coveredarea; G : the by daily net all-waveradiation absorbed snow in the open (langleys); 7-u* : the daily maximum temperature for Boise ("F); r = the coefficient of correlation; D : the coefficient of determination;S, = the standarddeviation of observedrunoff (in.); Sr-, : the standarderror of the stimated runoff (in.). (After U.S. Army Corps of

Engineers.2) for devekiped the Corps of Engineers2 the partly forestedBoiseRiver basin above by Twin Springs,Idaho, was

+ *Q : 0.00238G 0.0245(T^ 77)

(r4.s2)

area where Q = the daily snowmeltrunoff (in.) over the snow-covered in G = an estimatedvalue of the daily all-waveradiation exchange the open (langleys) ZLu, : the daily maximum temperatureat Boise (T) The equationis said to predict the daily snowmeltrunoff valueswithin 0.11 in. of observedvalues about 67 percent of the time. Figure 14.8 illustrates this relation. shouldseek for suitableindexes snowmelt,a hydrologist In attemptingto develop the approachmost closely resemblingthe thermal budget of the area, within the limitations of availabledata.

14.6 SNOWMELT RUNOFF DETERMINATIONS 291

Temperature Indexes
The atmospheric temperature an extremelyusefulparameterin snowmeltdetermiis nation. It reflectsthe extentof radiation and the vapor pressureof the air; it is also sensitiveto air motion. Frequently,it is the only adequatemeteorologicvariable regulady on hand, so widespreaduse has been made of degree-dayrelations in snowmeltcomputations. I A degreeday is defrnedas a deviation of 1ofrom a given datum temperature consistentlyover a 24-hr period. In snowmeltcomputations, referencetemperathe ture is usually 32"F. rf the mean daily temperatureis 43oF, for example,this is equivalentto 11 degreedays above 32'F.If the temperaturedoes not drop below freezingduring the24-hr period, there will be24 degree for eachdegree hr departure above32'F. In this examplethere would be 264 degree for the day of observation. hr Variouswaysof estimatingthe meantemperature haveenabledinvestigators to take severalapproaches. methodis simplyto average maximumandminimum one the daily temperatures. Basesother than 32"F are also used.Regardless the particular of attackemployed, degree a hour or degree is an index to the amountof heatpresent day for snowmeltor other purposesand has proved useful in point-snowmeltand runoff from snowmeltdeterminations. The standardpracticein developing snowmeltrelationson the basisof temperature is to correlatedegreedaysor degreehours with the snowmeltor basin runoff. In somecases, other factorsare introducedto defineforest covereffectsand/or other influences. Another approach often usedis to calculatea degree-day factor-the ratio of runoff or snowmeltto accumulated degreedaysthat producedthe runoff or melt.
2.8
1 A

; 1 A o

B
6 r-z

34

38

42

46 50 s4 58 Mean daily temperature, Z('F)

62

66

Figure 14.9 Mean temperature index. The equations are applicable only for the range of temperatures shown in the diagram. (After U.S. Army Corps of Engineers.2)

292

CHAPTER14 SNOW HYDROLOGY

and between factor has been found to vary seasonally Unfortunately,the degree-day values should be used with caution. Pointbasins;therefore, single representative rangefrom 0.015to 0.20 in. per degree basins factorsfor snow-covered degree-day point value of 0.05 can per day when melting occurs.Gartskastatesthat an average be used to representspring snowmelt, provided that caution is used. Linsley and factorsare usuallybetween0.06 and 0.15 othersstatethat basinmean degree-day and snowcover at meltingtemperatures. in./degreedayunderconditionsof continuous for index equations springtimesnowmeltfor clear Figure 14.9illustratestemperature areas.' and forested

BasinSnowmeltEquations Generalized
studiesby the U.S. Army Corps of Engineersat variouslaboratoriesin the Extensive West have produced several general equationsfor snowmelt during (1) rain-free When rain is falling, heattransferby convectionand periods and(2) periodsof rain.2a radiation is condensation of prime importance.Solarradiationis slight, andlong-wave rain-free periods When can readily be determinedfrom theoretical considerations. prevail,both solarand terrestrialradiationbecomesignificantand may require direct evaluation. Convection and condensationare usually less critical during rainless intervals.The equationsare summarizedas follows:2 1. Equations periodswith rainfall. for For open (coverbelow 10 percent)or partly forested(coverfrom 10 to a. 60 percent)watersheds,

+ M : (0.029+ 0.0084fto 0.007P)(7" 32) + 0.09 (over80 percent cover), areas forested b. For heavily M : (0.074 + 0.007P,)(7.- 32) + 0.0s

(14.s3) (14.54)

where M : the daily snowmelt(in./daY) P , : the rainfall intensity (in./day) T o : the temperatureof saturatedair at 10-ft level ("F) wind velocity at 50-ft level (mph) the average t ^. -- the basin constant,which includesforest and topographic of exposure the areato average effects,and represents from about 1.0 for clear wind. Valuesof k decrease plains areasto about 0.2 for denseforests ,, Equationsfor rain-free periods. a. For heavyforestedareas, , M:0'074(0.537'"+ O.47fi)

(14.ss)
(14'56)

areas(coverof 60-80 percent), b. For forested

+ 0.7870+ 0.0297" M : k(0.0084u)(0'22r'" areas, c. For partly forested M : k'(I - rx0.00401,)(l a) r) + 0.78r) + F(0'029 + k(0.0084o)(0.227'"

Q4.57)
)

14.6 SNOWMELT RUNOFF DETERMINATIONS 293 d. For open areas, M : k'(0.00s084)(1 a) + (1 -'N)(0.0zIzT: - 0.84) + N(0.02gr') + k(0.00S ao)Q.227'" 0.78Ti) + (14.58)

where M, a, k : as previouslydescribed T'" : the difference betweenthe t0-ft air and the snow surface("F) temperatures T'o: the difference between the 10-ft dew-point and ('F) snow-surface temperatures I, : the observedor estimatedinsolation (langleys) a : the observed estimated or meansnowsurface albedo k' : the basin short-waveradiation melt factor (varies , from 0.9 to 1.1),which is relatedto mean exposure open areascomparedto an unshielded of horizontal surface F : the mean basin forest-canopy cover (decimal fraction) Ti = the difference between the cloud-base and (oF) snow-surf,ace ternperatures N : the estimatedcloud cover (decimalfraction) Note that the use of equationsof the type given must be related to the areal extent of the snowcoverif realistic values are to be obtained. Presentmethods of determiningthis are not totally adequate. EXAMPLE 14.4 Use Eq. 14.53to estimatethe snowmeltat an elevationof 3000 ft in a partly forested arcaif the rainfall intensity is 0.3 in./day, the wind velocity is 20 mph, and the temperatureof the saturatedair is 42'F. b. Rework your solution for a denseforest cover and a saturatedair temperatureof 53"F. Solution

A , M : (0.029+ 0.0084ftr-rA.007P)(7, 32) + 0.09 + M _ (0.029 0.0084 0.5 x 20 + 0.007x 0.3)(42 32) + 0.09 + x
I'tl -

b . M = (0<074 0.007P)(7" 32) + 0.05 + M * (0.074 0.007 0.3x53- 32) + 0.05 + x M _ 1.65 in./day r:

I.24 in.lday

The Water Budget

The waterbudgetcanbe usedto estimate snowmeltrunoff from a watershed.2 the Such an approach particularmerit for areaswherehydrometeorologic has recordsare short. Difflculty with the methodis the usuallack of satisfactory datato quantify the various

294

CHAPTER14

SNOW HYDROI-OGY

properly. A hydrologicbudgetequationfor the earth's surface(Eq. 1.1) components can be written R : P - Z - A S where P : R: L : AS : the grossPreciPitation the runoff the losses the changein storage

(14.se)

For snowmeltcomputationsthis equationis modified somewhat' Grossprecipiiation for a given period P is now definedasthe sum of precipitations in the form of snow P" and rain P,, ot P:P,1-P" This may also be written as P:P,*P":P^+Li where P,: net precipitation L; : intercePtionloss A further refinementyields P: where P,n,P",: L,i, L"i: P-+ L,i+ P"nt L,;

(14.60) (14.61)

(r4'.62)

net rainfall and snowfall, respectively respectively the rain and snowinterception,

Figure 14.10indicatesthe nature of snow interceptionby forestedareas'Additional information on interceptioncan be found in Chapter 3' The total lossI is L : L,i + L,i + L" * Q"where loss L" : the evapotranspiration the changein availablesoil moisture Q,^: term AS is then given as The storage AS : (l7z - Wr) t Q, $4.63)

04.64)

whereWr,Wr:thefinalandinitialwaterequivalentsofthesnowpack' resPectivelY Qr : the ground and channelstorage Inserting uilo"r for p, L, and AS from Eqs. 14.62-14.64 in Eq. 14.59 gives - L" - Q"^ - (W, - Wt) - Q, R : P,n + L,i + P", + L"i - L"i - L,i and cancelingpositive and negativevaluesof L,i and lr; produces - Q, - L" R = P,n+ P", - (W, - Wt) - Q,-

(14.65) I
(14.66) 1 )

14.6 SNOWMELTRUNOFFDETERMINATIONS

295

4-')

Curv e A
.MC
1\MC o
6 J U

-<

(2) ,SP&
PPR

625 F B z^v ^ o
FQ 1<

6,
RF (1) tr RF -Curv B WFM WFM

(4)

PPM tr

o c /rt
:l

,%?r
,/,

r l 0 o
a h 5

./

-10 0 l0 20 30 40 50 60 Canopy density(%) 70 80 90 100

Figure 14.10 Engineers.2)

Snowfall interception loss. (After U.S. Army Corps of

The expression - (W, - Wr1represents snowmeltM; thercforc, P", the R:P,,+M-Q"^-Qr-L" (14.67) If reliable estimates the terms in Eq. 14.67 canbe secured,the basin discharge of R is computable.

Elevation-Band Procedure
Runoff from snowmelton a watershed can be estimatedfrom calculationsof excess watermadeavailtble on a seriesof contributingareas(bands)at variouselevations in the watershed. The practiceis as follows: divide the watershed into severalsubareas or bands;estimatethe quantity of snowmelt,rainfall, and lossesgenerated each on band during a prescribedinterval of time; and usethe weightedsum of thesecontributionsto provide an estimateof the excess wateravailable runoff. For eachband, for it is assumedthat snowmelt, rainfall, and lossesare uniform over the band. The subareas considered be either snow-covered snow-freeand melting or not are to or

296

CHAPTER14 SNOW HYDROLOGY

melting. For eachband, snowmeltis computedusing equationsof the type preSented or basedon expectations historic information, and losses earlier,rainfall is estimated havebeen in are estimatedas described Chapterb3 through 5. Ohce theseestimates made for each band, the following equation servesto provide a weighted value of excess water availablefor runoff from the basin.

)
M _

tte + Mt - Lt)jAt (14.68)

)a,
where : snowmelt water availablefor runoff (cmlday), Pr is the rainfall on the 'band, M loss,A, is the size of the M, is the snowmeltfrom the band, I, is the subarea subarea,and n is the total number of bands. EXAMPLE 14.5 water Given the data in columns 1-5 of Table I4.2, estimatethe amount of excess method(Eq. 14.68). usingthe elevation-band available runoff from the watershed for 14,5 14.2 DATA FOR EXAMPLE TABLE
(1) Elevation band no.
I

(2)
Subarea slze sq km

(3)
Snowmelt cmid

(4) Rainfall cm/d

(5)

(6)

(7)

Losses cm/d 0.40 0.50 0.70 0.60 0.30 0.10

( 3 )+ ( 4 ) + ( 5 )

(6) x (2)

z
J A

5 6 Totals

230 ))a 289 213 193 167 1316

0.02 0.40 0.60 0.70 0.3s 0.00

0.90 1.0 l 1.80 1.90 2.20 "2.40

1,.32 2.00 3.10 3.20 2.85 2.50

303.60 448.00 895.90 681.60 5s0.05 411.50 3296.65

Solution. The solution for the numeratoris the sum bf the productsgiven in is Table 14.2;the solutionfor the denominator the sum column 7 of spreadsheet of the subareas siven in column 2 of the table. : for The excess-water availabie runoff : 3269.6511316 2.5I cmlday'
II

Recessions Hydrograph
in curveshavebeen discussed Chapter 11 and take the generalform Recession Q: ' where Q : the dischargeat time t Qo : the initial rate of flow k:__afecessionconstant Qoe-o'

(r4.6e)

, : -

RUNOFFDETERMINATIONS 14.6 SNOWMELT

297

o u

Time

hydrograph' of Figure 14.11 Separation a snowmelt permit evaluationof the amountof runoff Studiesof daily streamflowby hydrographs of the derived from snowmelt.The t""hniqutnred is essentiallyone of separation illustratesthe proceFigure 14.11(not to scaleandoversimplified) daily hydrographs. fit peaks,respectively, snowmdlt dure. Assumeihat the first, second,and succeeding in time, at a point A the backward days.If the ultimate recessioncurve is extended from recessions curveft'om Hydrograph2 will intersectit. The areabetween recession is the melt attributedto Day Hydrograph1 and HydiogtipiZ (showncross-hatched) their can 1. In like -unn"r, a seriesofsnowmelt hydrographs be studiedto determine featuresas th height to By individual melt components. observingsuchhydrograph peak X, the height io trough i, andthe form of the recession,volumetric and rate treatmentof this forecastsof snowmeltrunoff can be made.A more comprehensive subjectcan be found in Ref. 25.

SYnthesis Hydrograph
with snowhydrologyareoftwo-types'The ofrunoffhydrographsassociated syntheses of The secondkind is the development flow distribution first is a short-termforecist. or a ' for a comPletemelt season forecastingis very helpful in prel of controls,while the synthesis part lating designfloods. To forecasta snowfieldand streamflowneedbe k parameters to ing, it is necessary havethe reliablepredictionof variousmeteorological can initial conditions.Known historicparameters be in additionto a knowledgeof the satparameters or flows whereasassumed generated usedfor reconstructinghistoric Figure 14.12 displayssomecommon hydrometeorologic isfy designflood syntlieses. data. severalfactors(not of great conceln where In snowmelthydrographsyntheses, First, a drainagebasin with snow"utlfully considered. only rainfall exists)mustbe system,since the areal extent of the as a homogeneous cover cannot be accepted

298

CHAPTER14

SNOW HYDROLOGY

6 5 ^ ! 9 > > v

B 5
(h

Short-waveradiationngtesj.^,-- .-

i hsolation obsened by USWB. of Boise Cily' lD 2. Basin albedo of snow surface ^ - = , | L Absorbed shon-wa\e ralialion compukd b) lomula 1ab. ll(

AbsorbedShort-wavefadiation

U
o

9.1 ' H 9 a t o 6

-8
Air Temperature Notes: 1. Veni;al bus show daily range of surface air temperature (max and min) at Idaho City' ID 2. aonnected points indicate daily (0700 hr) 700 mb temperature at Boise CitY, ID

90 80 70 60

t--

s
B

50 o 40

30 a 0.5 120
lrn

u 4 0.4
9 b
q d

gB
E

0.3 0.2

g >
7

2 0 MaY1955

31

20 10 1955 June

Figurel4.l2Hydrometeorologicdataandcomputationofwatergenerated. (AfterU.S.Army Corpsof Engineers.') the contributblanket is highly important. where only snowmeltflows are developed, If rainfall ing areaneel not U"itr" entiredrainage-only that portion with-snowcover' from bare areaswhile o"-"o., during the snowcoverperiod, contributionscan come may in may producecombinedrunoff' The natureof losses suchcases other expanses differ greatly for nonsnowoverlayedand coveredlocations' subThe altitude is an exceedinglypertinent factor in the hydrology of tracts due to a generalreducjected to snowfall.Ratesof .no*tnitid"crease with elevation tion in temperaturewith height. orographic effects and the temperature-elevation snowcover relations tend to raise the amount of precipitation with altitude' Greater melt rates' As a result' precipitation and reduced of depth occursbecause increased is apthe basin-wide melt and cover-area increase with height as the snowline

14.6 SNOWMELTRUNOFFDETERMINATIONS

299

proached,then diminish with elevationover the higher placesnormally completely snowcovereduntil late in the season. snowpack A exhibitsanotherimportant trait in relation to rainstorms.In the spring, relatively little runoff occurs from snow-free regionscomparedwith that from a snowfieldfor moderate rainfalls. During very cold weather,the situationduring heavyrains is often reversed, sincea dry snowpackcan retain significantamountsof water. Two basic approachesintroduce elevation effects into procedures for hydrographsynthesis.2 first dividesthe basin into a seriesof elevationzoneswheie The the snowdepth,precipitationlosses, melt are assumed and uniform. A second method considers watershedas a unit, so adjustmentsare made to accountfor the areal the extentof the snowcover,varying melt rates,precipitation, and other factors. To synthesize snowmelthydrograph,information on the precipitation losses, a snowmelt, and time distribution of the runoff are needed.Snowmelt is generally estimatedby index methodsfor forecasting,but in design flood synthesis heat the budget approach, the most used. Precipitation is determinedfrom gaugingsand is historicor generated data.Losses definid in two wayswheresnowmeltis involved. are For rain-on-snowhydrographs the water is considered lossif delayedvery long all a in reaching a stream. This is basically the concept of direct runoff employed in rainstormhydrographanalysis.For hydrographs derivedprincipally from snowmelt, only that part of the waterwhich becomes evapotranspiration, deeppercolation,or or permanentlyretained in the snowpackis consideredto be lost. Assessing time the distribution of runoff from snow-covered areasis commonly done with unit hydrographsor storagerouting techniques. rain-on-snow events,normal rainfall-type For unit hydrographs applied;for the distribution of strictly snowmeltexcess, are special long-tailedunit graphsare employed.Storage routing techniques widely exeicised are to synthesize spring snowmelthydrographs, perhapsdividing them into severalcomponentsand different representative storagetimes. The time distribution of snowmehrunoff differs from that of rainstormsdue mainly to large contrastsin the ratesof runoff generation. For flood flows associated with rainfall only, direct runoff is the prime concern,and time distribution of base flow is only approximated. errorsin estimates baseflow arenot generallyof any Big of practical significance where major rainstorm floods occur. In rainstorm flows, infiltrated water is treated as part of the base flow component and little effort is directedtoward determiningits time distribution when it appearsas runoff. In using the unirhydrograph approachto estimatesnowmelthydrographs, is customaryto it separate surfaceand subsurface the componentsand route them independently. Storagerouting has beenusedextensively routing floodsthrough reservoirs for or river reaches. is also applicablein preparingrunoff hydrographs. snowmelt It In runoff estimates, rainfall and meltwaterare treatedasinputs to be routedthrough the the basin, using storagetimes selectedfrom the hydrologic characteristics the of watershed. Two basichydrologicrouting approaches relatedto the assumption are of ( 1) reservoir-typestorage (2) storage or that is a function of inflow and outflow.These methodswere treatedin depthin Chapter13. Storagerouting techniques that separate runoff into surfaceand groundwater components, assign different empirically derivedstorage timesto each,and thenroute them separately havebeen employed.26 additional system An usesa multiple storage,

300

CHAPTER14

SNOW HYDROLOGY 120

,:

Rn

o 6 U
il

Three 6-hr stages

One 18-hr stage

Time (hr)

storagerouting' Figure 14.13 Example of multiple-stagereservoir-type routreservoir-typestorage iti. ngur" illustratesih" ut" of multiple-storage to runoff in u ttrunn". analogous unit ing for"evaluatingtime distribution of tiOrog.uptr. (Af1erU.S. Army Corpsof Engineers'2)

is routed throrrgh two or In reservoir-type storagescheduling.2 this method inflow suchan approach'Any more stagesof storageru"""rtiu!fy. Figure 14.13 illustrates the storagetime and.the desiredtravel time can be obtainedby properly selecting to reflectvarious Retentiontimes betweenstepsmay alsobe varied tf st4ges' use of single-stage that the. clark has suggested drologic characteristics. to irp", in time permits Jo=mputations be^^simplified.27 rfter translatirrg has snowmeltrunoff hydrographs method for synthesizing e most practicJd graphsdiffers primarily in ,,nir hrrrrrnoranh The characterof snowmeltunit 12' that of rainstorm unit plots' As disclssed in chapter time base length fiom storm events'In single isolated rainstorm unit hydrograph, ort"n are derived from snowmeltrunoff,ratesofwaterexcessaresmallandapproximate$continuous'Asa is result, the use of S-hydrographs indicated'2 utility, since it allows (1) adiustri"titoA has considerable The S-hydrogrupfr generationrates,(2) adjusting ments to the derivedoni, hydrographfor nonuniform of the areaunder iirn" p"rtoo to; d;ired interval, (3) ready adjustments il;;;t; to veragingseveralhydrographs get a unit

;?ilt1,"1*T,::t1il1""t"1':*"ff methodin adjustingfor nonunirydiograph

form generationrates of water excess' Onceap"r""nrug"S-hyatogtupftisderived'aunithydrographofanydesired are of the S-hydrograph m periodcanbe obtaineias indlcated nig. 14.15.ordinates

312

HYDROLOGY AND 15 CHAPTER URBAN SMALLWATERSHED havehad wide application;however, of Both categories peakflow determination in two relatively major difficulties are normally encountered applyingthe techniques. First, the rainfall-runoff formulas,suchas the rational formula, aie difficult to apply unless the return periods for rainfall and runoff are assumedto be equal. Also, of estimates coefficientsrequiredby theseformulas are subjectiveand havereceived criticism. The empiric and correlativemethodsare limited in application considerable to they are derivedfrom localizeddata and are not valid when extrapolated because otherregions. methods, The most fundamentalpeak flow formulas and empiric-correlativQ and severalof the the urban designscene, ilue to their simplicity,persistin dominating most popular forms are briefly describedto acquaintthe reader with methodsand are assu-ption*. Urban runoff simulationtechniques describedin Chapter25.

RationalFormula
The rational formula for estimatingpeak runoff rates was introducedin the United Sincethen it has becomethe most widely used by States Emil Kuichlingin 1889.18 Peak method for designingdrainagefacilities for small urban and rural watersheds. flow is found from QO: CIA

(1s.1)

where Qo: the peak runoff rate (cfs) to C _ the runoff coefficient(assumed be dimensionless) rainfall intensity(in./hr), for a stormwith a durationequal I _ theaverage to a critical period of time /" t" : the time of concentration(seeChapter Ii) A : the size of the drainagearea (acres) net cI : the average rain intensity (in./hr) for a storm with duratiofl: t, 1.0 because acre-in./hr is to The runoff coefficientcan be assumed be dimensionless for stormsof 5-10-year return periods equivalentto 1.008 ft3lsec.Typical C values are providedin Table 15.1. The rationale for the method lies in the conceptthat application of a steady, uniform rainfall intensity will causerunoff to reachits maximum rate when all parts are of the watershed contributingto the outflow at the point of design.That condition is met after the elapsedtime t", the time of concentration,which usually is taken as At the time for a waveto flow from the most remotepart of the watershed. this time, the runoff rate matchesthe net rain rate. Figure 15.1 graphically illustrates the relation. The IDF curve is the rainfall intensity-duration-frequencyrelation for the areaandthe peakintensityofthe runoff is Q/A: 4, which is proportional to the value of 1 defined at t". The constantof profottionatity is thus the runoff coefficient,C : (QIA)lL Note that QIA is a point yields nothing of the nature of the rest of the value and that the relation, as it stands, hydrograph. The definition chosenfor /" can adverselyaffect a designusing the rational channelvelocity is usedto estimatethe travel time from the formula. If the average most remote part of the watershed(a common assumption),the resulting design

WATERSHEDS 313 FOR 15.2 PEAKFLOWFOHMULAS URBAN TABLE15,1 ryPICALC COEFFICIENTS 5FOR TO 1O-YEAR FREQUENCY DESIGN
Descriptionof area Business Downtown areas ' Neighborhoodareas Residential Single-family areas Multiunits, detached Multiunits, attached Residential(suburban) Apartment dwelling areas Industrial Light areas Heavy areas Parks,cemeteries Playgrounds Railroad yard areas Unimproved areas Streets Asphaltic Concrete Brick Drives and walks Roofs Lawns; sandy soil: Flat,2Vo Avenge,2-7Vo Steep,TVo Lawns; heavy soil: Flat,2Vo Average,2-7Vo Steep,TVo Flunoffcoefficients

0.70-0.95 0.50-0.70 0.30-0.50 0.40-0.60 0.60-0.75 0.25-0.40 0.50-0.70 0.50-0.80 0.60-0.90 0.10-0.25 0.20-0.35 0.20-0.40 0.10-0.30 0.70-0.95 0.80-0.95 0.70-0.85 0.75-0.85 0.75-0.9s 0.05-0.10 0.10-0.15 0.15-0.20 0.13-0.17 o.r8-0.22 0.25-0.35

Time (min)

Figure L5.1 Rainfall-runoff relation for the rational method.

314

CHAPTER15

URBANAND SMALLWATERSHEDHYDROLOGY

dischargecould be less than that which might actually occur during the life of the project. The reason is that wave travel time through the watershedis faster than average discharge velocity (seeSection 13.1).As a result of using the slowervelocity I{ the peak time (/.) is overestimated, resultingintensityl from IDF curvesis too the small, and the rational flow rate p is underestimated. Rational Method Applications Most applications the rationalformulain deterof mining peak flow rates utilize the following steps:, 1. Estimatethe time of concentrationof the drainagearea. 2. Estimate runoff coefficient, the Table 15.1. 3. Selecta return period T, and find the intensity of rain that will be equaled or exceeded, the average, once every I years.To produceequilibrium on flows, this design storm must have a locally derived IDF curve such as Fig. 27.I3 or Fig. 15.2usinga rainfall durationequalto thetime of concentration. . 4. Determinethe desired peak flow Q,from Eq. 15.1. 5. Somedesignsituationsproducelargerpeak flowsif designstormintensities for durationslessthan /" are used.Substitutingintensitiesfor durationsless than t" is justified only if the contributingarea term in Eq. 15.1 is also reducedto accommodate shortenedstorm duration. the One of the principal assumptions the rational method is that the predicted of peak dischargehas the same'returnperiod as the rainfall IDF relation used in the
IDF curves for storms in vicinitv of example site

"

0 5 10 15 20

30

40

50

60

70

80

90

100 110 r20

130

Time (min)

Figure 15.2 Intensity-duration-frequency curvesusedin Example 15.1.

15.2 PEAKFLOWFORMULASFORURBANWATERSHEDS 315

prediction.Another assumption, onethat hasreceivedclosescrutinyby investigaand is tors,re'2o the constancyof the runoff coefficient during the progressof individual from a list stormsand also from storm to storm. The"coefficientis usually selected infiltration capacityof the drainagesurand basedon the degreeof imperviousness face. BecauseC : I,.rf I,the coefficientmust vary if it is to accountfor antecedent moisture,nonuniform rainfall, and the numerousconditions that causeabstractions rainfalls. In practice,a composite,weightedaverand attenuationof flood-producing agerunoff coefficientis computedfor the various surfaceconditions.Times of conof centration are determinedfrom the hydraulic characteristics the principal flow into two parts, overland flow and flow in defined path, which typically is divided channels;the times of flow in eachsegmentare addedto obtain /". Another assumptionwith the rational method is that the equation is most moisture conditions that exist for frequent storms,in the applicableto antecedent of interval,representative stormstraditionally used rangeof the 2- to 10-yrrecurrence Becausemore severe,less frequent for design of residential storm drain systems. moisture conditions,the rational coefficient is stormsoften have wetter antecedent by increased multiplying it by a frequencyfactor.The commonly usedmultipliers for lessfrequent stormsare:
period (yrs) Return
Multiplier

2-to 25
)U

100

1.0 1.1 t.2 1.25

EXAMPLE 15.I runoff ratesfor the area design and 5O-year to Usethe rational.method find the 10-year shownin Fig. 15.2arc applicable. shownin Fig. 15.3.The IDF rainfall curves Solution 1. Time of concentration: t,:tt*tz:15+5:20min

At = 3 acres

cr = o'3

tr = Az= Cz = tz =

15min 4acres o'7 5min

Figure 15.3 Hypotheticaldrainagesystem for Example15.1.

316

HYDROLOGY CHAPTER15 URBANAND SMALLWATERSHED

2. Runoff coefflcient: c : [(3 x 0.3) + (4 x 0.7)]lQ + 4) :0.53 for 10-yrevent C : 1.2(0.53): 0.64 for 50-yr event 3. Rainfall intensity-from Fig. 15.2: Irc : 4'2 in'/ht : 1so 5'3 in'/hr 4. Designpeak runoff: Q r c : C I A : 0 ' 5 3x 4 ' 2 x 7 Q s o : C I A : 0 ' 6 4x 5 ' 3 x 7 16 cfs 24 cfs rl

Rational Method Discussion The runoff coefficientin the rational formula is moisturecondition, recurrenceinterval, land on dependent the soil type, antecedent use, slope, amount of urban development,rainfall intensity, surface and channel of and roughness, durationof storm.Tablesand graphsgenerallyallow determination equationscan and regression C from only two or three of thesefactors.Nomographs providerelationsamongmore factors.One suchrelation, applicableonly in the region for which it was derived,is2r j 2(0.001CN1 ts-o't{(P+ 48)0 e1-s0 I)/zfo C : j .Z(t1-i)CN3To05[(0.01CN)o

n5.2)
where CN : SCScurvenumber(Chapter4) T : recurrenceinterval ( years)
land slope (7o) average I : rain intensity(in./hr) P : percentimperviousness Yet a The rational formula is a simplemodelto express complexhydrologicsystem. the by resultsimplying acceptance designmethodcontinuesto be usedin practicewith results. ers, officials, and the public. The methodis easyto apply and givesconsistent in the methoddemonstrates clearterms From the standpointof planning,for example, times of because increases surfaces runoff from developed the effectsof development: increase. and runoff coefficients concentrationdecrease is the For storm drainagesystems, designer normally askedto estimatethe peak at flow rate that might be equalledor exceeded leastonce in a given numberof years using the rational (described the frequency- see Section 10.4). For designs as equal to the frequencyof is assumed formula, the frequencyof the peak runoff event the rain event(an eventbeing deflnedas somerain depthin a given duration).Studies Figure 15.4 showscumulativelog-normal probabilhaveexploredthis assumption.z2 of ity functions (Chapter 26) fitted to observations rainfall and runoff on a 47-acre of imperviousness 0.44. The Surface area in Baltimore, Maryland, with an average and rainfall sequence the the observed to data are partial seriesfitted independently to correspond runoff doesnot necessarily Thus the largest runoff sequence. observed betweenany runoff lack of correspondence the largestranked rainfall, and a similar

s:

WATERSHEDS 317 FOR 15,2 PEAKFLOWFORMULAS URBAN

Percentage of sample values equal to or greater than indicated value . 99 5 9 9 9 8 95 90 80 7060504030 20 10 5 2 10.s

Rainfall frequency curve (rr = 7.5 min)

l"

./

!2 k

..t

tt

't'2'
" i/
oo

y"
o

o-'

2
a a

1 0.9 0.8 0.7 0.6 0.5 0.4 0.3

a o

Peak runoff ftequency curve

cd

5 1020
Recurrenceinterval (Years)

Figure 15.4 Distributions of recorded rainfall and runoff' (After Schaake.22)

in and the rainfall that producedit holds for the ranked position of the observations -yearrainfall frequencyof In the arraysof the two i"putut" sequences. Fig. 15.4, the 5 a the of4'0 cfs/acre; ratio indicates runoff io 6.5in./hr corresponds a runofffrequency eachcloselylogare coefficientof approximately0.6. Although the two sequences that the runoff coefficient increases normal, they tend to converge,which suggests slightly with more intense,iess frequent storms.In the designrange,however,the t"*ttr tend to support the assumptionof the rational method that the recurrence interval of the runoff equalsthe reiurrence interval of the rainfall. It shouldbe noted All in that the rainfall distributions Figs. 15.1and 15.4havesimilarproperties. IDF from many different rainfall intensitiesderived curvesare drawnthrough the average stormsof record; any single IDF curve dses not representthe progressof a single turns to the rational methodto storm.For lack of historicalrunoff data, the designer construct from the rainfall history what amounts to a runoff intensity-durationfrequencyrelation. by to The most critical (highestpeak) runoff eventis often assumed be caused If the watershed' the a storm havinga duration -qual to the time of concentrationof rainfall IDF curve is steepin the designrange,severaldurationsshouldbe testedfor the given frequency to assurethat no other storm of equal probability producesa higlier peak runoff iate. Most applicationsof the rational method do not includethis with the other thai ihe peak occursat /" is commensurate the testbecause assumption inherent assumptions.

31 8

CHAPTER15

URBANAND SMALLWATERSHEDHYDROLOGY

The rational method is used in the designof urban storm drainagesystems servingareasup to six hundredacresin size.For areaslarger than I mir, liydrograph or other techniques generally warranted.Considerabl" are 3udg-"nt is riquir-eAin selecting both the runoff coefficientsand times of concentratloi. a common procedure is to selectcoefficientsand assume that they remain constantthroughout the storm.As the designproceeds from point to point downstream, compositeweighted a C factoris computedfor the drainage areaaboveeachpoint. The time of concentration is composed an inlet time (the overlandand any channelflow timesto the first inlet) of plus the accumulated time of flow in the systemto the point of design. Figure 15.5 is an exampleof a designaid for prJdicting overland flow times. calculation of flow time in stormdrainscan readily be estimatid knowing the type of pipe, slope,size,and discharge.23 Generally,the pipe is assumed flow full foi this to calculation.(see Fig. 15.6.) Nomographs also are available solvethe Manning to equationfor flow in ditchesand gutters.The estimationof inlet time is frequentl| basedsolelyon judgment; reportedvaluesvary from 5 to 30 min. Denselyaevitopei areaswith impervioustractsimmediatelyadjacentto the inlet might be assigned inlet periods of 5 min, but a minimum value of 10-20 min is more uiual.

d o

10n
"-"

4U

Fo F

Figure 15.5 Surfaceflow time curves.(After FederalAviation Agency.23)

15.2 PEAKFLOWFORMUI.AS URBAN FOR WATERSHEDS 3.I9

1,500 1,000 800 600 500 400 300 200 2,400 2,000

1,500 1,000 800

0.004 0.005 0.006 0.007

0.4 0.3 o.2

0.7 0.8 0.9

s00
400

600

1.0

300 200

0.1 0.08 0.06 0.05 0.04 0.03 0.02 0.01 0.008 0.006 0.005 0.004 0.003 0.002
V) tl o o

6 cd bo

100 80 60 50 40 30

100 80

60 50 40
)n

po

10 8 6 5
i

r,-s-o.oto 0.008 0.006

a

0.10 0.08 0.06 0.05 0.04 0. 0.03 0.1 I , o,o2 0.015 3 g

h + a oo

3
a

10 8 6 5
A

0.001 0.0008 0.0006 0.0005 0.0004 0.0003 0.0002 0.000r 0.00008 0.00006 0.00005 0.00004 0.00003 0.00002

2 c
boo

3 2

: 8
o

1.0 0.8 0.6 1.0 0.5 0.8 0.4 0.6 0.3 0.5 0.4 0.2 0.3 0 . 1 3 0.2

o h o

0.00001 0.000008 0.000006 0.000005 0.000004

Figure 15.6 Flow in pipes (Manning's formula); (After Ref. 24.)

Most designers applyingthis methoddo not usethe time of concentrationin its strictestsense; rather, the largestsum of inlet time plus travel time in the storm drain systemis taken as the time of concentration.Caution is required in app$ing the method' Peak discharge not the summationof the individual dischapges, is b"Jarr* peaks from subareas occur at different times. The runoff from subareas should be recheckedfor eacharea under consideration. The average intensity / is that for the time of concentration of the total area drained. While I decreaiesas the design ploceeds downstream,the size of the contributing area increasesand normally e

320

CHAPTER15

URBANAND SMALLWATERSHEDHYDROLOGY

It continuously. shouldbe noted that the designat eachpoint downstream increases is a new solutionof the rational method.The only direct relation from point to point derivesfrom the meansfor determiningan incrementof time to be addedfor a new time of concentration.The effect is to provide an equal level of protection (i.e., an Example 15.2 is reproequal frequencyof surcharging)at all points in the system. referencesto illustrate the application of the rational duced from standarddesign method to an urban sYstem.2a EXAMPLE 15.2 of Basedon the storm sewerarrangement Fig. 15.7a,determinethe outfall discharge. Assume that C : 0.3 for residentialareasand C = 0.6 for businesstracts. Use a a 5-yearfrequencyrainfall from Fig. 15.7b andassume minimum 20-min inlet time. Solution. The principal factors in the designare listed in Table 15.2' Additional columns can be provided to list elevationsof manhole inverts, sewer This information is helpful in checkingdesigns inverts, and ground elevations. use and for subsequent in drawingfinal designplans.(see Table 15.3.) lI 'orational" that thepeak in Modified Rational Method Therationalmethodis truly flow rate is simply set equal to the net rain rate after sufflcient time occurs for the to entire watershed contributerunoff. This resultsfor any storm equallingor exceed-

design,requiring volume of runoff as well as peak flow rates.

IN OF TABLE 15.2 DEFINITION COLUMNHEADINGS TABLE15.3 Column Comment Line being investigated Inlet or manholebeing investigated Length of the line of Subarea the inlet Accumulatedsubareas Value of the concentrationtime for the area draining into the inlet Travel time in the pipe line WeightedC for the areabeing drained Rainfall intensitybasedon time of concentrationand a 5-yearfrequencyculve Uqitrunoff q: CI Accumulatedrunoff that must be carried by line Slopeof line Size of pipe Pipe capacity Velocity in full pipe Actual velocity in pipe

1 4 5 6 7 8 9 10 t1 T2 t3 14 15 t6 t7

15.2 PEAK FLOW FORMULASFOR URBANWATERSHEDS

o)

I
u\-q

_ 90lt
t
Legend

l
numbers Sewerlinewith manhole Storminlets Gutterflowline

MH1-1 r ,\e / --

contourline - areaoutline Drainage

\
a 5

N
\
J

2:i-yr averagefrequencY
l l r l ! l l

\ {

10-yr averagefrequencY
r

Ij-'ff:::i::::?

K:---t.-

10 20

30

40

50 60 70 80 Duration(min)

90

100 110 120

Figure 15.7 Sample storm drainage problem: (a) typical storm rainfall seier design plan anO (U) intensity-duration-frequency for Davenport, Iowa. (After Ref' 24.) curves

= E

pue lMo'l !y

6 r +.i6
6 A

o r 6ioi
6

n h oioi
@ @

n dt

o ;

puo roddnE
E o C o

1 q o 9 . l n
6 + i O 6 9 6 N O

\ \
6

N 6

! v ?
O @ @

puo JeMo'l di

? - : < : < x = x , x : V ( ] Y Y Y Y - ] : : c l o O O d + F t
O @ @

= o

puo reddnS
(g)]erresut tPl I (g)dorppenul"toqu"N $ qoqu"yl $ (4) sassol (u)^6reuaplol I

E X EE < i i
6 6 h 6 i O i @ o r

E q
* o @ @

? E qE q E q E f r
6 o i < l N : o o o

. n n

. q q

R -

=
c a o

( u0lrop;o qrdeOI (U)peeq,qcoten I (sdD',{lPoPn S

',qcoPn 11n1 $ 'r$cedeg 11n1 !

6 = = s :

F - R 3

\ O @

O \ r

6 r

A \ O

+r;ri
o - 6 $ n $

rid

d+
* : N

ris;
r

n n n

o c J c . ln n d ; 6 h =
N ! f , * - N i N N + O i

qcl S N S s

qq B H
N N

E
J

( uDrelauelc I (%)remes adols U to (sle)llounl lElol 3 ' (arceTsp) #orng $ (rq/'uI) ttBluleuI gounra6ere,rY @ lualclllooc
o
E ^
N

h h h r + d o o

o O o - i o

O O ' + O N 6 o 6 o o o d

k t
ul
T F t

s n + sq q -E q 3 N qE g
s 3
i i o a i

- q n ' 1 q . ' l \ 9 n \ 9 q \
0 0

r a c j c ' l
0 0

ua z

6 6 0

0 0

F } E
t!

+ 3 3 : 3 : $ .3i J A 3 A A+ : : :
n q c ' i ' 1 - ' 1 n
C Sq
O + N d N C ] +

u1 uolpes @ pualeddnoI F teloJ. 0 @ luer.uetcul (D q$uer I

i . ; N N

"

- :
O + N N N N N

09
6 O N N

3 uJ
o

6 6

E o
6

h O h O 9 c . . t. . 1 ' l q N Q N * i

qrl
d

*
N T \o$

6
J

qqo9
N O O

nn

qv-) oqn
o c l X X s s
O h d N a

n \ o

o E
f

: Y : X X + + < J ' * t h r
i i *

e
lf,'

oro
urorl O

o t
o

* t

o l

N l

d
o 3 N i *

o
g E

0)

lrl J
F

Y T i 1 J T l I T
: i : o * * * o o * N : i

eun e

W)

323 WATERSHEDS FOR FORMULAS URBAN FLOW 15,2 PEAK a In the modified rational method., full hydrographis developqdrather than the peak flow rate, using the following reasoning.If the storm simply estimating duration exceedslhetime of concentration,the runoff rate would dse to the rational At formula peakvalue,then stayconstantuntil net rain ceases. that point, runoffrates from the basin.Ifthe rainfall-excess rain is released to zero as excess would decrease releasetime (seeChapter 11) is equal to the time of concentration,the hydrograph rising to the peak al t : t",femaining flat trapezoldshape would havean approximate D until/: therainduration,D,andthenfallingalongastraightlineuntilt:I t". packagesfor urban hydrology incorporate the modified rational Many software method for hydrographanalysis.the method is approximateand shouldnot be apover 50 acresin size. plied to watersheds

SCSTR-55Method
for procedures estimatingrunoff volServicedeveloped The U.S. Soil Conservation They are known collectively as from urban areas.zs ume and peak ratei of discharge TR-55und indiuidually as thegraphical method,chart method, andtabular method. to in The threemethodsadjustrural procedures NEH-426 urbanconditionsby increasing the curve number CN foi impervious areas and reducing the lag time /1 for and imlperviousness channel improvements.Allowances are also made for various the The SCS designed first two slopes,and times of concentration. shapes, watershed complete for synthesizing methodsto be usedfor estimatingpeak flows,and the third up (usedfor small watersheds to The tabular method and chart method hydrographs. to help explainthe evolution here are wererevisedin 1986,21but described 2000 acres) for of the methods.All three were developed use with 24-hr storms'Use with other storm durationsis not advised. up watersheds, to 20 mr2 for The graphicalmethodwasdeveloped homogeneous by represented the runoff curve in size, on which the land use and soil type may be number. As shownin Chapter4, the runoff curve number is simply a third variable in a graph of rainfall versusrunoff. Tie SCS peak dischargegraph shown in Fig. 15,8 is limited to applications (see where only the peak flow rate ii Oisired for 24 hr, Type-II storm distributions expeof the 24-hr thunderstorm Chapter f 6l. A Type-II storm distributionis typical from rienced in all staiei exceptthe Pacific Coaststates.Figure 15.8 was developed in modeldescribed Chapter ofine SCSTR-20 eventsimulation numerousapplications time of concentrationin hours is enteredinto 24. To apply Fig. 15.8, the watershed per rate in cfs/mi2of watershed inch of net ttre grapir'to prJdu." the peak discharge from the 24-hr gross rain during tk Z+-nr period. The 24-hr net rain is estimated in amountusingthe scS curve number approachdescribed chapter 4. Fig. 15.9.Oncethe composite using the effectof urbanizationcanbe estimated for curvenumber(CN) hasbeenestimated the previousarea,a modifiedcurvenumber by is determined enteringFig. 15.9 with the value of the percentimperviousareaon to r"ading vertically to the curve corresponding the CN for the modified watershed, to determinethe modified the pervious watershed,and then reading horizontally compositerunoff curve numberthat would be usedin determiningthe net rain depth for the urbanizedwatershed' of Useof the 1975graphicalmethodis restrictedby the assumptions the tabular caseof the tabularmethod.This method.The methodii a -ompositeof resultsfor one

324

CHAPTER15

URBANNNO SUNU-WATERSHEDHYDROLOGY

'e 70n Ei 500 qF 400 E 300

o P,O R

E o

200

100 0.1

0.2

0.3 0.4 0.5 0.7 1.0 (hr) Time of concentration

2.5

3.0 4.0 5.0

(cfs/mi2lin.) runoff versus Figure 15.8 Peakdischarge of time of con(AfterU.S.SoilConsercentration for 24-hr,Type-IIstorm /. distribution. vation Service.25) restricts its applicationsto runoff volumes greaterthan about 1.5 in. (if the curve numberis lessthan 60). Time of concentration shouldrangebetween and 2.0hr, 0.1 and the initial abstractionshould not exceedabout 25 percent of the precipitation. The chart method allows determinationof peak flows for 24-hr Type-II storms over watersheds having a fixed length/width relation and no ponding areas.Three chartsareusedfor flat, moderate, steep or slopes approximately of I,4, or 16percent. Tablesof adjustmentsfor intermediateslopesare provided in the technicalrelease. Severalmicrocomputersoftwarepackages urbanhydrologyhavebeendevelfor oped.28 Over two-thirdsarebased SCSprocedures, cautionshouldbe applied on but

80
L) e .E

7n

u 6 0

l0

20

30

40 60 70 50 area(7o) Connected impervious

80

90

100

Figure 15.9 Percentage of impervious areas versus composite CNs for given pervious area CNs. (After U.S. Soil Conservation Service.2s)

325 WATERSHEDS FOR FORMULAS URBAN FLOW 15,2 PEAK that the commercialprogramsfully imitate TR-55 or other SCS handin assuming would carry would includeall three methods, An book methods. ideal TR-55 package would incorpoand would stateall assumptions limitations, and SCS endorsement, of percentage for rate all SiS adjustments peak coefflcient,percentimperviousness, and slope. length/widthratio variations, channelimproved,pondingor swampyareas, Its use shoutAako be cautionedfor other than 24-hr stormshaving a Type-II SCS not adheringto these limitations would not be qualified as distribution. Packages TR-55 procedures. is A significantproblem in someof the commercial softwarepackages the use for producehydrographs of a trianlular-shaped unit hydrographfor convolutionto to conceptualizethe stormsof-various durations.The SCS used a triangular shape use but peak flow rate of a curvilinear unit hydrograph, has never endorsed of other tabulatedhydrg; in than either the curvilinear shapediscussed Section I2.5 ot the graphsgiven in the TR-55 manual.For further reading,the SCS publisheda guide2e in for the useof the 1975TR-55 intendedto clarify procedures the original technical release.

SCSTR-55Method Prevailing
for is ratherthan the 1975version, recommended use. The 1986editionof TR-55,27 with the original . It incorporatesseveralyearsof resultsof researchand experiences edition. The revisionsinclude the following:

1. Three additional rain distributions(seeFig. 16'17). ,, Expansionof the chapteron urban runoff curve numbers' 3. A procedurefor calculatingtravel times of sheetflow' 4. Deletion of the chart method.
Modifications to the graphical peak dischargemethod and tabular hydrographmethod. 6. TR-55 computerProgram.

Tablet5.4 andFig.15'10 Ratherthanrelyingtotally on Fig 15.9,the newTR-55uses instancesindicated in the table' to provide urban runoff curve numbersfor certain For the new graphicalmethod,an urban curvenumberand the 24-hr designrain from the SCSrunoff 1, ih"n un initial abstraction is determined depth are estimated, peak flow is found from linear interpoequation(Chapter4) or from Table 15.5.The on lation of the curvesin Figs. 15.ll,15.I2, !5.13, or 15.I4, depending the rainfall ratio falls outsidethe curves,the distributiontype (Fig. rcn).If the computedI"f P nearestcurvi should be used. If the watershedcontains a pefcentageof ponds or swampyareas,the peak flow is multiplied by a reductioncoefficientfrom Table 15'6' EXAMPLE 15.3 concentration,CN: 75 A 1280-acreurbanTennesseewatershedhasa6.0-hrtimeof tainis 6'0 in' The25-year,24-ht of from Table15.4,and5 percent the areais ponded. Find the 25-yearpeak discharge.

TABLE 15.4 RUNOFFCURVE NUMBERSFOR URBANAREAS (see Sec. 4.9 foT other values)

Cover descriotion Average percent imperviousareaD

numbers Curve for hydrologic group' soil A

B

Cover type and hydrologic condition

c
86 79 74
98

Fully developed urban areas (vegetationestablished) Open space(lawns, parks, golf courses,cemeteries, etc.)" Poor condition (grasscover <50%) 68 Fair condition (grasscover 50-757o) 49 Good condition (grasscover > 757o) 39 Impervious areas Pavedparking lots, roofs, driveways,etc. (excluding right- of-way) 98 Streetsand roads Paved;curbs and storm sewers(excluding right-of-way) 98 Paved;open ditches(including right of-way) 83 Gravel (including righrof-way) 76 Dirt (including right-of-way) 72 Westerndeserturban areas Natural desertlandscaping(pervious areas only)' 63 Artificial desertlandscaping(impervious weed barrier, desertshrub with 1-2-in. sandor gravel mulch and basin borders) 96 Urban districts Commercial and business 89 85 Industrial 81 72 Residentialdistricts by average size lot 77 65 f acre or less(town houses) j acre 61 38 57 30 I acre 25 I acre I acre 20 5l 2 aqes l2 46 Developing urban areas Newly gradedareas(pervious areasonly, no vegetation)" 77 Idle lands (CNs are determinedusing cover types similar to thosein Table 4.7).

79 69 6l 98 98 89 85 82 77 96 92 88 85 75 72 70 68 65

89 84 80

98 92 89 87 85
96 94 91 90 83 81 80 79 77

98 93 9r 89

96 95 93 92 87 86 85 84 82

86

91

'Average runoff condition, and 1" : 9.25. 'The averagepercent impervious area shown was used to developthe composite CNs. Other assumptionsare as follows: impervious areas are directly connected to the drainage system,impervious areas have a CN of 98, and pervious areas are consideredequivalent to open spacein good hydrologic condition. CNs for other combinations of conditionsmay-becomputedusingFig. 15.9 or 15.10 " CNs shown are equivalent to those of pasture. Composite CNs may be computed for other combinations of open spacecovol type. dComposite CNs for natural desertlandscapingshould be comppted using Fig. 15.9 or 15.10 basedon the impervious areapercentage(CN : 98) and the pervious area CN. The pervious area CNs are assumedequivalent to desert shrub in poor hydrologic condition. eComposite CNs to use for the designof temporary measures during grading and construction should be computed using Fig, 15.9 or 15.10basedon the degreeofdevelopment (impervious areapercentage)and the CNs for the newly graded pervious areas. : Source U.S. Soil Conservation Service,2T

0.0

n5

> o

r.o I
I

70

60

50
Total impervious arca (Vo)

Composite CN

Figure 15.10 Graph of 1986 TR-55 composite CN with unconnected imperviousarea,or total imperviousarea,lessthan 30 percent.(After U.S. Soil . ConservationService.2T) TABLEls.s /aVALUES FORRUNOFF NUMBERS CURVE Curve number 40 4l 42 43 44 45 46 47 48 49 50
)l

L fin.)
3.000 2.878 2.762 2.65r 2.545 2.444 2.348 2.255 2.167 2.082 2.000 1.922 r.846 1.774 1.704 1.636 1.571 r.509 1.448 1.390 1.333 1.279 1.226 I.t75 t.t25 1.077 1.030 0.985 0.941 0,899

number Curve 70 71 72
t5

L (in.) 0.857 0.817 0.'778 0.740 0.703 4.667 0.632 0.597 0.564 0.532 0.500 0.469 0.439 0.410 0.381 0.353 0.326 0.299 0.273 0.247 0.222 0.198 0.174 0.151 0.128 0.105 0.083 0.062 0.041

74
IJ

52 53 54 55 56 57 58 59 60 6l 62 63
o+

65 66 67 68 69

76 77 78 79 80 8l 82 83 84 85 86 87 88 89 90 9l 92 93 94 95 96
o1 98

Source:U.S. Soil Conservation Service.

328

HYDROLOGY CHAPTER15 URBANAND SMALLWATERSHED

300

u o P0 R
I

tnn

3 S

100 R o 60

40 01

0.6

0.8 l

Time of concentration, Z; (hr)

(q*) for SCS Type-I rainfall distribution. (After Figure 15.11 Unit peak discharge Service.) U.S. Soil Conservation

9
Po

100
Ro

ci

60

30 0.1

0.2

0.4

0.6 0.8 1

810

Zr Time of concentration, (hr)

(q,) Figure 15.12 Unit peakdischarge for SCSType-IA rainfall distribution.(After Service.) U.S. Soil Conservation

15.2 PEAK FLOW FORMULASFOR URBAN WATERSHEDS 1000 800 600 500 400

329

^ E
o 90

{, s 300 2oo
6 d

= P 1oo
80 60 50
0.1

0.2

0.4

0.6 0.8 1

810

Z" Time of concentration, (hr)

(q,) for SCSType-II rainfall distribution. (After Figure 15.13 Unit peak discharge Service.) U.S. Soil Conservation

700 600 500 400 < 300

-*
d E
E

,nn

= 5 roo
80 60 40 0.1

0.2

0.4

0.6 0.8 1

810

Time of concentration, Z, (tn)

(q,) Figure 15.14 Unit peak discharge for SCSType-III rainfall distribution.(After Service.) U.S. Soil Conservation

330

CHAPTER15

URBANAND SMALLWATERSHEDHYDROLOGY

TABLE 15.6 ADJUSTMENTFACTOR(Fp)FOR POND AND SWAMP AREAS THAT THE ARE SPREADTHROUGHOUT WATERSHED Percentageof pond and swamp areas

0 0.2 1.0 3.0 5.0

Service. Source:U.S. Soil Conservation

1.00 0.97 0.87 0.75 0.72

From Solution. From Fig. 16.17, the Type-II storm appliesto Tennessee. q,: 96 csm/in. I":0.667.Thus I,fP: 0.11.FromFig. 15.13, Table15.5, From Chapter4, tberunoff from 6.0 in. is 3.28 in. Since5 percentof the area usingTable 15.6,giving 4 : 0.72. Thus is ponded,the peak flow is adjusted g : (96 csm/in.)(3.28 in.)(2.0 mr')(0.72): 453 cfs rr only. If a hydrographis needed The graphicalmethodprovidespeak discharges shouldbe used.The event subdivisionis required, the tabular method2T or watershed is simulationmodel TR-20 shouldbe usedif the watershed very'complexor a higher degreeof accuracyis required (seeChapter24). Assumptionsof the graphicalmethod include: The method shouldbe used only if the weighted CN is greaterthan 40. The ?i valueswith the method may range from 0.1 to 10 hr. that is, describable The watershedmust be hydrologicallyhomogeneous, by one CN. Land use, soils, and cover must be distributed uniformly throughoutthe watershed. The watershed may haveonly one main streamor, if more than one, the branchesmust havenearly equal times of concentration' The method cannotperform channelor reservoirrouting. that are not on the The Fofactorcan be appliedonly for pondsor swamps flow path. estimatedby this method will be reducedif Accuracy of peak discharge I"fP va\uesare usedthat are outsidethe range given. When -this method is used to develop estimatesof peak dischargefor use presentand developed conditionsof a watershed, the sameprogedure for estimating[. Both the graphicalandtabular methodsare derivedfrom TR-20 output.The use within the scopeof the curves of I permits them to be usedfor any size watershed watershedthat is or tables. The tabular method can be used for a heterogeneous for Hydrographs the subwatersubwatersheds. dividedinto a numberof homogeneous shedscan be routed and added.

15,3

PEAK FLOW FORMULASFOR SMALL RURAL WATERSHEDS

331

and is not detailedhere. in The tabularmethodis described the technicalrelease are employed: In using the method, the following steps and 1. Subdividedthe watershedinto areasthat are relatively homogeneous reaches' haveconvenientrouting in 2. Determinedrainageareaof eachsubarea squaremiles' in hours. The procedurefor estimatingI is 3. Estimate T"for eachsubarea outlinedinTR-55. 4. Find the travel time for eachrouting reachin hours' 5. Developa weightedCN for eachsubarea. 6. Selectan appropriaterainfall distribution accordingto Fig. 16.17. frequency(Chapter 16). e 7. Determin the 24-hr rainfall for the selected from CN andrainfall (Chapter4)' 8. Calculatetotal runoff in inchescomputed from Table 15.5. 9. Find I,fot eachsubarea selectone of the hydrographs 10. Usingthe ratio of I,f P andT,for eachsubarea, in tabulated TR55. 11. Multiply the hydrographordinates(csm/in.) by the area (mi2) and runoff subarea. (in.) of eachrespective 12. Route and combinethe hydrographs. The SCS recommendsthat TR-20, rather than the tabular method,be used if any of the following conditions apply: Travel time is greaterthan 3 hr. f is greaterthan2hr. differ by a factor of 5 or more. Drainageareasof individual subareas by havebeenincorporated SCSin a computerprogram.Copies The TR-55procedures are availablefrom the U.S. National TechnicalInformation Service.

'

FORSMALLRURALWATERSHEDS 15.3 PEAKFLOWFORMULAS Method SCSTP-149

TR-20 is the unit-hydrograph TR-55 is the SCS procedure for urban watersheds, (seeChapter24), andTP-149wasdevelprocedurefor larger agriculturalwatersheds bped to allow esiimation of peak flow rates from small (5-2000 acres)agricultural of It watersheds.3o consistsof a seriesof 42 charts from which the peak discharge a 24-hr ruinfall can be determined. watershedslope, storm Input to the procedure is the drainage area, average (I or II), watershed compositecurve number, and depth of rainfall. distributiontype chartsin the TP. Shownare type-I Figures15.15ind 15.16illustratethe numerous with CN : 70 for both. Similar slopedwatersheds, and type-Il curvesfor moderately given in Table 15.7.Applicationsof TP 149 for chartsare available the combinations havingcurve numbersother than the 5-unit incrementsof Table 15.7, to watersheds by or for slopesother ihan I, 4, or 16 percent, can be accomplished arithmetic or adjacentchart values' logarithmic interpolationbetween

332

CHAPTER15

URBANAND SMALLWATERSHEDHYDROLOGY

h\or@e

g ?3eR3=

^^^ e H =HEF==

'/,
1000 800 7N 600 500 400 300 200
o P!

,<\ 7
l
r!

,4

1000 800 700 600 500 400 300 200

.E

rno

'z ..\"\
L'//
l
'//.2

2\

8 0 & 7 0 3 a o 50 40 30

W
L

& .Mt/,,
7

100 80 70 60 50 40 30

t0 8 ,7 6 5
h

ffi
/ttl
9 r @ o o o o - N - + o n o o o o r @ o
O

a
v'

20 l$

H
l/,v ,/: R
O

10 8 7
o

) g?fiFAt
O O O O O O O

8
N

Drainage area (acres)

-tP-I49 peak Figure 15.15 ratesof discharge small watersheds, for Type-I storms:24-hr rainfall, moderate slopes, and CN : 70. (After U.S. Soil ConservationService, "A Method of EstimatingVolumeandRateof Runoff in Small Watersheds," U.S. Depaftment of Agriculture,Jan. 1968.)

TABLE 15.7 CHARTSAVAILABLEIN TP-149FOR PEAK FLOW RATESOF SMALL WATERSHEDS

Storm distribution rype

I,il

Slopetype Flat, TVo Moderate,4To Steep,167a

Slope range (%)

Curvenumber, CN 60,65,70,75,80,85,90 60,65,70,75,80,85,90 60,65,10,75,80,85,90

LU
I,il

0-3 3-8 8-30

333 RURALWATERSHEDS FOR FORMULAS SMALL FLOW 15,3 PEAK

nre
1000 800 700 600 s00 400 300 200

R se8sRsc

R ==FeF=- F

;90
E IUV

8 0 r.1 70

fi o o
50 40 30

100 80 70 60 50 40 30 20

10 8 7
o 5

10 8 7 6 5

nF-=

3?38RAE
Drainage area(acres)

= = x = = x x X ; + E 6 F d 6

Type-II storms: Figure 15.16 TP-149 peak rates of dischargefor small watersheds, "A 24-hr rainfalT.moderate slopes,CN : 70. (After U.S. Soil ConservationService, U.S.Department Method of EstimatingVolumeand Rateof Runoff in Small Watersheds," of Agriculture'Jan.1968.)

EXAMPLE 15.4 Comparethe peakflow ratesfrom Type-I and Type-II stormsusingFigs. 15.15and and that only stormtype changes all other conditionsare equal. 15.16.Assume with CN : T0lesultsin on Solution. A 4-in. rain over200 acres a watershed : II Qo : 52 cfs for a Type-I storm (Fig. 15.15) and Qp 9l cfs for Type in (Fig. 15.16).Thus the storm distributiontype makesa significantdifference results of peak flow estimationusing SCS techniques. I r

334

CHAPTER15

URBANAND SMALL WATERSHEDHYDROLOGY

FederalHighwayAdministrationscs PeakFlow DesignMethod

The FederalHighway Administration (FHWA) lists in their HydrologicEngineering ( Circular No. tq, F{ydrology 1995Ed.) a procedurefor estimatingpeak flow ratesfor havingtimes of concentrationbe,-ul1-to--"dium sizedwatersheds homogeneour, equationthat has coefficients an SCSregression tween 0.1 and 10 hours.It employs rainfall distributiontypes and ratios of the initial determinedfrom data on different in 1, abstraction (seeChapter4) and total precipitation,P. The peak discharge metric units is calculatedfrom
Qp: 4,AQ,

( 1s .3)

A in where qois the peak discharge m3/sec, is the drainageareain sq. km., Q is the from net rain depth in cm, andq, is the unit peak discharge
log qu: Co + Cl log /" f Crlogz t"

(1s.4)

coefficientsare in which /" is the time of concentrationin hours,and the regression obtainedfrom Table 15.8.
FOR FHWA HEC-19SCS TABLE 15.8 COEFFICIENTS PEAK DISCHARGEMETHOD Rainfall

rype

t,/P 0.10 0.20 0.25 0.30 0.35 0.40 0.45 0.50 0.10 0.20 0.25 0.30 0.50 0.10 0.30 0.35 0.40 *0.45 0.50 0.10 0.30 0.3s 0.40 0.45 0.50

co 2.30550 2.23537 2.18219 2.10624 2.00303 1.87733 t.76312 1.67889 2.03250 r.91978 1.83842 1.72657 1.63417 2.55323 2.46532 2.41896 2.36409 2.29238 2.20282 2.473t7 2.39628 2.35477 2.30726 2.24876 2.17772

c1 -0.51429 -0.50387 -0.48488 -0.45695 -0.40769 -032274 -0.15644 -0.06930 -0.31583 -0.282t5 -0.25543 -0.t9826 -0.09100 -0.61512 -0.62257 -0.61594 -0.59857 -0.57005 -0.51599 -0.51848 -0.51202 -0.49735 -0.4654r -0.41314 -0.36803

c2 -0.11750 -0.08929 -0.06589 -0.02835 0.01983 0.05754 0.00453 0.0 -0.13748 -0.07020 -0.02597 0.02633 0.0 -0.16403 -0.1t657 -0.08820 -0.05621 -0.0228r -0.01259 -0.17083 -0.13245 -0. l 1985 -0.1 1094 -0.11508 -0.0952s

IA

ilI

Source: Afrer U.S. Federal Highway Administration' Hec-19, Hydrology' FHWAIP-95, 1995.

15.3 PEAK FLOW FORMULASFOR SMALL RURALWATERSHEDS

335

The procedurehas the following limitations: ' use with homogeneous (CNsfrom zone to zone shouldnot differ watersheds CN shouldbe greaterthan 50 /. shouldbe between0. 1 and 10 hours I^/P shouldbe between and 0.5 0.1 /" shouldbe about samefor any of the main channels, watershed more if has than one main channel . no channelor reservoirrouting is allowed . no storagefacility on main channel ' watershedareain storage ponds and lakes shouldbe lessthan 5 percent . . . '

bYs)

SyntheticUnit-Hydrograph PeakRateFormulas
Peak flow rates from small watersheds can also be determinedusing the synthetic unit-hydrographtechniquesdescribedin Chapter 12. A storm having a duration definedby Eq. 12.22wirl produce,accordingto Snyder'smethodof synthesizing unit hydrographs, peak discharge 1.0 in. of net rain given by Eq. 12.17,or a for

o^ '

64oct'A
t m

( 15.s )

Similarly, the peak flow rate resulting from a storm with duration D given by Eq. 12.22or 12.23is, according the SCS methodfor constructing to synthetic unit hydrographs, equal to

484A ^ U^: -

(15.6)

where /o is the time from the beginningof the effectiverain to the time of the peak runoff rate, which by definitionis the watershed time plus half the stormduration. lag Both of Eqs. 15.5and 15.6 apply to 1.0 in. of netrain occurringin the durationD. Either can be multiplied by P"", for other storm depthswith equal durations.Peak flows for stormswith durationsother than D would need to be determinedby unithydrographmethods.

Discharge-Area Regression and Formulas

A multitude of peak flow formulas relating the discharge rate to drainageareahave beenproposedand applied.Gray3rlists 35 suchformulas,and Maidment32 compares many others.Most of theseempiric equations derivedusingpairsof measurements are of drainagearea and peak flow rates in a regression equationhavingthe form where Q: CA* : the peak discharge associated with a given return period Q A : the drainage area C, ffi : regression constants (15.7)

Popular discharge-area formulas in the form of Eq. 15.7 include the Meyers equation33 ( 1s.8 ) O : 10,000405

336

CHAPTER15

URBANAND SMALLWATERSHEDHYDROLOGY

where A : the drainagearea,which must be 4 mi2 or more Q : the ultimate maximum flood flow (cfs) This examplegives only one flow rate of unknown frequencyand is chosenonly to illustrate the form of flood flow equations.A program of determiningflood magnibasishas been completedby the on tudesfor atange of frequencies a state-by-state in discussed Chapters26 and 27 and techniques USGSusing the multipl- regression formulas are availablefrom the illustrated for Virginii in Problem 27.25. Similar for equations the Softwarecontainingall the USGSregression USGSfor other states. and Federal Highway United Statesis availablefrom the U.S. GeologicalSurvey Administratioh as part of the HYDRAIN softwarepackage'

Method SurveyIndex-Flood Geological U.S.

in Surveyindex-floodmethoddescribed Section27.4 is a graphThe U.S. Geological peak dischargerates. The ical regional c&relation of the recurrenceinterval with stepsinvolvedin the derivationof a regionalflood index curve are outlined in Section is to 27.4.The first stepin applyingthe technique a watershed to determinethe mean annual flood, defineOas ttre flood magnitudehaving a return period of 2.33 years. equations are Mean annual floods for ungaugedwatersheds found from regression and on flood magnitudes the similar in form to Eq. 15.7.For example, USGSreport3a in frequencies Nebraskagives,in cfs,
Qz'zt = CA0T

(1s.e)

where A : the contributing drainage arcain mi2 obtainedfrom Fig. 15'17 C : aresionalcoefficient Once the mean annualflood magnitudeis obtained,other annual flood magnitudes can easily be determinedfrom the appropriate index-flood curve (see Fig. the 26.4c).The usqof suchcurvesin urban hydrologyis limited because USGS data smallerthan 10 mi2. method seldomincludeswatersheds index-flood network for the in later, areapplicablefor watersheds the described equations, The USGSregression rangeand larger. 1-10 mi2

'

CyprusCreekFormula
includingestimates Extreme$ flat areasposeparticular difficulties to the hydrologist, runoff rates.Flooding in theseareastends!o runoff volgme,and,peak of infiltration, Flow velocitiesare low, and water standson the surface be shallow and widespread. by long periodsof time. Theseareasare often distinguished networksof for relative)y ihannels that have been constructedto store and eventually disstraight drainage rain. chargethe excess peak flow to a developed procedure3s calculatethe'instantaneous The Scs basedon first calculatingthe capacityof canalsthat would be flatland areas to limit flat-xeaflooding for the designstormto a duration that would prevent crop damage,and then to apply a multiplier to this rate to obtain the instant The procedureis illustrateil in Fig. 15. 18. peakfor the designof drainagestructures.

.oe

OOIfr

jtoo

.-> ' -46 *'-1

I' 1 |

'o?

' r l
.t\ t

6 . 1

lo.S

t "

:.9

9; E d
a =

x(.) s ! P>,

ra)

. Y v

- 92,

F\9
i : t6 r
9

I'o

a F = Y

;T{

x
U ',;
E O (J.n

20

x E O o ti= ;r o d c

ui dr :- ; < r t

: E.e .$e b"

338

CHAPTER15

URBANAND SMALLWATERSHEDHYDROLOGY

.--Duratio4

of Overbank Flow-->l InstantaneousPeak

o o

Drainage Ditch CaPacitY "Removal" Rate Based on

Iime

Figure 15.18 Illustration of relation befweenCyplys Creek flow. (After Soil ConservationService'35) stJntaneous

"removal" rate and peak in-

to duration was} hours,considered be the maximum allowabletime for The selected to called the CyprusCreekformula, wasdeveloped inundationof crops.An equation, flow rate, called the 24-hr removal rate. The equation, determinethe canal design curve based on rainfall depth,-contributingdrainage area,and the SCS composite number is
Q: CA5/6

( 1s .10)

channelcapacity fot 24-ht removal (cfs) where Q : reqrrired : drainagecoefflcient C areaGq mi) A: drainage bY coefficient,C,fot Eq. 15.10is found from an equationdeveloPed The drainage and Mi1ls36 Stephens (15.1) C : 16.39+ (14.75Q,",) where designeventfrom Fig' 4'I4' Q"", : the scS direct runoff (in.) for the 24-hr peak flow rate is Once Eq. 15.10is solvedfor the given frequency,the instantaneous from 1 to about areas drainage is from Fig. 15.19.The procedure limited to obtained rate peakinstantaneous to the 24-hr that 200 squaremiles.It is suggested ratios of the than or equalto 1.0.For flatland areas canaliemoval rateUetirniLO b values greater that the peak flows that havepart of the areain storm ,"wJts, the SCS recornmends 15.20.TheSCSfurther fromFig. ts.tqu.increasedbytheamountsindicatedinFig. that haveslopesthat are restrictingur" of thit procedureto watersheds recomniends methodssuchas TR-20, TR-55' other lessthan 0.002. For stJeperslopewatersheds, equationsare recommended' TP l4g, or regression

- <
I q

i)c<

o11

&

e&EEh+-il1e

t * .
4 q

z ^
;i^ - Y

3lj!.) = , = * d q

;>
=

N N

n $

Q O

o a - N

.e #
o .

(.) F
a

e f q + + t r
t < r o o .

E E b
c6

e
!

>

()
r ,
!

+ '; o c{
t q * c ' t 67)
9 t s
, L

9 H i 5
>, jl ,* -Q
v

- qca !t ,-:, l,o

. \ 4

:
N

o
N

0 ) d L r
: r I

3rX

eleX ;;oung e8ere^V JnoH-tZ runrxrxehtr 01 luod snoeupluelsuJJo oqed

340

CHAPTER15

URBANAND SMALLWATERSHEDHYDROLOGY

s40
o !l

o 9 ^ ^ i 4rl

0
Percentage Area Servedby Storm Sewers of

Figure 15.20 Effect of urban storm sewerson peak dischargefor urban areas. (After U.S. GeologicalSurvey.aT)

EXAMPLE 15.5 Usethe Cypro, Creek methodto determinethe peak 50-yr flow rate fron a 1.0 sq mi and drainage areathat hasa CN : 80,is 50 percent stormsewered, hasa50-yr,24-hr rainfall depthof 12.0inches. of Solution. From Fig. 4.I4, thedirectrunoff for 12 inches rain is 9.45in. The drainage coefficient,C, is found from Eq. 15.11,

: + C : 16.39 (14.75X9.45)155.t

FOR SMALLRURALWATERSHEDS 341 15.3 PEAK FLOW FORMULAS

The 24-hr removalrate is found from Eq. 15.10' : '0)5/6 155'8cfs 0 : 155'8(1 rate From Fig. 15.19the ratio of instantaneous to removalrate is 2.0, giving a From Fig. 15.20,it is existed. designflow rate of 311.6cfs if no stormsewers by shouldbe increased 35 percentfor areadischarge found that the unsewered a watershedwith 50 percent storm sewers.The final designflow is 1.35 x 31'1.6: 420.7 cfs. rr

Equations SurveyRegression U.S.Geological for UrbanAreas

The U.S. GeologicalSurvey,in cooperationwith the FederalHighway Administraa tion, conducted nationwidestudy of flood magnitudeand frequencyin urban waterincludat involved26ggaugedbasins 56 citiesin 31 states, The investigation sheds.37 Hawaii.The locationsare shownin Fig. 15.21.Basin sizesrangedfrom 0.2 to ing 100 mi2. parame(see,Chapter of a variety of independent 27) Multiple linear regression conductedto developpeak flow equationsthat could be applied to small, ters was throughoutthe United States.Similar USGS regression urban watersheds ungauged in for large rural basinsare described Chaptet27' equations involvesthe three most equations regression form of the developed The simplest variablesidentified.Thesewere contributing areaA (mi2),ba-sin^developsignificant peak flow RQ,(cfs)for the lth and ment factorBDF (dimensionless), the corresponding The from an identicalrural basinin the sameregionasthe urbanwatershed. frequency from can accountsfor regionalvariations,and estimates be developed latter vaiiable reports(seeSection27.4). The threeUSGSflood frequency any of the applicable flowsaregiven for parameter equations the 2-,5-,IO-,25-,50-, 100-,and500-year
as37

ez: l3.2Ao.zt(13 BDFl-o.azpnotz es : 10.6Ao.rz(13 BDF)-o.3eReo18 t0(13- BDnl-otuRQ?dn Qto : 9.5rAo ts(13- BDF)-o'z+P9o'to Qt5 : 8.68Ao ts(13- BDFl-o'zzR03o" Qso: 8.o4Ao ts(13- BOrT-ot'RQ?r!& Qno: 7.70Ao : A0'16(13 BDF)-'*RQ1i& Qsoo 7.47

(Ls.r2) (1 .1 ) s 3
(15.14)

(1s.1s) (1s.16)
(15.17) (1 .1 ) s 8

from data at 199of the 269 original sites.The other siteswere Thesewere developed detentionstorageor missingdata. All theseequaof the presence.of because deleted coefficientsof determinationabove0.90. tions have used values and of the Figure 15.22shows correspondence estimated observed Eq. 15.15.Forty percentof the valuesfall within one standarddeviation in devedping line. Graphsfor other recurrenceintervalsare similar to the 1O-year of the regression graphshownin Fig. 15.22.

O N o v N E > li p rii

'

h E

- ^ a s - _ . A

>'
q

,F

i .-r : t

'o 3 r {
g

o
o / o

()

B --l I I I I L,/ I
I

r d \Oa

l n
l V z

I 0".
) r.' U
I j A f
r E l

ol

E'J

,g

q g

--j

, l
: : F

: (,)
7

i I

PROBLEMS 353

15.5,

rainfall duration? is The 4-hr unit hydrographfor a 5600-acrewatershed

Time (hr) 0 (cf9
n 1

400

8 4 6 1000 800 400

10 200

t2 0

: discussthe Rework Example 15.2 basedon a C : 0.2 and C 0'4' Compare and at effect of C on the discharge the outfall' AwatershedhasareaA.Startingwithatriangular.shapedunithydrographwithabase : 484A/to' lengthof 2.67t, and a heightof [0, deriveEq. 15'9 (seealsoEq' t2'25)' Qo the derivation' State and carr units of eachterm usedin

UsingtheSCSdimensionlessunithydrographdescribedinChapter12,determinethe to peak peaidischargefor a net storm of 101n.in 2 hr on a 400-acrebasinwith a time ^of + nt and i lag time of 3 hr. Comparewith Eq' 12'17' with a 100-min time of concentrationreceivesrainfall at a A 10.00-mi2watershed rate of 2.75 in.lhr for a period of 200 min' 1f (cfs)from the watershed C : 0'4' a. Determinethe peak d-ischarge rate lcfs) 150 min after the beginningof rainfall. b. Estimatethe discharge rate from the watershed 40 min after the beginning of c. Estimate the dischar"ge rainfall.
F

E * n

{
*
F

F 6 0

9 4 0
o o

E - " ) i o
s? o

o
-=

100

-Time{min)-after-beianing-of rainfall

354

CHAPTER15

URBAN ANDSMALL WATERSHED HYDROLOGY

1s.13. A storm gutter receivesdrainagefrom both sides.On the left it drains a rectangular
600-acreareaof t" : 60 min. On the right it drains a relatively steep300-acrearea of t" : 10 min. The f index on both sidesis 0.5 in.ihr. Use the intensity-durationfrequency curves in Fig. 15.7 to determinethe peak discharge(cf$ with a25-year recurrenceinterval for (a) the 600-acreareaalone, (b) the 300-acrearea alone,and (c) the combinedareaassuming that the proportion of the 600-acreareacontributing to runoff at any time r after rain beginsis l/60. t5.14. A drainagebasin has a time of concentrationof 8 hr and producesa peak Q of 4032 cfs for a 10-hr storm with a net intensity of 2 in./hr. Determinethe peak flow rate and the time base(duration)of the direct surfacerunoff for a net rain of 4 in./hr lasting (a) 12 hr, (b) 8 hr, and (c) 4 hr. State any assumptions used.

15.15. A 1.0-mi2parking lot has a runoff coefficientof 0.8 and a time of concentrationof
(cf$ by the 40 min. For the following three rainstorms,determinethe'peakdischarge rationalmethod:(a) 4.0 in./hr for 10 min, (b) 1.0in./hr for 40 min, and (c) 0.5 in./hr for 60 min. State any assumption regarding area contributing after various rainfall durations. 15.16. The concentration time varies with dischargebut is relatively constant for large feel confidentin usingthe rational discharges. From this statement, why do engineers formula? 15.17. Determinethe 50-yearflood for a20-mi2 basin at the northwestcorner of Nebraska. that Fig. 26.4 appbes. Use the index-flood method and assume 1s.18. Determinethe entire frequencycurve for the basin in Problem 15.17 and plot it on probability paper.

15.19. Use the index-flood method to determinethe 10- and 50-yearpeaksfor a 6400-acre
drainagebasin near Lincoln, Nebraska.Assumethat Fig. 26.4 applies. 15.20. For the drainagebasin in Problem 15,19 determinethe probability that the 20-year peak will be equaledor exceeded leastonce (a) next year and (b) in a 4-yr. period. at Referto Section26.1.

15.2L. For a 100-mi2drainagebasinnearLincoln, Nebraska,usethe index-flood methodto

determinethe probability that next year's flood will equal or exceed3000 cfs. 15.22. UseFig. 26.4 to determinethe return period (years)of the meanannualflood for that region.How doesthis comparewith the theoreticalvalue for a Gumbel distribution? How doesit comparewith a normal distribution?Refer to Section26.6. for ts.23. Usethe Cyprus Creekmethodto determine 25-yr peak discharge the watershed the is describedin Example 15.3. Assumethat the watershed nearly flat.

15.24. You are asked to determinethe magnitudeof the S0-yearflood for a small, rural
drainage basin (nearyour town) that hasno streamflowrecords.Statethe namesof at leasttwo techniques that would provide estimatesof the desiredvalue. 't5.25. The drainageareas,channellengths,and relevantelevations(underlined)for several subbasins the Oak Creek Watershed Lincoln, Nebraska,are shownin Fig. 24.8. of at The watershed a SCScurve numberof CN : 75 which may be usedto determine has the direct runoff for any storm.Assumethat IDF curvesin Fig. 27.13 applyat Lincoln. Treat the entire watershed a singlebasinand determine 50-yearflood magnitude the as at Point 8 using: a. The rational method. b. The SCSpeak flow graph,Fig. 15.8. Eq. 15.5. c. Snyder'smethod of syntheticunit hydrographs, d. The USGS index-flood method. Figure 26.4 applies.

REFERENCES 355

15.26. RepeatProblem 15.25 with SubareaI excluded.compare the results with Prob-

event of the at lem 15.25 and comment on the effectiveness Point 8 for the 5O-year flood easilystorethe 100-year BranchedOak Reservoirat Point 9. (This reservoirwill from Area I.) 1s.27. RepeatProblem 15.25 for SubareaA' 15.28. RepeatProblem 15.25 for SubareaI' is 15.29. Describecompletelyhow the magnitudeof the 30-yearflood for a watershed determined by the USGS index-flood method. 15.30. A rural watershedwith a composite cN of 70 is being urbanized. Eventually' 36 percentof the areawill be impervious.Determinethe increasein runoff that can for be expected a 6.2-in.rain. unit peak flow for the SCS dimensionless hydrographin Ch. l2,_determine 15.31 Using the for the piak discharge a net storm of 10 in. in Zhr on a 400-actebasinwith a time to peak of 4 hr and a lag time of 3 hr. shown L5.32. A timber railroad bridgein Nebraskaat Milepost 27I.32 ontherailroad system in the sketchis to be replacedwith a new concretestructure.The 50- and 100-year are flood magnitudes neededto establishthe low chord and embankmentelevations, respectively.Determine the designflow rates using the scs TP-149 method. The bridge drainsthe zone marked,about45 acres.The moderatelyslopedbasinlies in a rainfall rype-n stormregion,the curve numberis 70, and the 24-hr 50- and 100-year g.4" respectively' depthsare 8.6" and

Bidge27l.32

method.The problem 15.32usingthe FHWA HEC-19 peakflow SCSdesign 15.33. Repeat from the relationships can be determined is 0.2 hrs. Valuesof 1o time of concentration in , in Fis. 414. Provide the answers both metric and English units.

REFERENCES
"A Critique of Current Methods in Hydrologic Systems 1 . J. Amorocho and W. E. Hart, Investigations Trans.Am. Geophys.Union 45(2),301-321(Jwe 1964)' ," ..NonlinearInstantaneous Unit-HydrographTheory," ASCE J. Hyd. Div. 2. f. p. jingh, -347(Mar' 1964). 90(HY2), Par I, 313

356

HYDROLOGY 15 ANDSMALLWATERSHED CHAPTER URBAN "ContinuousHydrographSynthesis with an and 3. W T. Sittner, C. E. Schauss, J. C. Monro, Res.5(5), 1007- 1022(1969). API:Iype HydrologicModel," WaterResources 4. J.E.Nash,"TheFormoftheInstantaneousUnitHydrograph,"Int.Assoc.Sci.Hyd.3@5), "Mathematical Models of CatchmentBehavior," Proc. 5. D. R. Dawdy and T. O'Donnel, ASCEJ. Hyd. Div.91(HY4), 124-127(Iuly 1965). J. 6. S. L. S. Jacoby,"A MathematicalModel for Nonlinear Hydrologic Systems," Geophy. Res. 7l(20), 48t | - 4824(0ct. 1966). Model," Proc. ASCE J. Hyd. 7. R. Prasad, "A Nonlinear Hydrologic System Response Div. 93(HY4)(1967). "Hydrology of Urban Runoff," J. ASCE 85, 418. A. L. Tholin and C. T. Keifer, 1959). 106(Mar. "Digital Simulationin Hydrology:StanfordWater9. N. H. Crawfordand R. K. Linsley,Jr., shedModel IV," Department of Civil Engineering,Stanford University, Stanford, CA, Tech.Rep.No. 39, July 1966. 10. JohnC. Schaake, "synthesisof the Inlet Hydrograph,"Tech.Rep. 3, Storm Drainage Jr., Project, JohnsHopkins University,Baltimore, MD, June 1965. Research "WaterPollution Aspectsof UrbanRunoff," Federal 11. AmericanPublic WorksAssociation, Water Pollution Control Administration, 1969. "Urban Water Resources Re12. Arnerican Society of Civil Engineers,First Year Report, search,"Sept. 1968. 13. W. Viessman,Jr., "Modeling of Water Quality Inputs from Urbanized Areas," Urban Water ResourcesResearch, Study by ASCE Urban Hydrology Research Council, pp. Sept.1968, A79-A103. "Characterization, 14. S. R. Weible,R. B. Weidner,A. G. Christianson,and R. J. Anderson, of Treatment,and Disposal of Urban Storm Water," in Proceedings the Third Interna(S. International AssociationonWater Pollution Researcft H. Jenkins, tional Conference, Press,1969. Elmsford,NY Pergamon ed.). "Pesticides and Other 15. S. R. Weible,R. B. Weidner,J. M. Cohan,and A. G. Christianson, Contaminants in Rainfall and Runoff," '/. Am. Water Works Assoc. 58(8), 1675(Aug. 1966). Departmentof Civil Engineering,University of Cincinnati, 16. Division of WaterResources, SeWater Pollution Control Research Cincinnati. OH. "Urban Runoff Characteristics," ries. EPA. 1970. Engineers, 17. Metcalf and Eddy, Inc., University of Florida, Gainewille, Water Resources "Storm Water ManagementModel," Environmental Protection Agency, Vol. 1, Inc., 18. E. Kuichling, "The Relation Betweenthe Rainfall and the Dischargeof Sewersin Populous Districts,"Tians.ASCE,20(1889). 19. W. W. Horner, "Modern Procedurein District SewerDesign," Eng. News 64,326(1910). "Relation BetweenRainfall and Runoff from Small Urban 20. W. W Horner and F. L. Flynt, Areas," Trans.ASCE 20(140),( 1936). 21. R. L. Rossnriller,"The Runoff Coefficient in the Rational Formula," EngineeringResearchInstitute, Iowa State University,Feb. 1981. "Experimental Examination of the 22. J. C. Schaake,Jr., J. C. Geye1,and J. W. Knapp, RationalMethod," Proc.ASCEJ. Hyd. Div.93(HY6) (Nov. 1967). "Airport Drainage," Advisory 23. FederalAviation Agency, Departmentof Transportation, PrintingOfflce, 1970. D.C.: U.S. Government Washington, Circular, AIC 150-5320-58.

r14-L2r(r9s7).

r971.

Chapter16

Design Hydrologic

Prologue
of The PurPose this chaPteris to: for usedin the United States designing . Introducethe hydrologist procedures to of structuresfor safe and effectivepassage flood flows' , Give sufficient information for the designerto selectthe applicablecriteria for hYdrauhcstructures' designing and provide methodsfor , provide a discussionof designstorm hyetographs design. selectingthe duration, depth, and distribution of plecipitation for precipitation without using . DemonJtrate how designiloods can be developed data' ' Discussparticular designmethods including airport drainage'urban storm sewerdeiign, and flood control reservoirdesign' . Describethe U.S. Federal EmergencyManagementAgency (FEMA) flood systemand piesent the hydrologicfundamentalsof flood plain management Plain analYsis. piotto studying 26 to Readersare encouraged reviewthe materialin Chapters and27 in presented this designprocedures rater discharge PredictingPeak minor and ma for use in designing aspects of engineering hydrology' I small crodsroadculverts, levees,dt akPort drainage structuresto the lumped together'with major structr design information. GenerallY,a h dischargefor a designfrequencY,a dischargehYdrograPhfor a design rates,low-flow frequencYanalysis, are often conductedas part of a designproiect'

360

DESIGN 16 CHAPTER HYDROLOGIC Mostdesignsinvolvinghydrologicanalysesuseadesignfloodthatsimulates are futirre eventorlmitates ime historicalevent.If streamflowrecords somesevere storm records from available are unavailable,designflood hydrographs synthesized are of procedures Chapters2, 12,and 15. Only in rare cases usingthe rainfall--runoff in small watersheds' particularly for streamflowrecords adequate complex designs, in and the empiric-coirelativemethodsdiscussed Chapter 15 are analyses Regional in Methodspresented chapsites' peak flow ratesat ungauged usefulfor determining for necessary hydrographs entire ter 12 andin this chapterare used for developing designs. manYengineering in minor and major structuresare described uyirologic;ethJds for designing levels,rnethodsfor this chapter.included are discussions;f data needs,frequency for designstorms,and hazard assessments floodplainsand dams' synthesizing

PROCEDURES DESIGN 16.1 HYDROLOGIC

in either the peak flow for Procedures estimatingdesignflood flows (interestcan be historical or projected rate or the entire hydrogiaph)includemethodsthat examine and methodsthat (flow-basedmethods), estimate flood flowsto arrive at i sultaute stormsto flood flow rates the evaluate stormsthat producefloods,andthen convertthe a on selecting can the In (precipitation-basedmethods). eachcase, analysis be based (callfrequency-basedmeth' flood i"rign rr"qo"ncy and determiningthe ass,ociated narrowing the final designsfor a ringe of flood frequenciesand_ ods), developing (called risk-basedmethods)'or choice on the basisof long-term c6sts and benefits designingonthebasisofanestimateoftheprobablemaximumstormormanmum RooAtnit could occur at the site (calledcritical-event methods). basedon frequencyMinor Structure Design Minor structuredesignis largely to Severalstepsin the hydrologicapproach risk]basedmethods. basedor sometimes to most designhandbooksand adoptedtechdesign are comrnon minor structtJre are: niques.The generalsteps(eachis illustrated subsequently) to the time 1. Determinethe duration of the critical storm,usually equated concentrationof the watershed. t Choosethe designfrequencY. frequencyand duration' on 3. Obtain the storri OeptltUasea the selected in 4. Qomputethe net direct runoff (severalmethodswere presented ter 4). 5. Selectthe time distribution of the rainfall excess' 6.Synthesizetheunithydrographforthewatershed(seeChapterl2). T.Applythederivedrainfall_excesspatterntothesyntheticunithydrograph get the runoff hYdrograPh. equalto t flood (usuallyassumed Establishthe frequen-cyiftn" calculated 8. designstorm frequencY).

DESIGNPROCEDURES 16.1 HYDROLOGIC

361

are of Major Structure Design Hydrologicdesignaspects maior structures considculvert, or urban drainage crossroad erably more complexthan thoseof a small dam, system.A designstorm hydrographfor a'large dam still is required but it is put to greater use. The designstorm hydrographis routed to determinethe adequacyof The economic spillwaysand outlets operatedin conjunction with reservoir storage. possibilities dictatesthe final designand ofthe spillwaysizefrom the various selection is a function of the degreeof protection providedfor downstreamlife and property, and policy andconstructionstandards, reservoiroperational project economy,agency largely basedon critical eventmethodspreMajor structuredesignis requirements. in sented Section16.5. in presented Water Resource System Design Most information and techniques protection aspectof small and large structhis chapterare directedtoward the flood for to tures.Needless say,a major structureis designed more than just flood protecfor irrigation, power, water supply, tion; it is multipurposeand may provide storage to The proper allocationof storage theseuses navigation,and low-flow augmentation. history in terms of the frequency of requiresan understanding the entire streamflow and yearly flows, as well monthly, seasonal, of occurrenceof low flows and average presentedin Part Five to provide a as the historical and designfloods. Material is hydrologistwith the tools to developcomplete streamflowhistories for a complex multipurposesysteminvolving various combinationsof minor and major structures, practlces. projects, and management water development

Methods Flow-Based
or For designlocationswhererecordsof streamflows are available, whereflows from to can be transposed the designlocation, a designflood magnitudecan anotherbasin be estimateddirectly from the streamflows by any of the following methods: ' 1. Frequencyanalysisof flood flows at the designlocation or from a similar basin in the region. from regresnormally developed 2. Useof regionalflood frequencyequations, gaugedflood data. sion analysis(seeChapter 26) of 3. Examination of the stream and floodplain for signs of highest historical of floods and estimationof the flow rates using measurements the crosssectionand slopeof the stream.

Methods Precipitation-Based
for or recordsare unavailable inadequate streamflow.estimaWhere stream-gauging floods can be estimatedby evaluatingthe precipitation that would protion, design ducethe flood, and then convertingthe frecipitation into runoffby any ofthe rainfallin runoff methodsdescribed Chapters10-15 or 2l-27. Typical methodsinclude: 1. Design using the greateststorm of record at the site, by converting the precipitation to runoff. in historicalstormfrom anothersimilar watershed of 2. Transposition a severe the region.

362

CHAPTER16

HYDROLOGIC DESIGN

3. Frequency analysis of precipitation and conversion of design storm to runoff. 4. Useof a theoreticalprobablemaximumprecipitation (PMP), or fraction of PMP, basedon meteorological analyses. methodsare Because flood flow rate is desiredin all cases, flow-based the the preferred over conversionof precipitation to runoff. Due to the relatively longer period of time and greaternumberof locationsat which precipitation amountshave been recorded, precipitation-basedmethods are used in the majority of designs, especiallywith small and very large basins.Flow-basedmethodsare typically used in the midrangeof basin sizes.

Frequency-Based Methods
most often proceed Regardless whetherflow or precipitation dataareused,designs of by selectinga minimum acceptable recurrenceinterval and using proceduresfrom Chapter27 to determine corresponding the worst condition storm or flood that could be equalledor exceeded recurrenceinterval. Criteria for selecting during the selected designrecurrenceintervals are summarized Section 16.3. Resultsfrom frequency in analysisof flood flow data normally provide reliable estimatesof 2-, 5-, 10-, and 25-yearflows.Extrapolationbeyondthe rangeof the period of flow recordsis allowed, but is lessreliable.

Risk-Based Methods
Recenttrends in designof minor (and major) structuresare toward the use of ecoThe risk methodselects the nomic risk analyses rather than frequgncy-based designs. costs.Tfreseare madeup of the structuresizeas that which minimizestotal expected with the particular structure. structurecostsplus the potential flood losses associated The procedureis illustrated in Fig. 16.1.The total expectedcost curve is the sum of

o b0 q

Optimal structure size, S* (least total expected cost) S^in Structure size, S

forstructure Figure 16.1 Principlesof economicriskanalysis size selection. (U.S. Federal Highway Administration, Hydraulic EngineeringCircular No. 17).

16.2 DATA FOR HYDROLOGIC DESIGN 363 the other two curves.Risk costs(flood damages, structuredamages, road and bridge losses, traffic interruptions)and structurecostsare estimated eachof severalsizes. for The optimal sizeis that with the smallestsum.Structuresselected risk analysisare by normally constrained sizesequal to or larger than thoseresultingfrom traditional to frequency-based methods.

CriticalEventMethods
Becauseof the high risk to lives or property below major structures,their design generallyincludesprovisionsfor a flood causedby a combinationof the most severe meteorologicand hydrologic conditions that are possible.Instead of designingfor somefrequencyor leastexpected total cost,flood handlingfacilities for the structures are sizedto safelystoreor passthe most critical storm or flood possible. Methodsfor designing critical eventtechniques by include: Estimating the probable maximum precipitation (PMP) and determining the associated flood flow rates and volumesby transformingthe precipitation to runoff. ) Determiningthe probablemaximum flood (PMF) by determiningthe PMP and convertingit to a flood by applicationof a rainfall-runoff model,including snowmeltrunoff if pertinent. 3. Examining the flood plain and stream to identify palaeo-floodevidences such as high-water marks, boulder marks on trees or banks, debris lines, historical accountsby local residents, geologicor geomorphologic or evidences. 4. In somecases, critical eventmethodinvolvesestimatingthe magnitude the of the 500-yr eventby various frequencyor approximatemethods.Often, suchas in mappingfloodplains,the 500-yr flood is estimatedas a multiple of the 100-yr event, ranging from 1.5 to 2.5. Due to lack of longer-term records, frequency-based estimatesare seldom attemptedfor recurrence intervals exceeding 500 years.

16.2 DATAFORHYDROLOGIC DESIGN

The designof any structurerequires a certain amount of data, even if only a field estimate the drainage of areaand a description terraintype and cover.The following of material identiflessomegeneraldata types and sources.

Physiographic Data
The hydrologic study for any structurerequires a reliable topographicmap. United StatesGeologicalSurveytopographic mapsusually are available. The mappingof the United Statesis almost completewith 15-minute quadrangles, and many of these areasaremappedby 7.5-minutequadrangles. County mapsand aerialphotoscan also be usedto advantage making preliminary studiesof the watershed. in

364

DESIGN CHAPTER16 HYDROLOGIC

drainage Based on an area map, a careful investigation of the watershed's information can be obtainedfrom USGS maps behaviormust be made. Additional and erosive that depict predominantrock formations.Soil types and the inflltration districtsor univerfrom U.S. Soil Conservation of characteristics soilscanbe secured divisions. sity extension of an The drainageareascontributing to large dams require stricter analysis The possibility of a minor structures. in designing area,shydrotogyitranis necessary for large uniformly intenserainfall over the entire basin is an unrealisticassumption shouldthus The watersheds. influenceof temporaland spatialvariationsof the rainfall "worst possible" rainfall values are the estimated For major dams, be considered. which is then usedin reservoir hydrograph, generallyconvertedto a designdischarge and storage, reservoii and spillway size, surcharge routing calculationsto propoition maintain power requirementsor sustaineddownany additional outlets neededto irrigation, or watersupply.The basicconcernin hydrologic streamflow for navigation, interestsusing a realistic estimatefor designof a large Aamis to protect downstream the designstorm hYdrograPh. purpose the Topographic'rnuf o"tuit necessarilyshifts with the type and .of of the pi"ta reconnaisiance alwaysincreases understanding structurebeing design"a. insignificantthe structule might be. matter how an area,shydrology*no

Data Hydrologic
datafor the regionunder one difficulty in hydrologicdesignis that of gettingadequate by issued published-reports from pr-eviously data study.ConsiOerabie canbe-acquired is a list of federal agencies and/or universities.The following agencies governmental hYdrologicdata: that Publish ' Service egricultural Research Service Soil Conservation Forest Service U.S. ArmY CorPsof Engineers National Oceanicand AtmosphericAdministration Bureauof Reclamation DePartmentof TransPortation Division U.S. GeologicalSurvey' Topographic Division WaterResources U.S. Geological.Survey, intergovernments, of Additional dlta often canbe procuredfrom departments state and regional and local agencies' statecommissions,

MeteorologicData
and Atmospheric The National WeatherService, couchedin the National Oceanic sourceof meteorologicdata publishedin a variety Administration, is the primary Figure 16.2 of forms, including their Hyirometeorologic Report (HMR) series. showstheapplicablereportsforvariousgeographicandtopographicregionsofthe

CRITERIA DESIGN-FREQUENCY 16.3 HYDROLOGIC

365

Figure16.2HydrometeorologicalreportseriescoverageofconterminousUnitedStates. of (U.S.Bureau Reclamation') collect and analyze state,and local agencies United States.lNumerousother federal inspect, or regulatelarge who design, precipitation information-"tp""iuriv-tr'rore

"*"8l;.H[1;:1;ce

in the region,maximum amount ttre meteoroioii" .huru"t"ristics of Jtot-t of total storm ;;r;r of precipitation,frequencies itable moisturein the atmospheJo snowmeltfor stormsover the of and influence. durationsbf U;t, depthsf*;;;i"", of major mountain chains,topography ", f;;;iii'*gion, region.tn ,o1n" areassuch has a very distinct impact on precipitation'

require:.5::S:1g:,:f stormhyetographs design forestimating or precrp-

CRITERIA DESIGN-FREQUENCY 16.3 HYDROLOGIC

S e l e c t i o n o f f r e q u e i s m o s t o f t e n b a s e d o n p o t e n t i a l d a m a g e t o p r o p e r r : l o s s e s s u c h a s i n t e r r u p t i o n o f A commefce' stanc the worst conditio involved,a greal a projects involve somerisks to property an A11 thror.t human tife li absent'the designcan proceed of the leastcost structuretn quencylevel and design alternativetoleastStructurecost,economicriskana] rather than the final designfrequencyis optimized for severaltrequencret storms would accommodate the actualconstfuction cost is u*"J. itr"r" costsincludenot only leasttotal expected to interruption of servicesand "o** du9 costsbut also the flood dama!;;irk una can be used' worth economicanalyses or p"resent commerce.Either annual

366

CHAPTER16

HYDROLOGIC DESIGN

MinorStructures
shown in Table 16.1 are typical of levels generallyencounThe designfrequencies tered in minor structuredesign.An exampleof variationsthat do occur is the design could effectivelyhalt backwater of frequencyof a culvert,which undercases excessive trafflc. the Servicerecommends use of a2l-year frequencyfor The Soil Conservation minor urban drainagedesignif there is no potential loss of life or risk of extensive damagesuch as first-floor elevationsof homes.A 100-yearfrequencyis commonly property damagemay occur.t recommended when extensive
DESIGNFREQUENCIES TABLE 16.1 MINORSTRUCTURE

Typeof minor structure Highway crossroad drainage" ADT' 0-400 400-1700 ADT 1700-5000 ADT ADT 5000Airfields Railroads Stormdrainage
Levees Drainage ditches

period, Return 4 10yr 10-25yr 25 yr 50 yr 5yr 25-50 yr 2-10 yr 2-50 yr 5-50 yr

= Frequency 1/7, 0.10 0.10-0.04 0.04 o.o2 0.20 0.04-0.02 0.50-0.10 0.50-0.02 0.20-0.02

'ADT : averagedaily traffic. (After Ref. 3).

Large Dams
of duringthe design the original structureandduring Damsrequirehydrologicanalysis when are periodic safetyevaluations. Significanteconomicand humanlosses possible from storage. quantitiesof water are rapidly released large the Initial heightsof retardedwater behind the dam, disregarding total volume water, can produce destructiveflood wavesfor a considerabledistance of stored Basedon two criteria, the TaskForceon SpillwayDesignFloodsrecomdownstream. in of the classification large damsas li,sted Table 16.2.The type of construcmerided has not been included in this grouping, althoughit affects the extentof failure tion resulting from overtopping. of haveadopteddefinitionsfor hydraulic elements Many of the federalagencies The following list is usedby the Soil ConservationService: dams. A spillwuy is an open or closedchannel,or both, used to convey excess water from a reservoir.It may contain gates,either manually or automatwater' of ically controlled, to regulatethe discharge excess to Theprincipal spillwayis the ungatedspillwaydesigned conveythe water for rates established the structure. from the retarding pool at release to spillway of a dam is the spillwaydesigned conveywater The emergency of in excess that impoundedfor flood control or other beneficialpurposes.

DESIGN-FREQUENCYCRITERIA 16.3 HYDROLOGIC

367

DAMS FORI-ARGE CRITERIA TABLE16,2 DESIGN danger lmpoundment Potential Category (1)
Major; failure cannot be tolerated Storage (acre-ft)o (2)

Failure damage Potential' Loss of life (4) Considerable

Height (ft) (3)

Damage (5)
Excessiveor as matter of PolicY

Spillwaydesignflood (6) Probablemaximum; most severeflood considered possible reasonably on the basin Standardproject; based on most severe storm or meteorological conditions considered reasonablycharacteristic of the sPecific reglon basis; Frequency 50-1O0-year recurrence interval

>50,000

lntermediate

1000-50.000 40-100

Possiblebut small

Within financial capability of owner

<1000

<50

None

Of samemagnitude as cost of the dam

and future potential oBased on consideration of height of dam above tailwater, stoarags volume, and length of damage reach, present of floodplain' population, and economic development tstorage at design spillway pool level. Sozrce: After SnYder.3

The retarding pool is the reservoir spaceallotted to the temporary rmpoundmentoi floodwater.Its upperlimit is the elevationof the crestof the emergencyspillway. Retardingstorageis the volume in the retarding pool' pool is the reservoir spaceallotted to the accumulationof The sediment incoming sedimentduring the life of the structure' storageis the volume allocatedto total sedimentaccumulation' Sediment sedimentpool elevationis the elevationof the surfaceof the anticipated sedimentaccumulationat the dam. Anearthspitlwuyisanunvegetatedopenchannelspillwayinearthmaterials. Avegetatedspillwayisavegetatedopenchannelspillwayconstructedof earth materials. spillway constructedon the downstream A ramp spillway is a vegetated dam. faceof an earth where The control section in an open channel spillway is that section through critical depth' flow passes accelerated spillwayis the channelupstreamfrom The inlet channelof an emergency the control section.

368

CHAPTER16

DESIGN HYDROLOGIC Top of dam

/ :Minimum crestJ spillway Emergency x r freeboard Surcharge poollevel Normal \ v I storage

Flood control retardingstorage

Emergency spillway

Reservoir Minimum Pool

Qo

zones' and poollevels storage reservoir Figure 16.3 Multipurpose The exit channelof an emergencyspillwayis that portion of the channel from the control sectionwhich conductsthe flow safelyto a downstream without jeopardizing the integrity of the point where it may be released structure. spiltway lrydrographis that hydrographusedto establish The emergency of the minimum designdimensions the emergencyspillway' Thefreeboard hydrographis the hydrographused to establishthe minimum elevationof the toP of the dam. Severalof thesefeaturesare illustrated in Fig' 16'3'

SmallDams
usingtwo or more levelsof frequencyto provide Small damscustomarilyare designed 16.3 an emergencyspillway and ensure an adequateallowable freeboard. Figure (MF)showsa iypicit small dam with normal freeboard(NF) and minimal freeboard The freeboardvaluesfor earth dams with riprap protection on the upstreamslope' 100-mph by shownin Tablq 16.3, atebasedon waverunup caused storm winds with The fetch wind velocities.Minimal freeboardpertainsto wind velocitiesof 50 mph. If is definedas the perpendiculardistancefrom the structureto the windward shore. values smoothconcrerc;atirerthan riprap is usedon the upstreamface' the freeboard shown shouldbe increased50 percent'"

CFIITERIA DESIGN-FREQUENCY 16.3 HYDROLOGIC TABLE 16.3 USBR RECOMMENDED NORMALAND MINIMUM VALUES,FT FREEBOARD Fetch (mi)

369

<1
I

4 6 8 10

3
+

2.5 5 10
Soarce:After Ref. 4.

5
o

of The U.S. Soil conservation Service designcriteria for principal spillways be No. 60 should small dams are given in Table 16.4. The SCS TechnicalRelease are of this table.sDesign frequency_requirements consultedfor full interpretation use of the structures.The SCS classifles selectedto fit the planned or foreseeable grouPs:u into structures three Class a. Structureslocated in rural or agricultural areas where failure mightdamagefarmbuildings,agriculturalland,ortownshiporcountry roads. ctass b. Structureslocated in predominantlyrural or agricultural areas where failure might damageiiolated homes, main highways.or minor railroads, o, "urrr]"interruption of use or serviceof relatively important public utilities. Classc. Structureslocatedwhere failure might causeloss of life, serious public damageof homes,industrial and commercialbuildings,important utilities, main highways,or railroads, generally ft The physicalsizeof a small dam can rangeto over 100^ in heightbut acre-ft of storageat the emeris restrictedto structuresretarding lessthan 25,000 are gency spillway crest. Small dams generally receive.special attention if they could causethe loss of life' Many ion.i.o"i"d in populatedareaswhere dam failure the by havebeencaused dam or leveefailure. When this possibilityexists, flood deaths of the probablemaximum precipidesignstorm for small damsis lstablishedby use maximization of the tatio'n, PMP. The PMP is generally defined as the reasonable maximum storm. Other definitions factorsthaioperate to producea meteorological including: havebeen proPosed,T canbe 1. The p1itp is the rnaximumamountand duration of precipitationthat expectedto occur on a drainagebasin' combinafrom the most severe 2. ThePMP is the flood that may be expected conditionsthat are reasonably and hydrologic tion of critical meteorologic possiblein the region. T'he pMp Las a low, but unknown, probability of depthat the designlocation the It occurrence. is n-either maximumobserved immune to exceedance' or region nor a value that is completely

370

CHAPTER16

HYDROLOGIC DESIGN

SPILLWAYS TABLE 16,4 SCS DESIGNCRITERIAFOR PRINCIPAL OF SMALL DAMS Precioitation data for maximum frequency2of use of spillwaytype: emergency

Class of dam
(a)

Purpose of dam
5lngle-

V"HI
Less than 30,000 Greaterthan 30,000

Existing or planned upstream dams

None

Earth

Vegetated

irrigation only

0.sDrJ 0.75DL
!D 0 5

O.5DL

None

0.75DL
D6 r25

SingIe or multiples

Lessthan 30,000 than Greater 30,000 All

All

None None Anyt None or any None or any

0.5(Pso + Ploo)
D r 100

0.5(Prs+ Pso)
rD 0 5

(b)

Single or multiple Single or multiple

100

(c)

All

Proo

I Product of reservoir storagevolume V, (acre-feet) times effective height of dam 11, (feet). 2Precipitation depths for indicated return periods (years). 3Applies to irrigation dams on ephemeralstreamsin areaswhere mean annual rainfall is less than 25 in. aDL = designlife (years). 5Class (a) dams involving industrial or municipal water are to use minimum criteria equivalent to that of Class (b). 6In the case of a ramp spillway, the minimum criteria should be increasedfrom Prr to Pt66. ?Applies when the failure of the upstream dam may endangerthe lower dam. Soarce.'Soil Conservation Servtce.

Estimatesof PMP are basedon an investigationby the U.S. WeatherBureau conducted establish maximumpossibleamountof precipitablewaterthat could to the of Figure 16.4providesestimates precipbe achieved throughoutthe United States.s'e itable water over watersheds betweensealevel and 8,000 ft. Figure 16.5 extendsthe estimatesabove 8,000 ft.1 Point valuesof PMP for the samelocale may vary with for duration of storm causingthe precipitation.Figure 16.6 providesPMP estimates 6-hr storms.These and similar publishedcharts for other durations are helpful in selectingthe PMP for any region in the United States. the deiign frequenlies-for principal spillwaysfor small SCS Class aob, or c damsare providedin Table 16.5.Theseare basedon 6-hr rainfall depthsfor ( 1) the for (Fig. 16.7) and(2) the PMP (Fig. 16.6).Designstormdepths 1OO-year frequency all watershedshaving a time of concentration less than 6-hr are establishedin are with greatertime of concentration,adjustments Table 16.5.For thosewatersheds madeto the 6-hr storm depth to accountfor the gteateramountsof direct runoff in a in longerperiod of time. Theseadjustmentsare discussed Section 16.4.

16.3 HYDROLOGICDESIGN-FREQUENCYCRITERIA 371

DEPTHSOF PRECIPITABLE WATER IN A COLUMN OF MILLIBARS AIR OF GIVEN HEIGHT ABOVE 1OOO
Assuming saturation with a pseudo-adiabatic lapse mte for the indicated surface temperatures Adapted ftom the U S Wther Bueau Hyalrometeorological Repon No 23 . 142230343842 46 505254 565860 TEMPEMrure 62 64 66 68 '10 '72 74 '76 78

80 .F

800
a

.: =
p
h

z
a
H d

Extended to 200 mbar on Figure 16.5

1.5 1.0 ABLE WATER IN INCHES

Figure 16.4 Diagram for precipitable water determination from 1,000 to 700 millibars.(U.S. Bureauof Reclamation.)

372

16 CHAPTER HYDRoLoGIc DESIGN

DEPTHS OF PRECIPITABLE WATER IN A COLUMN OF AIR OF GIVEN HEIGHT ABOVE 1OOO MILLIBARS lapse Assuming saturation a pseudo-adiabatic with
rate for the indicated surface temDeratwes Adapted from the U-S. Weather Bureau Hydrometeorolgical Report No 23 EMPEMTre

l8 26

66 70

74

200

2.0 2.5 3.0 3.5 1.5 WATERIN INCHES PRECIPITABLE

Figure 16.5 Diagram fbr precipitable water determination from 800 to 200 (U.S. Bureauof Reclamation.) millibars.From 1281. 103-D-1908.

16.4 DESIGNSTORMS

373

MaiorStructures

STORMS 16.4 DESIGN

the Once the designfrequencyhasbeenestablished, next stepin a structuredesignis of six ,io.- puru-eters: the stormduration,the durationof rainfall the determinaiion the excess, point depth, any ireal depth adjustment,the storm intensity and time distribution, and the areal distribution pattern'

Duration
The length of storm usedby the SCS in designingemergencyand freeboardhydrographsior small damsis of 6-hr durationor /c, whicheveris greater' Often, the minor cannotbe justified economicallyon the basisof this length it*"tu.. being designed of storm. Foi many minor structures, particularly urban drainage structures, a designflood hydrogiaph is basedon a storm duration equal to the time of concenfnis procedureusesthe rational method of Chapter 15 or trati-onof the wateisneA. the synthetic unit hydrographs of Chapter 12 along with a critical storm pattern pattern into the most critical sequence. produced by arranging the rainfall excess fn" SCS uies 24-hr durations for all urban watershedstudies' but for Durationsof approximately6hr or lessare satisfactory small watersheds, depthsfor periodsof up to 10 days' the lengthsof stormiln large areasrequire storm valuesare availablefor durationsof from 2 to I0 daysfor locations Freque-ncy-based Similar data are also availablefor other selectedareas within the United States.to outsidethe United States.Generally,however,designcriteria for large damsrequire of estimates storm depthsthat do not havefrequencylevelsassigned.

Durationof RainfallExcebs
and Initial rainfall during most stormsinfiltrates or is otherwiseabstracted, the durarainTf,is lessthan the actual rain duration by an amount.equalto the tion of excess time that initial abslractionsoccur. Excessrain duration 7s can be estimatedfor a 6-hr storm as a function of the curve number CN and precipitation P from

374

DESIGN CHAPTER16 HYDROLOGIC 125" 120"

i 7t?
"d, \
tv
,-o-

Bureau, Figure 16.6 The 10-mi2or less PMP for 6-hr duration (in.). (U.S. Weather NOAA.)

storm by Fig. 16.8. This family of curves was developed the Sclrlt y-h:I" P is the zero A CNof 100 represents una O""pttr CNis a losi parameterdefined1lhaptgr {. lossessothatZg:6hrforCN:l00.Table16.6isusedtofindthedurationof is the excess rain for any storm duration greater than 6 hr. The rainfall ratio (T'iUle 16.7) divided by the total precipitation absrractionP* losi before runoff to amount P. The time ratio from Table 16.6 is multiplied by the rainfall duration obtain Ze.

0 s0 r00
95"

2?0 3?0 490

90.

\ \

5P
85' 6u

Figure 16.6 Continued

Depth

available.

386

CHAPTER16 HYDROLOGIC DESIGN 1.0 0.9 0.8 X \ o 0.7 0.6

.=a

0.5

0.3
v.z

0.1

3 4 Time(hr)

Figure16.18 A6-hrdesignstormdistribution for SCSdam design.(After Ref. 12.)

country. For more preciseinformation on boundaries a statehavingmore than one in storrntype, contact the respectiveSCS State Conservation Engineer. The greatestpeak flows from small basinsare usually causedby intense,brief rains. Thesecan occur as distinct eventsor as portions of a longer storm. The 24-hr storm duration is longerthan neededto determinepeaksfrom small watersheds but is appropriatefor determiningrunoff volumes.In light of this, the SCS usesthem to studypeak flows, volumesof runoff, and direct iunoff hydrographs from watersheds normally studiedby the agency. Time distributionsfor PMP and other stormsusedin major structuredesigncan be constructed from Fig. 16.20.This family of curvesis usedby the U.S. Burlau of Reclamation6 threegeographical in zonesshownin Fig. 16.6.The corps of Engineers usesa distribution curve similar to Fig. 16.18for 6-hr SpS analyses. Triangular Distribution The simplest design storm distribution is a triangular shape. Because depth,P, andduration,D, of rain are alreadyestablished, peak the the intensity,l-u,, is 2PfD, foundby solvingfor the height of the triangularhyetograih as shownin Fig. r6.2L.The only remaining decision the time to the peak,40. is The ratio to/D has been investigated a large number of storms at locationsin California, for Illinois, Massachusetts, New Jersey,and North carolina. values range from about 0.3 to 0.5.17 Once the triangle is constructed, intensitiesat regulaiintervals may the be graphicallyor analytically determined input to the rainfall-runoff rnodelbeing for usedfor design. Blocked IDF Distributions A frequentlyusedprocedurefor developing design a storm distribution for short duration storms (up to about 2 hr) is to successively construct blocks of a design storm.histogramby using the appropriate intensity-

16.4 DESIGNSTORMS

Rainfall distribution

f--l ryp"r ffi rvwIe

i--l rypeu

llffil rvnerrl
rFooo

V v

(b)

for zonesI' Figure 16.19 SCS 24-hr rainfall distributions:(a) 24-hr rainfall distributions for scs rainfall distributions'(After Ref' 16') boundaries IA, II, and III and (b) approximate

388

CHAPTER16

HYDROLOGIC DESIGN 1.00 0.90

r ORO

;
E tr ;

o.7o
o6n nsn 0.40

E o.3o
o.20 0.10

Time (hr)

Figure 16.20 Distribution of 6-hr PMP for any area west of the 105' meridian. (After Ref. 6.)

Time, hr Figure 16.21 Triangular hyetograph. design duration-frequency curveto find therain intensities A/, 2 At,3 L,t,etc.,increments for of time and then to organizethese"blocks" of rain intensitiesin somepattern,usually symmetrical,.around the center of the storm, making sure that the area under the hyetographis equal to the designstorm depth, P, spreadover the designstorm duration,D. To apply the procedure, successive depths of equal-probability storms with durations A,t,2Lt,3A,t,4Lt,etc.,aredetermined of from the IDF curveandtabulated. Next, any of a variety of procedures,such as the alternating block method, the Chicagomethod,or the balancedmethod,are availablefor distributingtheseblocks and assuring that the total rain depth equals P. Most assumethat the highest

16.4 DESIGNSTORMS

389

highestoccursnext, and so on, intensityoccursin the middle of the storm,the second directions from the center block. The balanced method, for working out in both that exampli, assumes a Ar-hr stormwith intensityia,from the IDF curvecould occur, with equal probability, during the middle of the D-hr designstorm. This intensity is Next, the rain depth for plotted as the middle block of the designstorm hyetograph. obtained from the IDF curve. Its distribution is assumedto be a duration 2Lt is two-bar histogramwith the first half matchingthe intensity of the Ar-hr storm; the secondhalf intensity is calculatedby spreadingthe rest of the rain depth for the 2 for is Al-hr durationunifoimly over the secondAr interval.The process repeated rain a of with durations 3A/, 4Lt, . .. , up to D. The goalis to develop depthsfor storms suchthat a storm of any duration,centeredat the middle of the storm hyetograph will blockedIDF hyetograph, havea total rain depthmatchingthe rain depthfrom the duration. IDF curve for the siven

Areal Distribution
precipitation depths can and do vary from point to point during a storm. Areal exceptin major structure variation in designstorm depth is normally disregarded in The usual approach major structureanalysisis to selecta design(usually designs. isohyetalpattem for the PMP or SPS depth and elliptical) or historic (transposed) precipitation depthsio the isohyetsin a fashionthat givesthe desiredaverage assign depih over the basin. The averagedepth is determinedby the isohyetal method illustrated in ChaPter2. Four majorlypes of storm patterns are shown in Fig. 16.22 fot areasup to of 400 mir. Thesewere identifiedby iluff in his analysis midwesternstormpatterns.ll The letters H andL representareaswith high and low precipitation depths,respecas tively. The typical isohyetalpatternfor SPSstormshasbeenestablished generally Valley Tennessee elliptical in itrapeas shownin nt. 16.23.This patternis usedby the

?b

Figure 16.22 Major types of storm patterns: (a) closedelliptical, (b) open elliptical, (c) multicellular; and (d) banded.(After Huff.tl)

390

CHAPTER 16 HYDROLOGIC DESIGN

t t Scale: miles

Figure 16.23 Generalized pattern storm.

lsohyet

Area enclosed(mi2) l1 45 114 279 546 903 1349 2508 4458

B C D E F G H
(After Ref.18.)

Authority (TVA)18for areasup to 3000 mi2.Variationsin the rainfall depth found in a standardproject storm will divergefrom a maximum at the storm centerto a value considerablylessthan the average depth at the edgesof the watershed boundaries. This variation can be determinedand incorporatedin the designstorm. A slightly modified isohyetalpattern for SPS storms is used by the Corps of Engineersre shownin Fig. 16.24.The percentages as shownfor isohyets B, . . . , G A, are multiplied by the 96-hr SPS depth to give an elliptical pattern with the desired average depth.Similarmapsfor 24.,48-, or 72-hr stormscanbe obtained simplyby modifyingthe 96-hrpercentages Fig. 16.24.Thisis accomplished of usingthe deptharea-duration curves Fig. I6.24.For example, a24-hr stormis used, in if first notethat theA isohyetof Fig. 16.24encloses areaof 16nrr?. an From Fig. 16,25thecorresponding SPSpercentage a24-hr stormis 116 percentrather than the 140percentvalue for used with a 96-hr storm. Thereforethe pattern percentages vary with the selected designstorm duration. An additional aid for constructingdesignstorm distributionsover smallermid(up westernl8 watersheds to 400 mi') is presented Table 16.10.The ratio of maxiin mum point rainfall to mean rainfall over the basin is provided and can be used to estimatethe maximum depth occurring at a storm centerif the mean areal depth is

16.5 CRITICALEVENT METHODS

391

: miles

Figure 16.24 GeneralizedSPS isohyetal pattern for a 96-hr storm' The pattern may be orientedin any direction and may correspondto the depthby arearelation represented a 96-hr storm.

lsohyet
A

Area (mi2)

B C D E F G
(After Ref. 19.)

l6 100 320 800 1800 3700 7100

known. Ratios for 50-, 100-, and 200-nr2 areasare equal to those in Table 16.10 multiplied by 0.91, 0.94, and 0.97, respectivd. For uniform rainfall the 95 percent With extreme Variability the 5 percent ratio ratios of the table are recornmended. conditions' The 50 percentratios approximateaverage applies.

ULrnoos EVENT 16.5CRTilCAL

392

16 CHAPTER

DESIGN HYDROLOGIC 20,000 10,000

\
+ 72-hperiod

5,000 4,000 3,000 2,000 1,000

o

\\\

s=
,*Nr"

500 400 300 200 100 50 40 30 20

10 40

\\ \ \

o I

\ \ \

60

120 100 80 value of Isohyet Percentage SPS

140

Figure 16.25 SPS depth-area-duration curves by 24-hr storm increments. (After Ref. 19.)

future event, and then design accordingly. These methods include the use of the probable rnaximum precipitation PMP, probable maximum flood PMF, record floods, and high storm depths, record high floods, multiples of frequency-based paleohydrology.

MaximumPrecipitation Probable
Probablemaximum precipitation depthsfor drainagebasinsin the United Statesare identified in Fig. 16.2. providedin the respectiveNational WeatherServiceHMRs2o reasonstorm considered most severe The probablemaximum stormis deflnedasthe probablemaximumflood is customarilyobtained to ably possible occur.The resulting of and by usingunit hydrographs rainfall estimates the PMP preparedby the National rily'eather (seeFigs. 16.6 and 16.26). Service2l

EVENT METHODS 393 16.5 CRTTTCAL ON POINT MEANRAINFALL 400 miz TO OF TABLE16.10 RATIO MAXIMUM
Mean rainfall(in.) Rainfallperiod (hr)

1.0

1.5

2.O

2.5

4.0

5.0

5% Probabilitylevel ratios (Storms with extreme variation in intensity)

0.5 1 2 3 6 l2 18
)4

5.20 5.50 5.80 6.05

48

0.5
I
L

3 6 12 l8
JA

2.66 3.03 3.46 3.77

48

0.5
I

2
J

2.38 2.75 3.15 3.46

6 t2 l8
)4 48 Sorrce.'After Huff.lr

r.4r 2.r8 1.70 1.48 1.80 2.29 1.55 1.90 2.44 1.99 1.61 2.53 r.72 2.r2 J.t I 2.69 1.83 2.25 2.86 4.01 1.90 2.33 2.96 4.14 1.96 2.40 3.05 4.27 2.08 2.55 3.25 levelratios 50% Probability (Storms timedislributions) with average 1.22 1.32 1.5'1 2.02 r.27 1.65 L39 2.r5 1.32 1.46 r.75 2.29 1.38 1.85 2.42 1.52 r.43 r.63 1.98 2.59 1.50 r.75 2.12 2.78 1.57 1.81 2.20 2.89 1.60 2.28 1.87 3.00 2.44 3.r7 .1.68 ,1.99 levelratios 95%Probability (Storms intensities) with uniform r.16 r.28 1.18 1.53 1.20 1.23 1,38 1.72 1.24 1.28 1.47 1.90 1.27 1.53 1.33 2.02 1.31 r.43 2.24 t.67 1.38 1.78 2.50 1.50 r.4l 1.53 1.89 2.67 r.43 1.58 1.92 2.77 r.47 r.64 2.04 3.07 3.00 3.21 3.38 3.54

1.30 1.35 1.41 1.46 1.52 1.60 1.65 r.69 1.77

t.26 1.30 r.33 r.36 1.43 1.50 1.54 1.57 1.63

1.22 1.25 r.28 1.30 1.35 1.40 1.43 1.45 1.50

1.16 1.20 1.24 t.28 1.33 r.39 1.43 1.47 1.53

|.14 1.18 Lzr 1.23 r.28 1.32 1.35 1.38 r.46

l.l2 1.16 1.19 t.22 1.26 1.30 1.32 1.33 1.38

1.13 l.l7 1.20 1.22 1.27 1.3 I 1.33 1.35 1.40

1.11 1.15 1.18 r.20 1.24 1.28 1.30 t.32 1.36

1.10 1.r4 1.16 1.18 1.21 1.25 1.27 1.29 1.33

that suggests the by PMP advocated Hershfield2l methodto estimate A proposed 24-hr PMP at a point be computedby the equation
PMP24:P*KS,

(16.1)

where PMPZ : . F : K: , & :

the 24-hr probablemaximum precipitation the meanof the 24-hr annualmaximumsover the period of record a constant equalto 15 the standarddeviationof the 24-hr annual maximums

Adjustmentsto the value of F and S, for the record length are noted by Hershfield. However, for appraisalpurposesthese adjustmentsprobably will not significantly alter resultsmore than 5-10 percent.

394

CHAPTER16

HYDROLOGIC DESIGN

zo.t

,/

\\ - - - - - - " . . 20.1(3)

ptvtp(in.). (1) Alexandria, Figure16.26 Twenty-four-hour 2000-mi2 LA, (3) (2) June 13-17,1886. Eautaw, April 15-18,1900. Elba, March AL, AL, (4) (5) l1- 16,1929. Yankeetown, September , 1950. Altapass, NC, FL, 3=7 (AfterRef. (6) July13-17, 1916. Jefferson, September 10-13,1878. 18.) OH, The U.S.Bureauof Reclamation underwentconsiderable evaluationof its design criteria for new damsand for safetyevaluationof existingdams,following the Teton, Idaho, dam failure in 1976.The policy adoptedfor modification of existingdamsis first to determinewhether they will accommodate peak dischargeof the PMF the without overtopping.In addition, the dam and appurtenant featuresmust accommodateat leastthe first 80 percentof the PMF volumewithout failure. For embankment dams,failure is assumed occur if overtoppinglevelsare reached. to

Recordedbdremes-Creager, and Crippenand Bue EnvelopeCurves

Where frequency-based methodsof PMP/PMF studiesare unwarranted,designfor critical eventscanbe basedon the greatest recordedrain or flood flow for the location. Similarly, tablesor curvesof flood data can be developed give the maximum floods to of record in the region under study; seeCreagerflood envelope curvesinFig. 16.27.

16.5 CRITICALEVENT METHODS 10.000 5,000 3,000 2,000

1,000

395

$ zoo
q? 100

500 300

s 0
ii JU

20 10 5 3 2
I

= {6gtrQs9ae

o 048 -11

N-h

o R9-' 8 888 8 383 3 I i;;". . 6 -A N 6

Drainage area (mi2)

o o *

o o o o N o

o o h

o o o

Figure 16.27 Creagerenvelopecurves: O peak inflow for Harza Projects; 'recorded (1) unusualflood discharges. Congo at Inga, Congo.(2) Tigris at Samarra,Iraq. (3) (4) Caroni at Guri, Venezuela. Tigris at Eski Mosul, Iraq. (5) Jhelumat Mangla, Pakistan. (6) Diyala at DerbendiKhan, Iraq. (7) GreaterZab at Bekhme,Iraq. (8) Surinameat (9) Brokopondo, Suriname. Lesser Zab at DokenDam,Iraq. ( 10)PearlRiver,U.S"A.(11) Cowlitz at Mayfield, U.S.A. (12) Cowlitz at Mossyrock,U.S.A. (13) Karadj, Iran. (14) Agno at Ambuklao,Philippines. (15) Angat, Philippines. (16). Tachien, Formosa. (Nole.'Curves taken from Hydroelectric Handbook, by Creagerand Justin. New York: Wiley, 1950.)

In cases where estimates of PMP have not been made. volumes of rainfall to be expected can also be approximated from Creager rainfall envelope curves of the world record rainfalls as depicted in Fig. 16.28. Maximum flood flow data for 883 sites up to 25,900 sq km formed the basis for the Crippen and Bue envelope equation given by
4o : A)31 6rlosA+caGog A)2+ca(log lgfc r+

(16.2)

where qo is the maximum flow (m3/sec), is the drainage area (sq km) and the A coefficients from Table 16-11 usingFigure 16.29. are

StandardProjectStorm
The standardproject stormis anotherrainfall depththat is usedin the designof large dams. This value is usually obtainedfrom a survey of severestormsin the general vicinity of the drainagebasin. The storm selectedas the SPS may be oriented to produce the maximum amount of runoff for the SPF. Alternatively, severestorms experiencedin meteorologically"similar" areascan be transposedover the study
atea.

396

CHAPTER16

DESIGN HYDROLOGIC 2000 1000 600 400 200

/:fr>y%^"1",
ndia

? 1oo
; 6 0 s 4 0
& 2 0 10 6 4 2
Fr-

g
Ihra]l, TX imethport, I A -

runkiKO , Formo t a aquio, Pldl .ppineIr landr

ffi"*"[,,"1,

anlalca

I Unionville \4d. 2 4 6810 20 4060

tIT ri--ftTl- tft--J-f-1lHoit,MOreaDe Arses, Bglqeqla ?oint,Jamaica I | | I Plumb.

L{

-n"irpfii,W

ssen, Bal

3
j

6 Hours

t2

24

10 2030

9t2
Months

24

+.>+/>

DaYs
Dulatron

Figure16.28Creagercurvesofworld'sgreatestrainfalls.(AfterRef.18')

AND BUE PEAK DISCHARGE FOR CRIPPEN TABLE 16,11 COEFFICIENTS CURVES ENVELOPE Fig.16.29 Region
I

Coefficients

limit(sqkm) Upper 26000 7800 26000 26000 26000 26000 26000 26000 26000 2600 26000 18100 26000 26000 50 2600 26000 2600

c1
3.203865 3.4'10923 3.330746 3.258400 3.126412 3.500489 3.326333 3.236183 3.503734 3.314692 3.231389 3.596209 3.461373 3.07349'l 3,451746 3.s65536 3.389030 3.743026

c2
.8049163 .74'72908 .8443r24 .8906783 .796472r .9123848 .8503960 .9193289 .8054884 1.0386350 .8867450 .8806263 .8519276 .64'727rO .9718339 .9699340 .9445212 .7918884

c3

c4

2 3
4 5

6 '7 8 9 10 11 12
IJ

I4 15 16
L I

Nationwide

-.002975'7 -.0394382 - . 0 5 5 1 7 8 0 -.0000965 -.0642062. -.0021362 -.0870959 .0022803 -.0899000 .0022'744 -.1013380 .00496r4 -.0998'74'7 .0042129 -.0947436 .0029486 -.0890172 ,0018961 -.059'7463 -.0042542 -.102053s .0045531 -.0747598 .0000138 -.1094456 .0058948 -.0038285 -.0252243 -.00s7110 -.0617496 -.0034'776 -.0649503 - . 0 6 7 8 1 3 1 -.002'7647 .0244991 -.0192899

Source: AfIer Crippen,J. R., and C. D. Bue, PaPer1887, 1977. WaterSupPlY

..Maximum Flood flows in The ConterminousUnited States,''U.S.G.S

o !

3 F 6'* *
o F F : o o

r ll

k c
6 -

S B (h E
u)
f O

+
r N

+ Q E ( o

e a e
6 Fcr
F * ,'! o

h d 6 h

6 o d d n

+ $ i r N n

i h n i o N d d r ; O N N n h n

oa

3 .* @'a
9 -

q q q q q
@ n h a r : . f + d r r r r @ r
i O O i O

FA g 3 " ' o
A * = v O a ' o ^

: < = x x x
Y Y Y - -

!s E

Jjas5
:

6.?f)

o + s o @ 6 h n o

b.E 9q g 8d' ! o F
o 9

@ + d r * @ d O N N O

$ N O

>

f J

g =

E =

a ^ z +

o i n N O O N 6

* E -

r ro; + \i cr ; - ; d d
E
d N i h O o o d s

= ^^ E o e*v 9
d

r d o i v i . j

; s H o+ + . . , j . . l 6
. L ( E v
: d N i

F< a'

6 N r o n r 6 h o

= E ^
E E e

q *E5o ii:E
N d c n N d

6
t r

Eec'
-

o 6 n h n n n d l q d l

^ ^ ^ o d

F 6 -

E= P E* =' E 9 -

- . o

c^

d O @ d * : f + $ h

\ \ 9 n
o r

E PF a tr E g6 I 3 = E . Eg J
E o =

\ q c t \ q
N h o o a

4 . 2

? X

@ 9 0 - o * < i d + s o { 6

6 Pu "
. E o g S t
J @ E

* h h o n n r

s i

* i
i

d * o o
* * i 6

O * 6

9
t r F F {

o ^ c +

N : O O A E

5 5 - E d

408

DESIGN CHAPTER16 HYDROLOGIC

V:

i:::\

including storm Figure 16.36 Children's artist rendering of urban undergroundsysteT, A 1976by David Macaulay.Reprintedwith perdriins. From UNDERGROUND,Copyrighl missionof HoughtonMifflin Co' Al1 rights reserved'

and and the storm sewer systems, detention in gutters, house drains, catchbasins, locations' landscaped interceptionin extensively for normally accounted in urban storm drain designare: Two items many 1, Infil,tration. The ability of the soil to infiltrate water dependson given of characteristics the soil as noted in chapter 3. The rangeof values continuous in the following table is typical of variousbare soils after t hr of rainfall' RATEs rvircnr- |NFILTRATIoN
Soilgroup
High (sandy,oPen-structured) Intermediate(loam) structured) Low (clay,close-

(in./hr) Infiltration 0.50-1.00 0.10-0.50 0.01-0.10

thesevalues3 to 7'5 times' The influenceof grasscover increases

ANALYSIS 16.8 FLOODPLAIN

409

such to 2. Retention This is usually assumed be 0.10 in. for pervioussurfaces surfaces' as lawns and normal urban pervious Developmentof hydrologicparametersfor designof storm sewerpipes, street gurrers, detentionbasinsis by the rational method(or modified rational rnethodor or are hydrographs adequate, iee Chapter25) when peak flow ratesand approximate synthesismethodswhen greater unit hydrographand kinematic wave hydrograph The latter usually involveuseof public domainor vendor-developed detail is needed. stormwaterdesignsoftware.The hydrologicaspectsof computerizedhydrologicdetext' in signtools are deiailedin Chapter25. In addition to the material presented this modified rational method, ILLUDAS' descriptionsof usesof the rational method, TR-sj, SWMM, DR3M, and other tools in designingurban storrrtdrainagefacilities Additionally, in are addressed numerousurban drainagedesigntextsand handbooks. and county engineer'soffices have of many statedepartments transportationor city tocattyapplicabledrainagedesignmanuals.As well, the American Society developed a i'model" drainagedesignmanual for local adaptaof Civil Engineeishas developed of tion, availableby contacting ASCE in New York. Finally, the discussion urhan "shopper'sguide" to urban drainageanalysis modelsin Chaptlr 25 includesa useful and designsoftware.

16.8 FLOODPLAIN ANALYSIS over the years,the federalgovernDue to heavymonetary and other floodplainlosses and methof floodplainsof the nation's waterways ment hasbeenconductingstudies inthat causes prop"tty and pieventing overdevelopment ods of protecting life und flooding. Hydrology is a key ingredientin widespread water levelsand more creased these studies for identifying potential flow rates, studying effects of dams, open volumes and on water control structures hydrographs determining and.other channels, and by be safelystoredand conveyed the waterways of floodwatersthat will needto floodplains.

U.S.NationalFlood lnsuranceProgram(NFIP)
In 1968, the U.S. Department of Housing and Urban Development(HUD), later to called the FederalEmlrgency ManagementAgency (FEMA), initiated the NFIP of floodplainswith mappingof identify flood hazard areasand to provide occupants tolorv-costflood insurance'The NFIP requireslocal the flood-proneareasand access programsthat preventdevelgou"rn-"nt5, to adoptand implementflood management of opmentsin excess national standards' Since the inception of the National Flood InsuranceProgram, flood hazatd The areas have been mapped in over 18,000 communities in the United States. and has sinceconvertedto a maintenance programcost over $t.O Uittion to complete occur'-Each of these effort of updating and expandingthe maps as developments detailedevaluationof peak flow rates for studieshas required eithei approximateor has a range of recurrenceintervals. The 100-yeardischarge,called the baseflood, portion ofthe floodplainoccupiedby the baseflood The in beendetermined all cases.

4'10

oHAPTER HYDRoLoGIc 16 DESIGN

has been mapped,allowing communitiesto determinewhether a property is in the 100-yr floodplain,and in many cases, what water surfaceelevationwould be experiencedat the property during the baseflood. Figure 16.37illustrates typical NFIP mappingand floodplain the management procedure.Surveyedvalley and channelcross-sections used in determiningthe are 100-yr flow depth, allowing the hydrologistto delineatethe lateral extentof flooding duringthe 100-yrflood. Then afloodwaywidth is generallydetermined that portion as of the floodplain that is reservedin order to dischargethe 100-yearflood without cumulativelyincreasingthe water surfacemore than 1.0 ft. This procedureis illustrated in Fig, 16.38.The floodwayis most often centeredover the main stream channel,but can be offset or even split into severalzones. Development within the floodwayis allowed only if compensated relocating by the floodway or mitigating the water surfaceincreasedue to the development. The is that portion of the floodplain outsidethe floodwayin which developflood fringe ment is allowed,up to a point of full encroachment buildings,roadbeds, by berms,and so forth. As much as sevento ten percentof the total land area of the United States lies within the 100-year floodplain.The largestareasof floodplainare in the southern parts of the country, and the most populatedfloodplainsare alongthe north Atlantic coast,the GreatLakesregion,and in California. The floodplain mapping effort produced a large amount of data and analyses useful to designhydrologists. The productsof the program include: 1. The 10-, 50-, 100-, and 500-year frequencydischarge streams. for 2. The 10-, 50-, 100-,and 500-year flood elevations riverine,coastal, for and lacustrine floodplains.

Floodway Flood Fringe Flood Fringe

',100_year" Floodplain

Channel

Figure 16.37 Definition sketchof floodplain delineations.

ANALYSIS 16,8 FLOODPLAIN

100Year FloodPlain Floodway Encroachment Flood Fringe

411

width' the for Figure 16.38 Procedure determining floodway 3.The100-and500-yearmappedfloodplaindelineationsatscalesranging to from 1:4800 1:24,000' 4. The 100-yearfloodwaydata and mapping' wave haz5. Coastalhigh hazard irea mapping 1*"u* subjectto significant ards). 6. FloodwaYflow velocities' risk zones' 7. Insurance This information is provided in the form of three products: |.FloodlnsurancestudyReportsprovidegeneralprogramandcommunity floodway data,tabulated. flood dischatge information,tauutatea background information' tabusurcharge datalncluding velocity, floodwaywidth, and dati, and profi'lesof the 10-' 50-' 100-' and lated flood insurance'zone flooding' for 500-yearflood elevationversusstreamdistances riverine of provide delineations the 1002. Flood InsuranceRateMaps (FIRM maps) and500-yearfloodplains,basefloodelevations'coastalhighhazardareas' andinsuranceriskzonesonaplanimetricbaseatascalebetweenl:4800 and 1:24,000. 3.FloodBoundaryHazardMapsprovidedelineationsofthel00-and500floodplainand channelcrosssections locationsof surveyed yearfloodplains, floodwayon a of and delineations the 100-year usedin hydraulic analyses, 1:4800and 1:24'000' baseat a scalebetween planimetricor topographic

Studies for HydrologY FloodPlain

NFIP studiesare basedon Flood flow frequency estimatesfor gaugedlocationsin Annual peak Type III (seechapter 27; analysisof streamflowrecords. log-pearson by recommended FEMA' flows and historical data arcfitted accordingto procedures

412

CHAPTER16

HYDROLOGIG DESIGN

through For ungauged locations,flood flow frequency'estimates developed are regionalfrequencyanalysis throughrainfall-runoff modeling.Equationspublished or to by the U.S. Geological of Surveyrelatepeak dfscharges variousfrequencies various shape,andland use.These drainage basincharacteristics suchassize,slope,elevation, equationsare developedusing multiple regressiontechniques(see Chapter 26) at gaugedsitesthroughoutthe region. (Chapter24) use syntheticrainfall hyetoRainfall-runoff modelingtechniques graphs.Storm-eventmodels, such as the Corps HEC-I and SCS TR-20 packages, employdesignstormsof particular frequencies and then mathematicallysimulatethe physical runoff process.The resulting peak dischargeis assumed have the same to frequencyas the rainfall.

U.S.Flood Hazards
Despiteconsiderable effort and expenditurein identificationof floodplainsand flood damage hazardareas, continueto resultin severe dam failuresand other catastrophies life, property, and the environment.Floods from hurricanes,intenserainstorms, to and rapid snowmelt or structure failure have all contributed to the loss of life. A tabulationof eventscausingmore than 100 deathsin the United Statesis providedin Table 16.15. As indicated, the majority are hurricane related, principally concentrated in the east-coast and Gulf of Mexico regionsas sfown in Fig. 16.39. Monetary losses from floodsare also large.Table 16.16showsa numberof past each,given in 1966dollars. U.S. floodsproducingover $50 million in flood damages these floods have produced flood dam4gesin billions of dollars, disCollectively, I tributed through the yearsas shown in Fig. 16.40. The Federal InsuranceAdministration evaluatedthe floodplain areas in the and economicinformation, communitiesmappedby FEMA. By using demographic projectionsof future property at risk of flooding could be made.Results.suggest that property have occurred in floodplains. of investmentsin flood damageable billions value that Table 16.17lists the breakdown,by state,of estimated1990 development will be in harm's way.

Dam Break Hazards

' .-"-fabir- 16.18lists the outflow rates,peak depth, and storageat the time of failure for 18 significantdam failures in the United States.The death rate for dam failures is thosepeoplewho would relatedto the polpulation at risk (PAR). This term describes need to take someaction to avoid the rising water. Figures 16.41 and 16.42showthe losses functions of PAR for low (lessthan as 1.5 hr) and high (greater than 1.5 hr) advancewarning times, respectively.The high-warning-time losses significantlyless.This strongly supportsthe incorporaare tion of early warningand flood delayfeaturesin the designof any structure.Data used in plotting Figs. 16.41 and 16.42 are given in Table 16.19. resiTable 16.20providesa typical time line requiredfor alerting downstream dentsof a severestorm and potential dam failure. The valuesgiven are hypothetical, and apply to an assumed15-mi reach betweenthe storm center and the populated atea.

SUMMARY 413
1OO MOREDEATHSIN THE OR TABLE 16.15 FLOODSCAUSING UNITEDSTATES

Year
1831 1856 1874 1875 1886 1889 1893 1899 1900 1903 1903 1906 1909 1913 1913 1915 l9l9 l92l 1926 1927 1927 1928 1928 1928 1932 1935 1935 1936 1937 1938 1955 195'l 1960 1972 ' 1972 t1976

Streamor place BaratariaIsle, LA LA Isle Derniere, ConnecticutRiver tributarY TX Indianola, TX Sabine, PA Johnstown, Vic. GrandIsle,'LA PuertoRico TX Galveston, Central States HePPner,OR Gulf coast Gulf coast-New Orleans Miami, Muskingham,and Ohio Rivers Brazos River, TX Louisianaand TexasGulf coast Louisianaand TexasGulf coast Upper ArkansasRiver Miami and Clewiston,FL River Lower MississiPPi Vermont Puerto Rico FL Lake Okeechobee, San Francisco,CA PuertoRico Florida KeYs River, KS, NE RePublican United States Northeastern Ohio River New Englandcoast United States Northeastern Westcoast,LA PuertoRico Buffalo Creek, WV RaPid.Creek,SD Big Thompson,Co

Lives lost

Cause
Hurricane tidal flood Hurricane tidal flosd Dam failure Hurricane tidal flood Hurricane tidal flood Dam failure Hurricane tidal flood Hurricane tide and waves Hurricane tidal flood Rainfall-river floods Rainfall-river floods Hurricane tidal flood Hurricane tidal flood Rainfall-river floods Rainfall-river floods Hurricane tidal flood Hurricane tidal floo<l Rainfall-river flood Hurricane tidal and river flood Rainfall-river floqd Rainfall-river flood Hurricane tide and waves Hurricane tidal flood Dam failure Hurricane tide and waves Hurricane tidal flood Rainfall-river flood Rainfall, snowmelt-river fl oods Rainfall-river flood Hurricane tidal and river flood Hurricane rainfall-river floods Hurricane tide and river floods Hurricane rainfall-river floods Dam disaster Rainfall Rainfall

150 320 t43 176 150 2t00 2000 3000 6000+ 100+ 247 151 700 46'7 177 550 284 t20 350 100+ t20 300 2400 350 ,))\ 400 110

ro7
137 200 115 556 r07 t25 245 r39

414

DESIGN 16 CHAPTER HYDROLOGIC

Figure 16.39 Number by state of major hurricanesin the United States, NaI 899-1989.(Soarce.' tional Hurricane Center, National WeatherService, NOAA.)

Figure 16.40 Average in annual flood damages the U.S., 1916-85. (Source:National WeatherService,NOAA.)

Damages Damages (1985 $) Damages/200 million population (1985 $)

rg2o 25

30

35 40

45

50

55

60

65

70

75

80 8s

of Last Year five'Year Period

$50 MILLION TABLE 16.16 FLOODSRESULTINGIN DAMAGESEXCEEDING IN THE UNITEDSTATES Damage($ millions)

Year

Streamor place

dollars 1966 Dollars Contemporary

N.A."
.A

Cause
Rainfall-river flood Dam failure Hurricane tidal floods Rainfall and dam failure Rainfall-river flood Rainfall-river flood Hurricane rainfall-river floods Rainfall-river flood Hurricanetidal and river floods Rainfall-river floods Rainfall-river flood Rainfall-river flood Hurricane tide and waves Rainfall-river flood Rainfall-river flood Rainfall- snowmelt fl ood Rainfall-river flood Hurricanetidal and river floods Rainfall-river floods Rainfall-river floods Rainfall-river floods Huriicane tidal and river floodb Rainfall-river floods Rainfall-river floods Hurricane tidal and,rivet floods Rainfall-river floods Hurricanetidal and river floods Rainfall-river floods Rainfall-river floods Rainfall-river floods Rainfall-river floods Snowmeltfloods Rainfall-river floods Hurricane tidal floods Huiricane tidal and river floods Rainfall-river floods Rainfall-river floods Rainfall-river floods Rainfall-river floods Hurricanetidal and river floods Hurricane tidal floods Hurricanetidal and river floods Rainfall-river floods Rainfall-river floods Rainfall-river floods Hurricanetidal and river floods Rainfall-snowmelt river flood Rainfall-river floods Rainfall-river floods Hurricane tidal flood

River t844 Upper MississiPPi PA 1889 Johnstown, 1900 Galveston,TX and 1903 Passaic DelawareRivers 1903 Missouri River basin 1 9 1 3 Ohio River basin t913 Brazos and ColoradoRivers, TX 1921 ArkansasRiver 1926 Miami and Clewiston,FL t926 Illinois River 1927 New England 1927 Lower MississiPPi 1928 Puerto Rico and 1935 Susquehanna DelawareRivers r936 NortheasternUnited States t936 Ohio River basin 1937 Ohio River basin 1 9 3 8 New England sffeams 1938 California streams t942 Mid-Atlantic. coastalstreams 1943 Central states 1944 South Florida 1944 Missouii River basin 1945 Hudson River basin 1945 South Florida 1945 Ohio River basin 1947 South Florida t947 Missouri River basin 1948 Columbia River basin 1950 San JoaquinRiver, CA I 1 9 5 KansasRiver basin 1952 Missouri River basin 1952 Upper MississiPPiRiver 1954 New England streams 1955 NortheasternUnited States 1955 California and Oregonstreams 1957 Ohio River basin 1957 Texas rivers 1959 Ohio River basin 1960 South Florida 1 9 6 1 Texascoast 1964 Florida 1964 Ohio River basin 1964 California streams 1964 Columbia River -North Pacific t965 South Florida 1965 Upper MississiPPiRiver 1965 Platte River, CO, NE 1965 ArkansasRiver, CO, KS 1965 New Orleans and vicinitY
,N.A. : not available. Council, 1968. Source:IJ.S.WaterResources

25 25 50 150 t28 t3
'7rl

1,161 84 100 273 N.A, 516 349 64 130

]I

N.A, 50 284 50 36
zzl

178 N,A. 90 185 374

3tl

150 418 125 100 28 172 63 52 24 54 34 60 178 to2 883 180 198 180 684 271 65 144 lt4 78 300 32s 106 t'73 289 r39 158 19l 61 322

996 3'76 294 103 N.A. rt7 N.A. 75 98 6l 88 N.A. 226 57 N.A. N.A. N.A. 216 879 405 72 188 r20 86 336
3+Z

r12 183
3lL 144 r62 N.A, 65 338

FLOODING, AT VALUE RISKFROM PROPERry TABLE 16.17 ESTIMATED ON BASED 1990COSTS RANKED DECREASING IN ORDER. Rank 1 2 3 5 6 7 8 9 l0 tl 12 13 L4 15 16 l7 18 1,9 20 2l 22
z3
/)A

State
California Florida Texas Louisiana Arizona New Iersey New York Illinois Massachusetts Pennsylvania Virginia Maryland Washington Ohio Michigan North Carolina Wisconsin Georgia Connecticut Missouri Indiana Minnesota Nebraska Oklahoma Alabama South Carolina Tennessee Colorado Oregon Mississippi New Mexico Kansas Iowa RhodeIsland Kentucky North Dakota Urah Nevada Arkansas Delaware Maine West Virginia New Hampshire South Dakota Idaho Hawaii Vermont Wyoming Montana Alaska

value,X $1000 Property 163,323,1,92 131,548,814 72,376,950 45,402,322 45,094,183 38,945,265 32,005,900 26,880,755 23,8t3,115 18,888,390 17,441,420 16,330,448 t6,245,009 t5,273,r47 13,449,078 12,993,067 12,r8r,725 11,832,494 tt,1r7,290 11,654,861 r0,786,741 10,655,t64 t 10,360,574 9,501,778 9,274,903 9,220,305 8,037,425 7,137,757 6,861,790 6,134,073 5,519,278 5,279,t94 5,26r,678 4,312,117 4,r70,637 3,924,872 3,812,936 3,437,813 3,005,rs0 2,954,467 2,416,322 2,098,262 t,991,453 1,430,610 1,39t,498 1,323,90s 1,091,099 1,081,460 881,661 647,81,8

25 26 2'7 28 29 30 3l
JZ

33
3+

35 36
3 t

38 39 40 4I
4)
43

44
4f

46 47 48 49 50

"status of Floodplain Hazard Evaluation Under the National Flood Insurance Program," Source: B. R. Mrazik, Emergency ManagementAgency, Washington,DC, 1986. _ fed9r91

17.3 SUBSURFACE DISTRIBUTION OFWATER 429 locations it has become more important than overdrafts of groundwater supplies. Today,the hydrologistmust be concernedwith both the qualitli and quantity aspecrs of groundwater.Furthermore, there is emerging an increasing specialization in groundwaterquality modeling.This latter type of modelingis ge;erally beyondthe scopeof this text but information on this topic may be founOin'Refs. j-6.

17.2 GROUNDWATER FLOW-GENERAL PROPEFTIES

Understanding movementof groundwaterrequiresa knowledgeof the the time and spacedependencies the flow, the nature of thJporous medium and fluid, of and the ,boundaries the flow system. of Groundwater flowsareusuallythree-dimensional. Unfortunately, solutionof the suchproblemsby analytic methodsis complexunlessthe systemis symmetric.7,8 In othercases, space dependency oneofthe ioordinate direciions.uy 6" so slightthat in assumption two-dimensionalflow is satisfactory. of Many problemsof practical importancefall into this class. Sometimes one-dimensional flow can be assumed, thus further simplifying the solution. Fluid propertiessuchas velocity, pressure, temperature,density,and vis6osity often vary_intime and space.When timi dependency occurs,the issueis termed an unsteady flow problem and solutionsare usually difficult. On the other hand, situations wherespacedependency aloneexistsare iteady flow problems.Only nomogeneous (single-phase) fluids are consideredhere. For a discussion muliiple phase of flow, Refs.5 and g are recommended. Bou-ndaries groundwater to flow systems may be fixed geologicstructures free or water surfaces that are dependent their position on the stateol the flow. A hydrolfor ogistmust be ableto definetheseboundaries mathematicallyif the groundwater flow problemsare to be solved. Porousmedia through which groundwaters flow may be classifiedas isotropic, anisotropic,heterogeneous, homogeneous, severalpossiblecombinations or ofthese. An isotropic medium has uniform properties in all directions from a given point. Anisotropic mediahaveone or more propertiesthat dependon a given diiection. For example, permeabilityof the medium might be greateialong a horizontal plane than alonga vertical.one.Heterogeneoas mediahavenonuniform-properties umrotropy of or isotropy, while homogeneous media are uniform in their ihaiacteristics.

17.3 SUBSURFACE DISTRIBUTION WATER OF

4;

GEoLocy sotLs,AND 17 cHAprER GR.,NDWATER, and extends l. Soilwater zone.A soil water zonebeginsat the ground surface downward through the major root band. Its total depth is variable and

'

in may be encountered this region: hygroscopicwater, which is adsorbed from the air; capillary water, held by surfacetension; and gravitational soil water draining through the soil' water, which is excess from the bottom of the soil-watetzone 2. Intermediate zone.This belt extends to to the top of the capillary fringe and may ghangefrom nonexistence link The zoneis essentiallya connecting severalhundredfeet in thickness. between a near-ground surface region and the near-water-tableregion through which infiltrating fluids must pass. from the watertable (Fig.I7 '2) to 3. Capiltary zone. Acapillary zoneextends in by a height determined the capillary rise that can be generated the soil. The capillary band thicknessis a function of soil textureand may fluctuate not only from region to region but also within a local area. 4. Saturatedzone. In the saturatedzone, groundwaterfills the pore spaces completelyand porosity is thereforea direct measureof storagevolume. Part of this water (speciflcretention) cannot be removedby pumpirygor tensionforces.Specificretention ofmolecular and surface because drainage is the ratio of volume of water retained againstgravity drainageto gloss volume of the soil.

,:'##l":ff l;"lhr?ff :ffi fi',,nH#,:?'l*?,T#*.Ji'.i,'"1:f, ::J;

Waterthat can be drainedfrom a soil by gravity is known as the specificyield. as It is expressed the ratio of the volumeof waterthat can be drainedby gravity to the grossvolumeof the soil. Valuesof speciflcyield dependon the soil particle size,shape and distribution of pores, and degreeof compactionof the soil. Averagevaluesfor alluvial aquifersrange from 10 to 20 percent. Meinzer and others have developed procedures determiningthe specificyield.12 for

CONSIDERATIONS 17.4 GEOLOGIC

volumesand flow ratesrequiresa thoroughknowlof The determination groundwater basin.In bedrock areas,hydrologiccharacterisedgeof the geologyof a groundwater tics of the rocks,that is, their location,size,orientation,and ability to storeor transmit to basinsoften containhundreds rock areas, water,mustbe known. In unconsolidated fill to of thousands feet of semiconsolidated unconsolidated depositsthat originated quantitiesof areas.Suchfllls often contain extensive from the erosionof headwater of storedwater. The characteristics thesebasin fills must be evaluated. A knowledge of the distribution and nature of geohydrologicunits such as or to aquifurs,aquifugis, andaquicludesis essential proper planningfor development mustbe In supplies. addition,bedrockbasinboundaries of management groundwater locatedand an evaluationmadeof their leakagecharacteristics. An aquifer is a water-bearingstratumor formation that is capableof transmitAquifers may be considered ting waterin quantitiessufficientto permit development.

CONSIDERATIONS 17,4 GEOLOGIC

P^^L ...u,ritlge

431

arc?

Discharge area

E] F a

(/)
TT]

F z p

(b)

Figure 17.2 Deflnition sketches of groundwater systems and mechanisms for rechargeand withdrawal: (a) aquifernotationt0and (b) componentsofthe hydrologic II cycleaffectinggroundwater. as falling into two categories, confined and unconfined, depending on whether a water table or free surface exists under atmospheric pressure. Storage volume within an aquifer is changed whenever water is recharged to, or discharged from, an aquifer. In the case of an unconfined aquifer this may easily be determined as AS:SIAV

(17.1)

432

CHAPTER17

SOILS,AND GEOLOGY GROUNDWATER,

where AS : the changein storagevolume specificyield of the aquifer S, : the average the AV : the volumeof the aquiferlying between original watertable andthe water table at somelater specifictime For saturated, confined aquifers, pressure changes produce only slight modificationsin the storagevolume. In this case,the weight of the overburdenis supportedpartly by hydiostatic pressureand somewhatby solid material in the uqoif"t. When hydrostaticpressurein a conflned aquifer is reducedby pumping or with the result its causing compression, theioad on the aquiferincreases, othermeans, a alsocauses small the hydrostaticpressure that somewateris forced out. Decreasing an additional releaseof water. For confined expansion,which in turn produces in aquifers,water yield is expiessed terms of a storagecofficient S", definedas the per volumeof wateian aquifei takesin or releases unit surfaceareaof aquiferper unit of Figure 17.2 illtstrates the classifications changein head normal to the surface. aquifers. ratesofyield, thereare strataexhibitingsatisfactory In additionto water-bearing stratathat may contain large quantitiesof and impermeable also non-water-bearing rates are not high enoughto permit effectivedevelopwater but whosetransmission ment. An aquifugeis a formation impermeableand devoidof water; an aquicludeis an imPerviousstratum. In the following three chapters,the mechanicsof groundwaterflow and the presentedall elementsof groundwatermodelingwill be introduced.The techniques physical systemto be modeled.Before a numerical of dependon a=knowledge the a model can be developed, conciptual framework must be devised.This framework must take into accountthe region's topographyand geology;the types of aquifers, lithological variations,and characteristics; lateral extent,boundaries, their thickness, areas,their rates of dischargeand dischar_ge the nature and extent of rechargeand table'' recharge;and the elevationof the water

Topography
to it operates, is essential know something how a groundwatersystem To understand compiledshowingall surface map shouldbe A aboutthe region's surface. topographic water bodies,including streami, iakis, and artificial channelsand/or ponds,as well as land surfacecontours.Furthermore,an inventory of pumping wells, observation suchas identifying types of wells, and explorationwells shouldbe madefor purposes watertable locationsandrates;and determining soilsand rocks,pinpointingdischarge elevations.

GeologY Subsurface
governsthe occurrenceand movement The geologicstructureof a groundwater'basin of the grirndwater withinlt. Specifically,the number and types of water-bearing formations, their vertical dimensions,interconnections,hydraulic properties, and Once the before the systemcan be analyzed.u outcrop patternsmust be understood and lower conditions have been identified, contour maps of the upper subsuriace

SUMMARY 433 boundariesof aquifers,watertable contourmaps,and mapsof aquifercharacteristics can be prepared.Well-drillers logs, experimentaltest wells, and other geophysical explorationmethodscan be usedto obtakr the neededgeologic data.s-e'13'14

17.5 FLUCTUATIONS GROUNDWATER IN LEVEL

Any circumstance that alters the pressureimposedon undergroundwater will also factors,changes stream and in causea variation in the groundwaterlevel. Seasonal pressure river stages, changes, winds, tides,external evapotranspiration, atmospheric all loads,various forms of withdrawal and recharge,and earthquakes may produce fluctuationsin the water table level or piezometricsurface, depending whetherthe on aquifer is free or confined.eIt is important that an engineer concernedwith the development and utilization of groundwatersuppliesbe awareof thesefactors.The engineershould also be able to evaluatetheir importance relative to operation of a speciflcgroundwaterbasin.

17.6 GROUNDWATER-SURFACE WATERRELATIONS

Notwithstandingthat water resourcedevelopment often been basedon tL pr"has it that dominantuseof either surface wateror groundwater, mustbe emphasized these Changesin one two componentsof the total water resource are interdependent. and componentcan havefar-reachingeffectson the other. Coordinateddevelopment management the combinedresourceare critical. Linkage betweensurfacewaters of effects in and groundwaters shouldbe investigated all regionalstudiesso that adverse understood. can be noted if they exist and opportunitiesfor joint management In Part Three it was shown how surface stream flows are sustainedby the groundwater are resource,and it was also pointed out that groundwaters replenished by infiltration derived from precipitation on the earth's surface. Undergroundreservoirsare often extensiveand can serveto store water for a recharge, multitude of uses.If withdrawalsfrom thesereservoirsconsistentlyexceed results.By properly coordinating mining occursandultimate depletionof the resource optimum regional water resource the use of surfacewater and groundwatersupplies, development seemsmost likely to be assured.Several studiesdirected toward this use coordinated havebeeninitiated.r5'16

r Summary
The importance of groundwaterto the health and well-being of humans is well documented. Groundwateris a major sourceof freshwaterfor public consumption, industrial uses, and the irrigation of crops. For example, more than half of the freshwaterused in Florida for all purposescomes from groundwatersources,and about 90 percent of that state'spopulation dependson groundwaterfor its potable is this resource clear.Quantity and quality dimenwatersupply.The needto husband sionsare both important.

4g4

CHNPTENTZ

SOILS,AND GEOLOGY GROUNDWATER,

Groundwaterprotectionandmanagementpracticesmustbebasedonanunder. groundwateris^distributed standing of groundwatersources,the rianner in which of topogiaphic,and soil characteristics the region' geologic, below the earth's surface, andtheinterconnectionsbetweengroundwaterandsurfacewatersources.

REFERENCES
"Ground Water,"Mech'Eng' (Jan' 1960)' 1 . J. G. Ferris, ..GrouniwaterProtection," Final Report of the National The Conservationrounoaiion, L. D'C'' 1987' GroundwaterPolicy Forum, Washington, Gegpe-F Pirltder'.Groundwater e. F. Wood, Ra'ymond p"rru.u, diUiu- G' Gray' and E. CoitaminationfromHazardousWastes'EnglewoodCliffs'NJ:Prentice-Hall'1984' in the United States. R'pu,ri"t, unJ f. euurt" s, Grouidwater Contarnination ;. ffi;;; Press' 1983' ehitadelphia:University of Pennsylvania R.A.FreezeandJ.A.Cir"..y,Groundwater'EnglwoodCliffs'NJ:Prentice-Hall'19'19' York: McGraw-Hill' 1993' of llydrology' D. R. Maidment (ed.), ii"aLook ,,Niurnerical New Basins,"InternaModeling of Groundwater J. Boonstraand N. A. O" nlJa"r, The Netherlands'1981'' tional Institute for Land Reclamationand Improvement' New York: Wiley' 1965' R. J. M. DeWiest,Geolrydrology' Wiley' 1960" D. K. Todd, GroundwaterHydrology' New York: ..croonowu# n"gi#, of the United States," GeologicalSurvey water R. c. Heath, Printing Office' 1984' Supply PaperNo. 22a2' iishiniton, D'C-: U'S' Government DisposalPracproteciion ig"""y, "The Report to Congress:^Waste 1 1 . U.S. Environmentat summary'U'S' EPA' PB 265-364' tices and Their Eft'ectson Groundwater,"Executive 1977. States,"U'S' Geological tz. o. E. Meinzer, "The occurrence of Groundwaterin the united 1923' Survey,Water-SupplyPaperNo' 489' "A GeneralizedGraphical Method Evaluating -for 1 3 . H. H. Cooper, Jr. and i. n. lacoU, Trans'Am' Geophys'Union Formation constants una so*-u.izing well-Field History," 27, 526-534(1946). "Outline of Methodsfor EstimatingCrounlw11e1fupplies"' U'S' Geolog1 4 . O. E. Meinzer, D C:' 19?2ical Survey,Water-SupplyPaper638-C, Washington' "Conjurrliu" Op"tution of Dams and Aquifers"' Proc' ASCEJ' Hyd' Div' 1 5 . Nathan Buras, S9(HY6) (Nov. 1963). 16. F'B.Clendenen,..AComprehensivePlanfortheConjunctiveU{ilizationofaSurface water SupplyDevelopment:Solano Reservoirwith undergroun'Js,orug" for Basin-wide 1959. project california,,, po.to. oi Eni thesis,university of california, Berkeley,

1B Chapter

of Mechanics Flow

r. Prologue
The purposeof this chaPteris to: i I . . I Presentthe principlesof groundwaterflow. Describesoil propertiesthat affect groundwaterstorageand movement. Describethe relevanthydrodynamicequations. Relatethe mechanicsof groundwaterflow to modelingregional groundwater systemsand calculating flows to wells and other groundwater collection devices.

18.1 HYDROSTATICS
of Water locatedin pore spaces a saturatedmedium is under pressure(calledpore in by cin be determined insertinga piezometer the mediumat a point pressure), which it If LocationA (Fig. 18.1)is considered, can be seenthat pore pressure of interest. is given by P:h"l where p : the pore pressure(gaugepressure) from the point to the water table h o : the headmeasured the specificweight of water v: positive or negative,dependingon whetherthe pressure Pore pressureis considered the above(poJitive)or below (negative) point under consideration. headis measured the If an arbitrary datum is established, total head or piezometric head abovethe datum is (18.2) Po:Z'th wherePois known as the piezometricpotential. In Fig' 18.1 this is equal to ho 1- 7o zone.The zone andza - h6forPointB in the unsaturated for poiniA in the saturated

(18.1)

43r.

436

CHAPTER18

OF FLOW MECHANICS

preshydrostatic showing Figure 18.1 Definitionsketch medium. in sures a porous or tension vacuum(negative,pore of Iermh"is the pore pressure A while -hu denotes pressure) B. at

FLOW 18.2 GROUNDWATER

Analogiescan be drawn betweenflow in pipesunder pressureand in fully saturated confinid aquifers.The flow of groundwaterwith a free surfaceis also similar to that flow system in an open channel.A major differenceis the geometryof a groundwater pipe flow or channel systems'The channel as compared with common hydraulic problem "un eutily be recognizedby envisioninga dischargingcross section composedof a numbei of small openings,eachwith its own geometry,orientation, and iir" ,o that the flow velocity issuingfrom eachpore varies in both magnitudeand direction. Difflculties in analyzingsuch systemsare apparent. Computations are of usually basedon macroscopicaverages fluid and medium propertiesover a given area. cross-sectional flow problemsare density, in Unknown quantitiesto be determined groundwater la to are assumed exist'r In conditions pressure, and velocity if constanttemperature so general,water is consideredincompressible, the number of working variablesis later relative to the storagecoefficientfor ieduced.An exceptionto this is discussed here will be placed on the flow of water in a a confined aquifei. Primary emphasis porous medium. saturated

18.3 DARCY'SLAW
Darcy's law for fluid flow through a horizontal permeablebed is statedas' O: - K A dh dx (18.3)

LAW 18.3 DARCY'S areaincluding the spaceoccupiedby the where A : the total cross-sectional porous material K : the hydraulic conductivity of the material the Q: the flow across control areaA In Eq. 18.3 h : z * where h - the piezometrichead the elevationabovea datum p : the hydrostaticpressure C _ an arbitrary constant in If the specificdischarge = QIA is substituted Eq. 18.3, S q: -K*(,.t)

437

P-+c v

(18.4)

( 18 .s )

, Note that 4 also equalsthe porosity n multiplied by the pore velocity Vo.Darcy's law is widely used in groundwaterflow problems.Severalapplicationsare illustrated in later sections.

No is definedherein as
No: PQd lL

(18.6)

where q d P p

: : : :

the specificdischarge the mean grain diameter fluid densitY dynamic viscositY

For many conditions of practical importance (zones lying adjacent to collecting devicesare an exception),Darcy's law has been found to apply' to Of specialinterestis the fact that the Darcy equationis analogous Ohm's law

' : (*)u
where i:thecurrent R : the resistance E : the voltage

(18.7)

as current and velocity are analogous, areK andI f R, andE anddhldx.The similarity of the two equationsis the basis for electric analog models of groundwaterflow svstems.2'3

438

CHAPTER18

OF FLOW MECHANICS

EXAMPLE 18.1

: . Water temperaturein an aquifer is 60'F and the rate of water movement t .2 ft I day particle diameterin the porous medium is 0.08 in. Find the Reynolds The average number and indicate whetherDarcy's law is applicable. . Solution. Equation 18.6 gives the Reynoldsnumber as

X*: #
This may also be written as
qd N--'
t)

Converting From TableA.2 in Appendix A, o is found to be 12l X 10-s ft2lsec. gives4 : I.2/86,400 : 1.39 X 10-5' The the velocity4 into units of ftlsec thesevaluesin the meangrain diameterin ft : 0.08/12 : 0.0067.Substituting equation,we obtain
Nn:

r.39x10-sx0.0067
1 . 2 1X 1 0 - s 0.0077

SinceN,, < 1.0,Darcy's law doesapply' rl

18.4 PERMEABILITY
into The hydraulic conductivity K is an important parametrthat is often separated medium, the other to the fluid. The product one related to the two components,
k:Cdz

(18.8)

can be written

X:4
lL

(18.e)

of Dimensions intrinsic permeability areL2. Sincevaluesof k given asft2 or cm2 known as the darcy has beenwidely adopted. are extremelysmall, a unit of measure 1 darcy : 0.987 x 10-s cm2 or 1.062x l0-1r f* hydraulic conductivity are reportedin the literature. Severalwaysof expressing hasdefinedthe standardcoefficientof permeabilityK" as Survey The U.S. Geological - the nunber of gallons per day of water passingthrough 1 ft2 of medium under a

446

CHAPTER18

OF FLOW MECHANICS

" If Eqs. 18.46 are substituted into Eq. 18.50,then

#o;+frat:o

(18.s2)

must be a constant.A seriesof is The total differential dry' equal to zero,and ry' of curves {(r, y) equalto a succession constantscan be drawn and will be tangentat all points to the velocity vectors.Thesecurvestrace the flow path of a fluid particle and are known as streamlinesor flowlines. An important property of the stream function is demonstratedwith the aid of Fig. 18.2. Consider the flow crossinga across definedas f1 and tltr.lf the discharge vertical sectionAB betweengfueamlines as the sectionis designated Q, it is apparentthat

n: r,'
f*t

udy

( 1 8.s 3)

Q : and

I d,t' J,t,^

(18.s4)

Q:Qt-Q,

(18.55)

Equation 18.55illustratesthe important property that flow betweentwo streamlines is ionstant. Streamline spaclngreveals the relative magnitudesof flow velocities and vice versa. with narrower spacings, betweenthem. Higher valuesare associated as The curvesin Fig. 18.2 designated @rand Q,, calledeqaipotentiallines, ate determinedby velocity potentials Q@, y): constant. These curves intersect the flowlines at right angles,illustrated in the following way. The total differential d@is given by

a0 a+, d0: frdx, + fidt

substituting for terms aslax and 0Q/sy their equivalentsu and o gives us udx-lody:O
and

(18.s6)

(18.s7)

dy dx

u
t)

(18.58)

and sf The lineq Thusequipotential arenormalto flowlines. rystem flowlines equiponet. tentiallinesformsa flow

CONDITIONS 18.8 BOUNDARY

447

'

is One significantpoint of differencebetween{ and ry'functions that equipotenis irrotational. For two-dimensional flow the tial lines exist only when the flow condition of irrotationality is said to exist when the z componentof vorticity (. is
ZA[O; Ot

: t,: (x-,4) o
'
d-Q

(18.se)

substitutingforu and o in Eq. 18.59in termsof Proof of this is givenby Eskinazi.6 @,we obtain

_ a26:0
EY0x

(18.60)

6xdy

This indicatesthat when the velocity potential exists,the criterion for irrotationality is satisfied. or Once either streamlines equipotentiallines in a flow domain are determined, of the other is automaticallyknown because the relations in Eq. 18.48.Thus r

and

r: [(X*- H,o.) - H*) r: [(Xo.

(18.61a) (18.6lb)

It is enough then to determineonly one of the functions, since the other can be obtainedusing relationsEqs. 18.61aand 18.61b.The complexpotentialgiven by

(r8.62)
Of where l, the squareroot of -1, is widely used in analytic flow net analyses'a's specialimportanceis the fact that (18.63) y2w:V26+iYzt1r:g the satisfies conditions of continuity and irrotationality simultaneously. Equations presentedin this section have been limited to the case of twodimensional flow. Extension to three dimensionswould be obtained in a similar fashion.

CONDITIONS 18.8 BOUNDARY

surfaces). Impervious boundariesmay be artificial objects zuch as.concretedams, rock boundary In strata,oi soil stratathat arehighiy impervious. Fig' 18.3the impervious

448

CHAPTER18

OF FLOW MECHANICS

AB represents sucha limit. Sinceflow cannotcrossan imperviousboundary,velocity In components normal to it vanishand the imperviousboundaryis a streamline. other : constant. words,at the boundary,V Next look at the upstreamface of the earth damBC. At any point of elevation hydrostatic,or y along BC the pressurecan be assumed p:y(h-y) The definition of a velocity potential statesthat (18.64)

o : - * \ v *' r )* . ' (z /
in for Substituting pressure Eq. 18.65yields

(18.65)

o:-*l
and

*r]*t

(18.66) (18.67)

6:-Kh+C medium, levelh andan isotropic reservoir Thusfor a constant

d : constant

and surfaceBC, often termed a reservoir boundary, is an equipotentialline. CD The free surfaceor line of seepage in Fig. 18.3 is seento be a boundary zones.Since flow doesnot occur acrossthis unsaturated betweenthe saturatedand along this free surfacemust be boundary,it is obviouslyalso a streamline.Pressure constant,and thereforealong CD 6 + Ky: constant (18.68)

This is a linear relation in $, and therefore equal vertical falls along CD must be equipotential drops. One important groundwater flow associatedwith successive problem is to determinethe location of the line of seepage. the DE The surfaceof seepage of Fig. 18.3 represents location at which water downstreamface of the dam and trickles toward point E. The seepsthrough the is The surfaceof seepage neither a flowline nor an pressurealongDE is atmospheric. line. equipotential

ImPervious laYer

Figure 18.3 Some common boundary conditions.

NETS 18.9 FLOW

449

18.9 FLOWNETS
of Flow nets, or graphicalrepresentations families of streamlinesand equipotential lines, are widely used in groundwaterstudies to determine quantities, rates, and flow at directions of flow. The use of flow nets is limited to steadyincompressible media or for regions that can be constant viscosity and density for homogeneous Darcy's law must be applicableto segments. into homogeneous compartmentalized the flow conditions. The mannerin which a flow net canbe usedin problem solvingis bestexplained with the aid of Fig. 18.4. This diagram showsa portion of a flow net constructedso by that eachunit b.ounded a pair of streamlinesand equipotentiallines is approximately square.The reasonfor this will be clear later. A flow net can be determinedexactly if functions Q and $ are known beforeflow nets are hand.This is often not the case,and as a result, graphicallyconstructed widely used. The preparation of a flow net requires application of the concept of square elementsand adherenceto boundary conditions. Graphical flow nets ard practice an acceptable usually difficult for a beginnerto create,but with reasonable methodsfor graphicalflow net constructionare net canbe drawn.Variousmechanical here.5'e presented the literature and are not discussed in usingthe geometryof it After a flow net hasbeenconstructed, can be analyzed the net and by applylngDarcy's law. that the hydraulic that Remembering h : p/y + z, we find thatFig' 18'4 shows gradient G2betweentwo equipotentiallines is given by

o r: x
Lq: '*(*)

(18.6e)

Then by applyingDarcy's law, in the mannerof Todd,2the flow incrementbetween is adjacentstreamlines (18.70)

area for a net of unit width normal to the the whereLm represents cross-sectional plane of the diagram. If the flow net is constructedin an orthogonalmanner and
h- Ll
Equalpotential lines (0 = constant) Lm

r d " Qz

Figure 18.4 Segmentof an orthogonalflow net.

450

CHAPTER18

OF FLOW MECHANICS

composedof approximatelysquareelements, Lm: L,s and Lq = K Lh (18'71) Now if there are n equipotentialdrops betweenthe equipotentiallines, it is evident that h Lh: IL

sections If wherc his the total headloss over the n spaces. the flow is divided into m per unit width of the medium is by the flowlines,then the discharge

o:iu:Y

(r8.72)

can when the medium's hydraulic conductivityis known, the discharge be computed 18.72 and a knowledgeof flow net geometry' using Eq. the Where the flow net has a iree surfaceor line of seepage, entranceand exil of discussion these given in Fig. 18.3 are useful. A more comprehensive conditions conditionsis given in Ref. 10. Some trouble arisesin flow net construction at locations where the velocity Suchpoints are known assingularpoints andaccording infinite or vanishes. becomes In to De'Wiestmay be placed in three separatecategories.a the first classification lines do not intersectat right angles'Sucha situationoften and equipotential flowlines with a flowline; PointA in Fig. 18.5is-anelamnl-e. occurswhen a boundarycoincides has a discontinuityalongthe boundary that abruptly The secondclassification such the changes slope of the streamline.In Fig. 18.6 PointsA, B, and c represent B it is zero' At discontinuities. PointsA and c the velocity is infinite, while at Point direction insidethe flow in of discontinuitymeasured a counterclockwise If the angle the than 180o, velocity is zero; if larger than 180o'it is inflnite' The angle field is le-ss examPle. atA is 270",for net' The third categoiyincludesthe casewherea sourceor sink existsin the flow of the flow net Under these circumstancesthe velocity is infinite, since squares

Line of seepage Tangent Surface of seepage

tt

a<90o,0<90"
(a)

ft)

(After Figure 18.5 Some entranceand exit conditions for the line of seepage' Casagrande.lo)

18..!O VARIABLEHYDRAULICCONDUCTMry

451

discontinuities' Figure 18.6 Flowlineslope Wells and rechargewells approach zero sizeas the source or sink is approached. ' later. in a practical senseand are discussed representsinks and sources

CONDUCTIVITY HYDRAULIC 18.10VARIABLE

It is commonfor flow within a porousmedium of one hydraulic conductivityto enter another region with a different hydraulic conductivity. When such a boundary is crossed.flowlines are refracted.The changein direction that occurs can be determined as a function of the two permeabilitiesinvolved in the manner of Todd and this. Figure 18.7illustrates DeWiest.2'a of permeabilitiesK, and K, which are separatedby the consider two soils boundary lR shown in Fig. 18.7. The directions of the flowlines before and after crossingthe boundary are definedby angles0, and 0r'

Figure 18.7 Flowline refraction.

452

CHAPTER18

MECHANICS OF FLOW

"

For continuity to be preserved,the velocity componentsin media K, and Kt, areaat the which are normal to the boundary,mustbe equal,sincethe cross-sectional boundaryis AB for a unit depth.UsingDarcy's law and noting the equipotentialdrops h" and hr, we flnd

o#cos 92 K,L*cos o' : u, From the geometry of the flgure it is apparentthat AC : AB sin 0t BD : AB sin 0z

(18.73)

The headlossbetween andB is shownon the figure to be equalto both Lh" andA,hu, A and sincethere can be only a singlevalue, | Lh": Ah6 in Introducingtheseexpressions Eq. 18.73produces
K t _ K , tan 0, tan 02

(18.74)

For refractedflow in a saturatedporous medium, the ratio of the tangentsof anglesformed by the intersectionof flowlineswith normals to the boundary is given by the ratio of hydraulic conductivities.As a result of refraction, the flow net on the DB if K2 sideof the boundarywill no longerbe squares the equipotentialline spacing is maintained.To adjust the net on the K2 side,the relation
Lhu Kl

Lt%- It

( 18 .7s )

can be usedwhere Lhb + Lh". The Equipotentiallines are also refractedin crossingpermeability boundaries. relation for this is
K, K2 _ tan at tan a,

(18.76)

where a is the anglebetweenthe equipotentialline and a normal to the boundary of permeability.a

18.11ANISOTROPY
on In many cases hydraulic conductivity is dependent the direction of flow within a deposits often givenlayerof soil. This condition is saidto be anisotropic.Sedimentary fit this aspect,with flow occurring more readily along the plane of depositionthan acrossit. Where the permeability within a plane is uniform but very small acrossit as comparedto that along the plane, a flow net can still be usedafter proper adjustNonhomogeneous ments are made. A discussionof this is given elsewhere.a's'11'12

THEORY 453 18.12 DUPUIT'S 'sometimesbe analyzed by Yslng aquifers require special considerationbut may parameters'A detailid study is outsidethe scopeof this or representative average
book.3-5'r2

THEORY 18.12DUPUIT'S
free surfacecan be analyzed Groundwaterflow problemsin which one boundaryis a flow. This theory is foundedon two on the basis of Dupuit's theory of unconfined is only slightly madeby ptptl in tgO:.t' First, if the line of seepage assumptions equlp:tenand, correspondingly, horizontal r""y b" considered inclined, streamlines tl" slopesof the line of seepage -utfd tial lines will be essentiallyvertical. Second, at" known to be satisfactorily hydraulic gradient *r "fit. Whel fleld conditioos to Dupuit's theory by represented theseusru-ption*, the resultsobtainedaccording "o'-pur"veryfavorablywiththosearrivedatbymorerigoroustechniques' into a mathemdtFigure 18.8 is u,"fol in translatingthe foregoingassumptions 'ateadx dy th" figut" which has a base Consideran elementgiien in ical statement. the x direction and considand a vertical height h,Writing the cJntinuity equaiionin flow to be the case, ering steadY (18.77) infloqo : velocitSo X area,o The velocity at x : 0 is given by Darcy's law as
,ro: _K*

(18.78)

acrossthe elementat x : 0 is Thus the discharge - K - a h. d Y Qo: *h expansionas The outflow at x = dx is obtainedby a Taylor's series -K*.h - " + o- . . dv _ _ *(-**n ,_ Qor: dr 6x\ Ax,. aY\+ ' ' _ /

(18.7e)

(18.80)

'

Figure 1,8.8 Definition sketch for development of DuPuit's equation'

454

CHAPTER18

MECHANICS OF FLOW

Subtractingthe outflow from the inflow if K is considered constant,we obtain

I* - o,: K dxa, !(n*) "

dx\

6x/

tt*.

or

r,_ o.: rydr,A+fg) 2 Ax\6x/

'x

(r8.

wheredx anddy are considered in fixed lengths.A similar consideration the y tion yields

r,-o,:ryy*W)
i

, (18

; )

Assuminsthat thereis no movementin the vertical direction.theseare the components the inflow and outflow.Furthermore,still dealingwith steadyflow, of I changein storagemust be zero. As a result,

Kd:dy Kd:dy +fg) / * 2 *(#)y / : o 2 lx\Ex ay\6

and since (K dx dy)12 is constant,this reducesto A2h2 . Azhz - r -:0 0y' 6x' : Y2hz0

o8

or

(18.

Consequently,. Laplace's equation for accordingto Dupuit' s assumptions, functionh2 must be satisfied.la In the particular casewhererecharge occurring as a resultofinfiltrated is reachingthe water table, a simple adjustmentmay be made to Eq. 18.85. If recharge intensity(dimensionallyLT-t) is specified R, thenthe total recharge as to elementof Fig. 18.8 is R dx dy and the continuity equationfor steadyflow beco

*4!*(#.#).Rdxdy:s
or more simply,

(18

v2h2+?o:o
Now, applyingDupuit's theory to the flow problem illustratedin Fig. 18.9, assuming one-dimensional flow in the x direction only, we obtain the discharge unit width of the aquifer given by Darcy's law:

O: -Kh#
In this instanceh is the height of the line of seepage any position x along at imperviousboundary.For the one-dimensional exampleconsidered here, Eq. 18

18.12 DUPUIT'STHEORY

455

fto=50ft

free surface

(b)

Figure 18.9 Steady flow in a porous medium between two witer bodies: (a) free surface with infiltration and (b) free surfacewithout infi ltration. becomes d2h2 --;-;: clx'

(18.e0)

Upon integration,
h2:ax*b

( 18.e 1)

wherea andb are constants. Then for boundarv conditions at x : O,h : hs,

b=ht
Differentiationof Eq. 18.91yields ^. dh
Ztl';-

(18.e2)

ax

= A

( 18.e 3)

-QIK. Making this substitution,we obtain Also from Darcy's equation,h dhldx :

o:-T-

-2Q

(18.e4)

and inserting the valuesof the constantsin Eq' 18.91,we obtain

h2=-2f*+nt

(18,e5)

456

18 .CHAPTER

MECHANICSOF FLOW

(often calledDupuit' s parabola)' It This is the equationof a free surface. is a parabola andonotingthat at * : L, If the existenceof a surfaceof seepage.afBi. ignor.d, h : hr, we f,nd that Eq. 18'95becomes 2QL hL: _ x n ra

(18.e6) (18.e7)

or

o = L o l' - n ? ) 2L "

which is known as the Dupuit equation' EXAMPLE 18.3 RefertoFig.ls.ga.GiventhedimensionsshownandarechargeintensityRof : 1000 ft using Dupuit's equation'Assumerthat 0.01 ftlday, find the Oi."ft"tg" ut x K: 8. Solution. Note that

Q=n
dx or
Q=Rxi.C

Atx:0,
. Q: Qo

therefore,
Q: Rx -l Qo

Also,
Q dh -Kh=

- oh
IntegratingYields

nh

i:

Rx ' r Qo

- Kh,l,,o*,il' o^*1, : *
2lo" 2lo -"lo

and inserting the limits, we obtain

o L

-K(h'-Lh7) :ry
r

* QoL
RL 2

1)^:----:----

K(h?- h'r) z L

PROBLEMS

457

Then since Q : Rx + Qo,

K(h'^ h?\ . ,) 2L R 0 . 0 1 7 . 5 : 0.075gpd/ft2 _ o : 0.075(1000 soo)+ 8(50, 40r) 2000 0.075x 500 , 8 x 9 0 0 2000 3 7 . 5+ 3 . 6 4Ll gpdlftz T I

R(.

r summary
Understanding movementof groundwaterrequiresa knowledgeof the time and the space dependencyof the flow, nature of the porous medium and fluid, and the boundariesof the flow system. particular, groundwaterdevelopment In and management dependon understanding storage propertiesofthe associated the soils and rocks ' and the ability of thesesubsurface materialsto transmit water. Fundamentalto the mechanics groundwater of flow is Darcy's law (Eq. 18.3).Usingthis equationalong with a knowledgeof the hydraulic conductivity K, estimates flow can be had. The of hydrodynamicequationspresentedin this chapter serve as models for a variety of groundwatgrflow calculations.Applications are given in Chapters19 and 20.

PROBLEMS'
18.1. What is the Reynoldsnumber for flow in a soil when the water temperatureis 55oF, the velocity is 0.5 ftlday, and the mean grain diameteris 0.08 in.? 18.2. The water temperaturein an aquifer is 60'F, the velocity is 1.0 ftlday. The average particle diameterof the soil is 0.06 in. Find the Reynoldsnumberand indicatewhether Darcy's law applies. 18.3. ReworkProblem 18.2assuming temperature 65"F and the velocity is 0.8 ftlday. the is 18.4. A laboratory test of a soil gives a standardcoeff,cientofpermeability of 3.8 x 102 gpdlft2.If the prevailingfield temperature 60"F,find the field coefficientof permeis ability. " 18.5. ReworkProblem 18.4 assuming is 3.8 x r02 gpd/ft2and the temperature 65'F. K" is 18.6. Given the well and flow net datain the following flgure, find the discharge usinga flow net solution.The well is fully penetrating; : 2.87 X 10-4 ftlsec, a: 180 ft, K b = 43 ft, and c : 50 ft. 18.7. ReworkProblem 18.6assuming : 8.2 x 10-5 mlsec,a: 85 m, b : 2l m, and K c:26m.

458

CHAPTER18

MECHANICS OF FLOW
Stagnation flowline

Axis of

5 4 3 2 I

symmetry -1 -2 -3 /

K: 1 8 . 8 . R e w o r k P r o b l e 1 8 . 6 a s s u m i n g8 ' 4 X 1 0 - 5m l s e c , a : 1 0 0 m ,b : 2 2 m , a n d m c:35m. 18.9. A stratum of clean sand and gravel 15 ft deep has a coefficient of permeability of to K : 3.25 X 10-3 ftlsec, and is suppliedwith waterfrom achannelthat penetrates the bottom of the stratum.If the water surfacein an infiltration gallery is 2 ft above the bottom of the stratum.and its distanceto the channelis 50 ft, what is the flow into a foot of gallery?UseEq. 18.97.

REFERENCES
1. 2. 3. 4. 5. 6. 7. Henri Darcy, Lesfontaines publiquesde la ville de Dijon. Paris: V. Dalmont, 1856. D. K. Todd, GroundwaterHydrology. New Yorkl V/iley, 1960. Evaluation.New York: McGraw-Hill, 1970. William C. Walton, GroundwaterResource New York: Wiley, 1965. R. J. M. DeWiest,Geohydrology. New York: McGraw-Hill, 1962. M. E. Harr, Groundwaterand Seepage. SalamonEskinazi,Principles of Fluid Mechanics.Boston:Allyn and Bacon,7962"Flow of Groundwater,"in Engineering Hydraulics (Hunter Rouse,ed.) New C.E. Jacob, York Wiley, 1950.

Chapter '1 I

Wellsand Collection Devices

r Prologue Thepurpose this chapter to: of is

. Presentmethodsfor calculatingconfined and unconfinedsteadyradial flow toward a well. . Describeproceduresfor dealingwith unsteadygroundwaterflow conditions. . Describea method for estimatingflow to an infiltration gallery. Groundwater collectedprimarily by wells, althoughinfiltration galleriesare someis times usedwherethe circumstances appropriate.tOutflows from natural springs are are also amenable collection,but once thesewatersexit the ground, they become to surfaceflows and are handled as such. Wells are holes or shafts, usually vertical, excavatedin the earth for the purpose of bringing groundwater to the surface. Infiltration galleriesarehorizontalconduitsfor interceptingand collectinggroundwater by gravity flow. Problemsof groundwaterflow to wells and infiltration galleries can be solved applyingDarcy's law. , by

19.1 FLOW TO WELLS A well system be considered composed three elements-the well structure, can as of pump, and discharge piping.2The well itself containsan open sectionthrough which waterentersand a casingto transportthe flow to the ground surface. The open section is usually a perforatedcasingor slottedmetal screenpermitting waterto enter and at the sametime preventingcollapseof the hole. Occasionally,gravel is placed at the bottom of the well casingaround the screen. When a well is pumped,wateris removedfrom the aquiferimmediatelyadjacent to the screen. Flow then becomes established locationssomedistancefrom the well at in order to replenishthis withdrawal. Because flow resistance offered by the soil. of a head loss results and the piezometric surfaceadjacent to the well is depressed. producinga coneof depression (Fig. 19. 1), which spreads until equilibriumis reached and steady-state conditions are established.

RADIALFLOW TOWARDA WELL 19.2 STEADYUNCONFINED

461

Impervious

Figure 19.1 Well in an unconfinedaquifer.

by of The hydraulic characteristics an aquifer (which are described the storage permeability) can be determinedby laboratory or field tests' coefficient and aquifer "o111*onlyusedfield methodsare thp applicationof tracers,the use The three 111ort of A tests.3 discussion aquiferper!o.r. uni of field permeameterr, aquiferperformance for of given here alongwith the development flow equations wells.2'a's mancetestsis testsmay be either equilibrium or nonequilibriumtests.In Aquifer performance mustbe stabilizedfor a flow equationto be an equilibrium testthe coneof depression tesithe derivationincludesa condition that steady-state derived.For a nonequilibrium tests Adolph Thiem publishedthe first performance conditionshavenot beenreached. in 1906.6 basedon equilibriumconditions

A RADIALFLOWTOWARD WELL UNCONFINED 19.2 STEADY

Q: hrxYKrfi
where 2rrxy : . Kr: dyfdx : Q:

(le.1)

the areathrough any cylindrical shell, in ft2 with the well as its axis the hydraulic conductivity (ftlsec) the water table gradient at any distancex the well discharge(ft'lsec)
f,2 r..
n

Integratingover the limits specified,we find that

fo, I -gY : 2rrK1ly ay l r
Jhl

(re.2)

Jt

462

DEVICES CHAPTER19 WELLS AND COLLECTION

Q n f1 _ 2
and

2rrKr(h2z h?) 2 : nKr(hi hl)

^

( 1e.3)
g

ln(rr/r1)

(te.4)

converting K, to the field units of gpdlftz, Q to gpm, and ln to 1og,we can rewrite Eq. 19.4as
,ar: " ------;-;----;;-

1055Qlog(r2lrr) n;- ni

(1e.s)

one half of the original aquiferthickness If the drawdownin the well doesnot exceed by of estimates Q or Krcanbe obtained usingEq.19.4 or tr9.5,even ho,reasonable at if the heighth, is measured the well peripherywhere1L: r*, the radius of the well boring. EXAMPLE 19.1 an An 18-in. well fully penetrates unconfinedaquifer of 100-ft depth.Two observa: tion wells located100 and 235 ft from the pumpedwell areknown to havedrawdowns of 22.2 and 21.ft, respectively.If the flow is steady and K1 : 1320 gpdlft2, what would be the discharge? Solution. Equation 19.4 is applicable,and for the given units this is 1055Iog(r2/r1) :0.37rlt = rwQ35l1'00) log(r2/r') hz: 100-21=79ft
ht=

Q :

K(h?- h?)

100- 22.2: 77.8ft t32O(792 77.82) Q : 1055x 0.37107 gPm lr 634.44

A RADIALFLOWTOWARD WELL CONFINED 19.3 STEADY

The basic equilibrium equationfor a confined aquifer can be obtainedin a similar apply. Mathematimanner, using the notation of Fig. I9.2. The sameassumptions from cally, the flow in ft3lsecis found

o : 2nxmXr!x "d
Integrating,we obtain h"-h, O:2rrK'm:# " ln\r2/h)

(1e.6)

(re.7)

.I9.4 WELL IN A UNIFORM FLOWFIELD

--12+l

463

aquifer. Figure 19.2 Radialflow to a well in a confined The coefficient of permeability may be determinedby rearrangingEq. 19.7 to the form - - 528Q log(rz/r'\ nr _ *&; h) where Q : gpm K f : the permeability (gpd/ft') r , h : ft

(1e.8)

EXAMPLE 19.2 Determinethe permeabilityof an artesianaquiferbeingpumpedby a fully penetrating well. The aquifer is 90 ft thick and composedof medium sand. The steady-state pumpingrate is 850 gpm. The drawdownof an observationwell 50 ft awayis 10 ft; in a secondobservationwell 500 ft awayit is 1 ft. Solution
Kf:

52SQroe?Jr')
m(h2 - h')

528 x 850 x log 10 90x(10-1) : 554gpdlfr2 rl FIELD 19.4 WELL IN A UNIFORM"FLOW

well in a uniform flow field wherethe original piezometricsurface For a steady-state preis not horizontal, a somewhatdifferent situation from that previously assumed vails. Consider the artesian aquifer shown in Fig. 19.3. The heretofore assumed circular area of influencebecomesdistortedin this case.A solution is possibleby or, applyingpotentialtheory,by usinggraphicalmeans, if the slopeof the piezometric surfaceis very slight, Eq. 19.7 may be employedwithout seriouserror.

464

DEVICES CHAPTER19 WELLS AND COLLECTION

Original piezometric surface

Figure 19.3 Wellin a uniformflow fieldandflow net definition' . Figure 19.3providesa graphicalsolutionto a uniform flow field problem.First, an ortholgonalflow net consistingof flowlines and equipotentiallines must be constructed.This shouldbe done so that the completedflow net will be composedof a in number of elementsthat approachlittle squares shape.Once the net is complete, Uy if can be analyzed consideringthe net geomtry and using Darcy's law in the manner of Todd.3 EXAMPLE 19.3 to Find the discharge the well of Fig. 19.3by using an applicableflow net. Consider as the aquiferto be 35 ft thick, 4 : i.65 x 1d-4 fps, and other dimensions shown' Solution. UsingEq. 18'72,we find that

n : * *n'
whereh:35*25:60ft m : 2 X 5 : 1 0 , n:I4 3.65X10-4x60x10

t4 : 0.0156cfsper unit thlckness the aquifer of

The total dischargeQ is thus r r Q : 0.0156 X 35 : 0.55 cfs ot 245 gPm

19.5 WELLFIELDS 465

19.5 WELL FIELDS When more than one unit in a well fleld is pumped,there is a compositeeffect on the is This consequence illustratedby Fig. 19.4in which the conesof free water surface. are depression seento overlap.The drawdownat a given location is equalto the sum of the individual drawdowns. If, within a particular well field, pumpingratesof the pumpedwells are known, the compositedrawdownat a point can be determined.In like manner,if the drawdown at one point is known, the well flows can be calculated' 1,2, ' as 'n If the drawdownat a given point is designated m, andsubscripts particular well(e.g ., mt fefersto the drawdown are usedto relate this drawdownto a for I7,), for the total drawdownmr at somelocation'
n

lTlr:

sr
^Z

lTli

(1e.e)

The numberof wells, their rate of pumping,and well-fieldgeometryand charac' teristicsdeterminethe total drawdownat a specifiedlocation. Eq. 19.4,we obtain Again considering

h3- h'

: #*t"1

(1e.10)

lt can be seenthat the drawdownfor a well pumped atrate Q canbe computedif ho, ro, and r are known. It follows then from Eq. 19.9 that for n pumped wells in an unconfinedaquifer

h 3h ' : 2 * . " ? -

( 1 e .1 ) 1

Figure 19.4 Combinedeffect of pumping severalwells at equal rates'

466

CHAPTER19 WELLS AND COLLECTION DEVICES

where ho : the original height of the water table h : the combined effect height of the water table after pumping n wells r Q, : the flow rate of the ith well foi : distance of the lth well to a location at which the drawdown is considered negligible r, : the distancefrom well i to the point at which the drawdownis being investigated Todd indicatesthat valuesof rousedin practiceoften rangefrom 500 to 1000ft.3 The impact of this assumption softenedbecause inEq. 19.10is not very sensitiveto is Q are 16.Equation19.11shouldbe usedonly wheredrawdowns relativd small. for For flow in a confinedaquifer the expression combineddrawdownbecomes
n

h o - h : 2 =Q:, ,n'o'
z1rKm

f1

(r9.r2)

Equationsfor well flow covering a variety of particular well-field patternsare reported in the literature.3'7 Those given here are applicablefor steadyflow in a homogeneous isotropic medium.

19.6THEMETHOD IMAGES OF
Some groundwaterflow problems subjectedto boundary conditions negating the fitting these into infinite systems direct useof radial flow equations be transformed can equations applyingthe methodof images.2'8'e by When a streamis locatednear a pumped well and the streamand aquifer are interconnected, drawdowncurve of a pumpedwell may be affectedas shownin the Fig. 19.5. Another boundary condition often affecting the drawdownof a well is an of imperviousformation that limits the extentof the aquifer.The cone of depression After that, the shape a pumpedwell is not affecteduntil the boundary is intersected. of the drawdown curve will be changedby the boundary. Boundary effects can frequently be evaluatedby meansof "image wells." The boundary condition is replacedby either a rechargingor a discharging well that is pumped or rechargedat a rate equivalentto that of the pumpedwell. That is, in an infinite aquifer,drawdowns of the real and imagewells would be identical.The imagewell is locatedat a distance from the boundary equal to that of the real well but on the oppositeside(Fig. 19.5). are boundaries supplanted Streamsare replaced recharge by wells while impermeable for by pumpedimagewells.Computations the caseof a well and imperviousboundary directly follow the procedures outlined under the sectionon well fields.For the well to discharge. and streamsystem, recharge the imagewell is considered havea negative The headsare then addedaccordingto this sign convention. The procedurefor combiningdrawdowncurvesof real and imagewells to obtain an actual drawdown curve is illustratetl graphically for the example shown in Fis. 19.5.More detailedinformationon othercases can be found elsewhere.e'10

19.7 UNSTEADY FLOW

467

Cone of depressionof real well without sffeam

connectedto Figure L9.5 Drawdown in a pumping well whose aquifer is a stream'

FLOW 19.7 UNSTEADY

comesdirectly from , when a new well is first pumped alargeportion of the discharge u.nder thesecircumas volume released the cone of depression the storage _d91elons, permeabilityandthereforethe yield of overestimate the stances equilibrium equations conditionsare not encountered-as is usually the situathe well. When steady-state can,be approaches tion in practice-a nonequilibrium gOgltlgl must be used' Two suchasthat ratherrigorousmethodof C. V. Theisor a simplifiedprocedure taken,t^he by proposed Dcob.l1'12 that takesinto consideration In tgjS treis publisheda nonequilibriumapproach of characteristics the aquifer.ll His methodusesan analogybetween time and storage by the Biot-Fouiier law and groundwaterflow to a well' The heattransferdescribed boundary conditions' method providesa solution to Eq. 18.41 for given initial and constantthickness' Application of the methodis appiopriatefor confinedaquifersof flow, vertical componentsof flow must be For'use under conditions of unconhneO

468

CHAPTER19 WELLS AND COLLECTION DEVICES

negligible,and changes aquifer storage in through water expansion and aquifer compressionmust also be negligiblerelative to the gravity drainageof poresas the water table drops as a result of pumping.l3 Theis statesthat the drawdown(s) in an observationwell locatedat a distance r from the pumpedwell is given by

O ', s : -4rrT f * " - o" J, ;

(1e.13)

in which Q : (constant) pumping rate (L3T-r units), Z : aquifer transmissivity (LtT 'units), and a is a dimensionless variabledefinedby

, ,s" u : r--4tT

(re.r4)

where r is the radial distancefrom the pumpingwell to an observation well, S. is the aquiferstorativity(dimensionless), r is time. The integralin Eq. 19.13is usually and known asthewellfunctionof u andiscommonlywritten asW(u).It maybe evaluated from the infinite series

w(u): _'0.577216ln u * u - =+ - + -]-2x21 3x3l

Usingthis notation,Eq. 19.13can be written as

(1e.1s)

Qw(u) ( " : 4irT

(1e.i6)

The basic assumptiops employedin the Theis equation are essentiallythe sameas thosein Eq.I9.7 exceptfor the nonsteady-state condition. Some valuesof the well function are given in Table 19.1. In American practice,Eqs. 19.13 and 19.I4 commonly appearin the following form,

t:

rI4.6Qf* "-' , , J,;o'

I . 8 71 2 " S

(re.r7)
(1e.18)

,:

where 7 is given in units of gpd/ft, Q hasunits of gpm, and / is the time in dayssince the start of pumping. Equations 19.13and 19.14canbe solved comparing log-logplot of a versus a by I4z(z) known as i type curve, with a log-1ogplot of the observeddatar2ft versuss. In plotting type curves,W(u) ands are ordinates, and rt ft areabscissas. two curves The z are superimposed movedaboutuntil segments and coincide.In this operationthe axes must remain parallel. A coincidentpoint is then selected the matchedcurvesand on both plots marked.The type curve then yieldsvaluesof u and W(u) for the desired point. Corresponding from a plot ofthe observed valuesofs and r'ft aredetermined

FLOW 1g7 UNSTEADY OF VALUES U OF TABLE19.1 VALUES W(U}FORVARIOUS 'L0 5.0 4.0 3,0 2.0 XI x 10-r x 10-2 x 10-3 x 10-4 x 10-5 x 10-6 x 10-7 x 10-8 x 10-e x 10-10 x 10-11 x 10-t2 x 10-13 x 10-14 x 10-15 0.219 t.82 4.04 6,33 8.63 10.94 t3.24 15.54 17.84 20.15
aa A<

469'

6.0 0.00036 0.45

z,Jv

8.0 -' 0,00012 0.37 2.1,5 4.39 6,69 8.99 rt.29 13.60 15.90 18,20 20.50 22.8t 0,000038 0.31 2.03 4,26 6.55 8.86 I 1.16 t3.46 t5.'16 18.07 20.37 22.67 24.9'1 27.28 29.58 31.88

9.0 0.000012 0.26 1.92 4.14 6.44 8.74 1l.04 13.34 15.65 17.95 20.25 22,5s 24.86 27.16 29.46 3t.76

24.75 27.0s 29.36 31.66 33,96

0.049 1.22 3.35 5.64 7.94 t0.24 12.55 14.85 t7,15 19,45 21.76 24.06 26.36 28,66 30.97

0 . 0 1 3 0.0038 0.70 0.91 2.68 2.96 4.95 5.23 7.25 1.53 9.55 9.84 12.14 11 . 8 5 14.15 14.44 16.46 16.74 19.05 t8.76 2 r . 3 5 2r.06 23.65 2 3 . 3 6 25.96 25.67 27.97 28.26 30.56 30.27 32.86 32.58

0.0011 0.56 2.47 4.' 13 7,02 9.33 11 . 6 3 1.3.93 16.23 18.54 20.84 23.14 25.44 27.75 30.05 32.35

4.54 6.84 9.r4 r1.45 13.75 16.05 18.35 20.66 22,96 25.26 27.56 29.8'7 32.1,7

25.rr
2'1.41 29.71 32,02

..Methodsfor Determinlng Permeability of water Bearing M.aterialswith Special Referenceto Dischargt.fig Source: AfterL,K. Wenzel, Washington'DC' 1942' Well Methods," U. S. GeologiJSt.""V, W"t*-Supp-ly Paper 887'

values for data. Inserting these values in Eqs. 19.13 and 1,9.1'4and,rearranging, found' transmissibilityI and storagecoefficientS" can be "can be shortenedand simplified' when r is small and t often this procedure Thus terms in the seriesof large, Jacobfound that valuesof u arc generallysmall'12 for and the expression Zbecomes fql te.tS bpyondthe secondonebecomenegligible 264Q(loet2 - log tt) (1e.1e) ' h"- h which can be further reducedto

264Q T _ Lh
where Ah Q ho,h T : : : : drawdownper log cycle of time l(ho (gPm) well discharge as definedinFig' I9.2 (gPd/ft) the transmissibilitY - n)lloe t' - log t')l

(re.20)

paper' The Field data on drawdown(ho h) versust arc dtaftedon semilogarthmic a straightline drawdownis plotted on an arithmetic soale,Fig. 19.6. Thisplotforms and *to*" slopepermits computingformation constantsusing Eq' 1'9'20 0.3Tto ^ \ - : -r '

(1,9.2t)

Equation 19.21is obtainedthrough to with rothe time corresponding zerodrawdown. of manipulation Eq. 19'13.

470

DEVICES CHAPTER19 WELLS AND COLLECTION

, = t4

r\too'

= 4e,8oo sd/ft

tl

I I
I

Time since PumPing began (min)

Figure 1.9.6 Pumpingtest data, Jacobmethod. EXAMPLE 19.4

for Usingthe following data, find the formation constants an aquiferusinga graphical solution to the Theis equation.Dischargeequals540 gpm.
from Distance pumped well, r (ft) 50 100 150 200 300 400 500 600 700 800
Average drawdown, s (ft)

12 t /

1,250 5,000 il,250 20,000 45,000 80,000 125,000 180,000 245,000 320.000

3.04 2.16 1,.63 r.28 0.80 0.51 0.33 0.22 0.15 0 .l 0

FLOW 19.7 UNSTEADY

o.o ? 0.s E o.+

B

1.0 0.9 0.8 0.7

't I | , I W(u)vs. u

E 0.3

5 6 7 891x105

I rrzldav)
equation' to solution Theis's ' Figure L9.7 Graphical Solution.Plotsversusr2ftand}V(a)versusaasshowninFig.lg.T.Determine Eqs' 19'7 and 19'8:the match point as noted and computeS" and Z using T : lr4t6Q w(u) s X 114.6 540-

: --18c u" --- - -

^ 1.9 :

91,860gpdift

uT
1.87r'1t

_ 0.09x 91,860 0,22 tr : 1.87X 20,000

EXAMPLE 19.5

7"andstorage of in usingthedatagiv.en Fig. 19.6,findthe_coefficienttransmissibility : 300ft' Q: 1000gprn andr S, coeffrcient for an aquiflr, given Solution.FindthevalueofAhfromthegraph,5.3ft.ThenbyEq.|9.20 2640 264 x31000 --T -

T: -ff:

: 49,800gpd/ft we Using Eq. 19.21., find that

s":

O.3Tto r'

472

CHAPTER19 WELLS AND COLLECTIONDEVICES

Note from Fig. 19.6 that ts : 2.6 min. Converting to days, we find q6il)ti\--l becomes
, 0 -

and

s":

' x 1.81 lb-'days 0 . 3 x 4 9 , 8 0 0 X 1 .X 1 0 - 3 8

(3oo)'

: 0.0003 I T
EXAMPLE 19.6 point 200 ft awayfrom a pumping well. Given Find the drawdownat an observation that T : 3.0 x 10a gpd/ft, the pumping time is 12 days,S" : 3 X 10-4, and

gpm, 0 = 3oo

Solution. From Eq. L9.L8,u can be computed,

x 104 u : u.87 x (200)'x 3 x L0-41113.0 x l2l : 6.23x I}-s

we Referringto Table 19.1 and interpolating, estimateW(u) to be 9.1. Then, using Eq. 19.17 the drawdownis found to be , ft s : [114.6x 9.1 x 300]/[3.0x 104]: 10.41 rr EXAMPLE 19.7 Giventhat I : 0.0028m2ls, A well is beingpumpedat a constantrate of 0.0038 m3/s. : 90 meters, and the storagecoefficient : 0.00098, find the drawdown in the r observationwell for a time period of (a) 1,000 sec.and (b) 20 hours. Solution a. Using Eq. 19.14,u canbe computedas follows, x u :190 x 90 x 0.000981/[4 1000 x 0.0028] u : 0.71 Then from Table 19.1,W(u) is found to be 0.36. Applying Eq. 19'16,the drawdowncan be determined,

x n x r : [0.0038 0.367114 x 0.0028] s : 0.039m

b. Follow the procedureusedin (a) ' u : 1 9 0 x 9 0 x 0 . 0 0 0 9 8 1 / [x 7 2 , 0 0 0 x 0 . 0 0 2 8 ] 4 : 0.0098 a Then from Table 19.1, W(u) is found to be 4.06 Applying Eq. 19'16,the drawdowncan be determined,

x x s = [0.0038 4.06]114n x 0.00281 :0.44m rr s

GALLERY 19.10 FLOW TO AN INFILTRATION

473

AQUIFERS 1e.8LEAKY

havedealtwith free aquifersor thosecopfinedbe#een imperThe foregoinganalyses t"ulity, many casesexist wherein the confining strata are not comvious str-ata.1n PletelYimPerviousand water is at aquifer. The flow regime is altere< about 1930,leakYaquifershaveber De Glee, Jacob,Hantush, DeWies ers.t'to-'uA thorough treatmentc interestedreadersshouldconsulttheindicatedreferences.

a_)

WELLS PENETRATING 19.9 PARTIALLY

of the well' The ln many actual situations there is only partial penetration previouslyfor full developed question then arisesasto the applicability of procedures Penetration. ln 1957Hantush of studies this problemhavebeenconducted'7'2't'28 Numerous aquiferbecomes reportedthat steadyflow to a well just penetratingan infinite leaky from the well of about1.5timesthe aquiferthickness'28 very nearlyradial at a distance increasingly the As depth of penetrationincreases, approachto,radial flow becomes apparent.Therefore,computationofdrawdownsforpartiallypenetrating'wellsare provided that the made using equationsfoi total penetration with relative safety, aquifer thickness'At distancefr"o-ifr" pumped we11ls greaterthan 1.5 times the other relations points closerto the weli, it is frequJntly possibleto use a flow net or for develoPed this region.

GALLERY 19.10FLOWTO AN INFILTRATION

conduit constructed An infiltration gallery may be defined as a partially pervious part of this flow will be acrossthe path of t6" to"uigroundwater flow iuch that all or often built in a valley areaparallel to a streamso that Thesegalleries-are intercepted. location under gravity-flow they can convey the collected flow to some designated .ondition*.Figurelg.Sshowsatypicalcrosssectionthroughagallerywithone perviousface. ' pervious wall corrrputation of dischargeto an inflltration gallery with one assumptions Several inihe manneroutlinedby Dupuit-2e (Fig. f g.Sjis accomplished of the angleof tangent mustbe madelo effectthe solution.They are that the sineand that table are interchangeable; the velocity vectorsare evefyinclination ofthe water is incompressibleand where horizontal and uniformly distribuied; that the soil effects are negligible' isotropic; and that the gallery is of sufficient length that end do limit the utility of While permitting u ,oluion oi the problem, theseassumptions the results. Basedontheseassumptions,andfollowingtheproceduregiveninSecper unit width' using the the tion 18.12,Eq. 18.97canbe usedto calculate discharge

474

CHAPTER WELLS AND COLLECTIONDEVICES 19

Ground surface

Intersection of assumed and actual water tables

X+

Figure 19.8 Cross-section throughan infiltrationgallery. nomenclature Fig. 19.8,Eq. 18.97becomes of

s :;(hi

K . - "-

he)

This equationindicatesthat the computedwatertableis parabolic.This is often called Dupuit's parabola.Figure 19.8 showsthat the computedwater table differs from the actual water table in an increasingmanner as the gallery face is approached. is It thereforeapparentthat the computedparaboladoesnot accuratelydescribethe real water table. The differences,however,are small except near the point of outflow, providing the initial assumptions satisfled. The calculateddischarge approximates are the true discharge more closelyas the ratio of Z/h, increases.
EXAMPLE 19.8

A stratum of clean sand and gravel20 ft deep has a coefficient of permeability of K : 3.25 X 10-3ftlsec,andis supplied that penetrates to with waterfrom a channel the bottom of the stratum.If the water surfacein an infiltration gallery is 3 ft above the bottom of the stratum,and its distanceto the channelis 50 ft, what is the flow into a foot of gallery?UseEq. 18.97. Solution q : 0.5(3.25 10-3)(20 20 - 3 x 3)/50 x x : 0.072 cfs, the flow into one foot of gallery I r \

19.11SALTWATER INTRUSION
The contaminationof fresh groundwater the intrusion of salt water often presents by a seriousquality problem. Islands and coastalregions are particularly vulnerable. Aquifers locatedinland sometimes contaiq highly salinewaters as well. Freshwater

BASINDEVELOPMENT 475 15.12 GROUNDWATER

BASINDEVELOPMENT 19.12GROUNDWATER
permitting the maxiefficiently while simultaneously resources To use groundwater betweenwilhmum dJvelopmentof the resource,equilibrium must be established Economic, legal, political, social, and water quality drawals and replenishments. require full consideration. aspects only whenlong-termwithdrawls will of Lasting supplies groundwater be assured period. The potentialof a groundduringthe corresponding arebalanceJby recharge by water basin can be assessed employingthe water budgetequation,

)r-)o:As
where the inflow ) l includesall forms of recharge,the total outflow ) O includes duringthe accounting in the everykind ofdischarge,and AS represents change storage forms of recharge and dischargeare those listed in perioO. The most significant Table t9.2. A groundwaterhydrologistmust be able to estimatethe quantity of water that time basinin a specified canbe economicallyand safeiyproducedfrom a groundwater of the competentto evaluate consequences imposing period. He or sheshouldalsobe variousrates of withdrawal on an undergroundsupply. Developmentof groundwaterbasins should be based on careful study, since groundwaterresourcesare finite and exhaustible.If the various types of recharge balance the withdrawals from a basin over a period of time, no difficulty will be to drafts, however,can depleteundergroundwater supplies a Excessive encountered. The mining of waterwill ultimately is not feasible. point whereeconomicdevelopment depletethe entire supply. AND OF TONT"TSRECHARGE DISCHARGE 19,2 SdUE TABLE
Recharge from streams, Seepage Ponds,lakes inflows Subsurface Infiltrated precipitation Water rechargedartificallY Discharge to Seepage lakes, streams,spnngs outflows Subsurface Evapotranspiration Pumping or other artificial meansof collection

i-_

476

DEVICES CHAPTER19 WELLS AND COLLECTION

r Summary
primarily through the constructionof is The collection of groundwater accomplished influence the numerical estimation"of their performance. wells, and many Tactors to Some situation, ur" amenable solutionthrough the utilization of relatively simple applicationof the hyOthers dependupon sophisticated mathematicalexpressions. of undervariousConditions nonuniformity of aquifermaterials drodynamicequations and a variety oiboundary conditions.The readeris cautionednot to be misledby the and simplicity oi ,o-" of thqsolutionspresented to observethat many of theserl\ situations.(. are not applicableto all groundwater-flow to speciaiconditions and ,) of water through the ground is of a different magnr'tudd The rate of movement than that through natural or artificial channelsor conduits.Typical flow ratesrange of the from 5 ft/day ti afew feet per year. Theselow ratesof flow exacerbate impact groundwatersourcesand complicate cleanup since natural contaminani spills on flushing from the site may take many yearsto occur' in The methodsdescribed this chapterfor estimatingflows to collection devices principlesof fluid flow embodiedin Darcy's law. Applicafions arebasedmainly on the in arelimited to flowsin the laminar range,but undermostconditionsencountered the the describing mechanics Examplesof the ure of equations field, Darcy's law applies. of flow to wells andinfiltration gilleries were given in this chapter.Both steady-state as and unsteadvflow conditions were addressed well'

PROBLEMS
coefflcient of 19.1. A 12-in. well fully penetfatesa confined aquifer 100 ft thick. The permeabilityis 60d gpd/ft'. Two test wells located40 and 120 ft awayshow a difference in drawdown U-"i*een them of 9 ft. Find the rate of flow delivered by the well. of 1g.2. A l2-in. well fully penetratesa confined aqpifer 100 ft thick. The coefficient permeability is 600 gpd/ft2. Two test wells located 45 and 120 ft away show a the difference in drawdoin between them of 8 ft. Find the rate of flow delivered by well. well' The 19.3. Determine the permeability of an artesianaquifer for a fully penetrating pumping ft thick. The steady-state aquifer is composedof medium sand and is 100 in an observationwell 75 ft awayis 14 ft, and the rate is 1200 gpm. The drawdown well 500 ft awayis l.2ft. Find Kin gallonsper day drawdownin a secondobservation foot. per square ftj/day/ft' Lg.4. Considera confined aquifer with a coefficientof transmissibilityT of 680 Att:5min,thedrawdowns:5.6ft;at50min,s:23'Ift;andat100min's: well is 75 ft awayfrom the pumpingwell. Find the discharge 28.2 ftrThe observation of the well. 19.5. Giventhefollowingdata:0:59,000ft3lday'T:630ft3/day't,,3}days'r:1ft' the drawand s" : 6.4 x L6,4. Considerthis to be a nonequilibriumproblem.Find down s. Note that for a:8.0x10-e u:8.2 X 10-e

: W(a) 13'sa W(u)= 19.94

'

PROBLEMS 477

19.6. ' Determinethe permeabilityof an artesianaquiferbeingpumpedby a fully penetrating pumping of well, The aquifercomposed medium sandis 130ft thick. The steady-state is 1300 gpm. The drawdownin an observbtionwell 6{ft awayis 12 ft, and in a rate secondwell 500 ft awayis 1.2 ft. Find Ky in gpdlft2. 1g.7. Considera confined aquifer with a coefficientof transmissibilityT : 700 ft3lday-ft. : Atr : 5 minthe drawdown 5.1ft; at50min,s = 20.0ft; at 100min,s = 26'2ft. of The observationwell is 60 ft from the pumping well. Find the discharge the well. pumpedat atate of 300 gpm is confined and pumping 19.8. Assumethat an aquiferbeing testdata are given as follows.Find the cbefficientof transmissibilityTand the sttragq coefficientS. Assumer = 55 ft. | )
Time sincepumping started (min) Drawdowns (ft)

1,3 4.6

2.5 8.1

4.2 9.3

8.0 12.0

11.0 15.1

100.0 29.0

19.9. We are given the following dat4: Q = 60,000 ft3ldaY t:30days r:lft s"=6.4 x 1o-4

fcl(dar(fo r : 650
a : 8.0 x 10-e u : 8.2 X 10-e u : 8.6 x 10-e

Assumethis to be a nonequilibriumproblem. Find the drawdowns. Note for lV(u) : 18.06 W(u) = 19.94 W(u) : 17.99

19.10. An 18-in. well fully penetratesan unconfinedaquifer 100 ft deep.Ttvo observation wells located90 and 235 ft from the pumped well are known to havedrawdownsof 225 and20.6ft, respectivd. If the flow is steadyandKy: 1300gpd/ft2, what would be the discharge? lg.1l. A confinedaquifer 80 ft deepis being pumpedunderequilibrium conditionsat a rate ' the of 700 gpm. The well fully penetrates aquifer. Watpr levelsin observationwells Find the field 150 and 230 ft from the pumped well are 95 and 97 ft, respectively. coefflcient of permeability. lg.112. A well is pumpedat the rate of 500 gpm undernonequilibriumconditions.For the data listed, find the formation constantsS and Z. Use the Theis method.

r"/t
1,250. 5,000 tt,250 20,000 45,000 80,000 125,000 180,000 245,000 320,000

Averagedrawdown, h (ft)
3.2+

2.t8 .1.93 r.28 0.80 0.56 0.38 0.22 0.15 0.10

478

DEVICES ANDCOLLECTION 19 CHAPTER WELLS "We are given a well pumping at arate of 590 gpm. An observationwell is locatedat 19.13. r : 180 ft. Find S and Z using the Jacobmethod for the following test data.

Drawdown (ft)

Time (min)

Drawdown (ft)

Time (min)

0.43 0.94 l 08 1.20 r.34 |.46 1.56 1.63 1.68 1.71 l 85 1.93

26 78 99 t3r t'73 2t8 266 303 331 364 481 5'13

2.00 2.06 2.12 2.15 2.20 2.23 2.28 2.30 2.32 2.36 2.38

66r 732 843 926 IO34 r134 1272 1351 I4t9 r520 161 1

tg:t4.

the A24-1n.diameterwell penetrates full depthof an unconfinedaquifer.The original belowthe watertable and a bedrockaquifugewere located50 and 150ft, respectively, land surface.After pumping al arate of 1700 gpm continuously for 1920 days. and the original water levels in equilibrium drawdownconditions were established, 100 ft from the center of the pumpedwell were observationwells located 1000 and lowered 10 and 20 ft, respectively. a. Determinethe field permeability (gpdlftz) of the aquifer. b. For the samewell, zero drawdownoccurredoutsidea circle with a 10,000-ftradius drawfrom the centerof the pumpedwell. Insidethe circle, the average measured down in the water table was observedto be 10 ft. Determine the coefficient of storageof the aquifer. 19.15. A well fully penetratesthe 100-ft depth of a saturatedunconflned aquifer. The using drawdownat the well casingis 40 ft whenequilibriumconditionsare established of a constantdischarge 50 gpm. What is the drawdownwhenequilibriumis established of using a constantdischarge 66 gpm? by a long rainless period, the flow in Wahoo Creek decreases 8 cfs from 19.16. After an 8 Memphisdownstream mi to Ashland.The streampenetrates unconfinedaquifer, wherethe water table contoursnear the creek parallel the westbank and slopeto the streamby 0.00020, while on the east side the contours slope awayfrom the stream toward-theLincoln wellfield at 0.00095. Computethe transmissivityof the aquifer knowing Q : TIt, where1 is the slopeand I is the length. 19.17. The time-drawdown data for an observationwell located 300 ft from a pumped well (500 gpm) are given in the following table.Find the coefficientof storage artesian (ft3 of water/ft3of aquifer)and the transmissivity(gpd/ft) of the aquiferby the Theis method. Use 3 x 3 cycle log PaPer.

REFERENCES

479

Time (hr) 1.8 2.\

. A

Drawdown (ft)

Time

(h0
9.8 12.2 t4.7 16.3 18.4 21.0 24.4

Drawdown (ft)

3.0 3.7 4.9 7.5

0.27 0.30 0.3'l 0.42 0.50 0.61 0.84

r.09 t.25 1.40 1.50 l 60 t.70 1.80

level of the water table for an unconfined 19.18. Over a 100-mi2surfacearea, the average

of aquiferhas dropped I 0 ft because the removalof 128,000 area-ftof waterfrom the coefficientfor the aquifer.The specificyield is 0.2 and aquifer.Determinethe storage the porosityis 0.22. 19.19. Over a 100-mi2 surface area, the averagelevel of the piezometric surface for a confined aquifer in the Denver area has declined 400 ft as a result of long-te1m pumping.Determinethe amountof the water (acre-ft) pumpedfrom the aquifer' The porosity is 0.3 and the coefficient of storageis 0.0002. L9.20. Find the drawdownat an observationpoint 250 ft awayfrom a pumping well, given -: thatT = 3.1 x 104 gpdlft,the pumpingtime is 10 days,S" 3 x 10-4, andQ 280 gpm. 19.21. Find the permeabilityof an artesianaquiferbeingpumpedby a fully penetratingwell. pumpof The aquiier is 130 ft thick and is composed medium sand.The steady-state ft awayis 12 ft' and ing raie is 1300 gpm. The drawdownin an observationwell 65 well 500 ft awayit is 1.2 ft. Find Kyin gpdlft2. in a second 19.22, An 18 in. well fully penetratesan unconflnedaquifer 100 ft deep.Two observation wells located90 and 235 ft from the pumpedwell are known to havedrawdownsof 22.5 ft and20.6 ft respectively.If the flow is steady arld Ky: 1300 gpd/ft2, what would be the discharge? : 0.0028m2ls, 19.23. A well is being pumpedat a constantrate of 0.004 m3/s.Given thatT : 0.001,find the drawdownin the obser: 100meters,and the storage coefficient r vation well for a time period of (a) t hr, and (b) 24 hours' = 0.0028mzls, 19.24. A well is being pumpedat a constantrate of 0.003 m3/s.Given thatT = 0.001, and the time sincepumplng beganis 12 hours,find the storage"oiin"i"nt well for a radial distanceof (a) 150m, and (b) 500 m' the drawdownin an observation

REFERENCES
1 . J. W Clark, w. viessman, Jr., and M. J. Hammer,water supply and Pollution conffol, 2nd ed. New York: Thomas Y. Crowell' 1965. "Field TestsDeterminePotential Quantity, Quality of Ground Water 2. John F. Hoffman, Supply," Heating, Piping, and Air Conditioning(Aug' 1961)' 3 . D. K. Todd, GroundwaterHydrology, New York: Wiley, 1960'

Chapter20

Regional Modeling SYstems Groundwater

r Prologue
of The PurPose this chaPteris to: ' Describethe featureS large-scalegroundwatersystems. of to . Introduce the principlesof finite differenceapproaches modelingregional groundwatersystems. . Iilustrate the application of groundwatermodeling techniquesto the Upper Big Blue basin in Nebraska. flow of The analytical methodsdescribedso far havebeen applicablemainly to the regionalgroundwarcr rc individual wells. In this chapter,the conceptsof analyzing are are water systems introduced.Suchanalyses requisitesfor the wise development, management,andoperationofexpansivegroundwaterresources' mustbe dealt with jointly in most quality aspects Given that waier quantity an-d modelsmrrstoften prol"tt"t, t"gional groundwater decision-making waterresources of aspects groundThe to be designed includeboth of tfuse dimensions.l-2s fluid flow are complexand in water modelsare presented this chapter.Solutetransportmodels. this book, but a biief introduction to them is given at the end of the scope-of beyond and role of the chapter.It is important for the readerto understandthe importance thesewater quality-orientedmodels'

MODELS GROUNDWATER 20.1 REGIONAL

(mathematical) Groundwatersystemsmodels may be of the analog or the digital type most variety. The focus of this chapter is on the digital type of lodel, the by a set of equations commonly employedtoday. Such models are characterized be ,"pr"r"nting the ihysical i.o""*"* occurring in an aquifer. Thesemodels.may are discussed in deierministt or pio-babilistic nature, but only deterministic-models relations stemmingfrom known featuresof the They describethe cause-effect here. physicalsystemunder study'

482

CHAPTER20

SYSTEMS MODELINGREGIONALGROUNDWATER

Simplify equation so that solutions may be obtained by analytical methods

l _

Approximate equations numerically resulting in a matrix equation that may be solved using a computer

t_

model' a Figure 20.1 Logic diagramfor develqping mathematical OH') Worthington' WellAssociation, Water (Courtesy theNational of

a the Figure 20.1 characterizes procedurefor developing deterministicmathematical model. A conceptual model is formulated based on a knowledge of the of of characteristics the region and an understanding the mechanicsof groundwater flow. The next stepis to translatethe conceptualmodel into a mathematicalmodel, accompanied by usuallyrepresented a partial differential equationor setof equations boundaryand initial conditions.Conditionsof continuity and conserUy approplia@ uution of momentum, usually describedby Darcy's law, are incorporated in the and includeartesianor watertable condition designation model,Other modelfeatures quality and/or heattransIf dimensionality(one-,two-, or three-dimensional). water describing are fer considerations to be incorporatedin the model,additionalequations of energyare involvedand conservation of conservation massfor the chemicalspecies required.Typically usedrelationsare Fick's law for chemicaldiffusion and Fourier's law for heat transport. Once the mithematical model has been formulated, it can be applied to the situation at hand. This requiresconvertingthe governingequationsinto forms that through the use of numericalmethods facilitate solution.Ordinarily this is achieved partial differenthe to or suchasfinite differences dnite elements represent applicable the regionis divided for In tial equations. using a finite differenceapproach, example, as into grid elementsund th" continuousvariablesare represented discretevariables at the nodal points. In this manner,the governingdifferential equationis replacedby that a finite nu-b"r of algebraicexpressions canbe solvedin an iterativeway.Models of this type find wide applicationin the estimationof site-specificaquifer behavior. They havepro\iento be effectiveunderirregularboundaryconditions,wherethereare and heterogeneities, where highly variablepumping or rechargerates are expected'l Severil types of groundwater models and their applications are summarizedin Figure 20.2. region' A numberof stepsmustbe followedin modelinga targetedgroundwater They may be Figure 20.3 is illustritive. The f,rst step is to define the boundaries. ptiysical,suchas an imperviouslayer, or arbitrary, suchas the choiceof a politically,

Applications Watersupply
Regionai aquifer analysis Near-well performance Groundwatelsurface water interactions Dewatering operauons Seawaterintrusion Land fills 'Geothermal Thermal storage Land subsidence

Waste injection

Heat PumP Thermal poliution

Radioactive waste storage Holding pontls Groundwater pollution

Figure20.2Typesofgroundwatermodelsandtypicalapplications.(Courtesyof OH') Worthington, WellAssociation, theNationalWater by Next, the region is divided into discreteelements or otherwise,definedsubregion. grid (seeFigute 20'4)' a superimposing rectangularor polygonal ' (s. and z) and the once th-egrid is determined, controlling aquiferparameters If grid element. solutetransportis includedin the the initial conditionsare setfor each suchai hydrodynamicdispersionpropertiesmust also model, additional parameters hive beenmet, the modelcanbe operated Atter atl of thesespecifications be specifigd. comparisonsof and its output comparedwith recordedhistory (history matching). with counterpartmodelpredictionspermit recordedvaluesof headand other features and computeddata are considered to parameteradjustments be madeuntil observed by the modelerto be in closeagreement. to analyzea Upon completion of thelodel's calibration, it can be applied options. The model's prediction of the and/or development variety of management to decision-making outcomesof these alternative strategiescan be a valuable aid include: the problems that can be addressed Examplesof the types oI processes. of use; the impact on an aquifer of utilty of an aquifer to suppoit various levels storageof varying naturaiand artificial rechargerates; the effects on underground contamrate of movementof subsurface and spacing, pumpingrate; the weli lo-cation, intrusion. inants;and saltwater caution while numerical groundwatermodelshavemuch to recommendthem, Prickett and interpretedappropriately. to must be exercised enJurethat they are used three waysin notesthat overkill, inappropriatepiediction, and misinterpretationare To avoidthesepitfalls, both the modeler -#"t, can Uemisused.s which groundwater

484

CHAPTER20

MODELINGREGIONALGROUNDWATER SYSTEMS

illiiT3:':,",*itlffi,ffi:ff tlSl

(c water or the ar use'ourtesv Nation

upon which the model was and user must understandthe underlying assumptions founded,its limitations, andits sources errors.Usedwisely,modelscanbe powerful of decision-makingaids. Used inappropriately,they can lead to erroneousand sometimes damagingproposals.

20.2 FINITE-DIFFERENCE.METHODS
Digital simulation requires an adequatemathematical description of the physical processes be modeled.For groundwaterflow this descriptionconsistsof a partial to differential equationand accompanying boundaryand initial conditions.The governing equationis integratedto produce a solution that gives the water levels or heads associatedwith the aquifer being studied at selectedpoints in space and time. so The model can simulateyears of physical activity in a span of seconds, that the

METHODS 485 2A.2 FINITE-DIFFERENCE

(a)

A.r Ly ..,--V ^t ,
/_-/ | 1"

', / 1 . 1 , / 4Finitedifference
/
/

srld DlocK

(b) (b) boundaries. Finite-difFigure 20.4 (a) Map view of aquifershowingwell field and At' is the spacing t i"r""n"" grid{oi aquifir study,wGre Ax is the spacTqii lhu *ection' nodes;open in tt" y-at"rtion, and f is itre aquifer thicknesg.Solid dots: block-center worthingof the National Waterwell Association, circles:source-sink nooo tcoo*sy ton. OH.)

486

CHAPTER20

SYSTEMS MODELINGREGIONALGROUNDWATER

of consequences proposedactionscan be evaluatedbefore decisionsinvolving construction or social changeare implemented.The expectationis that the model runs * will lead to wiser and more cost-effective'decisions. of The finite-differencemethodis basedon the subdivision an aquiferinto a grid with zonesof the aquifer.The equationthat must and the analysisof flows associated and be solvedis derivedfrom continuity considerationq Darcy's law for groundwater motion. This yieldsthe followirigpartial differential equation(a versionof Eq. 18.41), aquifer.Note that the equationpresented flow through an areally extensive describing here describes two-dimensionalcase: the

s#P.N#ot:s!+w
where h : x : y : S: Z : V[ : total hydraulic head (L), x direction in a cartesiancoordinatesystem(L), y direction in a cartesiancoordinatesystem(L), specificyield of the aquifer (dimensionless), transmissivity the aquifer(I]lT) of sourceand sink term (L/T)

(20.r)

In the aboveequation,vertical flow velocitiesare consideredto be negligible are in everywhere the aquifer.The following assumptions implict in the derivation:the hydraulicconducfluid densityis constantin time and space; flow is two-dimensional; within the aquifer; flow obeysDarcy's law; and the specificyield of tivity is uniform the aquifer is constantin spaceand time. Equation 20.1 is nonlinear for unconfined variable. transmissivityis a function of headand thus the dependent aquifersbecause to integrateEq.20.I, initial valuesof head, transmissivity,saturated In order and sinksmust thickness the aquifer,andthe amountsof waterproducedby sources of for every point in the region of the integration.The specificyield and be identified location of geometric boundaries must also be defined. Unfortunately, analytic It solutionsto Eq. 2O.l are impossibleto obtain exceptfor the most trivial cases. is necessaryto resort to numerical integration techniquesto obtain the desired thus
answers.t'to-to

to Application of finite-difference techniques groundwaterflow problems reor quiresthat the region of concernbe divided into many small subregions elements valuesof all the variablesin (Fig. 20.5). For each of theseelements,characteristic which to Thesevaluesare assigned the centersof the elements, Eq. 20.I are specified. nodes.The headsin adjacentnodesare relatedthrough a finite-difference are called equation,which is derivedfrom Eq. 20.1. Thesedifferenceequationscan be derived The or by an appropriateTaylor's seriesexpansion by massbalanceconsiderations.8 to equationscan then be solvedsimultaneously yield the headsat resulting algebraic eachnode for eachtime stepconsidered. that the simulation methodspresentedin this chapter It should be understood pointed toward the analysisof regional rather than localized groundwaterprobare the lems suchas the prediction of the drawdownat a particular well. In suchcases, Here we,are in Chapter 19 are usually the most appropriate. methodsdiscussed that might occur over a large area with waterlevel or headchanges mainly concerned prescribedwater-usepractices. due to

2O.2 FINITE-DIFFERENCEMETHODS 487

t, I

o m,1
J

{
eAx'---

Node

Subregion

I Ay

l,,i

I
\ \ l, nd Boundary ofregion of integration a m,n

for a finiteFigure 20.5 Subdivisionof a region of integrationinto computationalelements differenceProblemformulation.

BoundaryConditions
In order to integrateEq.20.l,the governingboundary conditionsmust be-specified' in here.Otherswere presented ChapTwo typesof bJundary condition aie discussed ters 18 and 19. chosen Where the region of integrationis limited by a political or arbitrarily boundarycondition'10 constant-gradient boundary,it is oftei the policy to employ a change In this "ur", unar.umption is madethat the gradientof the watertable will not llvel may rise or fall. Where streamswith alongthe boundaryeventhoughthe water streamboundarycondiare system encountered, to interconnections the groundwater mathematicallyas are expressed boundaries tions are employed.Constant-gradient
a h l \

(20.2)

*:81x'!) g\x, y): a constantspecifiedat the location x, y throughoutthe period of where simulation (dimensionless) /z : hydraulic head (L) s : direction perpendicularto the boundary (L) as Streamboundariesare expressed h : f(x, Y, t) wheref(x,y,/)=anunknownfunctionoftimeatthelocationx'y (dimensionless) = hydraulic head (L) h

(20.3)

488

CHAPTER20

SYSTEMS MODELINGREGIONALGROUNDWATER

by boundariesdescribed The volumetric rate of flow acrossthe constant-head the Darcy equ-ation:ro Bq.20.2 can be modeledat eachtime stepusing

o: rfrtr

(20.4)

where h:head(L) Al = dummy variable denoting the length of the side of the subregion perpendicularto s (L) s : dummy variable denoting the direction of flow perpendicularto the boundary (L) (LilT) p : volumetricdischarge Z : transmissivityat the boundary (IllT) Use of this equationat a boundary is illustrated by the notation of Fig. 20.6. Considerthe flow from left to right in the x direction acrossthe left-handsideof the elementalregiondepicted.The node i - I, j lies outsidethe regionof integrationand may An that no information aboutit is available. assumption thus it may be assumed the across boundary be madeto circumventthis problem.It is that the transmissivity is uniform and equal to Tt,t. In finite-differenceform the head changeterm in Eq. 18.26 can be statedas
ah _hi.j-hi-l,j

6x

A,x
hi*r,j -

(20.s)

But the headh,-r,, doesnot exist, and anotherapproximationis required,

h,,i h,-r,i : hi,j

(20.6)

in are Thesetwo expressions then substituted Eq. 20.4 to yield

f,,,u#'LY Q;-r/",i:

(20.7)

Boundary where 6 h= r . dx

_ ^ 'i a 1 ' i'

Figure 20t6 Subregions adjacent to a constantgradientboundary.

2O.2 FINITE.DIFFERENCEMETHODS 489

At the beginningof each time step, a new volumetric flux is calculatedalong each by boundary.This is accomplished usingthe headsand transmissivconstant-gradient previoustime interval. . ities computedin the boundariesin groundtreatedas constant-head are sometimes Surfacestreams body wherethe waterlevel in the surface is waterproblems.The assumption adequate In during the time period of the modelingprocess. remain unchanged is expected to are flows,and henceheads, significantlyaffectedby however,surface many instances, They may then groundwatersystem. to withdrawalsor recharges the interconnected the To system. accommodate of water supplyfor the groundwater. be a limited source This expression linkage, a leakageterm may be applied.to water-groundwater surface may take the form

= -fi{n,.,.r leakage,,r,1
where

h,i,o)

(20.8)

b, , : thicknessof the streambed(L) hi,i.t : head in the aquifer at node i, j, at time k; k : 0 indicatesinitial conditions(L) t k;,; : hydraulic conductivity at node i,.i (LlT).

to When Eq. 20.8 is used,the streamis considered coverthe entire areareprenode. After eachtime step the leakagefrom the streamto the sentedby the related are aquiferis calculatedand streamflows depletedaccordingly.If the streamflowat a particular node becomeszero, the model can be madeto note that the streamis dry and break the hydraulic connection atthat point.l0

Time Stepsand ElementDimensions

of The success any finite-difference schemedependson the incremental valuesasand the time steps.In general,the smallerthe dimensignedthe elementdimensions sionsof elementsand time increments,the closerthe finite-differenceapproximation to the differential equation.However,as thesepartitions are madesmaller,a price in may computationalcostsand data needsmustbe paid. Furthermore,oversubdivision degree evenbring about computationalintractability. Thus the object is to selectthe of represehtation the systemwhile keeping of definition that results in an adequate data and computationalcosts at a minimum. There are proceduresfor making such but, exceptfor a brief discussionin the following section, they are not selections, presented here.lo-la

Flow Model One.Dimensional

To illustrate the finite-difference approachto groundwaterproblem solving, a onemodelsare Although mostpractical-scale is conceptualization discussed. dimensional of is in character,their development only an extension the two- or three-dimensional case.For details of some of the more complex modelsthe reader one-dimensional The should consult the appropriate references.6-8'i0-1s book by McWhorter and The treatment of to read and includesexcellentexampleproblems.8 Sunadais easy flow taken here follows the approachof that reference. one-dimensional

490

CHAPTER20

MODELINGREGIONALGROUNDWATER SYSTEMS Upstream headHn

Downstream head I{, H6*Huatt=0 Flow
hi*t

-----w

IlD-constantatt>0

Confined aquifer

l+Ay+l (After groundwater case. flow Figure20.7 Gridnotation a one-dimensional for McWhorter Sunada.8) and suchasthat Let us considera one-dimensional flow in a confinedaquifersystem illustratedby Fig. 20.7.lt is assumed that the flow is unsteadyand that the flowlines On are parallel and not time dependent. this basis,a unit width of the aquifercan be As madeaboutit can easilybe translatedto the total system. studiedand observations shownin the figure, the unit width of the aquiferis A.r. The flow regionis overlaidby a grid, and for eachgrid element,valuesof hydraulic conductivity Kr, elementlength y,, aquiferthicknessb,, storage coefficientSi, and the initial valuesof headh, must be specified.The massbalancefor grid elementI requiresthat the inflow (Qr-t-,) from I element - 1 to element minusthe outflow{Qt-,*r)fromelementlto element * I I I which occursin elementi, LV,/LI. must be balancedby the rate of changein storage To simplify the problem, let us further considerthat the aquifer is of uniform b, and thicknessand that it is homogeneous isotropic. Thus the valuesof K, ,S, and Ay are constant,and we shall considerthat from studiesof the aquifer properties,they are also known. Thereforeit mav be statedthat
K t : K z : ! . . : K ^ : K

Sr:Sz:..':S.:S bt:bz:',,:b*:b N r : L Y ,= ' ' ' : L Y - = N

(20.e)

Usingthis notation ln the wherethe subscript represents total numberof grid elements. and Fig. 20.7, we can seethat the flow from elementi - 1 to I is
. Oi-r-i: -UOO'| ;i '

(20.10)

where i : the elementnumber time n : the selected It in as Equation20.10is recognized Darcy's equation. is assumed this representationthat the head generatingthe flow at time n is the differencebetweenthe averageheadsat the two adjacent elementsdivided by the distancebetweentheir as exactness Ay diminishesto zero. centers(nodes).This approximationapproaches

2O,2 FINITE.DIFFERENCEMETHODS 491

area of flow and is in The ireaA appearing Eq. 20.10 is the cross-sectional obtainedasthe productof Al and b. Sincewe are dealingwith a unit width of aquifer, A.r : 1 and sinceb is a constantby definition here, Eq. 20.10 mtry be written
o,- -': hl - " i - t h,! -7"i '

Ay

(20.11)

for where 7 : Kb. A similar expression the flow from element ito i * 1 may be obtained:

(20.r2)
the Equations20.11 and 20.12 represent inflow and outflow from elementi. Considering that continuity conditionsmust be met, this changein flow acrossthe element which occursduring the time step.This is *urt b" balancedby the changein storage siven as

- outflow : in Now insertingthesethreeexpressions the continuity equation(inflow we changein storage), get

fr:'.(ry)

' (zo.tz)

t''i) (-rhi -,ni ) - f -rhi-,-- hi) : , 6,(h'lo'^. '\ ^t AY / / Av / \ \

becomes the By rearrangement, equation

1zo.r+i

(20.1s)
which is known as the explicit or forward difference form of the finite-difference as equationif n is designated the current value of time. If, on the other hand, n is definedas / + Ar, thin the equationis the implicit or backwarddifferenceequation. The Eachof theseforms hasits own solutiontechniques.8 explicit solutionto Eq. 20.15 here. will be discussed By letting n : t inF;q.2O.l and rearranging,one obtains

+ tur+^, ffirr;, +hi-) nllr ##]

(20.16)

derivativesare centeredat the beginningof the time stepand the In this casethe spabe singleunknownls h!*^'. Equation 20.16 canbe solvedexplicitly at eachelementfor only on a knowledgeof the the headat the next period of time. The solutiondepends

492

CHAPTER20

SYSTEMS GROUNDWATER MODELINGREGIONAI-

here, case discussed an unstable condition. Irt the one-dimensionalhomogeneous if stability is assured

<1 s/,vY t
TLt

(20.r7)

The equationshowsthat the choiceof time and spaceincrementsis not independent. an Satisfactionof Eq. 20.17 doesnot guarantee accurateapproximation,however;it for provides a stablesolution.8 only EXAMPLE 2O.T Refer to the one-dimensionalflow problem of Fig: 20.8. Let us assumethat the elementlength is 4 m and that the thicknessof the confinedaquiferis 2 m. It is further thit the headat the left and right sidesof the region is 8 m at / : 0 and that assumed : the head on the right side takes on the value 2 mfot all t greatet than zero. K Using the : 0.02. As shown in the figure, there are five elements. 0.5 m/day and S notation of eq. 20.16,the initial condition is hf; : 8.0 m. Use the explicit methodto determinefuture heads. Solution 1. First a determinationmust be made of the time step to use. This may be usingEq. 20.17. accomplished Lt <1 s(Ay)' ; W : :

0 . 1d a v s 6

was obtainedusingthe relation The valueof Zused in the aboveexpression T:Kb: Z:0.5X2.0:1.0 a To be on the safeside,we shall choose time stepof 0.1 days,althoughany stability. value lessthan 0.16 would have assured

8m

A)=4m Figtrre 20.8 Sketch for Example 20.1.

493 20.3 FINITE-ELEMENTMETHODS ^' 2."For the first time step,t : 0.1, we can calculateh'a* and corresponding headsfor the other elementsusing Eq. 20.16.Thus

: h,;",

* ,,411 -s(Ay),1 , + hg)" 0 f , - r f u l f# @?

and substitutingnumerical values,we get

,n, : 1.0(ol) + ^f. - 2(1.0)(0 /,1' (ffi(2.0 + 8.0) 8.olt drn+Fl. 1 ) l

: 3 . 1+ 3 . 0: 6 . 1 3 m Sinceft? andho2: 8.0 m and sinceh, : 8,0 by definition,it can easilybe from shownusingEq. 20.16 thatthe valuesof hlr andhl't ate not changed step. their original level of 8.0 during the first time 3. Now considerthe secondtime step,t + Lt = 0.2 days.For element4,

: h2,
For element3

: 0.31Q.0+ 8.0) + 6.13(0.37: 5.4 m )

#(h2,

+ + hg\ nZ'lt

#]

hg': ffi] ffi(h:'+h9\+l,l'[r

:) : 0 . 3 1 ( 6 . 1+ 8 . 0 )+ 8 . 0 ( 0 . 3 7 7 ' 4 m 3 Element 2 doesnot have a head changeuntil the third time step. until the headshavebeencalculated is 4. The processdemonstrated repeated for the total time period of interest.For this example,they will ultimately reachequilibriumconditions. rl This exampleproblem illustratesthe mechanicsof the finite-differenceprocedure. Problemsof practical scalewould require the use of a computer,but the approach would still be the same.

METHODS 20.3 FINITE.ELEMENT

flow problemsare for The mostwidely usednumericaltechniques solvinggroundwater The finite-elementmethodis similar the finite-differenceand finite-elementmethods. in lead to a set of N equations to the finite-differencemethodin that both approaches Nodesin the finite-elementmethodare Nunknowns that canbe solvedby relaxation.6 usually the corner points of an irregular triangular or quadrilateral mesh for twobricks or tetraheapplications, while for three-dimensional dimensionalapplications, are drons are commonly used.The sizeand shapeof the elementsselected arbitrary. to They are chosen fit the applicationat hand.They differ from the regularrectangular grid elementsusedin finite-differencemodeling.Elementsthat are closestto points of flow concentratiensuchas wells are usually smaller than those further removed

494

CHAPTER20

MODELINGREGIONALGROUNDWATER SYSTEMS

from suchinfluences. Aquifer parameters suchashydraulic conductivity may be kept constantfor a given elementbut may vary from one to another. To minimize the variational function, its partial derivativewith respectto head,,is evaluatedfor each node and equatedto zero.The procedureresultsin a set of algebraicequations that can be solved by iteration, matrix solution, or a combination of thesemethods.ra Finite-elementmodelersmust understand partial differential equationsand the calculus of variations.6 The finite-elementapproachoffers someadvantages over the finite-difference technique.Often, a smaller nodal grid is required, unA tnir offers economies in computer effort. The finite-elementapproachcan also accommodate one condition that the finite-difference approach is unable to handle.6When using the finitedifferencemethod,the principal directionsof anisotropyin an anisotroplcformation are parallel to the coordinatedirections.In caseswhele two anisotropicformations having different principal directions occur in a flow field, the finite-difference approach cannot produce a solution, whereasthe finite-element approach can. TLe finite-elementtechniquecan be used to simulate transient aquifei-performance. A detaileddiscussion the finite-elementtechniqueis beyondthe scopeof this boor<, of but there are many good references the interestedreader.6'rsts.zi for ,

20.4 MODEL APPLICATIONS

To illustrate how simulation modelscan be usedto provide insightsinto water managementschemes, model analysisof the Upper Big Blue basin aquiferin Nebraska a is presented' The studywasconducted the Conservation SurveyDivision of the by and University of Nebraskaunder the direction of Huntoon.l0 The useof groundwater irrigation in the Upper Big Blue basinwas observed for to be rapidly increasingand by 1972 about3.: wetlslmit iere in operation.At that time farmerswere becomingconcernedaboutthe progressive declineof water levels and were seekingguidanceaboutthe efficiencyof implementingsomeform of basinwide water management proram. The Universiry of Nebraski designeda model to evaluate situationand to explorevariousproposalsfor recharging:the the aquiferand for estimatingthe long-term consequences siveral scenariosof water use in of the basin. The study area is shown in Fig. 20.9. Generallythe water table is free in the regionof interest'Figure 20.10 showsthe configurationof the watertable asobserved in 1953.For modelingpurposes, water-levelcontoursshown were considered the to be representative predevelopment of conditions.This assumptionwas basedon the fact that groundw-ater withdrawals before this time were not extensive. was also It surmisedthat the Lonburs represented water table in which an equilibrium existed a between natural recharge and discharge the region.Transmissivities in wereestimated from drill-hole sample logs recorded in the area. These values are needed for modelingand are also important indicesof the potential yield of wells that might be constructed. As might be suspected, information of most concernto the local landowners the and water plannerswas the rate of decline of the water table. In particular, it was

.-/l

t-

z
()
I

6)

bo

o\
N

b0

,-

H H

3
d
C) C) a

o
C)

o
a

b
Q

?
a

ca ra

U
o
6 H

e F (.)

-.----d

e o rrl

a o d

N
ra

p R

.$E

20.4 , MODELAPPLICATIONS

497

where and resourcewould Le depleted, desiredto knbw how rapidly the groundwater on water use,and what when waterlevel declineswould posean economicconstraint would haveon thdrate of decline. and/or management impactsfuture developments represento The model developed explorethesefeatureswas a two-dimensional 20.1, alongwith the tation of flow through an areally extensiveaquifer.loEquation appropriate boundary conditions, constituted the model. The region shown in was divided into a finite-differencegrid and, after substitutionof the nodal n1g.Z0.S to ualu"t of Z and S, the model was operatedto predict water-levelchanges the year Calibraof development. 2020 for variouspolicies of rechargeand for severallevels The model was operatedover usinghistoricdata. tion of the model was accomplished the period 1953-1972 using the known distribution of wells and the averagenet pu-pug" per well to establisha match betweenobservedand estimatedwater-level the simulation of future trends proceeded. "ttung"i. Once this was accomplished, process. in achieved the matching Figures20.II and20.12showthe correspondence waterlevelsin the study that it On the basisof the model studies, wasdetermined was limited to the l912level. It areawould continueto declineevenif development groundwasevere wasfurther predictedthat someparts of the areawould experience by employingartificial ter shortageiby the year 2000. It was found, however,that the To rechargefrethods, permanentgroundwatersuppliescould be assured. assess modeled.Both of these were two water delivery systems effectsof artificial recharge, delivered water from Platte River Valley sourcesto rechargewells located in the were three rechargeschemes project area. Using thesetwo water delivery systems, was the cancellationof ii-ulut"d. The grosseffect of introducingthe rechargewells the effects of the proportionatenumber of pumping wells. Figure 2O-I3 showsthe plan computedwater-levelchangesat one location under a graduateddevelopment with no rechargeand.then (projected on the basisof the 1972rate of development) The continual for wlth graauateddevelopment eachof the three rechargeschemes. (curve 1) clear$ showsthe nature of downwardtrend in waier level with no recharge depictingthe threeartificial the problemin the UpperBig Bluebasin.The othercUrves such an approachis taken. ,""hurg" options show that stability can be achievedif it While the costs of implementing artificial rechargemight be excessive, is water table problem, short of apparentthat any long-terrn solution to the declining sourceof water. reducinguse, would require a supplemental Operationof the modelprovidedusefulinsightsinto the natureof the watertable that problem and suggested irrigators shouldbe making someimportant water managementdecisionsabout their future mode of operation..^ ^In is The modeling of groundwatersystems complex.lo-25 structuring models usuallybe made.Thesehave must simplifying assumptions suchasthat just discussed, to do with aquifer pbrameterssuch as transmissivity,specificyield (for unconfined Furthermore,the boundary aquifers),and storagecoefficient (for confined systems)' conditions are normally approximationsof what occurs in the physical system.and strataare someaboutthe uniformity of materialsin varioussubsurface assumptions to cannotbe expected yield models times crude.This doesnot meanthat groundwater must be cautiousabouthow useful results.It doesimply that the usersof the models aquifermodel suchasthat they interpretthe output.For example,an areally extensive to problem can be expected give by developed Huntoon for analyzingthe Blue River

o E

Z-u---sI

E.l(

F \ $

o
a
H H

()
o

e
a

:d65'
\ -o-

\=e 1t "-

r-O

t\-z
,
lo

U U
o

9pc
H ' A

o o i * O d O

E =
H E

d 9
: : 4 J

>.o
N X o ! l L ; hn:\ a;-Y

F I X ti ;i

2O.4 MODELAPPLICATIONS ^ 8 2 o

499

.8 s4
d 8 6 e 6 6
A 1/

rl--

Computed trend

Eno
P, 92 e 9 4 6

\/
V
Measured water levels

tse6
h q R

v
Y
1958 t959 1960 t961 t962 1964 1965 Year

F 100 1,02
1956

t967 1968 1969 1970 r97I 1972 1973

trends'(AfterHuntoon'10)' water-level and Figure20.1.2 Measured computed reliable information about water-leveltrends for various conflgurationsof developan ment. It shouldnot, on the other hand, be considered accuratepredictivetool for at monitoringthe water-levelchange somespecificpoint in the regionof concern.This type of information could be derived only from a more detailed modeling of the the lolafif surrounding point. The informationprovidedby the Blue River modelwas what the future might hold for severaldevelopment to targeted showlocal landowners options.The actual water levelspredictedby the levelsand for severalmanagement model were not of central concern;what was of interestwas the determinationthat water provided,or unless was unlessfuture development restrictedand supplemental current usescould be significantlyreduced,the outlook in the next 50 yearswas not good for irrigated farming. The model thus providedthe basisfor making somequantitativeobservations from aboutthe future. It also providedinsightsinto the relief that might be expected management Beyondthat, it could be usedto model other possible artificial recharge.

2010

2020

20.13 Computed water-level Figure under a plan of graduateddevelopchanges ment for conditions of (1) no techatge,(2) rechargeunder Scheme1, (3) rechargeunder Scheme 2, and (4) recharge under 3. Scheme (After Huntoon.lo)

500

CHAPTER20

MODELINGREGIONALGROUNDWATER SYSTEMS

options.A,model suchas this, carefully usedand properly interpreted,can thus add a powerful dimensionto decision-making processes.

20.5 GROUNDWATER QUALIW MODELS

quality hasbecomea major sourceof concernin recentyears.This has Groundwater comeaboutfrom the realizationthat many groundwater sources that wereat onetime consideredalmost pristine have now been degradedin quality by seepages from dumps, leakagefrom industrial waste holding ponds, and by other waste disposal and/or industrial and agricultural practices.To deal with suchproblems,'therehas been an expandingmovementto developquantitative techniques understandthe to mechanicsof groundwaterquality. Thesemodels,althoughnot as advanced their as surfacewater counterparts,are now beginningto play an important role in water quality management. The subjectof groundwater quality modelingis complexand underrapid development.Accordingly,a thoroughtreatmentof the subjectis beyondthe scopeof this book. The importance of this topic cannot be overemphasized, however, and the readeris encouraged consultthe references the end of the chapter,specifically to at Refs.6 and 26-30. In 1974, Gelhar and Wilson developed lumped parametermodel for dealing a with water quality in a stream-aquifersystem. The nomenclature and conceptualization of their model are shown in Fig. 20.14.2e The rationale for using a lumped parameterapproachwasthat when dealingwith changes groundwater quality over in long periods of time, temporal rather than spatial variationsare most important. Changes water table in the Gelhar-Wilson (GW) model are represented in by the following equation:
dh 'nd- t=

-q+e*q,-ep

(20.18)

where

h - average thicknessof the saturatedzone p : average effectiveporosity e : natural rechargerate natural outflow from the aquifer 4 , : artificial recharge/unitarea Q p : pumping ratelunit area T _ time

This is just anotherform of the continuity equationrelating inflow, outflow, and the (lefrhand term in Eq. 20.18).The changein concentration a changein storage of constituentis given by
,dc Pndt(e -l q, -t aph)c : ec. * q,c,

(2o.re)

QUALITYMODELS 20.5 GROUNDWATER

. Ct

(b)

model'(After of Figure 20.L4 Schematic the Gelhar-Wilson and Novotny Chesters.2a) where c: c; : c, : c: concentration coflcefltrationof the natural recharge concentrationof the artificial recharge of a first-orderrate constantfor degradation the contaminant

may be that dispersionis negligible.This assumption The GW model assumes objective of the model iS to estimateregional-average made on the basis that the The concentratiolrs.2e model also provides for the determinationof hydraulic and of Theseare measures the lag that occursin the times for the system. soluteresponse constituentinputs to the system.Gelhar and Wilson moue-"ni of both water and of that the response an aquiferto a specificinput can be likened to that of a assume well-mixed linear reservoir.Their studiesshowedthat the model's determinationof of leavingan aquifer is representative the average the concentrationof constituents that the model in the aquifer.On this basis,,itappears concentrationof the constituent

502

CHAPTER20

MODELINGREGIONALGROUNDWATER SYSTEMS

to stream, is well suitedto estimatingthe quality of groundwater discharging a surface providing-ihe aquifer is narrow relative to the length along which discharge occurs.

r summary
in is Groundwater a regionalaquifer system constantlyin motion. The amountstored evapotranspiration, flow to at any time is affectedby artificial and natural-recharge, springs and surface water courses,and by collection devices such as wells and infi ltration galleries. Natural hydrologic statesmay be significantly affected by human activities. Aquifer depletionshaving regional and national economicimplications are not uncorlmon. Depletionof the Ogallala aquiferin the central United Statesby long-term and extensive water withdrawalsfor irrigation is a good example.On the other hand, by inadvertently, humanintervention. waterlevelshavebeenmadeto rise, sometimes for Leaky irrigation canalsin central Nebraskawere at one time responsible groundwaterlevel risesin somefarming locationsof a magnitudesufficientto jeopardizeuse occur, they are of the land. Once major problemsof depletionor over-replenishment policy for groundwater has management not easilydealt with. In general,a safe-yield merit and shouldbe considered.6'30. Regional groundwater flow problems are usually modeled by an equation combiningDarcy's law and the equationof continuity.The resultingpartial differenthe tial equationo setof equations, or describes hydraulicrelationswithin the aquifer. To effect a solution to the governingequation(s),the aquifer's hydraulic features, geometry,and initial and boundary conditions must be determined.Unfortunately, many groundwaterproblems exist for which exact analytic solutionscannot be obtained.In suchcases, is necessary rely on numericalmethodsfor modeling.Under it to suchcircumstances, approximatesolutionis obtainedby replacingthe basicdifferan ential equationswith another set of equationsthat can be solved iteratively on a computer.Both finite differenceand finite elementmethodsare applicable. The finite differenceapproachdescribed this chapterreplacesthe governing in Thesecanbe solvedon partial differential equations with a setof algebraicequations. at the computerto producea set of water table elevations a finite numberof locations in the aquifer. Once the groundwatermodel has been calibrated,it can be usedto predict the prostrategies and/or management outcomes(impacts) of alternativedevelopment posed for an aquifer. Such analyses valuable adjuncts to decision-makingproare Models can, for example,simulate the effects of opening new well fields, cesses. analyzechangedbperating practices for existing well fields, explore schemesfor plans. artificial recharge,and predict the impactsof proposedirrigation development Groundwatermodelscan be applied to unconfinedaquifers,semiconfinedaquifers, large variaconfined aquifers,or any combinationthereof. They can accommodate coefficient,and tions in aquiferparameters suchashydraulicconductivityand storage they can be usedto analyzeunsteadyas well as steadyflow problems.

REFERENCES 503

PROBLEMS
20.t. Refer to Fig. 20.8. Assumethat the elementlength is 5 m and the thicknessof the
confined aquifer is 2.5 m. The head at the left and right sidesis 8.1 m at r : 0, and the head on the right is 2.5 m for all r > 0. K : 0.5 m/day and S : 0.02. Use the explicit method to determineheadsat future times. 20.2. Referto Fig. 20.8.Assumethe elementlengthis 10 ft andthe thicknessof the confined aquiferis 8 ft. The headat the left and right is 21 ft at t : 0, and it drops on the right sideto 8 ft for all t > 0. K: 1.5 ftlday-ands : 0.02. use the explicitmethodto calculatefuture heads. from studyingthis 20.3. Referto Fig. 20.12.Asidefrom the trend, what elsecanyou deduce figure? 20.4. Discusshow you would go about designinga grid for a regional groundwaterstudy. What types of boun{ary conditions might you specify?Why?

REFERENCES
Modeling.Worthington,OH: National Water 1. J. W. Mercer and C. R. Faust,Ground-Water 1981. Well Association, 2. C.A.AppelandJ.D.Bredehoeft,"statusofGroundwaterModelingintheU.S.Geological ey Survey," U.S, Geol. Surv Circular 737(1976). "Utilization of Numerical Groundwa3. Y. Bachmat,B. Andres,D. Holta, and S. Sebastian, ter Models for Water ResourceManagement,"U.S. Environmental Protection Agency ReportEPA-600/8-78-012. "Contribution of Ground-waterModeling to Planning," J. Hydrol' 43(Oct. 4. J.E. Moore,

r979).
5. T. A. Prickett,

"Ground-water Computer Models-State of the Att,'l Ground Water

t2r-r28(r979). r7(2),
6. R. A. Freezeand J. A. Cherry, Groundwater.EnglewoodCliffs, NJ: Prentice-Hall,1979. 7. G. D. Bennett, Introduction to Ground Water Hydraulics, book 3, Applications of HyD.C.: U.S. GeologicalSurvey,U.S. GovernmentPrinting Offlce, draulics. Washington, 1976. 8. D. B. McWhorter and D. K. Sunada, Ground Water Hydrology and Hydraulics. Fott 1977' Publications, Collins,CO: WaterResources Hydrology,2d ed.New York: Wiley' 1980' 9. D. K. Todd,Groundwater "PredictedWater-LevelDeclinesfor Alternative GroundwaterDevelop10. P. W. Huntoon, Rep.No. 6, Conservation in the Upper Big Blue River Basin,Nebraska,"Resource ments and Survey Div., University of Nebraska,Lincoln, 1974. "The Numerical Solution of Parabolic and and H. H. Rachford, Jr., 11. D. W Peacemen Indust.Afpl. Math. J.3,28-4I(I955)' Soc. Elliptic DiffetentialEquations," "Application of th Digital Computer for Aquifer 12. G. F. Pinder and J. D. Bredehoeft, 1968). 4(4), 1069 1093( Res. WaterResources Evaluation," Hydrol13. I. Remson,G. M. Hornberger,and F. J. Molz, NumericalMethodsin Subsurface ogy. New York: Witey, 1971. Techniques GroundwaDigital Computer for 14. T. A. Prickett and C. G. Lonnquist,Selected Evaluation,Illinois StateWater Survey Bull. No. 55, 1971. ter Resource

PARTFIVE

MODELING HYDROLOGIC

Chapter21

to lntroduction Hydrologic Modeling

r Prologue
The purposeof this chapteris to: . Introduce the types and classes hydrologicmodels. of . Illustrate the limitations, alternatives,steps,general components,and data needsof hydrologicsimulation models. . Presenta philosophicalprotocol for performing successful modelingstudies. . Give an overview of groundwatermodel types. . Distinguishthe need for separate, specificproceduresdetailedi4 subsequent 23, 24, and25. 22, Chapters ratesandvolumesof flow at any point of interestalonga stream Information regarding in is necessary the analysisand designof many types of water projects. Although many streamshavebeen gaugedto provide continuousrecords of streamflow,planinformastreamflow facedwith little or no available are nersand engineers sometimes must rely on synthesisand simulatlon as tools to generateartificial flow tion and regardingstructuresizes,the effects of for sequences use in rationalizing decisions water supplies,water quality, and the effects of control measures, land use, flood natural or inducedwatershedor climatic changes. making in both the designand operationof large-scale The problemsof decision have and canals,aqueducts, water supplysystems reservoirs, of systems flood control to suchas simulation and synthesis mathematicalapproaches resultedin a need for project. Simulationis definedas the mathematicaldescriptionof the investigate total the responseof a'hydrologic water resourcesystemto a seriesof eventsduring a time period.For example,simulationcanmeancalculatingdaily, monthly, or selected resulthydrograph streamflowbasedon rainfall; or computingthe discharge seasonal or hypotheticalstorm; or simply fllling in the missingvaluesin a ing from a known streamflowrecord. Simulation is commonly used in generating streamflow hydrographsfrom rainfall and drainagebasin data. The philosophiesand overall concepts used in simulation are introduced in this chapter. Chapter 22 summarizesconcepts of

508

21 To cHAprER rNTRoDUcroN HyDRoLocrc MoDELING streamdowsynthesis stochasticmethods.Chapters23-25provide detailsregardby ing determinislic continuousmodels, single-eventmodels,urban runoff and storm sewerdesignmodels,and water quality models. Stochastic techniques usedto extendrecords,either rainfall or streamflow,are classifiedas synthesis methods.This procedurerelies on the statisticalpropertiesof an existingrecord or regionalestimates theseparameters. overviewof synthesis An of techniques presented Chapter22. is in

21.1 HYDROLOGIC SIMULATION

In this chapter, simulation of all or parts of a surface,groundwater,or combined system implies the use of computersto imitate historical eventsor predict the future response the physicalsystem a specificplan or action. Physical,analog,hybrid, of to and or other modelsfor simulatingthe behaviorof hydraulic and hydrologicsystems imitating systemcomponentshave had, and will continue to have, application in prototype behSviorbut are not discussed here. A few of the numerous event,continuous, urbanruneff computermodelsfor and simulatingthe hydrologiccycle are comparedin Table 21.1. As shown in the tabfe, All mostof the modelsweredeveloped or by, universities federalagencies. have or for, from 1 to 10 percentof moderate-to-extensive and all have input data requirements,
TABLE 21.1 DIGITALSIMULATION PROCESSES MODELSOF HYDROLOGIC

Code name

Model name

Percentage of inputsby Date of original Agency organization judgmenf development or

Private ARS Stanford University EPA Corps USGS USDA 24
I 5
I

23 Continuous simulation models-Chapter streamflow

API USDAHL SWM-IV HSPF NWSRFS SSARR PRMS SWRRB HEC-1 TR-20 USGS HYMO SWMM UCUR STORM MITCAT SWMM ILLUDAS DR3M PSURM Antecedent Frecipitation Index Model 1970,1973,1974 Revised Watershed Hydrology Stanford WatershedModel IV Hydrocomp Simulation Program-FORTRAN National Weather Service Runoff Forecast System Streamflow Synthesisand Reservoir Regulation Precipitation-Runoff Modeling System Simulator for Water Resourcesin Rural Basins

I 10 10 10 3 5 10

1969 t970 1959 196l t972 1958 1982

r990 1973 I 965 t972 r972 t971. t972 r974 t970 I9'71 1972 1978 1979

Rainfall-runoff event-simulation models-Chapter HEC-I Flood HydrographPackage Corps Computer Program for Project Hydrology scs USGS Rainfall-Runoff Model USGS Hydrologic Model Computer Language ARS Storm Water ManasementModel EPA

10 1 5 z 3 5 5 I 5 5

runoff models-Chapter 25 Urban simulation

University of Cincinnati Urban Runoff Model Quantity and Quality ofUrban Runoff MIT Catchment Model Storm Water ManagementModel Illinois Urban Drainage Area Simulator Distributed Routing Rainfall-Runoff Model PennsylvaniaState Urban RunoffModel University of Cinci4natr Corps MIT EPA Illinois State Survey USGS PennsylvaniaState University

"Judgment percentagesare from U.S. Army WaterwaysExperiment Station.r

SIMULATION509 21.1 HYDROLOGIC trials Theseare normally validatedby repeated inputs that arejudginent pafameters. are primarily eventsimulationmodelsbut with the models.The urban runoff models modelsare deferred of the havebeenisolatedin Table2 1. 1 because descriptions urban 25. to ChaPter streamflowsimulationmodelsshown Severalof the major eventand continuous 23 and24.TheStanfordandHEC-1 models in Table2!.1aredescribedin chapters Table 2I't are briefly For are emphasized. further referince, most modelslisted in "Models and Methmodels,in the publication alongwith about 100 other described, Fleming's text plesentscomplete Studi-es."1 ods Applicable to Corps of Engineers models'2 descriptionsof the SSARR, S'[iM, HSP, USDAHL' and other

Models of Classification Simulation

the scienceof ct In recent decades has Passed water resourcesystems procedure.The variedni engineering has causeda proliferation of catego are classiflcations Presented. physical vs. Mathematical Models Physicalmodelsinclude analogtechnologies models'In contrast' mathematical and principlesof similitude appliedto smali-scale to representthe system'A laboratory flume models,"iy on mathematicafJtut"-.nt, hydrographtheory of rhay be a 1:10 physical model of a stream, while the unit to of response a watershed variouseffective modelof the Chapter12 is a mathematical rain hYetograPhs. by is classification achieved considerContinuous vs. Discrete Models A second the models as continuousbecause processes ing physical,analog,and some digital occur and are modeledcontinuous of necessityand advantages slicing qualifY as discrete models' A we indication method for routing a flo rz reservoirdischarge instantaneous time. over time and timethat involve changes Dynamic vs. static Models Processes dynamic models' In contrast' modelsthat by varying interactionscan be simrrlated static' Few hydrologic examine time-independentprocer*"* ui" irequently called simulation modelsfall into the latter category' havehad the greatest Descriptive vs. conceptual Models Descriptivemodels they are because appticaiionand are of particular interestto practicinghydrologists through empiricisrnand the useof basic phenomena to designed accountfor observed concepmomentumconservationassumptions' fundamentalssuchas continuity or heavilyon theory to interpletphenomenarather tual models,on the other hand, rely includemodelsbasedon of than torepresentthe physicalpto""tt. Examples the latter

'

510

CHAPTER21

INTRODUCTION HYDROLOGIC TO MODELING

probabiliiy theory.Recenttrendsin the useof artificial intelligence and expertsystems in water system modelingwould classifyas conceptual methods. Lumped vs. Distributed Parameter Models Modelsthat ignore spatialvariations in parameters throughoutan entire systemare classifiedas lumpedparameter models.An exampleis the use of a unit hydrographfor predicting time di;tributions of surface runoff for different stormsover a homogeneous drainagearea.The "lumped parameter" is the X-hour unit hydrographusedfor convolutionwith rain to givelhe storm hydrograph.The time from end of rain to end of runoff is also a lumped parameteras it is held constantfor all storms.Distributedparametermodelsaccount for behavior variations from point to point throughout the *yst"*. Most modern groundwater simulationmodelsare distributedin that they allow variationsin storage andtransmissivity parameters overa grid or lattice system superimposed overthe plan of an aquifer.More recently,surface water systems being analyzedthroughuseof are distributedparameterGeographical Information System(GIS) technologies. Black-Box vs. Structure-lmitating Models Both of thesemodelsacceptinput and transform it into output. In the former case,the transformationis accomplished by techniques that havelittle or no physicalbasis.The alchemist'spurported ability to transform lead into gold or plants into medicinewas accomplished a black-box in fashion- In hydrology, black-box models may sometimestransform "plants" into "medicine" even though the reasonsfor successare not clearly understood.For example, modelthat accepts sequence a a ofnumbers,reduces eachby 20 percent,and outputsthe resultsmight be entirely adequate predictingthe attenuationof a flood for waveas it travelsthrough a reachof a given stream.In contrast,a structure-imitating modelwould be designed useaccepted to principlesof fluid mechanics and hydraulics to facilitate the transformation. Stochastic vs. Deterministic Models Many stochastic processes approxiare mated by deterministic approachesif they exclude all considerationof random parametersor inputs. For example,the simulation of a reservoir systemoperating policy for water supply would properly include considerationsof unceitainties in natural inflows, yet many water supply systems are designedon a deterministic basisby masscurve analyses, which assume that sequences historical inflows are of repetitive. Deterministic methods of modeling hydrologic behavior of a watershedhave becomepopular. Deterministic simulation describes behaviorof the hydrologic the cycle in terms of ma[hematicalrelations outlining the interactionsof variousphases of the hydrologicclcle. Frequently, modelsare structuredto simulatea streamflow the value,hourly or daily, from given rainfall amountswithin the watershed boundaries. The model is "verified" or "calibrated" by comparingresults of the simulation with existingrecords.Oncethe modelis adjusted fit the known period of data,additional to periods of streamflowcan be generated. Event'Based vs. Contlnuous Models Hydrologicsystems be investigated can in greaterdetail if the time frame of simulationis shortened. Many short-termhydrologic modelscould be classifiedas event-simulation modelsas contrastedwith seauential

'

' 21.1 HYDROLOGIC SIMULATION

511

or coniinuousmodels.An exampleof the former is the Corps of Engineers-singlemodel eventmodel, HEC-1,3 and an exampleof the latter is the Stanfordwatershed which is normally operatedto simulatethree' by developed Crawford and Linsley,a might use four, five, oi *or" yearsof streamflow.A typical event simulation model perhapseven I min' a time incrementof t hr or have arisen water Budget vs. Predictive Models sevelal model classiflcations One important comparison of that distingtish betweenthe purposes the model types. precipito the model proposes predict future conditionsusing synthesized is whether tation and watershedconditions or model is definedas a model or set of in of inflows, outflows,and changes that simulationmodel studier advised that affirm the b use the Parameters example,meteorologicdat shed.For watershed'A water application amounts might be known for a given agricultural (ET) formula the budgetmodelwould be uied to determine correctevapotranspiration in the continuity equation parametersby testing a range of valuesuntil a balance or month-byoccursfor all time increments.This is often performedon a day-by-day the water budget model' month basis. once the ET parametersare derived from conditions'or farming predictive simulationsof diffirent crop patterns,meteorologic in relationships the model that the practicescould be performedwith the satisfaction corroboratehistorical water budget (precipitation outputsare measured dt require the simultaneous studies such as ET, infil ondarY Processes spatial distribution of water applications'

of Limitations Simulation
systems'some Becausesimulation entails a mathematicalabstractionof real-world systembehaviorcan^occur.The extent to which the degreeof *i,,"p,",entation of "depends on many factors' The test of a developed model and systemoutputs vary that the behavioris consisby demonstrating of simulationmodelconsists u"iifi"uiion the physicalsystem' tent with the known behaviorof resources Even verified simulation models have limitations in usesfor water of models will allow performanceassessments planhing and analysis.Simulation options,particularly optibut specificschemes cannotbe usedefficiently to generate once a near-optimal plan is formulated^bysome mal plans, for stated objectives. effectivefor testing a othertechnique, limited numberof simulationruns are normally combinationsof decisionvariablesusing ranand improving the plan by modifying for Techniques generatingoptimal plans are dom or systeriaticsamplingtechniques. in described Section21'3' proceAnother limitation of simulationmodelsinvolveschangingthe operating beingmodeled'Programming of the system duresfor potentialor existingcomponents for example,requires u "o*pot", to handle reservoir storageand releaseprocesses' reprogrammingis rerules, and considerable large portions to define the operatin! qoir"a if other operatingproceduresare to be investigated.

t-

512

oHAPTER INTRoDUoTIoNHYDRoLoGIc 21 To MoDELING A'fourth limitation of simulationmodelsis the potential overreliance sophison ticated output when hydrologicand economicinputs are inadequate. The techniques of operational hydrology can be used to obviate data inadequacies, these also but require input. Controversyover the use of syntheticdata centerson the questionof whetheroperationalhydrologyprovidesbetter information than that containedin the input.

Utilityof Simulation
Computersimulation of hydrologicprocesses severalimportant advantages has that shouldbe recognizedwheneverconsidering merits of a simulation approachto a the problem that has other possiblesolutions.One alternativeto digital simulationis to build and operateeither the prototype systemor a physically scaledversion.Simulation by physical modeling has been applied successfullyto the analysisof many components systems of suchas the designof hydraulic structuresor the investigation of streambank stability.However,for the analysis complexwaterresourcesystems of comprisedof many interactingcomponents, computersimulation often provesto be the only feasibletool. Another alternative to digital simulation is a hand solution of the governing equations.Simulation models,once formulated, can accomplishidentical results in lesstime. Also, solutionsthat would be impossibleto achieve hand are frequently by achieved simulation.In addition, the systemcan be nondestructively by tested;prdposedmodificationsof the designs systemelementscan be testedfor feasibility or of comparedwith alternatives; manyproposals be studiedin a shorttime period. and can An often overlookedadvantageof simulation includes the insight gained by gathering,organizing,and processing data, and by mentally and mathematically the formulating the model algorithmsthat reproducebehaviorpatternsin the prototype.

Stepsin DigitalSimulation
A simulationmodel is a set of equationsand algorithms(e.g.,operatingpolicies for reservoirs)that describethe real systemand imitate the behaviorof the system.A fundamentalfirst stepin organizinga simulationmodelinvolvesa detailedanalysis of all existingand proposedcomponentsof the systemand the collection of pertinent data. This stepis called the systemidentification or inventory phase.Includeditems of interestare site locations,reservoircharacteristics. rainfall and streamflowhistories, water and power demands, and so forth. Typical inventory items requiredfor a simulatiol study and data needsthat are specificto someof the modelsare detailed in subsequent paragraphs. The second-phase model conceptualiTation, is which often providesfeedbackto the first phasebf defining actual data requirementsfor the planner and identifying system components that areimportant to the behaviorof the system. This stepinvolves (1) selectinga techniqueor techniques that are to be used to representthe system (2) elements, formulating the comprehensive mathematics the techniques, (3) of and translatingthe proposedformulation into a working computerprogramthat interconnects all the subsystems algorithms. and Following the systemidentification and conceptualizationphasesare several stepsof the implementationphase. Theseinclude(1) validatingthe model,preferably

SIMULATION 513 21.1 HYDROLOGIC for the by demonstrating that the model reproduces any available observed behavior (2) modifying the algorithms as necessary to improve the actual or a similir system; the simulaaccuracy of the model; and (3) putting the model to work by carrying out tion exPeriments.

Model Protocoi
from recommendations adapted modelstudies, Five axiomsfor performingsuccessful by Friedrich,5are: 1. Evaluatethe data beforebeginning. t Document assumPtions. of 3. Plan and control the sequence computerruns' of 4. Insist on reasonableness output. 5 . Document,document,document. An Examining and evaluatingthe basic data are essential. annotated,bibliographic program \ record of the data ,orrr."* shouldbe maintained.It is alwaysgood adviceto of the numerical modelsthat output (echo)datavaluesasthey arereadin. Verification from the echo. valuesand proper entry of the data can be established Statisticssuchasih" 1n"un,mode,median,range,standarddeviation,skewness, for kurtosis, and rank order are often helpful in locating entry errors' Checking aheadof the rainfall? Are can inconsistencies identify errors. Didthe runoff occur characters Do waterlevelsgraduallyvaried,or are there discontinuities? alphabetical be interpretedas missingor appearin th! data?Will blank valuesin the data sets For hyzeros?Will zeros in the data sets result in overflows (division by zeto)? the limits recommended fall drographrouting, doesthe time interval selected between of for staUltityand convergence the numerical method usedto solvethe differential equations? ^ of Assumptionsare also important to the success a simulation' Assumptions and additional assumpthe were madeby the p.og.u--"iwhen developing mod91, deviation the tions aremadeby userJ.Fo, "*u-ple, a programthat calculates standard the sample size must be from an unbiased estimating equation assumesthat = 30 is often considered sufficiently lafge to validate ihe estimate.A value of N 24-hr storm being used' as minimal. For a TP-149 (Chapter 15) application, is a readingand by assumed the method?No computerprogram shouldbe usedprior to and becomingawareof the madeUy ttre programmer undershndingthe assumptions implicit in the hydrologicprocessthat was programmed' assumptions 'ih" lor cpst of simulationcan reiult in unnecessary runs and may enticeusers the information originally bstantively to purposeofruns) and working the plan can

r alproximate time and in a single useasa guideduringa simulationproject. combining severalinvestigations Some of the modelsavailable run is anotherway to conductarrefficient simulation. allowthis.Forexample,TR-20(seeChapter24)allowsthegenerationofflood

asmarr isonrY dtli|l,""i:#THfftime limits to monetary

514

21 CHAPTER

MODELING TO INTRODUCTION HYDROLOGIC

'

to from severalstormsat once.It is often desirable generatethe2-,5-, hydrographs location. at flood discharge a singlewatershed 10-, 25-, 50-,100-, and 500-year is able to generatefar more output than the hydrologistcan The computer Most modelsincorporateoptionsallowingthe userto specifyoutputquantity. analyze. shouldbe madeof which specific In additionto controlling outpul, a predetermination performed. A tabulation of key output data can be developedto analyseswill be compile and evaluatetrends (and make coursgcorrections)after eachrun. Because many opportunitiesexist in deterministichydrologyis about 80 percent acbounting, water budgelbalances.If the total rechargeto an aquifer is simulation for assessing less than the total outflow and withdrawls, but simulatedwater tables are rising, a shouldbe made.Writing important conclusions checkof input and model parameters runs helpsdocumentthe study and guiderevisions on the printed output of simulation in future runs. Documentation of simulation studies is generally deficient in practice. The record should communicatethe findings in a way that provides a later reviewer for made,andreasons each generalunderstanding the work plan followed,decisions of of made,provide samples the input stateassumptions should run. The documentation the statehow sensitive resultsare and output,explaininput preparationrequirements, and documentreasonsout-of-rangeparameand assumptions, to parameterchanges ters were accepted. Documentationis an ongoingand continual task. It is especiallycrucial if the model will be employedin regulatory proceduresor litigation. A comprehensive process would6: documentation Include an outline descriptionofthe problem being studied. and methodsused. Identify the equations,techniques, Demonstratethe model's validity to this problem. Discuss code. the usedin the code and in preparingthe input. Include all assumptions List publishedor known limitations or rangesof the applicability of the model. the 7. Characterize uncertaintiesin the model; describesensitivitytests. and data setsused. 8. Describeparameters 9. Statethe regulatory or legal criteria incorporatedin the model. 10. Describe the verification, whether with test data or analytical solutions. or 11. Include a narrative descriptionof the results,indicating any unexpected unusualoutcomes. 12.Presentany other details deemedrelevant. 13.Discusithe modelused. 14.Documentchansesmade in the model code. 1. 2. 3. 4. 5. 6.

Models Simulation of Components Hydrologic

for Numerousmathematicalmodelshavebeen developed the purposeof simulating A phenomena and systems. generalconceptualmodel including various hydrologic othersare described is mostof the importantcomponents shownin Fig. 21.1;several Irnportedwater in the lorygl leJtcould be input to reservoiror groundsubsequeht$.

SIMULATION 21,1 HYDROLOGIC

515

System outflow

Snow accumulation and melt Depression storage Overiand flow direct runoff

System outflow

SYsteminflow

System rntlow

Figure2l..l.Componentsofasurfaceandsubsurfacewaterresourcesystem. allocationson water storageor channelflow, or it might be guideddirectly to water The unnecessary. routing of or the far rlgtriif either storage distribuiion were deemed by flow could be accompdbhed simple lumped parameter channelflo* o, overland of for flow eq-uations discretesegments or techniques, solutionsof the unsteady-state algorithms and the channelcould be used.In other words,the selectionof techniques dependson the d9ere9of refinementdesiredas output to fepresenteachcomponent is justifled and also on knowledgl of the system.A distributedparameterapproach in described of Components modelsare information is adequate. only when available 22-25. ChaPters

Simulation DataNeedsfor Hydrologic

requiresa data inventoryaspart of the The simulation6f all or part of a water system (90 percentor mofe) are initial planningproc"ts. Most modelinput datarequirements or obtainedfrom engineering or map or tield ariaitable, canbe empiricallydetermined encompasses handbooksand equations.A generallist of data inventory topics that most hydrologic economicmodelingneedsfollows' Characteristics A. Basin and Subbasin times of concentration' l, Lagtimes, travel times in reaches, 2.Contributingareas,depressions,meanoverlandflowdistancesandslopes' \--

516

CHAPTER 21

INTRODUCTION HYDROLOGIC TO MODELING

3. Designstorm abstractions: evapotranspiration, infiltration, depression, and interception losses.composite curve numbers,infiltr_ationcapacitiesand parameters, indexes. @ 4. Land-usepractices,soil types, surfaceand subsurface divides. 5. Water-usesites for recreation,irrigation, flood damagereduction, diversions,flow augmentation, and pumping. 6. Numbering systemfor junctions, subareas, gaugingand precipitation statrons. 7. Imprevious areas, forested areas, areas between isochrones, irrigable acreages. B. ChannelCharacteristics 1". Channelbed and valley floor profiles and slopes. 2. Manning or Ch6zycoefficients variousreaches, hydrhulicor field data for or from which thesecoefficientscould be estimated. 3. Channeland valley cross-sectional data for eachriver reach. 4. Seepage information; channellossesand baseflows. 5. channel and overbank storagecharacteristics, existingor proposedchannelization and leveedata. 6. Sedimentloads,bank stability, and vegetative growth. c. MeteorologicData 1. Hourly and daily precipitation for gauges or near the watershed. in 2. Temperature, relative humidity, and solar radiation data. 3. Data on wind speedand direction. 4. Evaporationpan data. D. WaterUse Data 1. Flows returned to streamsfrom treatmentplants or industries. 2. Diversionsfrom streamsand reservoirs. 3. Transbasin diversionsfrom and to the basin. 4. Stream and ditch geometricpropertiesand seepage characteristics. 5. Irrigated acreages irrigation practices,including water useefficiencies. and 6. Crop types and water consumptionrequirements. 7. Pastconservationpracticessuchas terracing,insfallation of irrigation return pits, and conservationtillage. 8. Stock wateringpractices. 9. Presenceand types of phreatophytes stream valleys and along ditch in banks. E. StreamflowData 1. Hourly, daily, monthly, annual streamflow data at all gauging stations, includin$ statisticalanalyses. 2. Flood frequencydata ani curvesat gaugingstations,or regionalcurvesfor ungauged sites,preferablyon an annual and seasonal basis. Flow duration data and curves at gauging stations (also any synthesized _3. data for ungauged areas). 4. Rating curves; stage-discharge, velocity-discharge, depth-discharge curvesfor certain reaches.

SIMULATION517 21.1 HYDROLOGIC $, Flooded area curves. 6. Stageversusarea flooded' curves' frequencY 7. Stageversus basis. 8. Stageversusflooi damagecurves,preferablyon a seasonal curves' 9. Hydraulic radius versusdischarge sitesas fractions of gaugedvalues. 10. Sireamflowsat ungauged ll. Returnflows as fractionsof water-useallocationsdivertedfor consumptive use. distributionsof allocationsto users' 12. Seasonal 13. Minimal streamflowto be maintainedat eachsite' 14.Masscurvesandstorage_yieldanalysesatgaugedsites' F. Design Floods and Flood Routing temporaland spatialdistributionand l. Designstormand flood determination; mtenslty. 2. Maximum regional stormsand floods' to verificationof flood routing techniques be usedand neces3. Selection.and sary routrng parameters. during designfloods' 4. Baseflow estimates 5. Availablerecords of historic floods' G. ReservoirInformation L. List of potential sitesand location data' 2. Elevation-storagecurves. 3. Elevation-area curves. 4. Normal, minimal, other pool levels' loss 5. Evaporationand seepage data or estimates' 6. Sediment,dead storagerequirements' 7. Reservoireconomiclife. 8. Flood control operatingpolicies and rule curves' weir and outlet equations,controls. 9. Outflow characteristics, recreationbenefit functions' 10. Reservoir-based L1. Costsversusreservoirstoragecapacities' and benefits. of 12. Purposes eachreservoirand beneficiaries

Assessments Nonmodeling
can be the After researching availabledata and information, the needfor simulation model can be If assessed. a decisionis made to proceed,the appropriatesimulation planned,and data prepared' a selected, sequence Transformationof raw data into usabieform doesnot alwaysrequire a simulacan assessments for tion model.Much of the usualinformationneeded waterresources in microcomputer be preparedby hand or by using analytical proceduresavailable format.Typicalnonmodelinganalysesincludethefollowing: 1. Identify water-user groups and all basin sites for hydropower producirrigation, flood damagereduction from tion, reservoir-based-recreation,

518

CHAPTER 21

INTRODUCTION HYDRoLOGIC TO MODELING

reservoircapacity,industrial and municipal water supply,diversions,and flow augmentation. t Compile annualand seasonal and streamflows flood recordsat eachgauged site for the period of record at eachsite. 3. Determine the fraction of the allocation to each consumptiveuse that is assumed return to the streamat eachuser site in the basin. to of 4. Perform frequencyanalyses annual streamflowand flood valuesat each gaugesite in the basin. losses. and seepage Determinefor eachreservoirsite the eVaporation meanprobabilitiesto be usedin the firm and secondaryyield analy6. Select
SES.

7. Developflood peak probability distributionsat eachpotential flood damage center or reachin the basin. 8. Determinethe fraction of water to be allocatedduring eachperiod to each water-usesite in the basin. 9. Determine existing and proposed hydropower plant capacitiesand load factors. to 10. Identify any minimal allowablestreamflows be maintainedfor flow augreachin the basin. mentation at eachflow augmentation 11. Specify any maximal or minimal constraintson any of the annual or seaor sonal water allocations,storagecapacities, target yields. or minimal dead storage,active stor12. Specify any constraintson maximal age, flood control storage,or total storagecapacitiesat any or all of the reservoirsitesin the basin. 13. Determine annual capital, operation, maintenance, and replacement (OMR) costsat eachreservoirsite as functions of a rangeof total reservoir 'capacities or scalesof development. 14. Determinebenefitsas functions of energyproduced. L5. Determineannual capital and OMR costsat eachhydropowerproduction site as functions of variousplant capacities. 16. Determine benefit-loss functions for a variety of allocationsto domestic, commercial,industrial, and diversionuses. 17. Determine short-run lossesas functions of deviations (both deficit and surplus)in plannedor target allocationsto user sites. 18. Developbenefit functions at eachirrigation site in the basin.This analysis requiresinformation on the area of land that can be irrigated per unit of water allocated,the quantitiesof eachcrop that can be producedper unit areaof land, the total fixed and variablecostsof producingeachcrop, and the unit"pricesthat will clear the market of any quantity of eachcrop. 19. Developflood-damage-reduction beneflt functions at eachpotential flood site. This analysisrequiresrecords of historical and/or simulated damage floods, channel storagecapacities,and flood control reservoir operating policies. 20. Developreservoir-based recreationbenefitfunctions at eachrecreationsite in the basin.

SIMULATION 21.2 GROUNDWATER

519

21.2 GROUNDWATER SIMULATION Digital simulation models are used in a different manner to study the storageand movementof water in a porous medium. Distributedrather than lumped parameter models are used to imitate observedevents and to evaluatefuture trends in the the The equationsdescribing of and development management groundwatersystems. poroui mediumwerederivedin Chapter18 andmodelingof regional flow of waterin a in was discirssed Chapter 20. This section deals primarily with techniques systems used in solving the hydrodynamicequationsof motion and continuity, followed by of (2) of brief discussions (i) typical input requirements, techniques calibrating and (3) the sensitivity of groundwatermodels to parameter verifying the models, and modelis also of changeslAnexample the calibrationand applicationof a groundwater provided.

ModelTypes
Groundwaterstudiesinvolve the adaptationof a particular code to the problem at hand. Severalpopular public domain computer codesfor solving various types of becomemodelswhen flow problemsare listedin Table2L2.The codes groundwater geometryand to the codeby inputting the system beingsiudiedis described the system (aquifer and flow field parameters,initial and boundary known internai operandi conditions, and water use and flow stressesapplied in time to all or parts of the system).Codes have emerged in four general categories:groundwaterflow particle tracking codes,andaquifer testdata analysis codes,solutetransport codes, programs.to Groundwaterflow codesprovide the user with the distribution of headsin an aquifer that would result from a simulated set of distributed recharge-discharge any two points From Darcy's law, the flow passing at stresses cells or line segments. differential. The codes are used to model both can be calculated from the head confined and unconfinedaquifers.Eachcan be structuredto model regional flow, or and flow in proximity of a singG well or wellfield. Steady-state transientconditions and be barriers,full or partially penetratingstreams Boundaiiescan canbe evaluated. By heador constantgradientperimeters. application lakes,leaky zones,or constant after solving can velocitiesof groundwater be determined the of Darcy's 1aw, seepage for the head differentials. and velocitiesareknown, the advection,dispersion, seepage When groundwater can be modeled.Solutetransportmodelsbuild on in changes concentrationof iolutes groundwaterflow modelsby the addition of advection,dispersion,and/or chemical due or If reactionequations. the chemical,dispersion, dilution concentrationchanges particle tracking codesmodel transport by m groundwaterflow are not important, methodthan solutetransportmodelsto track the path advectionand providean easier moveunderthe influenceof headdifferentials.Aquifer that and traveltimes of solutes test data programsprovide userswith computersolutionsto many of the hand calculations (Ctrapter li) neededto graph and interpret aquifer test data for determining aquifer and well Parameters'
I

\.

520

CHAPTER 21

INTRODUCTION HYDROLOGIC TO MODELING

TABLE 21.2 GROUNDWATER MODELING CODES

Acronym code for PLASM MODFLOW AQUIFEM-1 GWFLOW GWSIM-II GWFL3D MODRET SUTRA RANDOMWALK MT3D AT123D MOC HST3D FLOWPATH PATH3D MODPATH WHPA TECTYPE PUMPTEST THCVFIT TGUESS

Description
Groundwater flow models Two-dimensionalfi nite difference Three-dimensionalflnite difference T!wo-and three-dimensionalfinite element Package of1 analytical solutions Storageand movementmodel' Three-dimensionalfinite difference from retention ponds Seepage Solutetransooftmodels Dissolvedsubstance transport model Two-dimensionaltransientmodel Three-dimensionalsolutetransport Analytical solution package Two-dimensionalsolutetransport 3-D heat and solutetransport model Particletrackingmodels Two-dimensionalsteadystate Three-dimensional transient solutions Three-dimensional transient solutions Analyticai solution package Aquifertest analyses Pump and slug test by curve matching Pumping and slug test Pumping and slug test c Specifi capacitydetermination

Source

ru.sws
USGS MIT IGWMC TDWR TDWR USGS USGS ill. SWS EPA DOE USGS USGS SSG Wisc GS USGS EPA

1971 1988 r979 1975 198 1 r991 1992 1980 1981 1990 1981 t978 r992 1990 1989 t991 1990 1988 1980 1989 1990

ssG
IGWMC IGWMC IGWMC

Note: IGWMC : International Groundwater Modeling Center; Ili. SWS : Illinois State Water Survey; SSG : Scientific Software Group; EPA : Environmental Protection Agency; USGS : U.S. Geological Survey; Wisc. GS = Wisconsin Geological Survey; MIT : Massachusetts Institute of Technology; TDWR : Texas Department of Water Resources;DOE : Department of Energy.

SolutionTechniques
With few exceptions,the hydrodynamic equations for groundwaterflow have no analyticalsolutions, groundwater and modelingreliesonfinite-dffirence and,finiteelementmethodsto provide approximatesolutionsto a wide variety of groundwater problems.The choice of method is normally driven by the systemto be modeled. Other numericalmethodsincludeboundary integral methods,integratedfinitedffirence methods,and analytic elementmethods. Thesesolutions,as with streamflowsimulation models,are facilitated by first subdividingthe region to be modeledinto subareas. Groundwatersystemsubdivision depends more on geometriccriteria and lesson topographiccriteria in the sense that the region is overlaidby a regular or semiregular pattem of node points at which (or betweenwhich) specificmeasures aquifer and water systemparameters input of are and other parametersare calculated.Approximate solutionsof simultaneous linear and nonlinearequationsare found by making initial estimates the solution values, of testingthe estimates the equationsof motion and continuity, adjustingthe values, in andfinally accepting minor violationsin the basicprinciplesor making further adjustments of-the parameters an orderly and convergingfashion. in

SIMULATION 21.2 GROUNDWATEFT

521

r '

or of Theorderly solutionof finite differenceanalogs the steady-state unsteadypartial difierential equation of motion for flow of groundwaterin a confined stare methods'An eady relaxaquifer or an unconfinedaquifer is obtainedby rela_xation For by of the equationis discussed Jacob.T two-dimensionalproblems, ation solution and by (ADI) methoddeveloped Peaceman the iterativealternating-direction-implicit is often adoPted. Rachfords prickett and Lonnquisteused the ADI techniqueto calculate fluctuations in at watertable elevations all nodesin an aquifermodelby proceedingthrough time in from a known initial state.Their modelis computationallyefficient small increments and readily appliedand is particularly attractivefor usewith problemsinvolvingtime are variablesind nu-"tous nodes.The primary aquiferparameters the permeability constantover the aquifer plan, result in a which, if assumed and storagecoefficient, and homogenJous isotropigcondition.For thosefamiliar with relaxationmethods'the (SOR)methodshavehad application over-relaxation Gauss-seidelandthe successive equations. in solvingdifference as modelsmay be classified spatialand temporal' Spatial system Input to groundwater thicknessdata over the initial oi projectedwater table maps,saturated input inciudes contourmaps,transmissivitymaps,regionalvariationsin storage ."gion, land surface coefficients,locationsand typei of wells and canals,locationsand types of aquifer actualor net pumpage both lateral and vertical, a nodecoordinatesystem, boundaries percolationand rechargerates for precipitation and other appliedwaters,logs rates, of drilled wells, geologicstratigraphy,and soil types and cropping patterns. Time-dependenidatarequirementsfor aquifer models involve principally the using a rangeof time incrementsfor suchvariablesas formulation of ti-" schedules, groundhydrographs, canaland streamflow precipitationhyetographs, pumpingrates, variablessuchasthe timing of added rates,and development waterevapotranspiration can apply only Becauseeachtemporal schedule wells or other systemcomponents. requirementsare also particular iubset of node positions,the time-dependent to a spatial. aquifer modelsrequire reliable estiIn addition to the listed input parameters, of percentages waters in the land phasethat actually percolateto the matesof the aquifer being modeled.Theseestimatescan be basedon knowledgeof the physical involved in unsaturatedflow through porous medium but are most often processes of that are modifiedduringthe calibrationphase the obtainedasjudgmentparameters stated,the lateral movementand the changesin piezometeror simulation. Simply (withdrawalratesor stresses watertable leneli are easilymodeledif the node-by-node by are aie known. The latter parameters governed the complexmoverates) recharge ment of water in the unsaturatedsoil zone and by the random precipitation and lies systems use consumptive patternsof the region.The art of modelinggroundwater to evaluatetheseparameters' in the ability

DataRequirements

Calibration
involvedin parameter modelcalibrationremovessomeof the guesswork Groundwater knowledgeof basedon available of Severalcombinations parameters, determination. physical syqtem,are testedin the model during a period for which records are the

I L,

522

21 CHAPTER INTRODUCTIoN To HYDRoLoGIc MoDELING availabi-e. Simulatedresultsare then comparedwith historical events.After structuring the model,calibrationis achieved operatingthe model during the studyperiod by by imposinghistoricalprecipitationamounts, evaporation and evapcanaldiversions, otranspirationrates, streamflowsand stream levels, pumping rates during known periods, and other stresses the aquifer. Calibration is achievedafter the flow, on storage, and other parameters lirnits to produce havebeenadjustedwithin reasonable the best imitation of recordedevents.

CaseExample
modelingin A typical finite-differencestudyinvolvingsurface waterand groundwater central Nebraskawas performed by Marlette and Lewis.tt The region involved is shownin Fig. 2I.2.In additionto the surface irrigation system represented the by severalcanalsand laterals,over 1200wells withdraw waterfrom the aquiferbetween the PlatteRiver and the Gothenburg DawsonCounty canals.The aquiferrecharge and and withdrawal amountsas percentages precipitation, snowfall, pumped water, of deliveredcanal water, evaporation,and evapotranspiration were estimatedusing a mix of judgment andphysicalprocessevaluations. The resultingsetthat producedthe wells showninFig.2l.2 bestcomparison with recorded events the six observation at is summarized Table 21.3. Samplesof the comparison in betweenrecordedand simulatedwaterlevelsin the DawsonCounty study during aZ-yearcalibrationperiod are shownin Figs. 21.3 and2l.4. The Prickettand Lonnquistmodelwasappliedin the DawsonCounty study.The by storage coefflcientfor this unconfinedaquiferwas established calibrationtrials as 0.25 and the adopted permeability was 61 mlday. Other trials were made using variouscombinations S andK, with S rangingfrom 0.10to 0.30andwith Kranging of between4I and I02 mlday. As with most unconfined aquifer models,water table elevations were most sensitive fluctuationsin the storasecoefflcient.Fisure 21.5 is to

tt'iorX
\F 'q;

Figure 21.2. Grid coordinatesfor Dawson County, Nebraska,aquifer model. o : observationwell.

SIMULATION 21.2 GROUNDWATER ALLOCATIONS FOR CRITERIA WATER RECHARGE TABLE21.3 ADOPTED .. AQUIFER THEDAWSON COUNTY OVER

523

component System
Rainfall Snowfall Pumpedwater Delivered canal water

and appliedamounts Allocation Recordeddepth if daily amount exceeded 0.25 cm at all nodes 25Voofrecorded depthsat all nodes Averagerate of 50 l/sec at-all well nodes during irrigation seasons Recordeddaily rates,applied to land surfaceone node laterally uphill and two nodesdownhill from canal Observeddaily lake evaporationdepth at all marsh and water surfacenodes I25Voof daily lake evaporation,applied at all alfalfa nodes

recharge/withdrawal Aquifer ot as a percentage amount applied 30 30 50 30 100 15

Evaporation Evapotranspiration

o o
6

B
d

0.999

0.998
A M J J A S O N D J F M A M J J A S O N

1970

1971

at waterlevels observation and Figure2L.3. Simulated recorded w e l l Di n F i g . 2 1 . 2 .:I8 2 ; i : 3 7 . a typical summary of the calibration results at a single observationwell located at F Position in Fig. 21.2. After vefification, the Dawson County model was applied to investigatethe Included among the schemes schemes. short-terminfluenceof severalmanagement involving the completeremoval or shutdownof the surfacewater were investigations canals,and other testsin which isolatedcanal contributionsto rechargewere determined by operatingthe model with singlecanals and comparingresults with water table fluctuationsfor identicalconditionswith all canalsremoved.Many other applicationsof the model are possible.This particular study revealedthat rechargefrom

524

CHAPTER21

MODELING TO INTRODUCTION HYDROLOGIC

o a
d

0.999

0.998 A M J J A S O N D J F M A M J J A S O N 1970 l97l

Figure 21.4. Simulated and recorded water levels at observation well F in Fig. 21.2. I : 97; i : 42.

lu3

P (J

o
q) 'F1

(.1

H 0)

i3 -

702

t9'70

r97|

permeability varying and with changes constant Figure 21.5. Water-level Well storage coefficients. 4 Fig. 21.2;K : 6Im/day;I : 97;J : 42. contributesto the waterbalancqof the aquiferbut is not the the existingcanal system from precipitationandfrom the dominantfactor\n the shortrun. The naturalrecharge Platte River accountfor the long-term water table stability in the region.

PROTOCOL SIMULATION 21.3 HYDROLOGIC

processis not new The useof hydrologicsimulationas a tool in the decision-making form. A model is a and but is of a different, more sophisticated more encompassing that permitsthe evaluationand-rmanipof rgplesentation an actualor proposedsystem

PROTOCOL SIMULATION 21.g HYDROLOGIC

525

useof ulation of-manyyearsof prototypebehavior.This is the featurethat makesthe of eventhe largest, thesetooii so attractiveand hoids suchpotential for the analysis so well It mostcomplexsystems. is alsothe prin-ipal featurethat makesthis approach suitedto water resourcessystemplanning and analysis' models, Apart from the useof cot1entionalhandmethodsand someelementary however,as judgment.This is changing, planninghastraditionally beena practiceof that permit the analysisof large numbersof alternaquantititive tools are developed is elementof the process, not ruled out but is tives and plans.Judgment,a-nessential to those in the planning through new insights that were not available strengthened professiona few years ago. "What if ?" Plannersare conti;ually required to anticipatethe future and ask suchas simulation and "What's best?" questions. Quantitativeplanning techniques, alternativesfor lesscost than can provide detailedinformation aboutmore planning principally at of uny oth", approachavailable.Development thesetools has occurred universitiesand federal agencies.

Models and Optimization CombinedUse of Simutation

at this An important secondtype of quantitativeplanningtool shouldbe mentioned information to selecta to modelsare designed utilize llmiPd system point.Screening Hence objectiveor setof objectives. for bestplan u*oni many alternatives a specified called,are orientedtoward modJs, or optimizationmodelsasthey are often screening plan formulation rn contrast to th Simulation models are suited to de reliable information on which to bi "If the modelsaddress question, out tion models,on the otherhand,ask, look like after we arr will the system special merits of each, these two Completeder planningtechnologies. ing, oPtimization,and simulation r Final designvaluesshouldbe using a and to the systemel-ements operatinga detailedsimulation model over time while at and/or streamflows, precipitation-amounts of sequence known or synthesized benefiisovertime for flood control, reservoirand streamthe sametime accumulating siderecreation,wateryields,strean and anY other factorsnot consider model. Severalsimulationruns wi resultin a plan that bestmeetsthe c by generated coilventionalmethods' inforResultsof optimization modelswill provide readily obtainedand useful in orderto testthe mostpromismation for initiating more refinedsimulationanalyses and arrive at final plans for the design,construction,and operationof ing measures Even thor a water resourcesystem. regardingboth the de for decisions postoptimizationsimulation is rec often requiredin prel assumptions

526

CHAPTER21

MODELING TO INTRODUCTION HYDROLOGIC

alternativesowing to time and cost limitaof for preliminary screening development currenttime and sizelimitations evolves, of tions.Unlessa new generation computers by do not allow screening simulating all iilternativesunlesssubstantialsacrificesin followedby detailedsimularealismare made.For the present,preliminary screening to tion appears be the most effectivemeansfor arriving at optimal waterdevelopment and management Plans.

MODELS SIMULATION OF 21.4 CORPS ENGINEERS

a In 1964,the U.S. Army Corps of Engineersdeveloped specialtybranch locatedat Center(HEC) in Davis,California. The facility provides the HydrologicEngineering researchresultsto practical needsof the Corps fleld a centerforipplying academic to providestraining and technical assistance governoffices.In addition, the center hydrology,hydraulics,and reservoiroperations. in menr agencies advanced Over the years, a large number of analytical tools were developedat HEC. river/reserof the Table 2I.4 summarizes computerprogramsin categories hydrology, operations, stochastichydrology, river/reservoir water voir hydraulics, reservoir
TABLE 21.4 HEC WATER RESOURCECOMPUTERPROGRAMS Name Date of latest version HydrologyModels 1980 September

Purpose
Simulatesthe precipitation runoff river process any comPlex in

HEC-I, Flood HydrograPhPackage

Basin Rainfall and SnowmeltComputation

July 1966
many subbasilsof a river basin using gaugedata and weightings (includedin HEC-1).

Unit Graph and HydrographComputation

Iuly 1966

Unit Graph Loss Rate OPtimization

August 1966

HydrographCombining and Routing

August 1966

Computessubbasin interception/infiltration, unit baseflow, and hydrographs runoff hydrograph(included in HEC-1). Estimatesbest-fitvaluesfor unit graph and lossrate Parameters from given precipitation and runoff (included in subbasin HEC-1). at Combinesrunoff from subbasins and routes confluences througha river hydrographs network using hydrologicrouting methods(includedin HEC-1).

MODELS SIMULATION 21.4 CORPSOF ENGINEERS TABLE21.4 (Continued)*

527

Name
Streamflow Routing OptiffIization

Dateof latestversion
July 1966

Purpose
Estimatesbest-fit valuesfor hydrologic streamflow routing parameterswith given upstream' downstream,and local inflow (includedin HEC-l). hydrographs gravitY and Computes seepage, pressue flow, pumping and for overtopping discharges Pond areasbehind leveesor other flow obstructions.Main river elevation and ponding area elevation-area-capacitydata are usedin computingdischarges. Simulatesthe precipitation runoff processfor a single, usuallY urban, basin for manY Yearsof hourly precipitation data. Simulates qualitY of urban runoff and dry weather sewageflow. EvaluatesquantitY and qualilY of overflow for combinations of sewagetreatmenl plant storage and treatment rate.

Interior Drainage Flood Routing

November1978

Storage,Treatment, Overflow, Runoff Model (..STORM")

JuJy1976

hYdraulics River/reservoir
HEC-2, Water SurfaceProfiles August 1979

super-critical. AnalYzes for allowableencroachment a given rise in water surface. Gradually Varied UnsteadyFlow Profiles Iarnary L976 Simulatesone-dimensional, unsteady,free surfaceflows in a branching river network. Natural and artificial cross sectionsmaY be used. Uses an exPlicit centered difference computational scheme. Computestables of hYdraulic elementsfor use bY the GraduallY Varied UnsteadYFlow Profiles or other programs.InterPolates values for area, top width, n value, and hydraulic radius at evenly spacedlocations along a reach.

Geometric Elements from Cross Section ("GEDA') Coordinates

June 1976

534

CHAPTER21

TO MODELING INTRODUCTION HYDROLOGIC

PROBLEMS
in are 21.1. Simulation and synthesis treatedseparately Chapters'22and 23. List the most of distinguishingcharacteristics eachmethod and give an exampleof each. 21.2. Listatleastthreereasonsmanyofthedevelopedmodelsoftherainfall-runoffprocess might not be usedby hydrologists. 21.3. You are askedto determinea designinflow hydrographto a reservoirat a site where I,!st generalstepsyou would take as a hydrolno recordsof streamfloware'available. ' ogist in developingthe entire designinflow hydrograph.

REFERENCES
ExperimentStation,"Models and MethodsApplicableto Corps of 1. U.S. Army Waterways PaperH-74-8, National TechnicalInformation Engineers Urban Studies,"Miscellaneous Service,Aug.1974. in 2. GeorgeFleming, ComputerSimulation Techniques Hydrology. New York: American Elsevier,1975. "HEC-I Flood HydrographPackage,"Users and ProU.S. Army Corps of Engineers, grammers Manuals,HEC Program723-X6-L20I0, Jan.1973. "Digital Simulationin Hydrology:StanfordWater+ . N. H. Crawford and R. K. Linsley,Jr., shedModel IV," Department of Civil Engineering,Stanford University, Stanford, CA, Tech.Rep.No. 39, July 1966. 5 . A. J. Friedrich, "Managementof ComputerUse in SolvingEngineeringProblems,"U.S. Army Corps of Engineers,Hydrologic EngineeringCenter,Davis,CA, 1979. Council, Ground WaterModels-Scientific and RegulatoryApplica6 . National Research Mathetions. Water Scienceand TechnologyBoard, Commissionon PhysicalSciences, D.C., 1990. Washington, National AcademyPress, matics,and Resources, ed.)' New Hydraulics (HunterRouse, 7 . C. E. Jacob,"Flow of Groundwater,"in Engineering York: Wiley, 1950. "The Numerical Solution of Parabolicand Elliptic and 8 . D. W. Peaceman N. H. Rachford, Indust.Appl. Math.3' (1955). J. DifferentialEquations," Soc. "selectedDigital ComputerTechniques Groundfor 9 . T. A. Prickett and C. G. Lonnquist, Ilinois StateWaterSurveyBull. No. 55,197I. Evaluation," waterResource 1 0 . D. R. Maidment, (ed.), Handbookof Hydrology.New York: McGraw-Hill, 1993. "Digital Simulationof Conjunctive-Useof Groundwater 1 1 . R. R. Marlette and G. L. Lewis, in DawsonCounty, Nebraska,"Civil EngineeringReport, University of Nebraska,Lincoln, 1973. 12. W. K. Johnson,"Use of SystemsAnalysis in Water ResourcePlanning," Proc- ASCE J. Hyd. Div. (1974). Planning and Design CaseStudiesin Modeling 1 3 .R. deNeufvi[e and D. H. Marks, Systems Cliffs, NJ: Prentice-Hall,I914. . Englewood Optimization and EvaluatioiT 14. D. P. Loucks, "stochasticMethodsfor Analyzing River Basin Systems."Cornell Univer', Center,Ithaca, NY, Aug. 1969. and Marine Sciences sity Water Resources MA: HarvardUniversity System*Cambridge, 1 5 . A. Maass,(ed.),Designof WaterResources Press, 1962. t6. A. F. Pabst, "Next Generation HEC Catchment Modeling," Proceedings,ASCE HyHydrology,SanFrancisco,CA, July 25-30' on draulicsDivision Symposium Engineering 1993.

Chapter22

Time":?J5: Hydrologic

r Prologue
The purposeof this chaPteris to: . Show how time seriesanalysisis used for generatingsynthetic hydrologic records. of aspects hydrologic . Give definitionsof termsusedto describe stochastic the series. including masscurve analy. Introduce fundamentalsof streamflowsynthesis and sesequences, serial-dependent sis, random generationof sequences, quences havingprescribedfrequencydistributions' Time-seriesanalysisof hydrologicvariableshas becomea practical methodolof ogy for generatingsyntheticsequences precipitation or steamflowvaluesthat can ui-usedlor u tung" of applicationsfrom filling in missingdata in a gaugedrecord to, extendingmonthiy streamflowrecordsl, and from analyzinglong-term reliability of floodsor snowmeltrunoff quantito or yields of-watersheds2 reservoirst'o forecasting Thesesynthetic hydrologytechniques iies from syntheticprecipitation sequences.5 widein augmentttre simulationtools described Chapter21. Both have,experienced generationof a Synthesisinvolves the spiead use by hydrologistsand engineers.6 or ,"qu"n"" of valuesfor somehydrologicvariable(daily, monthly, seasonal, annual). for are The techniques most often applied to produce streamflowsequences use in generaterainfall sereservoir designor operation studiesbut can also be used to be quences that can subsequently input to simulation models' of to If historioalflows iould be considered be representative all possiblefuture during its lifetime, there would be little variationsthat someproject will experience for The hiitorical record is seldomadequate predicting needfor synthetichyOroiogy. patternis unlikely to recur, future eventswith certainty,however.The exacthistorical of sequences dry years (or wet years) may not have been as severeas they may beiome, and the singlehistorical record gives the planner limited knowledgeof the magnitudeof risks involved.

536

CHAPTER22

TIME SERIESANALYSIS HYDROLOGIC

particularly if Syhthesisenableshydrologiststo deal with data inadequaci.es, Short historical records of hydrologic record lengths are not sufficiently extensive. hydrologicsynto variables*"fu u, streamfloware extended longer sequenceslrsing scienceknown as operational hydrology'1 of thesisand other techniques the broad either preservethe statisticalcharacterof the historThesenew, syntheticsequences ical records or follow a prescribedprobability distribution, or both. When coupled provide hydrologistswith imthe with computer simulation techniques, techniques proved designand analysiscapabilities. The rnethodsdescribedin this chapterare basedon probability and statistics. The material presentedin Chapter 26 should be reviewed prior to studying this chapter.

HYDROLOGY 22.1 SYNTHETIC

techniquesare classifiedas (I) historical repetition methods, Hydrologic synthesis that historical recordswill repeatthemwhich assume suchas masscurve analyses, sevlesin as many end-to-endrepetitionsas required to bracketthe planningperiod; which assume such as Monte Carlo techniques, (2) random generationtechniques, events,any of which random,independent that the historicalrecordsare a numberof could occur within a definedprobability distribution; and (3) persistencemethods, such as Markov generationtechniques,which assumethat flows in sequenceare dependentand thit the next flow in.sequenceis influencedby some subsetof the are previousflows. Historical repetition or random generationtechniques normally flows for shortertime intervals flows. Successive applied only to annual or seasonal analysisby the Markov generationmethod. are usually correlated,necessitating of As with most subfields hydrology,a number of computerprogramsfor timeOne ofthe first, and havebeendeveloped. and seriesanalysis hydrologicdatasynthesis model HEC-4 Army Corps of Engineers wasthe U.S. one of the most widely app1i"d, Its use is limited, though, to synthesizing in (seeSection21.a) pubfished I97IJ in monthly streamflows a river reach.Other codess'e of sequences seriaitydependent Additional models and descriptionsof however. are avarlableto thi hy-drologist, theory and applicationsof time-seriesanalysisof precipitation and streamfloware detailedin a number of availabletexts and publications'10-l3

MassCurveAnalysis
by was devised Ripplla to investechniques One of the earliestand simplestsynthesis that the future His analysisassumes tigate reservoirptoragecapacityrequirements. in inflowsto a reseivoirwill be a duplicateof the historicalrecord repeated its entirety to many times end fo end as is necessary span the useful life of the reservoir. as Sufficient storageis then selectedto hold surpluswaters for releaseduring critical Reservoirsize selectionis easily accomperiods when inflows fall short of demands. pmfr"a from an analysisof peaks and troughs in the mass curve of accumulated Future flows can be similar, but are unlikely to be syntheticinflow versustime.15-17 produce identicalto pastflows.Randomgenerationand Markov modelingtechniques of, t-hat seqUences are difJerentfrom, although still representative historical flows.

HYDROLOGY 537 22.1 SYNTHETIC

EXAMPLE 22.1 Streamflowspast a proposedreservoir site during a 5-year period of record were, acre-ft.use Rippl's 8000,and 12,000 10,000,6000, 14,000, in respectively, eachyear method to determinethe size of reservoir neededto provide a yield of ,nur. Crrru" 9000 acre-ft in eachof the next 10 years. is of solution. A lo-year sequence syntheticflows,usingRippl's assumptions, equal to the historical record repeated shown in Table Z2.l.Inflows are set twice. rl FOR 22.1 STREAMFLOWS EXAMPLE22'1 TABLE
Flows (thousandsof acre-ft) Year Inflow Cumulative inflow

r 1 14

2 4 1 24

3 4 0 6 8 38 30

5 6 1 2 1 64 50

7 8 9 1 0 4 1 0 6 8 1 2 100 88 80 74

draft of 9000 acre-ft per year for 10 years.

90,000 80,000

^
I

70,000 60,000 50,000 40,000

4000 acre ft storagerequired Storagerequfued for 9000-acre-ft/Yr draft is maximum ,' , Cumulative draft, slopeof9000 acre-fl/Yr

30,000 . 20,000 10.000

1 2 3 4 5 Year 6

Eigarc 22.1 Mass curve for ExamPle 22.1: inflow; --- cumulativedtaft.

cumulative

538

CHAPTER HYDRoLoGIc 22 TIME SERIES ANALYSIS

RandomGeneratiOn
one method of generatingsequences future flows is a simplerandom rearrangeof ment of past records.If the streamis ungauged and recordsare not available, a probability distributioncan be selected and a sequence future flows that follow the of distribution and haveprescribedstatisticalmomentsis generated. Wheneverhistorical flows are available,a reasonable sequence future flows of can be synthesized first consultinga table of randomnumbers,selecting number, by a matchingthis with the rank-in-file numberof a pastflow, and listing thecorresponding flow as the first value in the new sequence. The next random numberwould be used in a similar fasion to generate next flow, and so on. Randomnumbershavingno the corresponding flows are neglected and the next randomnumberis selected. Table-B.3 in Appendix B is a table of uniformly distributedrandom numbers(eachsuccessive numberhas an equalprobability of taking on any of the possible values). illustrate To the use of Table B.3 in the random generation process, the first three yearsof a syntheticflow sequence could be generated selecting 53rd, 74th, and23rdfrom by the the list of past flows. Alternatively, the flows in 1953, 1974, and 1923 could also be selected the new randomsequence. as Most computershave random number generationcapabilitiesin their system libraries. Rather than storing large tables of numberssuchas Table B.3, successive random integersare usually generated the computer. by EXA]I/IPLE22.2 Annualflowsin Crooked Creekwere 19,000, 14,000, 21,000, 8000,11,000, 23,000, 1 0 , 0 0 0 , a n d 9 0 0 0 a c r e - f t , r e s p e c t i v e l y , f o,r2 ,e a 4s 5 , 6 , 7 , a n d 8 . G e n e r a t e a l y 3,r , 5-year sequence annual flows, O,, by matching five random numberswith year of numbers. Solution. Randomintegers between0 and 9 are generated from the computer. The Q, valuesin Table22.2 areselected from the eight given flowsby matching the respective year numberwith the random number.The digit t has no correspondingflow in the 8-year sequence, the next random number, 2, places so the 14,000-cfs flow in Year 2 in the first position of the synthetic 5-year sequence.rI
TABLE 22.2 DEVELOPMENT OF s.YEAR SYNTHETIC SEQUENCE Randomdigit
I o

Q(acre-ft)

2 3
I

5 6

2 5 8 I
I

Skip 14,000 11,000 9,000 19,000 8.000

544

CHAPTER22

TIME SERIESANALYSIS HYDROLOGIC

'

coefficient of the standarddeviation for the daily flow loga3.- a regression rithms within each month of record to the logarithm of the monthly total flow., the Given the calculatedstatistics, simulationof daily flows could be structured in the following manner: from Eq. 22'1,0. standardized variates 1. Generate 2. Use the logarithm of the monthly mean flow as an initial estimateof the mean of the logarithmic daily flows for the month' 3. Calculatethe standarddeviationof flow logarithmsby the previouslydetermined regression equation. 4. Apply the inverseof Eq. 22.1I, k--

?lz('
-:) .

'l-;
13 1

(22.r2)

variatesto flows,multiplying by the approprito transformthe standardized ate standarddeviationand addingthe mean. and 5. Add the differencebetweenthe total monthly flow generated the given given monthly flow, and repeatthe simulation. monthly flow to the 6. Multiply daily results of the secondsimulation by the ratio of the given monthly total. monthly total to the generated Each simulation techniquecommonly requiresmodificiationswhen applied to individual problems.Methodsoutlined thus far can be utilized as guidesin establishflows.Simulationof flows for a given station ing a procedureto follow in synthesizing means,and standarddeviations with serial correlation,skewness, hasbeenpresented the for rnaintained.When generatingstreamflowsequences an entire system, preserbetweenstationsbecomesa significantfactor. vation of cross-correlation are runoff sequences employedto determinethe capacSyntheticallygenerated flow magniities of reservoirsto satisfy specifieddemands.Individually generated of sequences flows. Hydrolotudesare uncertain, as are the syntheticallygenerated severalequally likd sequences probabilitiesof flowsby generating gistscanestimate of recurrences certain values.Herein lies one of the most of flowsand then evaluating useful applicationsof Markov generatingtechniques.

i Summary
as Hydrologicmodelingis often presented comprisingonly the deterministicmodels in described Chapters2I,23,24, and 25. The fully of the rainfall-runoff process equippedhydrologistincorporatesthe synthetichydrology models describedin this A chapterin the analysisand designof water resotrces systems. growing numberof basis of synthetichydrology and timeprojects are constructedor operatedon the sedesanalysiseachyear.

PROBLEMS

545

PROBLEMS
record in Example22'2 and detet22.1. plot cumulativeinflows versustime for the S-year yield of neededto provide a mine by mass curve analysisthe size of the r_eservoir What is,the maximumyield possible? 12,000acre-ft in eachoithe next 24years. of sequence synthetic 'r, t use the annual rainfall trom Table 26.2to generatea lO-year assumption' annual rain depthsfor Richmond using Rippl's masscurve curve methods'Use 22.3. RepeatProblem 22.2 wingrandom generationrather than mass digits of numberJfrom taUl'n.: and matchthesewith the last two i*o-Olglt.unaom in the yearnumbers Problem26.32.

22.4. RepeatProblem22.2wingrandomgenerationtogenefateal0-yearsyntheticsestandard
tr<

and qu"n"" of annual rain defths that has.a normal CDF with a mean 1 equal to that of ihe annual rain data from Problem 26'32' deviation follow a log-Pearson Repeat Problem 22.4 assumingthat the annual rain depths statisiicsusethe mean,standarddeviation,and skewofthe Type III distribution.For logarithmsof annualrain at Richmond' (c) Pearson using (a) normal distribution, (Uitog-nbrmal distribution' and distribution.

22.6. SelectagaugedStreaminyourgeographiclocationandprepareaquarterlymodel Type III

)a1

CanyouconvertthesimulationprobleminExamp|e22.4toalog_normaldistribution given in the example? simulation?What difficulties aie encounteredwith the data duration of rainfall data, flt a distributionto the time betweenstorms,and month. Preparea computerpfogram data coveringthe selected 20 yearsof recorded of storms' the to randomly generate times betweenstormsand the durations

22.8. Selectamonthofthunderstormactivityinyourregion.FrompublishedNoAAhourly storms'for

22.9. Flowsduring6yearsofrecordwereusedinsynthesizingthemasscurveshownonthe
following page. a.UseRippl'sassumptionandthegraphtodeterminethemissingmagnitudeofthe flow for the 12th Year. yield of 2000 acreb Determinethe reservoircapacityrequired to allow an annual acte-ftlYr' for ftlyr. RePeat 500 does this value relate c. Determine the maximum yieid possibleat the site. How statisticallYto the flows? how you would developa table of randomprecip22.10. Describewith words and equations of 4 in' and a standard itation depthsthat follow a normal distributionandhavea mean deviationof3in.Useyourmethodtocalculatethefirstthreedepths. is given below' use random of 22.LI. A sequence uniformly distributedrandom numbers annual rain depths that will follow a of generationto generatei 5_y"u, sequence in'' a standarddeviation F"u..on fype iU distribution and will havea mean of 25'8 of4'0in.,andaskewcoefficientof_2.2o.Randomnumberstobeusedare20,0I' 9 0 . 0 3 .a n d 8 0 . Type III to a Pearson 22.L2. Total July runoff from a basinis randomly distributedaccording the the is 10,000 acre-ft', standarddeviationis 1000 acre-ft' : distribution. The mean is 0.50. Start with Q1 coefficient skewis -0.6, and the lag-oneserialcorrelation of randomly selected 10,000and find nu" -oiJuu.kov-generated flows if a sequence and50 years' return periodsgives2, 100, 10,2,

546

CHAPTER22

HYDROLOGIC TIME SERIESANALYSIS 30,000

25,000

20,000
9l B

15,000

10.000

101112

Year Figtre 22.9 Mass curve.

REFERENCES
1 . U.S. Army Corps of Engineers,*HEC-4 Monthly Streamflow Sirnulation," Hydrologic EngineeringCenter, 197l. 2. R. M. Hirsch, "synthetic Hydrology and Water Supply Reliability," Water Resources Research, 15, no. 6, L979. v. "The Valueof Stochastic StreamflowModelsin Overyear J. R. M. Vogel,andJ. R. Stedinger, v. Research, 25, no. 9, 1988. ReservoirDesign Applications," WaterResources A Hydrologyin ReservoirOperation," J. Irrigation D. K. Frevert,et al., "Use of Stochastic ASCE,v. 115,no. 3, 1989. and DrainageEngineering, "An Evaluationof the Practicality and Complexityof SomeRainfall 5 . J. W. Delleur, et al., v. Research, 12, no. 5, 1976. and Runoff Time SeriesModels," WaterResources 6. J. D. Salas, et al., "Applied Modeling of Hydrologic Time Series," Water Resources Littleton, CO, 1980. Publications, "Operational HydrologyUsing Residuals," Hydraulics J. 1 . G. K. Young,and W. C. Pisano, Division,ASCE,v.94, no. HY4, 1968. "Applied StochasticTechniques, PersonalComputer Ver8 . U.S. Bureau of Reclamation, Division,Denver,CO, 1990. sion 5.2. User'sManual," Earth Sciences "SPIGOT, A Synthetic StreamflowGenerationSoft9 . J. C. Grygier, and J. R. Stedinger, ware Package," School of Civil and Environmental Engineering, Cornell University Ithaca.NY. 1990.

Chapter23

Models Simulation Continuous

r Prologue
The purposeof this chaPteris to: , Introduce and describecomputercodesavailablefor performing contrnuous simulation of surfacerunoff and streamflow' model. Presentin detail how one of the programs-the Stanfordwatershed "otnponentsof the hydrologiccycle' simulatesthe miscellan"ou, .Showthemajorsimilaritiesanddifferencesoftheleadingmodels. from are . Providea detailedcasestudyof how the modelparameters developed availableinformation. . Illustrate, by using the casestudy,the stepsinvolvedin calibratinga continuous model and verifying the results' . show how well the moleh are able to replicate gaugedstreamflows. with tools for estiin Simulation modelsdescribed this chapterprovide hydrologists in accounting tittt" for precipitation'direct runoff ' by mating streamflow continuously interflow, deeppercolation' baseflow' and streaminfiltration, evaportranspiration, simulationmodelstrack flow. During rain-free intervalsbetweenstormi, continuous deeppercolation,and baseflow, to evaporation, of the storage water.andits depletion until the next rain or snow eventoccurs' ThemodelsarebasedonthephysicalprocessesdescribedinChaptersl_14.As

yasdeterministiciools't1;:H':1;ffi ff;ff t: such,theycrassif :T:TLi""ll':'ff

rain and snow into runoff and streamflow' can be usedto or other similar procedures' r are then input to continuoussimulation in Table2l'l ate Severalof the continuoussimulationmodelsidentifiedearlier model, version IV (swM-IV)' is presented here.The Stanford watershed described indetailastypicaloftheothermodels.Manyoftheothersare,infact,basedon

MODELS 549 SIMUIATION STREAMFLOW 2g.1 CONTINUOUS in the SWM{V, and several simulate various componentsof the hydrologic cycle casestudiesof independent and comparestwo samemanner. Section 23.2 ptesents were {etermined and how the Stanford model studies,showinghow the parameters modelswere calibratedand applied to the problemsbeing assessed'

MODELS SIMUI-ATION STREAMFLOW 23.1 CONTINUOUS API Model

simulation of a This model was one of the earlieststructuredto give a deterr.ninistic of It was originally testedon watersheds 68 and continuousstreamflowhydrograph. g:i miz and must be calibrated^ each watershedto obtain a reliable method of r to givenin Fig. 23' 1' simulatint the streamflow.lA flow diagramshowingthe structureis ttie interrelations pertaining to .this .model of Four basic components describe precipitation index, streamflowin a river: a unit hydrograph,an API (antecedent iilustrated in Sec. 10'3), a relation for groundwater introduced in chapter 2 and as flow hydrograph a function and recession, a relation for computingthe groundwater both groundwaterflow and This model generates of the diiect runoff hydrograph. from precipitation values.The API model continuesto enjoy direct runoff discharge popularity and use in simulation modeling' widespread

Direct runoff hydrograph method) lunit-hydrograPh

Groundwater outflow hydrograph

Figure 23.1 Schematic diagram of API-type hYdrologic model. (After Sittner et al'r)

550

CHAPTER23

SIMULATION MODELS CONTINUOUS

StanfordWatershedModel lV (SWM-|V)
Crawford and Linsley designedthis digital computerprogram to simulateportions (the land phase)of the hydrologic cycle for an entire waftished.2The model has from the muchdevelopment sinceits conceptionand is currently available undergone U.S. Environmental Protection Agency under the name HSPF, which is a public of subsequently) the original program. The domain FORTRAN version (discussed of hydrograph a as SWM-IV hasbeenwidely accepted a tool to synthesize continuous parameterapproachis A hourly or daily streamflowsat a watershed,outlet. lumped used and data requirementsare much less than for alternative distributedmodels. data,and a variety of watershed Hour$ and daily precipitationdata,daily evaporation parameters input. are of The relationsand linkage of the variouscomponents SWM-IV are shownin Fig.23.2. Hydrologicfundamentalsare usedat eachpoint to transformthe input data data are into a hydrographof streamflowat the basinoutlet. Rainfall and evaporation first enteredinto the program.Incoming rainfall is distributed,as showninFig.23.2, amonginterception,imperviousareassuchas lakes and streams,and water destined to be infiltrated or to appearin the upper zone as surfacerunoff or interflow, both of evenwhich contributeto the channelinflow. The infiltration and upper zone storage storage. and to activeand inactive groundwater tually percolateto lower zone storage parametersgovern the rate of water movementbetweenthe storage User-assigned zonesshownin Fig. 23.2. condiThree zonesof moistureregulatesoil moistureprofiles and groundwater for is in encountered smaller watersheds accounted tions. The rapid runoff response in the upper zone,while both upper and lower zonbscontrol suchfactorsas overland for The lower zone is responsible longerflow, infiltration, and groundwaterstorage. term inflltration and groundwaterstoragethat is later releasedas base flow to the flow, and is stream.The total streamflow a combinationof overlandflow, groundwater interflow. of Model Structure The SWM-IV is madeup of a sequence computationroutines in for eachprocess the hydrologiccycle (interception,infiltration, routing, and so on). of Separatediscussions each componentare provided in the following paragraphs. Actual calculationsproceedfrom processto processas ilfustrated by the arrows in or Fig.23.2. All the moisturethat was originally storedin the watershed wasinput as precipitation during any time period is balancedin the continuity equation P : E + R + A S where

(23.r)

P : precipitation E :.evapotranspiration R : runoff AS : the total change in storagein the upper, lower, and groundwater storagezones The changein storagefor each zone is calculated as the difference between the volumesof inflow and outflow. Furthermore,all hydrologicactivity in a time interval to is simulatedand balancedbeforethe programproceeds the next time interval. The simulationterminateswhen no additional data are input.

MODELS SIMULATION STREAMFLOW 2g.1 CONTINUOUS

551

maximumsare providedin Table23.1. MAXIMUM 23.1 ryPICAL TABLE

RATES INTERCEPTION Watershed cover Glassland Moderateforest cover Heavy forest cover
Source: After Crawford and Linsley.2

rate (in./hr) Interception

0.10 0.15 0.20

(ET) is assumed occur at to Evapotranspiration In SWM-IV evapotranspiration "upper" storage zone' The upper and the the potential rate from interceptionstorage soils.The lower soil zone surface and the zonesimulates depressions highly permeable storagezone' , simulatesthe linkage to the groundwater from the lower zone is set equal to the ET opportunity, Evapotranspiration defined in fig. ZZ.Z.W opportunity is defined as the maximum amount of water availablefor ET at a partiiular location durrng a prescribedtime interval. In the modelinglogic, ET occursfrom severallocations(seeFig. 23'2) includingthe interand streamand lake surfaces, lower zonestorage, upperzonestofage, ceptionstorage, is frominterception arrdlpper zone storage ttotug". Evapotranspiration groundwater to is assumed be the lake evaporationrate Jet equal to the potential rate, Eo, which calcuiatedasthe productof a pan coefficienttimes the input valuesof the evaporation to wateris assumed occur at a rate equal of pan data.The evaporation any intercepted has when the interceptionstorage and ceases rate io the potential evapotranspiration been depleted. alsooccursat the potentialrate. The from streamand lake surfaces Evaporation surfaceareaof streamsand lakes(ETL) defined by total volgmeis governed the total to and as the ratio of the total str-eam lake areain the watershed the total watershed groundwaterstoragealso occurs at the potential rate from area.Evapotranspiration and is caiculatedin a similar fashion using a surfacearea equal to a factor K24EL the area.Thus the parameterK24EL represents fraction multiplied by the watershed from the groundwater evapotranspiration of tfre total watershedarea over which at set will occur.Most investigators this parameter avalue equalto the fraction storage Its phreatophytes. value is normally small but can area coveredby of the watershed be large, for eiample, in an agri-ultural area that has many acresof subirrigated alfalfa. the is If interceptionstorage depleted, modelwill attemptto satisfythe potential at for ET by drawingfrom the upper zone storage the potential rate. Once the upper lower zonebut not at the potential rate; zone sto;ageis dJpleted,ET ociurs from the the ET ratb from the lower zone is alwayslessthan Eo. When interception and the entersas EoinFig-23.3, do uppef zone storage not satisfythe potential,any excess

MODEI.S 23 SIMUIATION CHAPTER CONTINUOUS

/-------'-------\
/r \.
\

Actual evaootransoiration -'-r--\--------1-------J I

t., /
/

l* - - - - - - - - - - - - - -----\

model IV flowchart. (After Crawford and Linsley.2) Figure 23.2 Stanford watershed

MODELS SIMULATION STREAMFLOW 29.1 CONTINUOUS

553

F".""I
(: "iPrf) @
Channel inflow

@
Channel inflow Channel inflow

zi
\

-k^-*ff^." \ - - - - ; - - J

Simulated streamflow

tt ,'
t

Eigure 23.2 Continued

"7
554
CHAPTER23 CONTINUOUS SIMULATION MODELS

''E. h a'.: va' x x

n l -

25

50

75

100

Percentase oraretxTi* SfttaffiTffi,:'JTl?J."pportunitv

Figure 23.3 Evapotranspiration relation used in the Stanfordwatershed model. (After Crawford and Linsley'1 and the rate of evapotranspiration from the lower zone is determined from the shaded afea, or

E:Eo_% Q3.2) 2r The variable r is the evapotranspiration opportunity, deflnedas the maximum water amountavailable ET at a particular location during a prescribedtime period. This for factorvariesfrom point to point over any watershed from zero to a maximumvalue of -.^ LZS (23.3) r : ,''J LZSN where LZS : the current soil moisture storagein the lower zone (in.) LZSN : a nominal storage level, normally set equal to the medianvalue of the lower zone storage(in.) K3 : an input parameterthat is a function of watershed cover as shown inTable 23.2 The ratio LZSILZSN is known as the lower zone soil moisture ratio and is usedto comparethe actuallower zone storage with the nominal value at any time. Valuesof ET opportunity are assumed vary over a watershed to from zero to r along.the straight line shownin Fig. 23.3. This assumed linear cumulativedistributionof the parameter over an area is also usedin evaluatingareal disftibutionsof infiltration rates. Infiltration Like the erapotranspirationopportunity, the infiltration capacity of a watershedis highly variable from point to point and is assumedto be distributed accordingto a linear cumulativedistributionfunction shownas a line from the origin to Point b inFig.23.4. TABLE 23.2 ryHCALLOWER ZONE
EVAPOTMNSPIRATION PARAMETERS

Watershed cover
Open land Grassland Light forest Heavy forest

0.20 0.23 0.28 0.30

MODELS SIMULATION STREAMFLOW 23.1 CONTINUOUS

555

Percentage of area with an infiltration capacity equal to or less than the indicated value

Figure 23.4 Assumedlinear areal variation of inflltration capacity over a watershed. (After Crawford and Linsley.2)

Infiltration into the lower and groundwaterstoragezones is determinedas a function of the moisture supply 7 available for infiltration. Steps to determine infiltration for a given moisture supply7 are: "infiltration" in 1. The net infiltration is determinedfrom the area labeled to Fig. 23.4.This wateris assumed infiltrate into the lower and groundwater by zones.The areaenclosed the trapezoidis given by the equations storage the in the first row of Table23.3.If themoisturesupply7 exceeds maximum infiltration capacityb, the maximum allowed net infiltration is b/2, which is the median infiltration capacity. 2. Some of the moisture supply contributesto an increasein the interflow detentionduring any time increment and is calculatedas the region indicatedby an arrow in Fig. 23.4.Eqtationsfor this areausingvariousranges in x are providedin the secondrow of Table 23.3. The volume of water in as a stateofbeing transported interflow at any instantis calledthe interflow detentionor detainedinterflow. to 3. Any remainingmoisturesupplied,AD in Fig. 23.4, contributes increasing Equations for this the surface detention during the time increment. area are included in Table 23.3 for various valuesof 7. triangular-shaped
TABLE 23,3 EQUATIONSFOR THE SHADEDAREAS IN FIG. 23.3

Component
Net infiltration Increasein interflow detention Increasein surfacedetention of Percentage increaseddetention to assigned interflow
Searce; After Crawford and Linsley.2

x<b

b<i<cb

x>cb
b 2

x-

zb

t
b 2 i2 2cb

i2/ r\ - l l - - f c/ 2r\ 2rb

u-t, - ts
2' - c b

2rb
/ r \

x-t
/ , t \

r oo(- :) r

r oo(- "* r

_ r tr )

too2- _ j ilb

c - l

556

MODELS CHAPTER23 CONTINUOUSSIMULATION

ThEquantity of net infiltration is contrblledlargelyby the maximuminfiltration because the capacityb, while the pbrameterc significaltly affectshydrographshapes increment. The parametercoritrols the amount of water'detainedduring thelime valuesof b andc foi ahy time intetval dependon the soil moistureratio, LZS/LZSN, and on the input parametersCB and CC; CB is an index that controls the rate of on infilffation and depends the soil permeabilityand the volume of moisturethat can be storedin the soil. Valuesin the rangefrom 0.3 to 1.2 arecommon.The parameter CC is an input value that fixes the level of interflow relative to the overland flow. Values CC rangefrom 1.0 to 5.0. of If the soil moistureratio is lessthan 1.0, the variable b is found from

u "

cB
2ALzs/LzsN)

(23.4)

and whenLZS/LZSN is greaterthan 1.0,the equationfor b is

. 0 :

,n-rr"tvttt*,

*n

rr?5)

by Theseequations were developed Crawfordand Linsley from num0rousttials using a When the soil moistureratio reaches value SWM-IV in many different watersheds. The parameterc is of 2.0, the variable b reachesits minimum value of *r of CB. determinedfrctm
c : .(CC)ZQzslLzsN)

(23.6)

Variations in parametetsb and c with changesinLZS/LZSN aie shown in Figs. of 23.5 and23.6.Midrangevalues CB : 1.0and CC : 1.0wereusedin developing thesecurves. Figure 23.7 is a graph of distributionof water amonginfiltration, intefflow,'and overland flow for various valuesof the moisture supply 7. Different valuesof b and c would producea different set of curves. Water stoied as overland flow surface detention will either contribute to in as'depicted Fig. 23.2.The portion that streamflowof enterthe upper zone storage

0.0

0.2

0.4 0.6 0.8

1.0 1.2

r.4

1.6 1.8 2.O

ratt" zone moisture (-!ZL'J Lower soil b Figure 23.5 Variationin patarneter for variousvaluesofthe soil moisture ratio. (After Crawfordand Linsley.2)

0.2

0.4 0.6

0.8

1.0

1.2

1.4

1.6

zone soll molstufe rauo Lower \-I_ZSN'

/ Lzs)

Figure 23.6 Variation in parameterc for various values of the soil moisture ratio. (After Crawford and Linsley.2)

o o I
E

lncrease in overland flow surface detention

b=L.0 c=1.5

0.2 0.4 0.6 0.8 1.0 1.2 1,.4 1.6 supply 7 Moisture

to Figure 23.7 Typical SWM-IV response moisture supply variations.(After Crawford and Linsley.2)

100
; Y

8F
o 9 ^

80 60
, Inflection

bR

0 0

0.5

1.0

1.5

2.0

2.5

3.0

rati" zone moisture (ffi) soil Upper Figure 23.8 Delayedinfiltration as a function ofupper zone soilmoisture ratio. (After Crawford and Linsley.z)

558

CHAPTER23

CONTINUOUS SIMULATION MODELS

entersthe upper zone storageis called delayedinfiltration and is a function of the upper zone soil moistureratio,\JZS/UZSN, as shownin Fig. 23.8. The inflection point occursat a soil moistureratio of 2.0. If the ratio is lessthhn 2.A, thepercentage retainedby the upper zone is given by

e: roolr'(ffi)(--ri-)"'' ]
where UZI| is determinedfrom

(23.7)

r J Z r r : r r l # k - r . o ]r . o +
The curve is definedto the right of the inflection point by

(23.8)

P': lool(tt . uzLz/ r

whereUZI2 is determinedfrom

f/

t.0_)r,',-l

(23.e)

rrzr2:rrlffi-z.+] .o or

(23.r0)

zone,as shownin Fig. 23.2; receives a Upper Zone Storage The upper storage large portion of the rain during the flrst few hours of the storm, while the lower and groundwaterstorage zonesmay or may not receiveany moisture.The portion of the upper zone storage that is not evaporated transpiredis proportionedto the surface or runoff, interflow, and percolation.Percolation(upperzone depletion)from the upper LZS/LZSN. zoneto thelowerzonein Fig.23.2 ocurs only whenUZS/UZSN exceeds When this occurs,the percolationrate in in./hr is determinedfrom

: o.oo3(cBxuzs$(ffi PERC ffi)'

(23.rr)

where CB is an index that controls the rate of infiltration. It rangesfrom 0.3 to 1.2 depending the soil permeability and on the volume of moisturethat can be stored on in the soil. The variablesUZS and UZSN are definedas the actual and nominal soil moisturestorageamountsin the upper zone.The nominal value of UZSN is approximately a function of watershed to topographyand cover and is alwaysconsidered be much smaller than the nominal LZSN value.The initial estimatesof UZSN relative to LZSN are found ftomTable 23.4.
TABLE 23.4 VALOESOF UZSN AS A FUNCTIONOF LZSN FOR INITIAL ESTIMATESIN SIMULATION WITH SWM-IV Watershed Steepslopes,limited vegetation,low depression storage Moderateslopes,moderatevegetaion,moderatedepression storage Heavy vegetalor forest cover, soils subjectto cracking,high depression storage,very mild slopes
Source: After Crawford and Linesley.2

O.O6LZSN O.OSLZSN O.14LZSN

MODELS SIMULATION 29.1 CONTINUOUS STREAMFLOW

559

The parametersLZSN and CB must also be estimatedat the beginningof a simulation study.The combinationthat will most satisfactorilyreproduceboth longto and short-termhistorical responses hydrologic inputs can be determinedby the following procedure:2 1. Assumean initial value for LZSN equal to one quarterof the mean annual rainfall plus 4 in. (usedin arid and semiaridregions),or one eighth of the annual mean rainfall plus 4 in. (usedin coastal,humid, or subhumidclimarcsJ. ) Determinethe initial value of UZSN fromTable 23.4. 3. Assumea valuefor CB in the normal rangefrom 0.3 to I.2. 4. Simulatea period of record usingthe streamflow,rainfall, and evaporation data and systematicallyadjust LZSN, UZSN, CB, and other parameters is and betweensynthesized recordedstreamflows satisfacuntil agreement CB LZSN is adjusted; is do tory. If the annualwaterbudgets not balance, and recorded adjustedon the basis of comparisonsbetween synthesized flow rates for individual storms. zonein storage Lower Zone Storage and Groundwater The lower groundwater The percolaFig.23.2 receives waterfrom the net infiltration and from percolation. of from F;q.23.I1.Thepercentage net infiltrationthat reaches tion rateis determined groundwaterstorage depends the soil moisture ratio LZS ILZSN as shownin Fig. on P, 23.9.If this ratio is lessthan 1.0,the percentage is found from

P,:1oo[#(-#t,"',
is andif LZS/LZSN is greaterthan 1.0,the percentage

(23.r2)

tn",*:r* rn equations, both :, _ i'E;il,

+:roofr'-(--rg ] l"
] .,.,

(23.r3)

(23.r4)

LZS LZSN equalsthe lower zonestorage Notefrom Fig.23.9 thatthe nominal storage when half or 50 percent of all the incoming moisture enters groundwaterstorage. The outflow from the groundwaterstorage,GWE at any time is basedon the versustime. This commonly usedlinear semilogarithmicplot of baseflow discharge in technique delcribedin Section11.4andillustrated Fig. 11.8.In modifledform was the baseflow equationis GwF : (LKK4)[1.0 + KV(Gws)](sGw) where LKK4 is definedby LKK4:1.0-(KK24)t/e6

(23.rs)
(23.16)

in which KK24 is the minimum of all the observeddaily recessionconstants(see rate discharge to the 1 Secti,on 1.4), whereeachconstantis the ratio of the groundwater

560

CHAPTER23

MODELS SIMULATION CONTINUOUS

Inflection 9oo
oF \ i ltt

0.5

1.0

1.5

2.5

*u" moisture (;R) Soil of Figure 23.9 Percentage infiltratedwaterthat reaches and (After Crawford Linsley'2) storage. groundwater (K n groundwaterdischargerute 24 hr earlier. Thus the recessionconstantKK24 hasvalues The variableGWS in Eq.23.15 usingt : I day. Eq. 11.1)is determined Its on the long-terminflows to groundwaterstorage. value on any given that depend day'svalue,adjusted previous as (e.i., the lth day)is calculated 97 percenLofthe day : 0'97 (GWS'-1 * inflow to groundor storage, GWS; foi any inflow to groundwater during daYl)' water storage In Eq. 23.J{, SGW is a groundwater storage parametel that reflects the term fluctuationJin the volume of water storedand rangesfrom 0.10 to 3.90 in' The groundwater KV in Eq. 23.15 allows for changesthat are known to exist in the to when KV is zero, E,q.23.15reduces Eq. 11.1and ratesas time passes. recession follows the linear semilogrelation- If the usual dry season recessilon groundwater the are slofaggs being rate KKz4is too largefor wet periods(whengroundwater recession sothat the KV ihe from the streams) parameter is hand-adjusted by recharged seepage +-KVaGWS) will reduce the effective rate to some desiredvalue during term i0 periods.Table 23.5illustratesthis computationby showingeffectiverecesrecharge to 1'0' sion rates for variouscombinationsof KK24 and GWS when KV is set equal lost to deepor that is either The fraction of activeor deepgroundwaterstorage or is diverted as flow acrossthe drainage groundwaterstorage Gi;.b.D inactive to basin boundary is input ur pu.u*tt"r K24L. This fraction is the total inflow FOR RATES RECESSION 23.5 EFFECTIVE TABLE
OF VARIOUSCOMBINATIONS KK24 AND GWS WHEN l(/ : 1'0 GWS

0,5 0.99 0.98 0.97 0.96 0.99 0.98 0.97 0.96 0.985 0.970 0.955 0.940

1.0
0.98 0.96 0.94 0.92
n07

0.94 0.91 0.88

Source: After Crawford and Linsley.2

MODELS 561 SIMULATION S1REAMFLOW 23.1 CONTINUOUS groundwaterand representsall the active groundwaterstoragethat does not contribute to streamflow. Overland Flow The overlandflo* pro".r. hasbeenstudiedLy -uny investigators.

hydraulictechniques. of Averagevalues of lengths, slopes,and roughnesses overland flow in the calculatethe Manning and continuity equationsare usedin SWM-IV to continuously rate q is then relatedto D,. D".The overlandflow discharge detentionstorage surface As the rain supplyrate continuesin time, the amountof water detainedon the The amount of surface surfaceincreasesuntil an equilibrium depth is established. detentionat equilibrium estimatedby SWM-IV is D":
6 0.000818i0 no'6L1'6 s0'3

(23.r7)

where D" : the surfacedetention at equilibrium (ft3lft of overland flow width) j : the rain rate (in./hr) S : the slope(ftlft) L : the length of overlandffow (ft) coefficient ru : Manning's roughness The overland flow dischargerate is next determinedas a function of detention storage from

0 , : # y,,(?)"'[, * o,o(r2)']"'
where e : the overland flow dischargerate (cfs per ft of width) D : the averagedetention storagedurrng the time interval

(23.18)

but that occurs after rain ceases, the The equationalso appliesduring the recession Typical overland flow roughnesscoefficientsafe to ratio blD" is assumed be 1.0. providedin Table23.6. The time at which detention storagereachesan equilibrium is determirtedfrom
t

.e

: -

0.94L3/5n3/s
i2/s S3/ro

(23.te)

where /, is the time to equilibrium (min). Crawford and Linsley show that these overlandflow hydrographs.2 equationsvery accuratelyieproducemeasured For eachtime interval Lt, an end-of-intervalsurfacedetentionD, is calculated from the initial value D, plus any water addedAD (Fig. n.q to surfacedetention from during the time interval, lessany ovedandflow dischargeQthatescapes storage
.4.

562

23 CHAPTER

MODELS SIMULATION CONTINUOUS ryPICAL MANNINGEQUATIONOVERLANDFLOW NN PARAMETERS, ROUGHNESS n Manning's for overlandflow

TABLE;3.6

Watershed cover Smooth asphalt Asphalt or concretePaving Packedclay Light turf Denseturf DenseshrubberYand forest litter
Source: After Crawfordand Linsley.2

0.012 0.014 0.03 0.20 0.35 0.40

of continudetentionstorageduring the time interval. This is simply an expression itv. or Dz:Dt+LD-4Lt

(23.20)

(D: + D)/2' EquaThe discharge is found from Eq' 23.18usinga valueo! D . @ of overland flow using easily tions 23.17- 2i.20 allow the completedetermination overlandflow' length, slope,and roughness found basin-widevaluesof the average same is Interflow The watertemporarilydetainedasinterflow storage treatedin the inflow to interflow detention was fashion as overland flow detention storage.The similar constant usinga daily recession is in defined Fig. 23.4.Theoutflow simulated constantIRC is the recession The interflow discharge. to that definedfor groundwater 24 at ratio of the interflow discharge any time to the interflow discharge hr average is outflow from detentionstorage earlier.For each15-min time intervalmodeled,the

rNTF : LIFC4(SRGX)
where

(23.2r)
()1n\

LIFC4:1.0-(tRc;'inu

Its The variableSRGXis the water storedin the interflow detentionat any time. applied to each time value continuouslychangeswhen the continuity equation is accordingto continuity, on the interval. The end-of-inteival value of SRGX depends, from the interflow valueat the beginningofthe interval and any inflow to or discharge detentionduring the interval. modelutilizesa hychannel Translation and Routing The Stanfordwatershed to routing technique translatethe channelinflow to the watershed drologicwatershed as in outlei. Clark's IUH time-ar"a to"ihod described Section 12.6is adoptedalmost l2.Inplaceofthenetrainhyetograph,theStanfordmodelviews presentedinchapter "inflow" hyetograph. This inflow is the sum of all channelinflow componentsas an whereit is next routed then translatedin time through the channelto the basinoutlet, by to system accountfor the attenuationcaused storage through an equivalentstorage (linear in the sensethat Roulingthrough the linear reservoir in the-channelsystem. is accomto storageis assumed be directly proportional to the outflow' Eq. 12.35) plishedfrom (23.23) I or:7-KS1(1 -01) )

MODELS 563 SIMUI-ATION STREAMFLOW 29.1 CONTINUOUS where Oz: the outflow rate aI the end of the time interval

o t : the outflow rate at the beginning I _ the average inflow rate during the time interval
Also,

KSI

K - Ltlz :v*6*

(23.24)

from watershed parameters Examplesof the determinationof K and othernecessary in Section23.2. data areincluded Applications of the SWM-IV Applicationsof the model typically beginwith data for a three- to six-yearcalibration period for which rainfall and runoff data arc of adjustments severalparameters Thesedata are usedto allow successive available. of recordedhydrographs the streamflowagree.If sufficient until the simulatedand data areavailable,a secondperiod of record may be reservedfor use as a control to checkthe accuracyof the paiametersderivedfrom a calibration with the first half of the data. in modelwasoriginally developed 1959and hasunderThe Stanfordwatershed translatedthe Crawfordand Linssincethat time. James3 goneseveralmodifications ley model from ALGOL to FORTRAN. Several modifications of the FORTRAN Included amongtheseare the versionhave evolvedfrom a variety of investigations. model (KWM),4'5 the Kentucky self-calibrating_version Kentucky watershed (OpSETi,4the Ohio State University version, the Texas version,6the Hydrocomp nonproprietary SimulationProgram (HSP) written in PL/1, and EPA-produced, HSP called HSPF, and the National WeatherServicerunoff FORTRAN u"rrion of of forecastingmodel. Brief descriptions severalof theseare includedbelow.

ARS RevisedModelof WatershedHydrology(USDAHL)

and farming Growing interest in the effects of soil types, vegetation,pavements, growth of the USDAHL practicei on infiltration and overlandflow hasresultedin the continuoussimulation model. The 1974 versionTof this model was developedby ServiceHydrographLaboratory. at investigators the Agricultural Research Input data to the model are relatively extensive.Continuousrecords of the pan-evaporation the temperatures, weekly average precipitation,the weekly average and cultural practicesare land use, amounts,and detaileddata onioils, vegetation, required' is The studywatershed initially dividedinto asmany asfour distinctland-useor and a main calling routine computefor each soil-type ,on".. Fourteen subroutines zone the snowmelt, inflltration, overland flow, channel flow, evapotranspiration, recharge,and return flow' groundwaterevaporationand movement,groundwater to coefficients by potentialsare estirnated app$ing assigned Evapotranspiration zone is computedusing a pun-"uuplrution data. Infiltration for eachsoil or land-use Waterstoredin cracksin dry soilsis simulatedasa function modifiedHoltan equation. of soil moisture.Manning's equationand the continuity equationare used to route solution of the continuity overlandflow. The streamflowis routedby a simultaneous

564

CHAPTER 23

MODELS CONTINUOUS SIMULATION

movementsare calculatedby Darcy's function. Groundwater equdiion and a storage zoneis calculatedas a function of the temperequation.The daily snowmelton eachdensityfor the zone, vegetative ature at which snowmeltstarts,the weightedaverage the weekly average temperature,and the potential snowmeltper day in the zone air snowpack. Precipitation falling during a snowmelt day also contributes to the snowmeltequation. Service Among otheruses, modelhasbeenappliedby the Soil Conservation the Figure 23.10 showsthe results of in preparing environmentalimpact statements. and widely applying the 1974 version to annual runoff'from four widely separated In diversifiedARS experimentalwatersheds. addition to the runoff, the model computes the evapotranspirationamounts, soil moisture changes,return flows, and groundwaterrechargedepthsfor eachof the zones. Although other modificationsare possible,the USDAHL model is specifically generallyunder 20 squaremiles. for designed relatively small rural watersheds,

$-i

E U

Cumulative computed runoff, O (in.)

Figure 23.10 Chart showing the accuracy of USDAHL-74 model for estimatingthe cumulativecomputedrunoff as comparedwith the cumulao OH: A W- 11, tive measured runoff at four watersheds. W-97, Coshocton, Hastings,NE; I W-3, Ft. Lauderadale,FL; x W-G, Riesel, TX (After Holtan andLopez.l)

MODELS SIMULATION STREAMFLOW 2g.1 CONTINUOUS

565

NationalWeatherServiceRiver Forecast System(NWSRFS)

by modelwasdeveloped the Hydrologic yet anotherversionof the Stanfordwatershed The WeatherServiceOffice of Hydrology.8 Laboratory staff at the National Research by river flowsand stages the use in forcecasting for NWSRFS model wis developed to severalriver National WeatherService.The model has been applied successfully in North Carolinato 1000mi2in Oklahoma'River basinsrangingin sizefrom 70 mi2 in SWM for forecastin! in"largeriver basinsdoesnot require the detail incorporated the NWS model includestwo major changes For this reason' smaller watersheds. fewer process involvingthe useof a longertime increment,simplifiedprogramming, parameters for determiningoptimal watershed comput;ions, and a rapid procedure historical flows accurately. that allow the model to reproduce inputs and A 6-hr time incrementis usedby the model, allowing fewer rainfall suchas overlandflow that of processes ,nor" i.po.tant, fewer detailedcalcuiations rapidly than with occur in'shorter time periods.Iterations are thus completedmo_re the National WeatherService optimization the SWM. As with the OPSETmodel, available procedurefor determiningparametervaluesgivesthe model a strengthnot with the SWM-IV. Other modificationsincludeh uPPerand lower zone retentionand the uPPersoil zone to groundwate al groundwaterevapotranspiration jointlY comPuted the NWS versi in runoff, interflow' and groundnatedand is replacedby three types ofrunoff: surface and slow response' water flow-representing fast, medium, and instantaInput data for modll calibration consistof meandaily discharges runoff events.Rainfall is input as a continuous for neoushydrographs a few selected Betechniques. record of 6-hr basin-widemeansdeterminedfrom areal averaging and in the detail of processsimulation'the in causeof the changes routing increment outputfromtheNwsiersionissimilarinmakeuptotheSWM-IVoutput.

Regulation and Reservoir Synthesis COEStreamflow Model (SSARR)

for large Another widely used continuous streamflow simulation model designed The SSARR model was developed basinswas devllopedby the corps of Engineers.e operation primarily for streamflowand flood fo.ecastingand for reservoirdesignand hydrologistsat the National Weather of i,rior to the development NWSRFS, studies. both rain and Service used.the SSARR model. The model has been applied to snowmeltevents. basin into Applications of the model begin with a subdivisionof the drainage and characterconsistentwith subdivides, hydrologic units of i size homogeneous rJservoirsites,diversionpoints,soil types,and otherdistinguishchannelconfluences, points throughoutthe are ing features.The streamflows computedfbr ail significant river sYstem.

566

MODELS 23 SIMUI-ATION CHAPTER CONTINUOUS duintutt data can be input at any numberof stationsin the basin.The part that or will run off is divided into the baseflow, subsurface interflow, and surfacerunoff. The division is based on indices and on the intensity of the direct runoff. Each and componentis simply delayedaccordingto different processes, all are then comrunoff is then outflow hydrograph.This subarea bined to producethe final subbasin routed through stream channelsand reservoirsto be combined with other subarea hydrographs, of which becomepart of the output. all by Routingsthrough channelsand reservoirsare accomplished the sametechallowances and of nique.This requiresan assumption shortstreamreaches, occasional in for backwatereffectsare necessary the channelrouting process.Streamflowsare on synthesized the basisof rainfall and snowmeltrunoff. Snowmeltcanbe determined on the basis of the precipitation depth, elevation,air and dew point temperatures, albedo,radiation, and wind speed.Snowmeltoptions include the temperatureindex method or the energybudgetmethod. indicesfor subdiInput includesthe precipitation depths,the watershed-runoff initial reservoir elevationsand outflows, flow among the three processes, viding drainageareas,bounds on usablestorageand allowabledischargefrom reservoirs, total computationperiods,routing intervals,and other specialinstructionsto control plots, prints, and other input-outputalternatives. This model was one of the earliestcontinuousstreamflowsimulation models and has its primary strengthin its verified using a lumped parameterrepresentation indicatedby testsconductedin severallarge drainagebasinsincluding the accuracy ColumbiaRiver basin and the Mekong River basin.

Program(HSP) Hydrocomp Simulation

at A commercial version of the Stanford water model was developed Hydrocomp, incorAmong severaladvantages Inc., namedthe HydrocompSimulationProgram.to chanand poratedin HSP are hydraulic reservoirrouting techniques kinematic-wave Other major changesinclude the addition of water quality nel routing techniques. simulation capabilities.Due to theseadditions,the model is often referredto as the Hydrocomp water quality model. The HSP model has been usedroutinely for severaltypes of hydrologicstudy including floodplain mapping,water quality studies,storm water and urban flooding studies,urban drainagefacility design,and water quality aspectsof urban runoff. The model consistsof three computerroutines: to disk storage handleinput datawith l, Library allowsthe useof direct access routines. efficient data management in 2. Lands handlesthe usual SWM lands phasealong with addedprocesses calculating soil moisture budgets,groundwaterrecharge and discharge, inflow to streamchannels,and eutrophication. 3. Channel is responsiblefor assemblingand routing all channel inflow through channelnetworks,lakes, and reservoirs. water balanceby trackingprecipitaThe HSP model incorporatesa continuous of tion through all possibleavenues the hydrologicand water resourcesystem.The

29j

MODELS SIMULATION STREAMFLOW CONTINUOUS

567

inflow, interflow, and surfacerunoff are individually simulated,lagged, groundwater ind combinedat appropriatetimes as the channelinflow. The routing of computed model.Water inflow through the ihannel networkutilizes a modified kinematic-wave ratesand other waterparameters relatedto variabledischarge are quality constituents so that the coupling of quantity and quality of the runoff is accomplished. radiation,wind, and Inputs foisimulating waterquality includethe temperature, of the factors under study which form the basis for humidity and observedvalues calibration.At lest two yearsof data are preferablefor calibration; however,calibrawith lessthan one year of data. Other requiredinput includes tion hasbeenachieved and potential in hourly precipitation 5-, 15-, or 30-min or greater time increments or water quality simulation is desired,the temperature, if evapoffanspiration; snow radiation, wind, and humidity factorsare needed' Outputsfrom HSP can be obtainedfor any desiredpoint within the watershed. Included in the output options are valuesof quality dataat outfalls or other points, rates,streamand lake ,"r"ruoi, levels,hourly and meandaily discharge river stages, and total dissolvedsolids,algaecounts,phosphorus' dissolvedoxygen remperarures, BOD, colipH, carbonaceous nitr;te, nitrite, ammonia,iotal nitrogen, phosphate, the usual daily, monthly, and annualwater budgets, metals,and forms, conservative snow depths,and end-of-periodmoistureequivalents' Typical simulation periods in HSP applicationsrange from 20 to 50 years. of Hour-by-hour data are not viewed as an exact sequence future flows. Rather, the of probability of occurrences rangesin the factorsof data arcusedfor analysisof the the model is functioning with a purposesimilar interest.When usedin this -unir"r, in described Chapter22. to that of someof the operationalhydrologytechniques

Program-Fortran(HSPR Simulation EPA Hydrocomp

of Following development the HSP versionof the Stanfordmodel,the U.S. Environmental Piotection Agency contractedin 1980 to have public-domain version made for available continuousstreamflowsimulationand water quality modeling.The origThe HSPF code inal program,written in ALGOL, wasconvertedto FORTRAN 77'11 portions of water quality modeling alis availa|le for PC applications.Substantial gorithms were addedio HSP in developingHSPF. The hydrologiccycle processes, however, are essentiallythe same as in SWM-IV and HSP' One exceptionis the addition of severalrouting proceduresnot previouslyavailable. for parameters that drive ihe routinesin HSPFmustbe estimated all the hydroof difficult because the numerous making verification of the model logic processes, and caliestimatingtheseparameters combinationsof paramiter values.Methodsof in Section23.2 (Table23.8 defines bratingthe modeiareillustratedin the casestudies usedin the Stanfordmodel)' over 35 parameters

ModelingSystem(PRMS) USGSPrecipitation'Runoff
their urban storm-eventmodel, DR3M (seeChapter 25), the U.S' After developing severalother computercodesto model continuoushyGeologicalSurv"y developed for pRUS performssimulationof daily streamflows a variety drologicpro""rr"r. fne from the USGSin is available It of precipitation,climate,and land usecombinations.
I

L.

568

CHAPTER23

MODELS SIMULATION CONTINUOUS

During storms,the model givesoutput on PC, riinicomputeroor mainframeformat.12 any prescribedtime interval. Betweenstorm periods,the model tracks soil moisture zones on a daily basis until the next storm interval. and other storage/depletion Streamflowis output as mean daily flow rates. of is approach utilized in PRMS.The smallestsubdivision A lumped-parameter to the study watershedis a hydrologic responseunit that is assumed. behaveas a HRUs in most areasof hydrologicelement.The USGS has delineated homogeneous and reservoirs, R3M model,the streams,storm sewers, the United States.Like the D. detentionponds in the watershedare modeledas nodes and interconnectinglinks' kinematic-wave Hydrographrouting in channelsand reservoirsis accomplished.by methods,respectively. and storage-indication

in for ARS Simulator WaterResources Rural Basins(SWRRB)

Through a cooperativeprogram with TexasA&M University, the Agricultural Research Service, U.S. Department of Agriculture, developeda continuous daily streamflow simulation ptog.a- for use in modeling ungaugedagricultural areas.t3 This FORTRAN 77 water budgermodel is availablefrom the ARS in PC format. Its focusedon a model that would allow the user to predict impacts of development practicessuchas crop rotation, fall plowing, urbanmanagement variouswatershed ization, conservation tillage, terracing, fallowing, and floodwater detention on monthly and annual water and sedimentyields from rural basins. and Sedimentyields for each of the subwatersheds, for the total basin, are hydrolThe methoduseswatershed usingthe universalsoil lossequation.la calculated ogy outputsfrom the rainfall-runoff portion to estimatesedimentyield, using inputs of runoff volume,peak flows,soil type anderodibility, crop types,erosionmanagment factors,and watershedslopeand length. style.Oncerainfall or snoware The subbasins modeledin a lumped-parameter are algorithms linked as shownin Fig. 23.11 fall recordsare input, physical-process sedimentprolosses, for solar radiation, snowmelt,surfacerunoff, ET, conveyance percolation,and soil from watersurfaces, evqporation ductionfrom individual storms, basedon crop types,temperCrop productionis alsocalculated moistureaccounting. ature, consumptiveuse of water, and irrigation practices. by are for Hydrologicabstractions eachsubwatershed estimated the SCS curve number(CN) method(Chapter4, Section4.9). Soil waterbudgetingis performedby adding the net moistureinput to the soil profile from precipitation after subtracting from soils),percolause direct runoff, ET (consumptive by the cropsand evaporation runoff is Direct surface tion, and retu{n flow (groundwaterflow back to the streams). set equal to the net rain from the CN method. For input to the sedimentyield component, a peak flow rate for individual storms is estimatedusing a modified rational method (Chapter 15). Snowmelt is calculatedby the degree-daymethod ' in described Chapter14.

ExpertSystems
of analysisare leadingto developmenl expert system(ES) Recenttrends in systems use in planning and designof water that rely on artificial intelligencefor techniques
-, .oo^,,r^^o -r^io;id--7[mrfian-.m h-c-ciw?i'- ii-FiYmqfinn nhfqined frnm extencive

MODELS SIMULATION STREAMFLOW 29.1 CONTINUOUS

C*.')
' /READTNPUTDATA/
INITIALIZE PARAMETERS

READ OR GENERATE PRECIPITATION AND MAX/MIN TEMPERATURE GENERATE SOLAR RADIATION

CoMPI.ITE SOIL TEMPERATTTRE COMPUTE SNOWFALL AND SNOWMELT

ACE

COMPUTEPEAK RATE TRANSMISSIONLOSSES. OUTI-ET SBDT,M}{T YIEIO ENO ROUTE SEDIMENT TO BASIN

COMPUTEANDPRINT FINALBASIN STATISTICS

Figure 23.11 Flowchart of the SWRRB hydrologicpr6""*, algorithms.(After Arnold't3)

570

CHAPTER23

MODELS SIMULATION CONTINUOUS

interviJws of one or more experts in some field. The computer can then make "decisions"in muchthe sameway asthe,experts, applyingtheir judgment and experito othersthrough the expert systemmodel. ence and making theseavailable simulation,incorStreamflowmodels,especiallythosethat perform continuous judgment. Developers and usersof parameters require considerable that porateinput have accumulated decades of experience in assigning ahe watershed models and judgment can be extractedby an Their experience coefflcientsand parameters. hundredsof questionsto build an expert systemmode'. interview processinvolving about uncertainties but direct answers alsoaddresses The modelnot only incorporates promise.ls modelingtechniqueshow considerable with this each.Early applicdLtions have the potential to be In addition to streamflowsimulation, expert systems of of complexriver basin systems dams,reseruseful in the designand management Operationsfor such flood control structures. voirs,powerplants,diversioncanals,and systemsinvolve independentand collective decisionsby dozens of professionals' contact with numerousother conTheseexpertsare normally in radio or telephone from theseteams,the could be developed If trollers and decision-makers. ES data A prime incentiveof implementingexpert exists. potential for improvedmanagement profesinvolvescapturinginsightsof experienced systems iystemsin waterresources positions. sionalsbefore they retire or move into other

STUDIES MODEL SIMULATION 23.2 CONTINUOUS

applicationsof the Kentuckyversion in This sectiondescribes detail two independent model to small basinsin Kentucky and Nebraska.Results watershed of the Stanford in obtainedby Clarkel6in modelingthe CaveCreek (CC) watershed Kentuckyand by in using KWM for the Big BordeauxCreek (BBC) watershed Nebraska the authorslT watershedshaving relative$ good Both are small, homogeneous are compared. simultaneously are recordsof precipitationandrunoff. The two casestudies described input data requiredto develop how different analystsdealt with the decisions to show parameters. and

Selectionof WatershedSize and Study Period

size and time period to be subarea Severalguidelinesexist for selectinga watershed watersheds use of relatively small, homogeneous modeledin a simulation study.The is recommendedto minimize any difficulty or subdivisionsof larger watersheds causedby ignoring spatial variations in precipitation over larger areas.This also such as soil types, soil minimizes the'effects of lumping watershedcharacteristics the representing entire usesinto singleparameters and profiles,imperviousareas, land drainage upper limit of 25 fii12for study-watershed an catchment.Ross suggests areas.18 sizearisesfrom the fact that few strearns One difficulty in restrictingwatershed continuouslygauged,and refiable streamflow data are for small drainageareasare it because was one was selected difficult to obtain. For example,the BBC watershed in Nebraskawith sufficientrecords.The CaveCreek waterof few small watersheds on shedwas selegted the basisof the following criteria'1e

MODELSTUDIES SIMULATION 2g.2 CONTINUOUS

571

minimum of 10 yearsof continuousrunoff recordsin order to establish the existingrainfall-runoff relation. 2. A drainas" ur"u of less than 5 mi2 so as to be reptesentativeof small drainage6asinsfor which better runoff coefficientsare needed' 3. A locaiion in closeproximity to a rain guagefor which hour$ precipitation dataarc available. under study. 4. The availability of soil surveysfor the watershed 1.

'A

DataSources
and Hourly precipitation data were availableat sites L20 fiil from the CC watershed runoff data were 8.0 mi fromthe BBC gaugingstation.Soil surveyrecordsand Daily pan evaporationdata were availableapproxiavailablefor both watersheds. mately 30 mi south of the BBC watershedand 25 mi south of the CC watershed. d Drainageareasof 2.53 mi2for the CC watershe and9 .22 mi2for the BBC watershed for maps.Input parameters the were found from U.S. GeologicalSurveyquadrangle 23.7. Numerical Stanfordand Kentuckyversionsare comparedand definedin Table names,aretabulatedfor both watersheds usingthe Stanfordversionparameter values, in in Table 23.8. Each parameteris described detail in the following sections.

Time-Area HistogramData
The time-area histograms for the BBC and CC watershedsare developedin for Traveltimes and times of concentration the Figs.23.12 and23.Ii, respectively. larger found from the Kirpich equation(for watersheds model are Stanfordwatershed s1 than 15 acres),ie

a : ooo78(#)"

(23.2s)

where T" : the time of concentration(min) L : the horizontal projection of the channellength from the most distant point to the basin outlet (ft) : the slopebetweenthe two points S the In developing time-area histogramfor Big BordeauxCreek, 30 points and times (min) were plottedas shownin Fig. 23.12.The dashed correspondingi.auet by (lines of equal travel time to the basin outlet) were constructed linear isochrones paris of isochroneswere interpolation between the plotted points. Areas between from planimeteiing.To contrast,the time-area histogramfor CaveCreek determined was construcrcdfas shown in Fig. 23.13, by assumingthat the flow velocity was velocity obtainedby dividing f streamflow constant"u"ryrli"r", equalto the average into the channel length Z. This procedure simply places all points along each to isochroneat equal distances the basin outlet.

WatershedParameters
The watersheddrainage area (AREA), the impervious fraction of the watershed surfacedraining directly into the stream (A), the fractional streamand lake surface ground slopeof overlandflow perpendicularto the contours area(ETL), the average

572

CHAPTER23

MODELS SIMULATION CONTINUOUS FOR KENTUCKYAND PARAMETERNAME COMPARISONS STANFORDWATERSHEDMODELS

TABLE 2i,7

Parametername

Kentucky verston
NCTRI CTRI RMPF RGPMB AREA FIMP FWTR VINTMR BUZC SUZC LZC ETLF SUBWF GWETF SIAC BMIR BIVF

Stanford verston

descriPtion Parameter Integernumber of elementsin the time-area histograq Time-area histogramordinate data are to be printed (cfs) DischargevalueLelow which no synthesized factor for precipitation data for a distant statton Multiplication drainagearea (mi2) Watershed areathat is impervious Fraction of watershed area in lakes or swamps Fraction of watershed Maximum rate of vegetativeinterceptionfor a dry watershed(in'/hr) capacity storage surface lndex for estimating Ind,exfor estimatingsoil surfacestoragecapacityduring summers Index of moisture storagein soil profile abovewater table (in') parameterfor lower zone soil moisture Evapotranspiration flow from the basin Subsurface by Groundwaterevapotranspiration phreatophytes Factorvarying infiltration by season Index of infiltration rate Index of rate and quantity of water enteringinterflow Averagebasin ground slope (ftlft) Averageoverland flow distance(ft) coeff,cientfor overland flow Manning roughness coefficientfor flow over impervious areas Manning roughness constant Daily intelflow recession Streamflowrouting parameterfor 1owflows Streamflowrouting parameterfor flood flows Index capacity of existingchannel,bank-full (cfs) routing of Exponent flow for nonlinear Daily baseflow nonlinearrecessionadjustmentfactor Daily baseflow recessionconstant (in.) storage Indexof groundwater storageat beginningof year (in') and depression Depth of interception Cuirent equivalentdepth of moisturein the soil profile (in') Cunent value of BFNLR Interflow storage Control options for input, output, and internal branching Title of computer simulation output, alphanumericinput Mean daily diversion into or out of the basin (cfs)

Z C MINH K1 AREA ETL EPXM CX EDF LZSN K3 K24L K24EL EF CB CY

SS L NN NNU IRC KSC KSF CHCAP RFC

oFss
OFSL OFMN OFMNIS IFRC CSRX FSRX CHCAP EXQPV BFNLR BFRC GWS UZS LZS BENX IFS CONOPT

t<r/24 KK24
GWS UZS LZS

DKN

aoa
DIV

Qaa
DIV

(SS),and the nban length of overlandflow (r) for the Big Bordeauxcreek watershed and lengths measuredfrom 7.5-min series were determinedfrom-areas,elevations, parameter usGS topographicmaps (see Table 23.8 for BBC values).Other BBC gri : 0.0d5 (determinedfrom ETL : half the product of the stream values utl from Fig '23'14 length and channelwidth at the outlet), SS : 0.088 ftlft (determined 20-ft contoursat eachgridline intersegvaluesbetween u, ih- -"un of'140 measured of from Fig.23.14 as the average l40lengths 183.2ft (determined tion), and L: lines from gridline intersection points to the measuredperpendicularto contour

FOR BIG BORDEAUXCREEK TABLE 23.8 SUMMARYOF INPUT PARAMETERS CREEK WATERSHEDS AND CAVE

Parameter name
TCONC TINC Z C

Description Time of concentration Routing interval Number of elementsin the time-area histogram Time-area histogramordinates

a^ BBC value(s) Units value(s)

AREA

^ ETL

ss
L

CHCAP IRC KK24 KSC KSF MINH K1 NN NNU EMiN EPXM CX EDF LZSN K3 K24L K24EL EF CB CY KV24 SGW UZS LZS GWS VOLUME EVCR

drainagearea Watershed Impervious fraction of the watershedsurface stream and lake surfacearea as fraction of Watershed 0.005 watershedarea 0.088 ft/ft Averageoverland flow ground sloPe ft 183.2 of overlandflow Averagelength cfs 39 Bank-full flow in channelat gaugingstation 0.485 constant recession Daily interflow 0.977 \ constant Daily baseflow recession 0.989 parameterfor low flows Streamflowrouting 0.989 Streamflowrouting parameterfor flood flows cfs 1.0 rate to be printed Minimum hourly flow 1.0 Precipitation adjustmentfactor for distant gauge 0.3'7 coefficientfor overlandflow Manning roughness 0.0i3 coefficientfor impervious areas Manning roughness 0.5 infiltration by seasons Factorfor varying 0.15 in./hr Maximum interceptionrate for dry watershed 0.80 capacitYindex Surfacestorage 1.10 capacity moisturesiorage Soil surface ln. Il./d profile moisture storageindex Soil 0.28 Soil evaporationparameter 0.0 Index of inflow to deepinactive groundwater 0.0 Fraction of watershedareain phreatophytes 1.0 infiltration rates for seasonal Factor allowing 0.'75 Factorcontrolling infiltration rates 3.0 Index controlling water enteringinterflow 1.0 Parameterfor allowing nonlinear recesslon in. 0.1 storagevolume parameter Groundwater in. 0.0 Equivalentdepth of upper zone storage in. 7.0 of lower zone storage Equivalentdepth in. 0.2 moisture conditions Index of antecedent acre-ft 0.0 water in swamPstorage Volume of 0.911 Jan. Monthly evaporationpan coefficients 0.911 Feb. 0.911 Mar. 0.911 Apr. 0.552 May 0.677 June 0.654 July 0.651 Aug. 0.642 Sep. 0.911 Oct. 0.911 Nov. 0.911 Dec.

min 105 min 15 7 0.129 0.158 0.221 , 0.151 0.126 0.145 0.070 mi2 9.22 0.0

60 i5
4

0.18 0.29 0.31 0.22

2.53 0.0 0.0 0.075 300.0 40 0.75 0.94 0.90 0.90 0.2 1.0 0.10 0.015 0.10 0.90 1.25 4:85 0.25 0.0 0.0 0.15 0.65 3.50 0.99

574

MODELS SIMULATION 23 CHAPTER CONTINUOUS 0.3

I

15-min isochrone

0.2

"15 rs 30 4s 60 90 105 Time(min)

-l

3li

Time.area

a5-min {

Ll--:Ja
68o -l \t

$.(-t

15 30 45 60 75 90 105 Time (min) 30-minisochrone

45-min isochrone

60-min isochrone N

--.1-^

t-\-75-minisochrone
90-min isochrone
n . , u miles r IJJJJ-JJJJJ 1 l

for Figure 23.12 Time-areahistogramdevelopment Big BordeauxCreek' The valuesrepresenttravel times (min.) to the outlet.

overlandflow distancescan also be estimatedas the nearestchannel).The average

resulting in an estimateof the bank-full BBC flood of 39 cfs. Another techniquefor rating curveif th CHCAP involvesthe useof the stream-gauging-station determining maps. from topographic gaugeheightfor bank-full flow is known or can be estimated the cross of from a hydraulic analysis thJCave Creekvalueof 40 cfs wasdetermined to spctigl and profile of the main channel.As a rule of thumb, CHCAP is selected )

.../

3l.ovo _

Time-arca histogtam Time (min) Contributing area (Vo)

0 -1 5 15-30 30-45 45-60

18.0 29.0 31.0 22.O

Figure 23.13 Derivation of the time-area histogramfor thecave creek (After Clarke'16) Dashedlines representisochrones. Watershed.

0
l r r

miles
r r r r t

1
r l

0 1000
f""t

h -w" -q
:h &"{

, TGrid r v a l : f m = t l Z O f t inte
\I

Figure 23.14 Grid overlay for determinationof the mean overland - slopeand distancefor Big BordeauxCreek.

576

CHAPTER23

MODELS SIMULATION CONTINUOUS

twice the value of producewell-definedoverbank and floodplain flow if Q exceeds wheneverQ is lessthan half of Similarly, low, shallow flows are assumed bffCAp. CHCAP.

StreamflowRecessionand Routing Parameters

The rate at which the model allows water to passthrough the upper soil zonesto the channelsis controlled by the daily interflow recessionconstant(IRC)' A graphical semilogarithmictechniqueof hydrographanalysisdevelopedby Barnes is used to constantfor The estimatethis parameter:21 determinationof the interflow recession which is a semilogarithmicplot of Big BordeauxCreek is illustrated in Fig. 23.1.5,

I o.s E o.+
o

K 0.3
E o
d

V'L

p9

o.ro
q

lnterflow Llog Q = -0;725 Lt = 3.2 days K=0.485=IRC

& 0.04 0.03

S o.os o

1 4 1 6 1 8 2 0 2 2 2 4 2 6 2 8 3 0 1 3

(Ju1Y-August' 1969) Date Figure 23.15 Determination of the interflow recessionconstant (fnC) for th9 Big BordeauxCreekflood eventof July 20, 1969.The K: IRC, C' : QoK'andr: 1day. determinationof

MODEL STUDIES SIMULATION 2g.2 CONTINUOUS

577

!2
F

o b0

5
logQ,=1ogQ6+tlogK logK= LlogQlLt

d o o

f f i z o

' 3 " ' s1 9 1 1 3

1969) (July-August, Date Figure 23.16 Determination of the base flow recession constant (K"K24) for the Big Bordeaux Creek flood event of July 20, 1969' The determinationof K : KK24.

578

MODELS SIMULATION 23 CHAPTER CONTINUOUS

'

alternative method for determining Kl<24 using data collected several days after an severalToodpeaksis illustratedfor CaveCreek in Fig. 23.I7. The line represents points. envelopedrawn to the right of the data Two streamflowrouting parameteisare usedby the sinlulation model to route to inflow hydrographs the point of interest for the basin (in this caseto the stream garglng;tatiin ror compa;isonwith measuredflows). The first parameter 1_519t_lt (CHCAP), thechannelaapacity usedinroutingonlyif channelflowsarelessthanhalf during flood flows for and the second(KSF) accounts channeland floodplainstorage capacity).Respectivedata for Big _B-ordeaux (flows greaterthan twice the channel shown in Figs. 23.18 and Creek were determinedusing the two runoff hydrographs the Because inflection point on respectively. Z3.lg,representinglow and flood flows, the represents time at which a reversalof portion of eachhydrograph the recession KSC and KSF flow directionihrough the channelbanks occurs,then the parameters water in storagein the channel. The representdaily recessionconstantsfor the is equationused in determiningboth parameters derived from a hydrologicrouting inflow and outflow during in the average equatingl/K timei the difference technique of the outflow in the channelreach.As the routing period-withthe time rate of change for shownin the figures,the low flow and flood flow routing parameters Big_Bordeaux a similar result was observedfor Cave Creek are identical. Table 23.8 showsthat Creek.

and Parameters Data Hydrologic

severalhydrologicparameters characteristics, watershed In addition to the described and an impressiveamount of hydrologicdata are required as input to the simulation of program. bue to the excessive'bulk hydrologicdata for daily evaporation,hourly pr"-lpitution, and daily streamflowfot a 4-yearperiod in BBC and a 10-yearperiod

4.0
o
d

3.0

1.0

2.0 3.0 4,0 5.0 24 Discharge hr later(cfs)

Figure 23.17 Determinationof KK24 for CaveCreek.(After Clarke.l6)

STUDIES 579 MODEL SIMULATION 23.2 CONTINUOUS

0.36 days t = routinginterval= 15min = 0.0104 0.32 o?o ,, ',/ --# = - 0'28at inflection Point 0.264 K = - f foi , = - : 6 6 = 0 e 4 3 d a l s --

x S C =K _ o . s t = 0 . 9 8 9 K + 0.5t ,r'
In{lection

E o.rt
o P0

pomr

6 0.26
E

! b0

8 0.22
0.20
0.18 0.16 0.14
' Date (October 1968)

Figure23.1sDeterminationoflowflowstreamflowroutingparamecreek low flow eventof october3, tei(rsc) for the Big Bordeaux 1968. in CC, the actual hydrologicdata valueshavebeen compiled but'are not includedin this text. , usedasinput only in the BBC applicationis the Oneusefulhydrologicparameter is flow at the gaugingsiation below which no printing of simulateddischarges hourly of desired.For the initial simulation of runoff from Big Bordeaux Creek, a MINH in to 1.0 cfs was selected minimize the printed output. This can easilybe decreased sirnulationtrials" after the uncertaintiesin someof the other parameters successive reduced. havebeen A precipitation-weightingadjustmentparameter(K1), representingthe longterm ratio of av"rage precipitation over the basin to averageprecipitation at the might ue greateror lessthan 1.0 if the gaugewere locatedat any precipitationgaugp, were within for di*tun." frornthJstudy basin.The recordinggauges both watersheds were set to 1.0. proxirtrity, and both K1 parameters close pan ivlonthly-evaporation coefficients(EVCR) are usedby the simulationprodata datato lake evaporation in convJrtinginput pan-evaporation gram asmultlpters evaporation rates)'The nearest potentiale.'upottuntpiration for determining iand also 20 pan for Big BordeauxCieet is approximately mi southat the Box Butte Reservoir,

580

CHAPTER23

CONTINUOUS SIMULATION MODELS

t = routing interval = 15 min = 0.0104 days

do =-t'' at lnrlectlon Polnt

al K=o a9 =0.873days *n - ).) du/dt KSF= K-0.5t = 0.989 K + O.st

'? !) o E

5.0
,r'

Inflection pomt

4.0
d o

oo 6
J.U

20 21 22 23 24 25 (March 1969) Date Figure 23.19 Determination flood flow streamflow of routing parameter (KSF)for theBig Bordeaux Creekfloodflow event of March17,1969. providing records of daily and monthly pan-evaporation depthsfor the months between and including May and September. Estimatesof corresponding monthly lake evaporationamountsat Box Butte Reservoirwere obtained from maps and charts developed Shaffer.22 by Valueslisted in Table23.8 are assumed apply to eachyear to of the simulation even though the simulation program allows changes from year to year. Manning's roughness parameters flow over soil and impervioussurfaces for are both required as input to the program. For the Big BordeauxCreek area,the initial estimates qverlandflow (NN) and impervioussurfaceflow (NNU) were 0.37 and for 0.013,representing coefficients dense for shrubberyand forestlitter for overlandflow and smoothconcletefor impervioussurface flow. The later n is significantonly if the fraction of impervious area (A) is nonzero. The Cave Creek analysisincorporated NN : 0.1Qfor light turf and NNU : 0.015for concrete pavement.

t6

I7

18

t9

ModelCalibration
Severalof the following parameters determined trial and adjustmentuntil the are by comparisonbetweenthe simulatedand recordedstreamflows satisfactory. is Guidelines-forestablishing initial valuesexistfor only a few of the parameters, whereas most

MODEL SIMULATION STUDIES 581 2g,2 CONTINUOUS The BBC datain Table23'8 represent ranges. from suggested are initially determined initial, unmodified estimates;those shown for CC are the optimal result of many repetitive runs. (EMIN) rangesbetween0'1 and One factor for varying infiltration by seasons and simu1.0 and has been shownby Briggs to be signiflcantin matchingmeasured parameter are for no Because guidelines estimatingthis lated winter peak flow rates.23 Bordeaux midvalue of 0.5 was selectedfor Big presently u*ilubl", the suggested Creek. requireinitial input valuesthat are Severalsoil moistureandrouting paraineters improve successively data.Calibrationprocedures from available difficult to estimate parameters For Big BordeauxCreek,the by the initial estimates trial and adjustment. initial estimatesare as follows. and EPXM-the ma*imum interception rate (in./hr) for a dry watershed. trial valuesof 0.10, 0.15, and 0.20 fot crawford and Linsley2suggest moderateforest covers,and heavyforest covers,respectively. grasslands, for ifr" O.ts-in./hr valuewas selected the moderateforestcoveralongBig Creek. Clarke used0.10 in the CaveCreek study' Bordeaux CX-an index of the surfacecapacityto store water as interceptionand from 0.10to 1.65,and normallyranges This parameter storage. l depression the midvalu" of o.go was selectedfor Big Bordeaux creek although a greaternumber might be indicative of the forest cover. Clarke independently.arrivedat a final, similar value (0.90) for CaveCreek' the capacity,representing moisturestorage EDF-an index of soil-surface during warmer months due capacityavailable additional moisturestorage to vegetation.Dependingon the soil type, the index ranggsfrom 0.45 to 2.00. Sandysoilsiimilar to thosein the BBC areareadily give up moisture capacity.lnitia,l valuesof 1.10 storage resultingin increased to vegetation, for the Big Bordeaux and cave and 1.25 were independentlyselected Creek areas. index (in.) approximatelyequalto LZSN-a soil-profile moisturestorage abovethe water table and below the ground the volume of water stored surface.This parameteris a major runoff-volume parameter,inversely related to the basin yields, interflow, and groundwaterflow. The LZSN from on index,depending porosity andthe specificyield of the soil, ranges plus half the meanannualrainfall can be usedas an 2.0 to 20.0,and 4 in. rainfall' By the use of a initial estimatein areasexperiencingseasonal 15.55 in. of annual precipitation, the Chadron, lg3I-Ig52 averageof : 11.78in. for Big Nebraskl, precipitation station givesan initial LZSN BordeauxCreek. A similar analysiswas usedfor CaveCreek' that controlsthe rateof evapotranspiparameter K3-a soil evaporation from the lower soil zone.The parameterrangesfrom 0.2 to losses ration cover.As an initial on 0.9 depending the type and extentof the vegetative which agreed for the light forestcoYerin BBC, 0.28 was selected, estimate by with the estimatis suggested Crawford and Linsley.2 Also, K3 is and _ approximatelyequal to ihe fraction of the basin covered by forest

582

CHAPTER23

MODELS SIMULATION CONTINUOUS

grassfor dep-rooted vegetation.Recommendations2 barren ground, lands,andheavyforestsare'respectively,0'20'0'23'and0'30;0'25was oPtimal for the CaveCreek studY' controllinl K24L - a Parameter from active groundwater stori It basinboundarY' also drainage groundwaterthat Percolatesto K}4LPatameter canbe estimatr are thesel0sses small to waterlevels,or it is often assumed be zerobecause and runoff' of comparedto the magnitudes rainfall K24FjL-'the friLctionof the total watershedover which evapotranspirarate' to tion from groundwaterstorageis assumed occul at the potential zerounlessa significantquantity of vegetation d is This pararneter assume water, drawswater directly from the watertable.Plantsthat seekphreatic or alfalfa, are called phreatophytes' suchas cedar or cottonwood EF-afactorrangingfrom0.lto4.0thatrelatesinfiltrationratesto evaporationrates.Thisparametersimplyallowsamolerapidinfiltration 1'0 was A rate recoveryduring raim"t seasons. normal starting value of "nig Boideaux Creek, whereasthe CaveCreek value was set for selected much lower. on the soil CB-an index that controls the rate of infiltration, depending permeabilityandthevolumeofmoisturethatcanbestoredinthesoil.A Big midvalue oi O.ZS(0.3-1,.2) was selectedfor the sandy soils around BordeauxCreek;asmallervalueof0.65wasoptimalfortheCaveCreek study. moisture cY-an index controlling the time distributionand quantitiesof a moderately enteringinterflow' This index rangesfrom 0'55 \" 4':' and the waterfor high valueof 3.0 was selected Big BordeauxCreekbecause a moderately shedcontainsmany pine needfemats. clarke also selected high CY for CaveCreek. KY24-adailybaseflowrecessionadjustmentfaclorusedtoproducea for Big An simulated"oruilin"u, baseflow recession. initial value of 1'0 the base flow recessionfor the Bordeaux creek was selectedbecause hydrographinFig.23.16islinear.Lateradjustmentsmightberequiredin matchingsimulatedand recordedbaseflow recessions' SGW-agroundwaterStorageincrement(in.),reflectingthefluctuations for Big in storaqelolume. Usually, an initial estimate(0.10 was selected and adjusted after several simulation trials; Bordeaui creek) is made parameters SGW rangesfrom 0.10 to 3.90 in. Thib and the following four studY' were usedonly in the BBC UZS-thecurrentvolume(in.)ofsoilsurfacemoistureasinterception anddepressionstorage'Becausethesimulationbeginsonoctoberlofthe as zero first calibration yeai the parametermay initially be designated unlessprecipitation occutt during the last few daysof September'

23.2 CONTINUOUS SIMULATION MODEL STUDIES

583

LZS-the current volume (in.) of soil moisturestorage betweenthe land surface and the watertable.Sixty percentof LZSN, or 7.0 in., wasselected to initiate the Big BordeauxCreek iimulation. GWS-the current groundwaterslopeindex (in.). This index providesan indication of antecedent moistureconditions.Suggested initial valuesfall between0.15 and 0.25, or the value assigned SGW can be used.A to midrange0.20 in. was selected Big BordeauxCreek. for VOLUME-the volume (acre-ft) assignedto swamp storageand dry groundrecharge, accounting the runoffrequired to recharge for swamps in late summer.Since no swampswere visible on the USGS 7.5-min topographicmaps,an initial valueof 0.0 acre-ft was selected Big Bordeaux for Creek.

HydrologicData
In addition to the parametervaluesof Table23.8, the Stanfordmodelrequiresa large volume of hydrologic data for each water year of the simulation. The following description of requried input data for each water year illustrates the data that were compiledand reducedto necessary input form for 4 wateryearsbeginningon October 1, 1968andendingon September 30,1972.Only the input for BBC is described. The input includesthe following data for eachwater year: 1. The new year identificationentry containsthe water year and the recorded annual streamflow.The valuesof the annual streamflowfor Big Bordeaux Creek are listed in Table 23.9. 2. A descriptionofthe streamgaugingsite. 3. The title to be appliedto the ordinatefor graphicalplotsof the simulatedand recordedrunoff hydrograph;namely,the daily average flow rate (cfs). 4. Pan-evaporation data, read as 365 or 366 single entries containing daily evaporationamounts(in.). 5. The monthly evaporationpan coefficients,comprising the data listed in Table 23.8, beginningwith October. (average 6. Recorded daily streamflows flow for the day,cfs)readas 365or 366 entriesfor October I through September 30. 7. Hour$ rainfall data, read for each water year. Two entries per day, each containingan identificationofthe gaugeanddate,areusedto providehourly TABLE 23,9 ANNUAL BORDEAUX BIG CREEK STREAMFLOWS
Water years Recorded annual streamflow (acre-ft)

1968*1969 1969-1970 r970*1971 r97t-1972

434.0 465.4 296.4

584

23 SIMUI.ATIoN MoDELS cHAPTER CONTINUOUS depthsin inchesbefore noon on the first and after noon on the second.Valuesare precipitation.Because the variablenumof requiredonly for half-daysexperiencing ber of possible rainfall values,a sentinelentry with the year setequalto 2001 is placed at the end of the data, indicatingthat all the precipitation data for the wateryear has beenread.

Version Outputfrom the Kentucky

are a Dependingon which optimal input, output, and branchingparameters selected, variety of output data are availablefrom the Kentucky version, including plotted graphs of measuredand synthesizeddaily streamflow rates. Options include the following: . 1. A table of synthesized average daily streamflowrates (cfs). daily flow rates. 2. A table of monthly and annual totals of synthesized monthly and annualtotals ofrunoffin equivalentinchesover 3. Synthesized the watershed. 4. Svnthesized monthlv and annual interflow amounts (in.) over the watershed. 5. Svnthesized monthlv and annual baseflow amounts(in.) over the water-

s[eo.
streamflowrunoff from the watershedfor the 6. The volume of synthesized entire water year (acre-ft). 7. A summationof all the recordeddaily streamflowrates(cfs)for eachmonth and year. 8. The recordedannual total of runoff (in.) over the watershed. 9. The recordedvolume of runoff from Novemberthrough March (in.) over the watershed. 10. The amountof synthesized snowfrom Novemberthrough March (in.) over the watershed. 11. The volume of the recordedannual streamflow(acre-ft). 12. The sum of the recorded precipitation for each month and for the year. (in.). 13. The synthesized monthly and annual totals of evapotranspiration 14. The monthly and annualrecordedlake evaporationamounts(in.). (in.). moisturestorage 15. End-of-the-monthlevelsof UZS, the current surface 16. End-of-the-month levels of LZS, the current soil moisture storage(in.). 17. End-of-the-month values of SGW the current groundwater storage fluctuation(in.). 18. An annual moisture balance (in.), which representsthe moisture not accountedfor within the program.This is illustratedin the CaveCreekoutput.

GaveCreekModelCalibration
Synthetic and actual flow rates at the Cave Creek gauging station are shown for a singleday in Fig. 23.20. Other typical output for portions of one water year of the an is in simulation presented Tables23.10and23.ll. The former provides hour-byof MINH, Table23.8. Note value of hour listing of all flow ratesin excess the specified

SUMMARY ^ t20 Syr thetic -r r 100

585

a 8 0 {, o s 6 0 E 4 0

t/

l0

12 Time (hr)

I4

16

18

for and hydrographs Figurp 23.20 Comparison synthesized recorded of (After C1arke.t6.1 Cave Creek. MINH that the streamflowwas zero from Octoberthrough Decemberand exceeded only during two daysin January. in Table23.ll containsmostof the informationdescribed the 18 items of outpuf for the Kentuckyversion.The daily flows are followedby the syntheticand recorded monthly totals, monthly interflow and base flow amounts, monthly precipitation in'the totals, monthly actual and potential ET amounts,and end-of-month storages soil profile and groundwaterzones in inches. Of particular interest is the annual summaryin the lower right. Of the 37.5 in. of precipitation,23.8 in. went to ET, and 11.6in. ran off or was discharged from storage, the remaining2.1 in. recharged for the soil profile. Moisture not accounted during the year was 0.0844 in.

of to Changes Sensitivity ModelResponse Parameter

One interestingand useful aspectof simulation is the easewith which changesin watershedparameterscan be evaluated.Clarke testedthe sensitivity of KWM by rangeswhile holding varying severalof the parameters Table23.8 over reasonable in all other parameters constant.The resultsof his analysisfor CaveCreek on a typical were taken from day in March are summarizedin Table 23.12. Theseobservations graphssuchasFigs.23.2l and23.22,which illustratethe sensitivity flood magniof tude and timing to changesin Z and KSC. These results and the summary in Table 23.12 are applicableonly to the CC watershedand should not be viewed as generally applicable.

r Summary
precipitation recordsare available,one of the most effective If actual or synthesized flowsunderchanging meansof analyzinghistoricalflowsor evaluatingfuture possible land use patterns is through any of the continuousstreamflow simulation models described this chapter.Hydrologicproblemsand applicationsthat can be analyzed in

=
u a/) u r m
@ F S N

r <

Yi

rt

=
r
F O @ s
h N

z g
U'

z
t/) q

r'o

r s

r o

6 0

? o
r a
t!
@ F

s o

u
F

r @
@ @

r <

o
E

il

o t!
z

s
6
tr
l! a @ r

6 o N

- Y n u rlf
T

trul

o r a @
@

u
F

e ! q

oo
6-> z < d o
cc lL >ul
+ u llj

F qt

o o

@ r
F @

U)
L

u, r

N T + a i

o
L

m 6

U'
L

>z
Y Y F

O-r
= O

=
iE o U

=
J 4

=
E U r

=
J

o r

N Q

=
J

E> 9o 6H
) a

u u N T ll

c.l \ + m

N 6
D

z
CE

2
F

U
N T

ll

<tr ( )cu g < f =

o

:z u
u u

z x u J
u

=
x

o 6 @ o

+ei

= x =

< L

<o-

<o-

>o = < -

at N ut o

z >z U
u u E
o

(-)

u
h N N

u
4

U
L

@ o o

o
N

u U' E] k

O N

F S

+ o

N O @ F @ <

s o
O N

o r
N F

o @
@ N

r o @ o s(9

=
r

ri
N

qt
O

z.
U)

F @ @ +

+ct N

rtri

N @ o q i

z.
U) u \z
N

+ o
@ N

o r

bts O N

a N @ N F

c? r?

o r o o

I o o U u

N @ o @

<t=
6io F 5
4 F O N N
N

@
F

qq? < @

o r e @ + @

cq aj

N +

+ s 6i ci r D

r < f @ O N

oq\
N N @ -

\ a

o +

k =
o? e?
@ r o o d o i F @
@ o t s o O N

c!q o o
N -

qo?
@ h

< o
F

< o

< N O N

a q

9 O

r N

@ o

O F @ld)

ddi

=
CL

q q b s

o r

cqa < 6
6 6

r o

o
N

-tO

d -

o o

o < Ei+
@

u
F

o q O F

ctF @ h

< i o

qq o? + o
N F

u
r F @ q

o @ o m

d + u !z CE
J

v) r

o O N

U) r

\ q
F O

(tt r

<

tn
g

< f s

o? oj s e
N N

v)
L

@ t l

e? 6{ O F

CJ''

Y F

!z

rt+

F M

O N

F N N < f m

o @ @ o N N

qo{ rc9

t\

- h < f o

z
q)

r Y

d
F

=
iE - q 6 <
e N

il

O N

s
8
o
I (D N

z
v u u u

o r

@ o ri ct N N

N T

r +

o?F

- f O

E
u < L

=
x

>=
<o-

= >
< L

= =
<o-

x <o<o-

= =
x

!) 't (J

z. :z u
u r U
q

ul o

E61

'

H<
6 = > z o <
U
T
q

cn

=
Y F F Z.L U U YCn @ o

o A A
Q

a a

3-

a
u

? z
o a

- o

@ @

il>

F <

tr
lJ-

@ r o o

@ +

< m

o o

N N

F l

f (E

@ r

r @ o o

t s o

N F

O O

o o

r @

O N

m F a r

M N o o N O

r o r @
N F

o F @

O F @ F

6 e

:)

U) ]U TE l r q)

o @

@ F

ai 6i

r N N O

+ o

9o r :
I > > a Ful
U Y
J

o o

o r

@ A

F F

o r

+ o

+ N

=E
O(r zlu

>- (-.) - U

3o
z.
u :z> -o 1 Z
F

eid

1k
-= { a OIE -r9 <= > a F
C\| IJJ

z.

o o

o ci et

?e
dt

E :z u r cc u
O

z . >< ir
E u

e o

o o

o o

o
o

o o

z. u>u < u
N O

+ 6

r @

N N

O N

a B u) t)

CD L)

u u

(N

H =
u

a z= >
o <
U' u

b z

z z

L)

tr

u v, a)

o
F f

U) u

U) u

v) v) U U

v, u

3 z

< =

z.

z.

z. z.

E =
F Z u Y, F L u Vt

m r r

+ Oo F
@ F

@ 6

<t F

r
N

@ r

<f o

o @ N N

- a

+ =

6 N N

o h
N F O N

o o

o o O J

N O

O N O N O N

+ o
N O

+ I

? z
U

r + o
N

r r
F

o o o r O N

+ o
N O

F >

tr< =
cc r @ r
F N

+ N N O

@ F @

(Q o

F O < i e
O F

m + o o +
@ F

6 F

F O

F @ N F N N F O N

o
o
N

N O N S r N

r o-

o +

@ F

o + o $ + o

o
N

o o *

o m

o o

r @ d d
F

o r
@

r
E

6 N

E N + 6

r r

F @ @ N O F

o o

z.
N q F

o r
F O

@ h

o r o +

r
N

o + @ d6i

o + + a

:z z

-H 3o
z v > z. z.
u
F

N
Q

r
o

F F O

N O F

o o

r o
N

r r
F O

^cn 6 -

'2
F

z .
u U u

x5
u z
Y

o +
o

e o o

@ N

o q
@ e o

HB
- o t 4 U

v) u)
N N t/) u F

u>u < o u

+ D
N N

r @ N N

=
N O

q
L

u
c) u

z z.
L

=
@

z.
u F

Ct) 7 (/j= N ;

zc) - z

Q q

U) U a
p 4

z + > o ant

= d

z.
6

E
- X ; i - t

Hg esBE.g sEBB-A' E
ut E
rj'tr
d o O

U U E

6 d d 6

A22:EAfr22EEEA

t55t
A a

I g 3 .bE !8 3 3 ' d sb.o;F I 9 E t e 0 h h EE

o
tll

u = - 6 U g o o o = o : : &&,&thl,E'ji/oZqMrh

o o
E E ! ! ! - -

I I I I93'ts

! E ! E

333E SKf r " , $$S$F

I I I I I

o o IU

AAAAAi!;EAAAA
ttErc o o 0 ) o ! ,r

(E

o) (t>

(u
o

|r
U)

tr
lll

.EQ,ae ilf H !
o c (t E

t-

ti vj I I +6Nati i r i #; t 'Ai c |i N | c . l l -6 -6 - 6- 66- tr e o . A o o !> Q ' <

E^,3E

ut
E. n q Y fuJ

r,J i dZZZZ-Z=a=

oH
> lrJ YA < 4
o o o o IE

9r. DE

8> 6 z

E (u
N I N

o o

c,

E
6 o

. 7

tu o
F

= Y

23.2 CONTINUOUS SIMUI-ATION MODEL STUDTES

591

6000 5000

4000

5 rooo
2000 1000 0
Time (hr)

Figure 23.21 Sensitivityof modelresponse the length-of-overland-flow to parameter.(After Clarke.16)

5000 ^ 4000

KSF = 0.85

5 :ooo
2000 1000 0
10

I
12
Time (hr)

\ \
S F = (t.99-

t4

16

18

20

22

Figwe 23.22 Sensitivity model of response thechannel to routingparameter. (After Clarke.16) usingcontinuous modelingincludewatershed yield studies, reservoirdesignand operation studies,sedimentyield estimatingfor reservoirdesignor analysisof impactsof erosion controls on water quality, water supply studiesfor municipal, industrial or agricultural demands, litigation over impactsof wellfieldsor direct diversions, determination of hydropower production potential, evaluationsof flows that will pass through critical in*stream or riparian habitat reachesof a stream,identification of flows that will be availablefor recreationalusesof a stream,and, amongnumerous other applications,analysisof water quantity and quality impacts of removingdams or rnaking other major upstreamchanges. The modelsdescribed this chapter,and other similar continuoussimulation in codes,are availablefrom the federal or state agencies that originatedthe code, or from numerouspublic outlets or vendorswho havebeen authorizedto distributethe software.

592

MODELS SIMULATION 23 CONTINUOUS CHAPTER

Problems
percentages the period. Reconstruct block diagram ofFig. 23.2 and plot approximate io show,for averageconditions,how the rain would be distributed(a) initially and (b) after 30 days. rain 23.2. A sloping,concreteparking lot experiences at afate of 3.0 in.i hr for 60 min' The ftlft. If the water detentionon the lot is ft tot is SOO deepand has a slopeof 0.000-1 for completeoverlandflow hydrograph 1 ft zero at the start of the storm,calculatethe the SWM-IV equations.Use a 5-min routing interval and continue of width using computationsuntil all the detainedwater is discharged. 23.3. Calculatethe SWM-IV overlandflow time-to-equilibrium for the lot of Problem23'2 and compareit with the Kirpich time of concentrationfor the lot. Should thesebe equal? 23.4. Cornpare,by listing traits and capabilitesof each,the SWM-IV with its more sophisticated offspring HSP and HSPF. 23.5. Discussthe primary differencesamong the four versionsof the Stanfordwatershed in model described this chaPter. 23.6. Verify Eqs. 23.23 and 23.24by starting from Eqs. 13.4 and 13 '33' "typical" 23.7. Discuss the watershedbehavior that is depicted in Fig. 23.7. Is this a watershed? betweenthe two U.S. Departmentof Agriculture continuous 23.8. Comparethe differences that would be the simuiationmodels,USDAHL and SWRRB,and discuss applications best suitedto each. in 23.9. Review the differencesbetweenwater budget and simulation models discussed here simulationmodelsdescribed which of the continuous Chapter2l anddetermine could be usedto perform water budgetcalculations. 23.10. For what applicationsmight the following be best suited? API model USDAHL HSPF PRMS SWRRB by 23.1r. For the continuoussimulationmodel selected your instructor,describefour different types of problemsthat could be analyzedif you were given the ful1, calibrated model. h in that 23.r. Assume a 30-mi2rural watershed your localereceives 3-in. rain in a 10-day

REFERENCES
"ContinuousHydrographSynthesis with an and 1 . W. T. Sittner, C. E. Schauss, J. C. Monro, Res'5(5), 1007'1022(1969)' API-Type Hydrologic Model," WaterResources "Digital Simulationin Hydrology:StanfordWaterL. N. H. diawford and R. K. Linsley,Jr., shedModel IV," Departmentof Civil Engineering,Stanford University, Tech. Rep. No.
J.

3 9 .J u l y 1 9 6 6 . "An Evaluationof Relationship and Watershed BetweenStreamflowPatterns L. D. James, the Use of OPSET," ResearchRep. No' 36, Water Resources Through Characteristics Institute, University of Kentucky,Lexington, 1970'

Chapter24

Simulation Single-Event Models

r Prologue
The purposeof this chaPteris to: . Describehow storm event models are structuredand how they are used to for simulatedirect runoff hydrographs singlestorms. models(note . Describethe five most widely usedfederal agencysingle-event popular single-eventurban runoff simulation models are describedin that Chapter25). ' Provide a detailedcasestudy using one of the models,HEC- 1' . Introducethe emergingtechnologyof storm surgemodeling-the simulation of hydraulic surgesresulting from wind energy acting on the ocean surface. high-intensityrainfall events. by floodsare caused short-duration, Many severe model simulatesrunoff during and shortly following these watershed A single-event in modelsare normally interested the peak Usersof single-event discreterain events. flow rate, or the entire direct runoff hydrographif timing or volume of runoff is modelssimulatethe rainfall-runoff processand makeno special Single-event needed. to account for the rest of the hydrologic cycle. Few, if any, simulate soil effort occurring beinterflow, baseflow, or other processes moisture,evapotranspiration, discreterainfall events. tween that to in Modelsdescribed this chapterare applicable studiesof watersheds are that are primarily rural in makeup.Urbanizedsubareas allowed,but for watersheds principally urbanized, the single-event and continuous models described in ire createdby Chapter2i aremoreapplicable.Coastalfloodingthat is inducedby surges action on the ocean surfaceis modeledby a different class of single-event wind in described Section24'3. models,

SIMULATION EVENT 24.1 STORM

Event simulation model structurescloselyimitate the rainfall and runoff processes suchasunit-hydrograph approaches, Lumpedparameter in developed earlierchapters. methods,are generally incorporated even though some use distributed parameter

SIMULATION595 EVENT 24,1 STORM begins simulationstudies for Preparation implementingmost single-event techniques.. as subbasins illustrated inFig' 24'1' subdivisioninto homogeneous with a watershed Computationsproceed from the most remote upstream subbaslnin a downstream direction, modelfor a typical basin(Fig. 24.I),the runoff hydrographs In any single-event and A, for eachof subbasins B, . . . , E are computedindependently, then routed and through(callednodes)to obtain designhydrographs cornbinedat appropriatepoints for parameters the storm; then appliesthe storm out the basin.The modelieads input Bf to the first upstreamsubbasin, computesthe hydrographresultingfrom the storm A; event; repeatsthe hydrographcomputationfor subbasin combinesthe two comroutesthe hydrographby conventional into a singlehydrograph; puted'hydrographs t[rough reachC to the upstreamend of reservoirR, whereit is combined iechniques with the compuied hydrographfor subbasinC; and so on through the procedure inFig.24.l. detailed are Hydrographcomputationsfor subbasins most often determinedusing unit24.2. The precipitation hyetographis hydrogiaph proceduresas illustrated in Fig. leaving are input iniiormly overthe subbasinarea,andptecipitationlosses abstracted, (see Chapter 12) with the un "*..r, prelipitation hyetographthat is convoluted

Outlet (point of concentration)

reservoirsites,damagecenters' 1. Subdividebasinto accommodate divides,gaugingstations' points, surfaceand subsurface diversion features' land uses,soil types,geomorphologic precipitationstations, in sequence eventsimulationmodels: 2. Lomputation B' a. Computehydrographfor subbasin A. b. Computehydrographfor subbasin forA and B. c. Add hydrograPhs end of reservoirR' d. Routecomtined hydrographto upstream b. Compute hydrograph for subbasinC. D. f. Computehydrographfor subbasin at g. CombinethreehYdrograPhs R. h. Route combinedhydrographthroughreservoirR' i. Route reservoir outflow hydrograph to outlet' E' j. Computehydrographfor subbasin at k. Combinetwo hydrographs outlet. Figure 24.1 Typical watershed subdivision and computation sequence for event-simulation models.

Precioitation

- ?l H g J'VhYetograPh a
Time

s
/
Outlet
/ Gross precipitation Excess (net

Time

I c"-bt". I - -

\Z

hydrograph Total streamflow

modelof the rainfallevent-simulation Figure24.2 Typical lumpedparameter runoff process.

MODELS 597 AGENCY SINGLE.EVENT 24.2 FEDERAL prescribedunit hydrographto producea surfacerunoff hydrographfor the subbasin. lossesare divided among the loss componentson the basis of preThe abstracted flows and waters derived from groundwaterstorage Subsurface scribedparameters. which when combinedwith the runoff hydrograph, are transformedinto a subsurface the total streamflow hydrographat the subbasin surfacerunoff hydrographforms outlet. This hydrographcan then be routed downstream,combined with another contributing hydrograph,or simply output if this subbasinis the only, or the final, subbasin beine considered.

'

MODELS AGENCY SINGLE-EVENT 24.2 FEDERAL

by in depicted Figs.24.1and24.2 arerecognized most The rainfall-runoff processes techniques of the eventsimulationmodelsnamedin Table21.1. Specificcomputation river routing, reservoirrouting, and baseflow are comunit hydrographs, for losses, paredin Table24.1for five of the major federalagencyrainfall-runoff eventsimulaBrief descriptechniques. amongavailable tion models.All the modelsallow selection of an application tions of eachof thesemodelsare followedby an illustrative example of the HEC-I model to a single storm occurring over a 250-mf watershednear Lincoln, Nebraska.

Model SurveyRainfall-Runoff U.S.Geological

The USGSmodel can be usedin evaluatingshort streamflowrecordsand calculating basins.lThe programmonitorsthe daily moisture peak flow ratesfor natural drainage soil and can be used as a continuousstreamflowsimulation content of the subbasin its model.The model is classifiedas an eventsimulationmodelbecause calibrationis during a few evaporation,and discharges based on short-termrecords of rainfall, into the USGS severaltimes and hasevolved floods.It hasbeenmodified documented in urban continuoussimulation model, DR3M, described detail in Chapter25. of initial estimatesof 10 parameters,which are Input to the model consists modified by the model through an optimization fitting procedurethat matchessimulated and recordedflow rates. Other input includesdaily rainfall and evaporation, close-interval(5-60 min) rainfall and dischargedata, drainageareas,impervious areas,and baseflow rates for eachflood. hyetograph, Phillip's2infiltration equationis usedto determinea rainfall excess and then routed through a linear reservoir, outlet which is translatedto the subbasin in routing techniquedescribed Chapter 13. using the time-.areawatershed for can be usedto simulatestreamflows relaThe USGSrainfall-runoff model linear storage-outflowcharapproximately tively short periodsfor small basinswith in acteristics regionswheresnowmeltor frozen groundis not significant.Output from storm runoff volumes, storm the model includes a table showingpeak discharges, of printout of magnitudes parameters rainfall amounts,and an iteration-by-iteration peak flow rates. and residualsin fitting volumesand

598

CHAPTER24

MODELS SINGLE-EVENT SIMULATION

TABLE 2II"1 HYDROLOGY PROCESSES AND OPTIONSUSED BY SEVERAL MODELS AGENCY RAINFALL-RUNOFFEVENTSIMULATIONS Table 21.1) Model code names (slee HEC-1 (Corps)

TR-20

Modeledcomoonents Infiltration and losses Holtan's equation Horton's equation Green-Ampt Phillip's equation SCS curve number method Exponentiallossrate Standardcapacitycurves Unit hydrograph Input Clark's Snyder's Two-parametergamma response SCS dimensionless unit hydrograph River routing Kinematic wave Full dynamic wave Muskingum Muskingum-Cunge Modified Puls Normal depth Variable storagecoefficient Att-kin method Translation only Reservoirrouting (Puls) Storage-indication Baseflow Input Constantvalue equation , Recession Snowmeltroutine

(scs)

USGS (USGS)

HYMO (ARS)

SWMM (EPA)

x
X

X X X X X

X X X Yes
No

X X
No Yes

for ComputerProgram ProjectFormulation Hydrology(TR-20)

A particular$ powerful hydrologicprocessand water surfaceprofile computerprogram was developed CEIR, Inc.3and is known by the codename TR-20, which is by Nuntber20. The ServiceTechnicalRelease an acronymfor the U.S. Soil Conservation model is a computerprogram of methodsusedby the Soil ConservationServiceas presented lhe National Engine in ering Handbook.a 'is recognized as an engineer-orientedrather than computerThe program with easeof use as a purpose.Input data oriented package,having been developed for sheetsand output data are designed easein use and interpretationby field engi-

MODELS 24.2 FEDERALAGENCY SINGLE-EVENT

599

neers,and the programcontainsa liberal numberof operationsthat are user-accomof modating,even at the expense machinetime. The TR-20 was designedto use soil and land-use information to determine for runoff hydrographs known stormsand to perform reservoirand channelrouting of model,with no provisionfor additional It hyOrograpns. is a single-event the generateO events.The programhasbeenusedin all or losses inflltration beiweendiscretestorm for 50 statesby engineers flood insuranceand flood hazardstudies,for the designof planning. reservoirand channelprojects, and for urban and rural watershed historicalor syntheticstormusingthe SCS Surfacerunoff is computedfrom an The standarddimenlosses' in described Chapter4 to abstract curvenumberapproach for is usedto developunit hydrographs each hydrogiaphshownin Fig. 12.13 sionless hyetographis constructedusing the subareain the watershed.The ixcess rainfall effectiverain and a given rainfall distributionand is then appliedincrementallyto the runoff hydrographfor th9 storm' unit hydrographto obtain the subarea method to route As shown in Table 24.I, TR-20 usesthe storage-indication 13.2).The baseflow is addedto the direct (seeSection throughreservoirs hydrographs .onoffhyatographs at any time to producethe total flow rates.The programusesthe routing the flows logic depictel in fig. Z+.iby computingthe total flow hydrographs, with those the routedhydrographs combining and thiough streamchannels reservoirs, outlet' to hydrographs the watershed from6ther tributaries,and routing the combined by hydrographs the convex method Prior to 1983, the model routed stream inflow (Secby (Section13.1),which has sincebeenreplaced the modifiedatt-kinmethod tion 13.3).Asmanyas200channelreachesandggreservoirsorfloodwater-retarding in can structures be accommodated any singleapplicationof the model.To add to this the capability,the programallowsthe concurrgntinput of up to 9 different stormsover area. watershed is Subdivisionof the watershed facilitatedby determiningthe locationsof control points. Control points are defined as stream locations correspondingto crossor sectionaldata, reservoirsites,damagecenters,diversionpoints, gaugingstations, data requiremay be desired.Subarea where hydrographdata tributary confluences ments include the drainagearea,the time of concentration,the reachlengths,strucin-Section 21.1, andeither routing coefflcientsfor eachreach ture data as described data along the channels.Whenevercross-sectionaldala are proor cfoss-sectional peak flow in vided, the model calculatesihe water surfaceelevations addition to the Subareasizes are dictated by the rates and time of occurrence at each section. it is locations of control points. To provide routing and flood hazatd information, of so that the hydraulic characteristics the to necessary defineenoughcontiol points instreamaie definedbetwJencontrof sections.Applications with TR-20 normally few hundredfeet to 2 mi or more apart' The between-a points spaced corporatecont*rol mi2' that ."rolting subarLas contributerunoff to a control point are usually lessthan 5 less than 25 mi2 even though there is no Common subareasizesfor structuresare limitation on reachlength or subareasize within the program' characteristics, to Minimal input daia requirements TR-20 includethe watershed including the depth, duration, and distribution; at leastone actual or syntheiicstorm routing the discharge,capaciiy, and elevation data for each structure; and the

600

CHAPTER24

MODELS SIMULATION SINGLE-EVENT

according data for eachreach.Input can be described coefficientsor cross-sectional to the following outline: characteristics. 1. Watershed a. The area(in mi2)contributingrunoff to eachreservoirand crosssection. (SeeChapter4.) b. Runoff curve number CN for eachsubarea. coded with eachsubarea, moisturecondition associated c. The antecedent as dry, normal, or wet. (hr). d. The time of concentrationfor eacfi subarea mainstream. and subarea e. The length of eachchannelrouting reach 2. Velocity-routingcoefficienttable. a. A table containingrouting coefficientsfor a rangeof velocities(ftlsec). This table is containedwithin the program and need only be enteredif the user desiresdifferent velocities. hydrograph. 3. Dimensionless a. This table is containedwithin the program and need only be enteredif the user desiresa different hydrograph. 4. Actual hydrograph. a. Actual hydrographscan be introduced at any point in the watershed. rates (cfs) spacedat equal Hydrographordinates are read as discharge time incrementsapart, up to a maximum of 300 entries. b. Baseflow rates (cfs)can also be specifiefl' 5. Baseflow. with a. In additionto the option of specifyingthe baseflow ratesassociated can be specifiedor modified a hydrographthat was input, the baseflow at any other control Point. 6. Stormdata. a. Stormsare numberedfrom 1 to 9 and are input as cumulativedepthsat equally spacedtime increments. b. As an alternativeto specifyingcumulativedepthsat varioustime increstorm can be input, and up to 9 stormscan be ments, a dimensionless by synthesized specifyingeachstorm depth and duration' data. 7. Streamcross-sectional a. Up to 200 crosssectionsmay be input for a singlerun. Cross-sectional versusflow area. data consistof up to 20 pairs of valuesof the discharge routing coefficientsare deterb. If cross-sectionaldata are provided,the of mined from them. In the absence suchdata, the user must specify a routing coefficientfor eachreach. 8. Structure.data. rates(cfs) a. The rbservoirdataconsistof up to 20 pairsof outflow discharge versusstorage(acre-ft). b. A maximum of 99 structuresare allowed in a run. The desiredoutput from TR-20 must be specifiedby a set of input file control at Hydrographs eachcontrol point for eachstormcanbe printedby specifyvariables. ing the control point identification in the control cards. Any combination gf the following items can be producedat eachcontrol point:

MODELS 601' AGENCY SINGLE-EVENT 24,2 FEDERAL * . 2. 3. 4. rate, time of peak, and peak watet'surfaceelevations. The peak discharge rates in tabular form for the entire hydrograph' The discharge for Water surfaceelevations t'heentire duration ofJunoff' The volume of direct runoff, determinedfrom the area under the hydrograph. 5. Hydrographordinatesin any specifiedformat. 6. Summary tablescontainingwater balanceinformation'

from field surveys' by Basicdata needed the computerprogramare determined Rainfall frequency data are input from daia in the U.S. Weather Bureau TP-40 information for presentand future condiand report.s Peak-discharge area-flooded tions for severalreturn periodsare output by TR-20 in a form suitablefor direct use in an economicevaluationmodel.

for ComputerLanguage Hydrologic Problem-Oriented Modeling(HYMO)

who haveno convenfor A unique computerlanguagedesigned use by hydrologists experiencewas developedby Williams and Hann.6 tional computir programming oJ Once the progru- has been compiled,the user forms a sequence commandsthat plot, or add hydrographsfor subareasof any watershed. synthesiz{ route, stoie, to commandsare availableto usein any sequence transformrainfall data Seventeen through streamsand.reserroute thesehydrographs and to into runoff hydrographs voirs. The HYMO model also computesthe sedimentyield irt tons for individual stormson the watershed. Watershedrunoff hydrographsare cornputed by HYMO using unit-hydroaccordingto can either be input or synthesized Unit hydrographs graph techniques. shown in Fig. 24.3. Tetms in the equationsare unit hydrograph the dimensionless
1.0
. i l - I

q = q' " ( - \
\'p

ttr-'trtrtr-tt
I

t6 inflection point

s ls'o

q = qos(to-t)tx

tp

unit hydrographused in HYMO. Figure 24.3 Dimensionless (After Williams and Hann.6)

602

CHAPTER24

SINGLE-EVENT SIMULATION MODELS

4 : flow rate (ft3lsec)at time t q, : peak flow rate (ft3lsec) te = time to peak (hr) n : dimensionless shapeparameter q6 : flow rate at the inflection point (cfs) /o : time at the inflection point (hr) K : recession constant(hr) Once K and to and 4oare known, the entire hydrographcan be computedfrom the three segment equationsshownin Fig. 24.3. The peak flow rate is computedby the equation
u ^ - -

BAO

(24.r)

parameter,related to n as shown inFig.24.4 where B : a watershed A : watershed area (mi2) O: volume of runoff (in.), determined by HYMO from the SCS rainfall-runoff equationdescribed Chapter4 in t The duration of the unit hydrograph equated is with the selected time increment.The runoff Q for the unit hydrographwould of course be 1.0 in. The parametern in Fig. 24.4 is obtainedfrom Fig. 24.5. Parameters and to for ungaugedwatersheds K

500

50

10
0 2 4 6
n

Figure 24.4 Relation between dimensionless shape parametern and watershedParameterB. (After Williams and Hann.6)

MODELS SINGLE-EVENT 24.2 FEDEMLAGENCY L2

603

10

. K tp

' shape Figure 24.5 Relation betweendimensionless to parameter and ratio of recession constant time to n (AfterWilliams Hann.6.1 peak. and located equationsbasedon 34 watersheds are determinedfrom regional regression ranging in in Texas,Oklahoma, Arkansas,Louisiana, Mississippi,and Tennessee, size from 0.5 to 25 mi2, or

i.'t sLp-o (#,)"K : 27.oAo 231

and

(24.2) (24.3)

! Ao tp : 4.63 +"5rr-o''u(

)t'"

where SLP : the differencein elevation(ft), divided by floodplain distance(mi), betweenthe basin outlet and the most distant point on the divide LfW : the,basinlength/width ratio River routing is accomplishedin HYMO by a revised variable storqge The continuity equation,I - O : dS/dt, and the storage cofficient (VfC) method.T : KO, are combinedand discretizedaccordingto the methodsoutlined equation,S the the in Chapter 13. The VSC methodrecognizes variability in K as the flow leaves floodplainand valley area.Relations the confinesofthe streamchannelandinundates data, or by betweenK and O are determined HYMO from the input cross-sectional equation if the floodplain and HYMO will calculatethe relation using Manning's The bed slopeand reachlength are also coefficientsare specified. channelroughness part of the required input.

604

CHAPTER24

MODELS SIMULATION SINGLE-EVENT

method(seeChapter 13) is usedto route The Widelyadoptedstorage-indication curvemustbe determined The storage-outflow throughreservoirs. inflow hydrographs externallyby the userand is input to the prolram asa table containingpairedstorages and outflows,using storagedefinedas zero wheneverthe outflow is zero. for and the datarequirements eachcommandare comrnands The user-oriented as follows: 1. Sta;t: the time rainfall beginson the watershed. 2. Storehydrograph:the time incremenfto,be used,the lowestflow rate to be flow rates spacedat the stored, the watershed atea, and the successive specifledtime increment. 3. Develophydrograph:the desiredtime increment, the watershedatea, the channellength and maximum SCSrunoff curve numberCN, the watershed differencein elevation,and the cumulativerainfall beginningwith zero and at accumulated the end of eachtime incrementuntil the end of the storm. identificationnumber,numberof 4. Computethe rating curve:cross-sectional points in the crosssection,the maximum elevationof the crosssection,the main channel and left and right floodplain slopes,Manning's n fot each segment,and finally pairs of horizontal and vertical cpordinatesof the points describingthe crosssection. the 5. Reachcomputations: numberof crosssectionsin the routing reach,the time increment to be used in routing, the reach length, and the discharge rates for which the variable storagecoefficientis to be computed' 6. Print hydrograph:the idpntification nurnberof the cross sectionat which are hydrographs to be printed. at the 7. Plot hydrograph: identificationnumberof the crosssections which the hydrographs to be plotted. are 8. Add hydrographs:the identification numbers of the hydrographsto be added. 9. Route reservoir:the identificationriumbersof the locationsof .the outflow relation for the reservoir. and andinflow hydrographs, the discharge-storage L0. Compute travel time: the reach identification number, reach length, and reach slope. 11. Sedimentyield: severalfactors describingsoil erodibility, cropping management,erosioncontrol practices,slopelength, and slopegradient. Output from HYMO includes the synthesizedor user-providedunit hydrothe graphs,the storm runoff hydrographs, river- bnd reservoir-routedhydrographs, Hydrographs and the sedimentyield for individual storms on each subwatershed. from the 34 test hydrographs computedby HYMO comparedcloselywith measured watersheds.

Model (SWMM) Storm Water Management

The EnvironmentalProtectionAgency model,SWMM,s is Hstedin Table21.1 in two to locationscorresponding rainfall-runoff eventsimulationand urban runoff simulation. The model is primarily an urban runoff simulation model and is describedin detail-in Chapter25.

MODELS 605 SINGLE-EVENT AGENCY 24.2 FEDERAL numefousmodiflcationsand Lilie mostothers,the SWMM modelhasundergone was a single-event improvementssinceits first releasein 1972.The initial version8 allow its ri'sein continuousmodeling of urban storm model, and newer versionse,lo includesa new snowmelt The latestrelease waterflows and waterquality parameters. scour and routine, a new storm watei tto.ugt and treatment package'a sediment depositionroutine, and a revisedinfiltration simulation' androuting routinesarehydraulicratherthan hydrologic' swMM's hydrograph consisitingof single A distributed parameier approachis used for subcatchments as parking lots, city lots, and so on. Accumulatedrainfall on theseplots is first routed or closed overlandflow to gutter or storm drain inlets,whereit is then routed asopen Of the five channelflow to the receivingwatefsor to sometype of treatmentfacility' detail greatest modelscoripared in Table 24.l,the SWMM givesthe event-simulation in simulation,but cannotbe usedin large rural watershedsimulations. step using overland flow depths and flow rates are computedfor each time depth over a Manning's equation along with the continuity equation. The water a will increale without inducing an outflow until the depth reaches subcatchment subover the specifieddetentionrequirement.If and wheneverthe resulting depth than the specifieddetentionrequirement,Da, an outflow rate D., c^atchment, is largei is computedusing a modified Manning's equation
V:-J1-(D,-

r .49.

D,)2/3St/2

(24.4)

and

Q*:

vw(D, - D)

(24.s)

where V:thevelocity n : Manning's coefflcient S : the ground slope W : the width of the overlandflow rate Q. : the outflow discharge have been computed,they are After flow depthsand rates from all subcatchments to form the total combinedalong with the flow from the immediateupstreamguttef gutter. flow in eachsuccessive gutter and PiPeflows are routed The to any points of interestin the network wh ordinatesfor each time step in the routir incrementsuntil the runoff from the storm ters of the gutter shape,slope,and length availJoef{cienti ior ttre pipes or channelsmust also be suppliedand are roughness able in most hydraulicstextbooks' includethe Other iniut requiredfor a typical simulationwith the SWMM model following: percent imsuchas the infiltration parameters, characteristics 1. Watershed pervious area, slope, area, detention storage depth' and Manning's coefficientsfor overlandflow' for 2. The rainfall hyetograph the storm to be simulated'

606

CHAPTER24

MODELS SINGLE-EVENT SIMULATION

and 3. The land-use data, averagemarket values of dwellings in subareas, populationsof subareas. 4. Characteristics gutters such as the gutter geornetry,slope, roughness of coefficients,maximum allowabledepths,and linkageswith other connecting inlets or gutters. 5 . Street cleaningfrequency. and their sizes. 6. Treatmentdevicesselected 7. Indexesfor costsof facilities. 8. Boundary conditionsin the receiving waters. 9. Storagefacilities, location, and volume. and invert elevations. 1.0. Inlet characteristics suchas surfaceelevations pipes such as type, geometry,slope,Manning's n, and 11. Characteristics of downstreamand upstreamjunction data.

(HEC-1) Package HEC-I Flood Hydrograph

a HydrologicEngineeringCenterdeveloped series The U.S. Army Corpsof Engineers universiof comprehensive computerprogramsascomputationalaidsfor consultants, ties, and federal, state, and local agencies(see Section 21.4). Programsfor flood analyprofile computations, reservoirsystem hydrograph computations, watersurface ses,monthly streamflowsynthesis, and reservoir systemoperation for flood control package, HEC-1, is described The single-event flood hydrograph comprisethe series.
here.tt

Two of The HEC-1 model consistsof a calling program and six subroutines. these subroutinesdeterminethe optimal unit hydrograph,loss rare, or streamflow The other hydrograph values. routing parameters matchingrecordedand simulated by subroutinesperform snowmelt computations, unit-hydrographcomputations,hydrographrouting and combiningcomputations,and hydrographbalancingcomputations. In addition to being capable of simulating the usual rainfall-runoff event processes, HEC-1 will also simulatemultiple floods for multiple basin development by plansandperform the economicanalysis flood damages numericallyintegrating of conditions. curvesfor existingand postdevelopment areasunder damage-frequency HEC- 1 underwentrevisionsin the early 1970sand againin the 1980s.Several features were added (e.g., SCS curve number method, hydraulic routing), and a in microcomputerversion was developed 1984. The 1985 releaseexpandedearlier versionsto include kinematic hydrographrouting, simulation of urban runoff, hyimpactsof drographanalysisfor flow over a dam or spillway,analysisof downstream and flood control sysdam failures,multistagepumpingplants for interior drainage, for The 1990versionof HEC- 1, available PCs or Harris minicomputfem economics. ers, incorporatBsyet other improvements.It adds report-quality graphic and table capability, storageand retrieval of data from other programs,and new hydrologic procedures including the popular Green and Ampt infiltration equation(Chapter4) and the Muskingum-Cungeflood routing method (Chapter 13). In addition to the unit-hydrographtechniquesof the earlier versions, the by modified HEC-I allows hydrographsyntheses kinematic-waveoverland runoff for similar to those developed use in SWMM. The runoff can either be techniques, or concentratedat the outlet of the subarea uniformly addedalong the watercourse distributingthe inflow to the channelor gutter in linearly leogth through the subarea,

MODELS 24.2 FEDERALAGENCY SINGLE-EVENT

607

direction. The 1990 versionallows the use of increasingamountsin the downstream the Muskingum-Cungerouting method in a land surfacerunoff calculation mode. Precipitation can be directly input, or one of three synthetic storms (refer to A Chapter 16) can be selected. standardproject storm (SPS) is availablefor large basins (over 10 mi2) located east of 105' longitude, using proceduresdescribedin but manuals.A 96-hr durationis synthesized, the stormhasa 6-hr Corpsof Engineers peak during eachday. A secondtype of storm is the probablemaximum precipitation (PMP), using reports availablefor estimatesfrom National Weather Service hydrometeorologic (Chapter 16 describes these).A minimum duration is 24ht, and different locations of The third method allows synthesis any stormsup to 96 hr long may be analyzed. only specify the desiredduration and duration from 5 min to 10 days.The userneed the depth,and the programbalances depth aroundthe central portion of the duration usingthe blockedIDF methodof Section16.4. The later versionsof HEC-I include all the precipitation loss, syntheticunit hydrograph,and routing functions developedfor earlier versions.Additional loss methodsincludeboth the SCS curve numbermethodand Holtan's lossrate equation (an exponentialdecayfunction). the of Because the popularity of SCS techniques, HEC-1 now includesTR-20 The duration ofthe SCS dimensionprocedures losses and hydrographsynthesis. for lessunit hydrographis interpretedby HEC-1 as approximately0.2 times the time to g peak, but not exceedin 0.25 times the time to peak (this convertsto 0.29 times the lag time). For routing through streamsand reservoirs,the newestversion of HEC-1 inchannelroutcludesall previousmethods,and additionally performskinematic-wave shapes. ing for severalstandardgeometriccross-section models,HEC-1 is relatively compact In comparisonto other event-simulation and still able to executea variety of computationalproceduresin a singlecomputer there is no provirun. The model is applicableonly to single-stormanalysisbecause periods of little or no precipitation. sion for precipitation loss rate recovery during and routing reachesas shown in After dividing the watershedinto subareas by can be determined one of four methods: Fig. 24.6,the precipitation for a subarea precipitation station data,(2) basinmeanprecip(1) nonrecordingand/or recording itation, (3) standardproject or probablemaximum hypotheticalprecipitation distributions,or (4) syntheticbalancedstormmethodusingIDF data (Section16.4).Either actual depthsor net rain amountsmay be input, dependingon the user's choice of if The HEC-1 lossrate function is easily bypassed losses. for techniques abstracting the net rain is availablefor direct input. such The programlogic for HEC-1 is shownin Fig. 24.6.Hydrologic processes hydrographcombining, as the subarearunoff computation, routing computation, subtractingdivertedflow, balancing,comparing,or summarizingare specifiedin the illustratedinFig.24.L input usingthe sequence One lossrate in the HEC-1 modelis an exponentialdecayfunction that depends loss rate, in losses.The instantaneous on the rainfall intensity and the antecedent in./hr, is

L, = K'Pf

(24.6)

608

CHAPTER24

MODELS SIMULATION SINGLE-EVENT

READINPUTDATA; REFORMATDATAAND WRITE TO WORKINGFILE

REA,DANDPRINT JOB SPECIFICATION;

READ ANDPRINT DATA

COMPUTE RL]NOFF HYDROGRAPH

1 Figure 24.6 HEC-1 Program OperationsOverview.l

where

L, : P, : E : K' :

loss rate (in./hr) the instantaneous intensity of the rain (in./hr) the exponentof recession(rangeof 0'5-0'9) with time as lossesaccumulate acoefficient, decreasing K, : KoC-cuMlllo+ AK Q4.7)

where

Ko : the loss coefficient at the beginning of the storm (when value of 0'6 CUML = 0), an average loss CUML : the accumulated (in.) from the beginningof the stormto time t of C : a coefficient,an average 3.0

If AK is zero, the loss rate coefficientK' becomesa parabolicfunction of the loss,CUML, and would thus plot asa straight-linefunction of CUML on accumulated graphpaperif K were plotted on the logarithmic scale.The straightsemilogarithmic in ln fine reLtion t aepiCteO Fig.24.7, showingthe decrease the lossrate coefficient much lossratestypically decrease during any storm.Because accumulate asthe losses above the loss tate K' is increased more rapidly during the initial minutesof a storm, the straightJineamountby an amountequal to AK, which in turn is madea function before theK' value is againequalto the th;t will accumulate of the amountof losses

MODELS 24.2 FEDERALAGENCY SINGLE-EVENT

609

0.2 cuMLl Ko

A,K=Oif CUML> CUMLy

(in.) loss Accumulated CUML ' K' of Figure 24.7 Variation the lossrate coefficient with the CUML. loss accumulated amount

loss,CUML,, is user-specifieda-nd-rs straight-linevalue,K. This initial accumulated rela6d to AK in sucha fashionthat the initial lossrateK' is 20 percenttimes CUMLI Ko greaterthan Ko (seeFig. 24.7).Initialloss coefficients are difficult to estimate,and to determineinitial infiltration itandard purvesin Chapters4 and 23 are available 26. rateS, For gaugedbasins,HEC- 1 allows the userto input rainfall and runoff data are from which the lossrate parameters optimizedto give a bestfit to the information provided. Estimatesof parametersfor ungaugedbasinsfall in the judgment realm lossrate function is availablein noted in Table 21.1. An alternativeto the described a constantloss rate, similar to a HEC-1, which is an initial abstractionfollowed by @index. The HEC-1 model provides separatecomputations of snowmelt in up to to 10 elevationzones.The precipitationin any zoneis considered be snowif the zone 32'F, plus 2".Thp snowmeltis usually temperatureis lessthan abase temperature, is the or computedby the degree-day energybudgetmethodswhenever temperature zonesareusually considThe elevation "qou1to or greaterthan the basetemperature. ered in incrementsof 1000 ft althoughany equal incrementscan be used. unit hydrographsfor eaph subareacan be provided by the user, or clark's unit hydrograph (IUH) can be used. methodl3 o{ synthesizingan instantaneous Clark's method is more commonly recognizedas the time-area curve method of determined in described Section12.6.The time-area histogram, synthesis hydrograph with a unit design-storm from an isochronal map of the watershed,is convoluted in usingEq. 12.38,as illustratedin Fig. t2.l9.The methodsdescribed hyetograph through linear reservoir arc Section 1.3.2 then usedto route the resultinghydrograph coefficientK. Input data for Clark's storage usingEq. 12.35with a watershqd storage

610

24 MoDELS cHAPTER SINGLE-EVENT SIMULATIoN

method eonsists the time-area curve ordinates,the time of concentrationfor the of Clark unit graph,and the watershed storage coefficientK. If the time-area curve for the watershedunder considerationis not'available,the model-providesa synthetic time-arca curve at,the user's request. Because Corps of Engineers the commonlyusesSnyder'smethod(Chapter 12) in unit-hydrograph synthesis large basins,the Snydertime lag from Eq. 11.5 and for peakingcoefficientCofrom Eq.12.17 can be input, and Clark's parameters Snyder's will be determinedby HEC-1 from the Snydercoefficients.The actual or synthetic time-area curve is still required. recession usingan exponentof Baseflow is treatedby HEC-1 as an exponential 0.1 in the following equation:

Or: #

(24.8)

where Q, : the flow rate at the beginningof the time increment Q2 : the flow rate at the end of the time increment R : the ratio of the base flow to the base flow 10 time incrementslater The base flow determinedfrom this equation is added to the direct runoff techniques. startingpoint The hydrograph ordinates determined from unit-hydrograph for the entire computationis the user-prescribed baseflow rateat the beginningof the simulation, which is normally the flow severaltime incrementsprior to any direct runoff. Ifthe initial baseflow rate is specifiedas zero, the computerprogram output containsonly direct runoff rates. "storageThe HEC-I package allows the user a choiceof severalhydrologicor All and routing" techniques routing floodsthroughriver reaches reservoirs. usethe for relation; some are decontinuity equation and some form of the storage-outflow scribed in more detail in Chapter 13. The five routing proceduresincluded in the programare the following: or 1. Modified Puls-this methodis also calledthe storage storage-indication for normally reserved usewith methodand is a level-pool-routing technique reservoirsor flat streams.The techniquewas describedin detail in Section I3.2. . 2. Muskingum-describedin detail in Section13.1. 3. Muskingum-Cunge-a blendedhydrologicand hydraulic routing method detailed Section13.1. in 4. Kinematicwave-described in Section13.3. lag average method.The 5. Straddle-stagger-also known as the progressive techniqpesimply averages subsetof consecutiveinflow rates, and the a averaged inflow value is lagged a specifiednumber of time incrementsto form the outflow rate. of Input to HEC- 1 is facilitatedby arrangingthree categories datain a sequence summarizedin Table 24.2. The compatiblewith the desiredcomputationsequence, individual input records are precededby a two-charactercode. The first character

MODELS 611 SINGLE.EVENT AGENCY 24.2 FEDERAL DATAFORHEC-111 OF TABLE24.2 SUBDIVISIONS INPUT Jobcontrol
I -, Job initialization V-, Variableoutput summary O , Optimization J -" Job type

and Hydrology hydraulics

K-, Job step control H-, Hydrographtransformation Q-, Hydrographdata B_, Basin data P _, Precipitation data L-, Loss (infiltration) data U_, Unit graph da-ta M-, MeIt data R_, Routing data data S-, Storage D_, Diversion data W-, Pump withdrawal data

and Economics endof job

E_, etc., Economics, data ZZ,End of Job

in Example24.1.

EXAMPLE 24.1 a In June 1963 the Oak Creek watershedshown in Fig. 24.8 experienced severe

watershedareaA-H within the boldface 1. over Subarea For the remainingelongated at border,usethe Junestormto simulatehydrographs eachof Points1 8 usinga single at Points 3 and 8' run of HEC-I and comparepeak flows with recordedvalues

SCS runoff equationaand a basin-widecompositecurve number of 73 (see Chapter4).

612

CHAPTER SINGLE.EVENT 24 SIMULATION MODELS

-o+fr; rsog

paralso

KA\
\: \%

, 1236

Scalein miles

;"'F4 d"fs;
Miles 2 Net rain (in) 7.8
+.-)

r--l
l Rafmond Nebraska I 190
\o

\ @ \8.

34
I

,,,G 'I-t
t).

dt\ -

B C D E F G H
f

33.4 26.9
. 27.3 9.2 28.3 17.0 5.0 28,0 82.9 258.0

65

4.1 2.8
A A

t.7 t;7 1.0 N.A.

TOTAL

Figure 24.8 Oak Creek watershed subarea map.anddata sheet,June 1963.

MODELS 613 SINGLE-EVENT AGENCY 24.2 FEDERAL The computation logic to simulate runoff for this storm consistsof the following22 steps: L. Computethe hydrographfor AteaA at Point 1. 2. Routethe A hydrographfrom Point 1 to Point 2. 3. Computethe hydrographfor Area B atPoint 2. 4. Combine the two hydrographsatPoint 2. 5. Route the combinedhydrographto Point 3. 6. Computethe hydrographfor Area C-at Point 3. at 7. Combine the two hydrographs Point 3. 8. Route the combinedhydrographto Point 4. 9. Computethe hydrographfor Atea D at Point 4. at 10. Combinethe two hydrographs Point 4. 1L. Routethe combinedhydrographto Point 5. 12. Computethe hydrographfor AreaE at Point 5. at 13. Combinethe two hydrographs Point 5. 14. Routethe combinedhydrographto Point 6. 1.5. Computethe hydrographfor Atea H at Point 6. at 16. Combine the two hydrographs Point 6' 17. Route the combinedhydrographto Point 7. 18. Computethe hydrographfor Area F at Point 7' at 19. Combine the two hydrographs Point 7. 20. Route the combineAhyJrographto Point 8. 21. Computethe hydrographfor Area G at Point 8' at 22. Combinethe two hydrographs Point 8. are simulatedby convoluting the net Runoff hydrographsfor subareas by with unit hydrographssynthesized Clark's method using storm hyetograph (seeChapter I2). A Co vabteof 0.8 is applied for Oak Snyder'i coiffiiients Crlek becauseof the moderately high retention capacity of the watershed. are for time lag values eachsubarea found from Eq. (11.5) usinga C' Subarea 2.0. valueof Hydrographstreamrouting is performedusing the Muskingumtechnique reachtraveltime, using (chapter 13) with x : 0.15 andK: the approximate velocity determinedas velocity. A Ch6zy average by the avefage length divided three reachslopeis used.If K exceeds the squareroot of the average 100 times the reach is further subdividedby HEC-1 into shorter routing increments, lengthsto ensurecomputationalresolution. A sampleof the input and output for this job is shownasTable24.3.Each Only Steps 1-5 are in of the 22 computationalstepsare separated sequence. in the sample.Note that the HEC-1 lossrate function was not usedso included that the end-of-perlod excessand rain depths are equal. Note also that hydrographrouting of the A hydrographfrom Point 1 to Point 2 was facilitated using three equal reachlengthseachwith a K-value(AMSKK) of l.2hr. flow ratesfor eachof steps A summaryof HEC- 1 peak and time-averaged is given in Table 24.4.Note that the simulatedpeak at Point 3 is 27,539 l-22 of very well with the recordedval:ue 27,500' The correspondcfs, which agrees -ing simulated peak flows at Point 8 arc 22,453and 21,600cfs, and observed -' - - - -respectively. Il *

o
N N

H
N

. E @ o i >

--

z x ci z
6 q
N

E E

o r F
F

7 z +

F'qEE . o co R E
L
N J : :

@ L F - = < t < / > E @ Ho

I I

U :

c o

ll o
F

> l x , u z
!l

= 5gs=

q
E

:
@

E
6 N

ct

c!

:z u
u
)l

J ; Z

gEcj

rt

< =

tsN
u q
g

- a o q * = ! e

N o

s^Niil;.1

_^ _fr3eflfo o o t : $ f l F

*-

z
F

3
F , i U N A

o o e F e v s E c c 3 V E V f f R c a 3V H v E v S B c c 3 Y ? .
I t Eo o o o < - +
-U_=

= tr
F N O t h @ F O O O F N O

F E

= 9 9 t r P 3

R F N R I R R R F R

B E

z.

226 S FtrE
e
o o Y
f,

z u e
J

F
F

s u
6

F E N

O N t s N q
N

z
o

t 3

t
' i

r
*

= u

h
P
=
<

EE
* *

tr

E A

: J
z

a ss cq ri 6i
4 d
=

N
o
F

dt c;
N

il F F

il

!!

;l
6

F - 69 1 t
F

EE

eirt

u
4 F

RE
]

z. u

=
r
q
F N

z u E r u
Y

6' u

=
z
m
a

$r

R =
m x E

c |r

=
E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E
q

d
o o
J

- - - - - Y Y - - - - - - - - - - - - Y - - - - = - - - - - - - -

F N 6 + o @ t s o o - N m + 6 @ F @ o o - N O < @ @ F O O - N O 6 h 6 n 6 6 b 6 6 6 @ @ @ @ 6 @ @ @ @ F F F F t s F F F F F @ @ @

<
>

E 1 6 E i 6 E t 6 E t 6 E F 6 E i 6 E b 6E F 6 E t d 6 ; 6 E b 6E i 6 6

=
o z
F F q F

h 6 @ @ F F @ @ O O O O F - N N O Q O O - F N N O O < < h s @ @ r - N N N N N d N O O O O O O O O O O O O e O O

i6 E;6 E i6 5;6 E

=
< o

z . z . z z z z . z z z z z z . z . z .z z z z zzzzz z z zzzzzz.z ------------------------------a-+ + + + < < + a + + + + + + + + + + 4 4 b h 6 6 A h b 4 h r a r a N N N N N N N N N N N N N N N N N N N N N N N N N N N N N N N N N

z. .

* * *

* * * * * * * * * *

+ * * * * * *

* * * * * * * *

* * * *

d
=

- s d s s 3 E 5 S R R S 3 = S S 3 ei fd F s 3 N 3 e g + e R 5 fr F h + F @ O - @ - e + - - 6 - F F - - @ b r N @ q N F F
- O @ N O 6 F + < N O O @ N o F h O N - - - N o o o o N N - F F

.
ll o q

s
<

a E a

= @ E R q +N f i q B q s q q * E E q E e E e e E e e q E e E e e e ex q e * q
U >< u

6i!P+

d F
F

;s H
>+
= N x (
T

aEs

q
o

BAA8AAt83888EB8ttt333AAEtt33EA=EA

A'

u 46

F E +
@ r { 8

z.

O O O F O o @ N F F o O F o O o o O O O O O O O a A Q Q A Q a Q A d ? q q e q < ? c q q q q q q q q q q q q q 9 - N f r + F @ F @ + o o N 6i

rr q O
J

O3=S3:=9=S9RFNRTRKRRRgESA
E

a E
F

nz.z
- L

z u z. z.
o H o

=
z.

- 6 - O - 6 6 6 0 0 0 0 0 0 0 0 0 0 0 0 0 < < h h @ @ r r @ @ o o E i 3 = = i < N ; 6 U E i i = = F n l 6 o + s ' A a N N N N d N N N O e e O e Q O O O O O O

ll

= +

E=

zz,zz

z zz

zzzzzz.

z .z z z . z . z . z , 2z z z z . z . z . z z z

zzz.

a - - - - - - - - - - - - - - - - - - - ? - - - - - - t - - - - o o m o o o o o o o o o o o o o o m o m q + + s s + + t $ + $ s +
N N N N N N N N N N N N N d N N N N N N N N N N N N N N

z
E

3
tr

=
n6;
Ylt
F

= ii

d d d d d d o d

d d d d d d o d e ; c j d d o o o

o o

a FtrR R E 6S S S S S E S s B s S S S S H FE E S o
z.
> = E o o o o o o o o o e i6 E ;6 E i6 F E6 E o o o e o o o o o d ;6E 6 + + 6 6 6 6 F N 6 6 6 d o O F F N N O S S + 6 6 a6 E Fi E

7 >
{ 6

z. z. z. z. z. z z z z z. z. z. z. z z z. z. z z z z z z. z. z - - - - - - - - - - - - - - - - - - - - - - - - n N o 6 N o N h o N 6 F r 6 o o 6 6 6 h 6 - - - N - - N N N N 6 N 6 N r h b o o o N N N N N N

* =

cj

ct

ei

ci

c;

<j

ci

ct

cj

ii

E 8 i l f rs s H h E a E E E S S A A E S S R F S R F F
2

= EBagBBaata=R===B=sBB33EEE F S E E E E = T E E = 5 ! i = H N N F K = 3 =S S E =
N

V
= = { O

z. z. z. z. z. z. z. z. z z. z. z. z. z. z z. z. z z. z. z. = = z z
= = = = = = = = = = = = = = = = = = = = = f - - - - - - - - - - - - - - - - - - - - - - - a = = = + + < + + + + + + + + + + + + { $ 9 e ! Q o o h o h d N N d N N d N N N N N N N N N N N N N N N N N N

o U

= z.
z.
o

z.
L

N F

z. o

E
F E

u
J

u
E

6 F

= < 9 d

ot + Gt ci 6i - N - m N 6 6 F o + O + O @ O @ O = - t s h + o N N F

r; r; Kj + ci + o o o @ O @ o O O @ + N - + O F h + N - F F

qt

gi

ot

ct

di

<t

q u

o E o

= =

@ t s o o F N O + h @ F @ O Q - N O S 9 @ r o +o { + + s + + N N N o o o 6 o d s N - o r i

+ +

t o

o'i tP +

n E ra oGi!P+

z.
6

= =

u F E

? Y x q E E frzz
o = = > o a

EE EEEEEE EEEEEEE=EFFEE=E E
z. z. z. z. z. z. z. z z. z. z. z. z z z z. z z. z. z. z. z z z z.
- a ? - - ' t - - - - - a - a q q q = 9 q g g = g s s s s s < s < s < s d N N d N N N N N N N N N N N N N N N N N N N N N

ifE

r
e

8 -E9 A o
+!Pct hH!2
@ F o

u
{ c

=== =>>
'

=x
x > =
@

- z z.\ tr-g?Yx = * = 6 EZ>O> - = ' <

N

=
= J

c t d d d F +

o o r

3 $ B S g E R S N J H E F E E J = a s 3 f f i $ K 9 A A S 5 = 3 6 D
N N N N N N F F F

a . ar '+ t r r
a = Z =

az. E V
E 5 o -

o'
E

o - N o + 6 @ F @ o Q - N A + E r

= - - -

=
u ^

+ v i Y

z.
E

>+

tr

E E E E E E E E E E E E F F = F IR F F E E E = E E =
z. z. z z z z z z z. z. z. z z z z z. z z. z. z z z z z z.
- - - - - - - - - - - - - - - - - - - - a - - - N

=
o

t d.6
*

$
t
T

v B
U A

N I
T

o N oN o N m o N oN o N oN o N eN e R R R R R R R R R N N F t t N N N N N N

=
J

-i

ct ci

ct

ct ct

ci

ei e; ci

et ci

ct

ci

c;

P FFFREESASSFbBtsA6SESeSSssE o
z.

EE S g E E E g E E E E E E E E E E F F E E F F E E
?
>

z z . z z . z . z . z . z . z . z . z . z . z zz 1 1 1 1 4 z
J

z zz
-' -

z
-

-J -'

3 n n K R R K K R K K FR .R S R K K R R R FR R R R

* * * * * * * * *
= E "i "i.i ci d d ei ct ci d ci a I ci d <j ct d d d d ct d cj

E 8 i l f iS A a h t s g A E D A g A A b S S R E S R f ,
z.

E EEEEEEEEEEFF=RFRFEEBE=EEE
7 U

= z
z.
dl

z.
o

= = = - - - - = = D = > = = = = = 1 = = = = 5- == = - - -

z 1 z. z. z. z z. z. = z. z. z z. z = z 1 1 = ) = t = 4 ) ) z 2z z z' 3 2 =

---

---

---

z.

a
lo F

3 x x x x * N t N * x * * F * R * N N R R K R R K K
* x * * * * * *

u
J

E F

a 3 $ p F $ HEE 5 E F H= F E P
@ r U

+ s 3 n j6 '6 id jq ' @
i B ot E;\u M F 9

e R N R R E E S E S B H h S g S S S S $ s 9 + s g B
z
E
J L

E a

U =

6 ? Y x s E E

E E E E = EE EgE EE EEE EEEgFFEEEF-9 I

-

==
,,,=
U

r!6=

giH

E<n= = > q
< F
- z z +

z. z. z. z z z z. z z. z z z. z z z z z 1 1 1 z < = z z z' = = = = = 5 = = 5 = = 5 - 5 -= i = = =r ? 5 - ? = ? -r = - ? - - - ? = - - = = - = - - O

z>> - 3 5
;

*N X * * N * * X * X tN t* tfi * R x * fi N tx

;+ sHm
= N

E= EHii
\ o

"o-oNEH
- N

E - F oT V X = t
E > @ >
N

FEHffEPE$nHHFEE=EEE
q

a l

a V= G
o @ E 5 o -

o + 4 @ r @ o o F N 9 S P 9

= = E R F N , R X

R
U ^

= ES
F -

o <

z.

EgEEEEEEEEEEFF=FFFFREE=EE = -^
z .z z z z z 1 1 1 = 1 z = 4 = 4 = = < = = = = = <
= - = - i - i - - - - - - - - - - - - - - - - - n o :ZO u
N N N

Y Y

=
o

R R R R R R R R R R R R S IR R F K IR R R F S S N N

a >
E
F

=
@ E

z
u
F

s
U ' :

(5 o U F

6 9
u z. 2
@

< = , 6 @

u
Y

a
Y

@ E
@ F

x C
C

5 E
ts

O
F

ci

ct ct d

ci

ct

ci

ct d

ct d

ci

c; ci

c;

=
q @ U

ceeeEaeeeEeEEeeccaEEeeeEeEEeeeEE
EEEEEEEEEEEEEEEEgEEEEEEEECqEEqqEE

d
6 q

z =- = -=- = -=- = -F = e E C q e e e e e e E e e e e e e E e a e E E s a - E F N o + 4 @ F @ q o F N o + h @ r - o F S R S F R F R R B E S g 6 @ @ F l 6 @ @ @ @ @ @ @ ; e h h 6 h 6 6 h h 4 o

z.

EEEEEFFEEEEFFFFRRRFEEEgFEEEEgEEEEE
4, z z z z. z. z z z z z. z. z. z z z z. z. 1 1 1 z z z = 1 1 1 1 1 2 z z z
> = = r 5 ? -t 1 a -t -. -

a
q

E x x * * R f i N S t t N N S S N F t F K K K R K R K R R F R K K R K
* * * * * * * * * * * * * * * * * * * * * * * * * * + * * * * * * x

L <

? - s e f i s s i i * B $ = R i E E s B 6 R $D E EB s D R s
d r o 6 ; 6 o F c 6 m - N - 6 - s t a f N + F - + @ F @ o F o r o o N F F r o 6 o F F

o P + g 4 i 5

@ q

u x u

ee EqNfrqqEfifrITqiEEEEE eqEEeeEEeEeee

ec 3 =.AeEeeqeeE eeeEeEeeeeeeEeEeEeee
Z 8 = N K t s B 9 S K =f
G

iqiqEEeqqqEt

eqeEeeeqeeE
R * K F N R R B E S A

,l , x i < _ 6 5 i+ 6
J

tr i<

o
e

- N o + h @ F @ @ o = s P 3 5 9 = = = R F N

g 9 F

z
U

z. F z

EFgEEEEEEEEEEFE=RRRFFEE=gEEEEEEEEE e =s =
tl
J
4

U ^
L v

7 >

2 z. - z z z. z. z z - - z. z. =z = z5 z z. - =z - ? = =1 =1= 1 = = =z = z= 1 -1 =1= 1 1 z z z z' = = z. - = r = = = = - - - - - - - = = = - - - - - i - - - - - - - - - - - - - -

2. E =

s R 3 * * R R R R R R R R RR R R R R R S T R s x * s s * N * * *
N r a

s t

4. 4^
ts yE
u

:
R

=
E

ct cj

cj

e; ct

ct

ci

c; d

et ci

ct d

cj

ct ct d

ei d

t
f

a
z

F F F R E 5 S S S 3 S b E t sE ; s a s 3 E s I E F

E EEgEEEgEEEEEEEE=PFEEEEEEE
= = g z z .z . 1 1 z = z z .Z z .z = = = = < = = = = = - = = = = = = = = = ; = = = = i = = - = = - = - = - = = = K F R R R K R K K K R K RiI+AR N K K R R R R K R

= iE

6i

6i -

ci

cj

ci

ct d

ei d

ci

ci

cj

ct

E
z

E i l s s a H h a g E E 3 3 3 3 A E E t sR F S R * R

N q

EEEEFFEEEEFF=FRRFREEEgEEE
= = = = = = = < z- = = 4 = = = = = = - - = 1 = = = = = = = 1 = = = = = = = = = = i ; = 444 4 = = = R S * * * x N f, * * * s * s tx $tR R R K R K R t I
i

z.

z .<E o 3 P
' u z,
o o
F . 6 f = >
* *

tr

< =

N
F

P>

$ R 6i ri di d +/ E E H H E il H d g E F H F F E E s 3
J F O F 6 + O N N F -

g g a s a E gs s s $ s s s s g B E K NRR3 6 s s
z.

==
L E

d r d
= N =

<

o <

o U
6

E EEEEgEEEEEEEEEEEEEFFFEEEE s
d P d
EC'I6 + @ =

z, z
F

=-= = = = = = = = == = = = = = == = = = = = == = 3 fi * R* t F s Nt x t s * sx t N* t * * * * NN

u < c

E F A
= F x

==
@

ffi85

z. o E

E
E

OEggF$FFF
r

6 F

u + + o = <

u z F z

U ^ = G

==
N

z.

z.
Y v O N

= = = = = = z. 4 = = = = = 2- z-.= = = = = = = = = = = = = = = = = = = = 4 = = = = 4 = = 1 = - = = = R R R R R R R N R R R R R R R R R R R R T S T S R

i a Y O u

5 o
F

=
J L J T

et ci

et cj

ci

ci

cj

ci

ct o

ci

ct d

ci

ct d

ci

R F R R B E S S $ s A b B g B s s a g S t D E g
zz.

==
z.z

E=gEEEEEEEEEEEEEEFFEEFFEE
z z. z. z z z z z z - -z - z= z z. z. z z z z 1 1 = = Z. Z Z.
= = - - - - - - - - - - - - - - - - - - - - - - ?

R R K R R R K K K R R K R K K R K K R K R K R R R

==
I L

O F

ci

ot r{.J
+ N

ot F -

Fi ct F

ct 6i ri

-.j

ci

6i e; Gi o

i 6 i l f f if i A t s h f f g E E 3 3 3 3 E E 3 3 R = S R f ,
z z

EEEEEEEFEEEEFFFFFRFFEEEgEEE
z.
F

== ========================= 3E xtRNx*tF*fiR*ttttstKRKKRRR * * * * * * * * * * * * * * * * * * * * * * * == 39

77

z. z z. z. z. z. z. z z. z z z. = 1 1 1 1 z z = z. z. z z z

@ F

z.
c F

I r

EnFEss$FEER=$ssFEEiEE
g
+

Fi d; ci O F O

ct

E E R N R R t E S g $ E E h S g g E $ e $ 5 s $ s e B
z. z.

d N =d P =

+.dE

>

g EEEEE=gEEEEEEEEEEEEEPFFEEE + d E E s
L -

= =
, , , = = + t u

9 = ! ! = N =

e
2 2 R
F J

= == =========,===== ========== 33 xssNN*sNtsF*tK*NFsNstN*tF

* * * * * * * * * + * * * * * + * * * * * * * * * * *

? 7

z. z z z. z. z. z z z. z z. z. 1 1 1 1 1 1 = z z z z. z z

Vivx u E E
* O O

=q EFs
= R = @ P

E
E
oz. z =

E z z

=_ ==
@

=E{
==
; '
J J 4 (

E=E
+ N O F F @ L 6 l U q

cicic'di= B Eg

E$ g * F EF $ F HF E

t s N
@ @ @ N N N

o o u
F

F N o + 6 @ F @ o = = s P = 9 9 5 = S R F N R N K
U ^

9<
=F
o

= z
z.

z z

EE=EEEEEFEEEEEFF=FRRFEEE=gE =^ :
v Y T

== ====-=====----=-=-=======
3 3

7 7

z z. z. z z z. z. z. z. z. z. z z. z 1 1 1 = 1 z z z z z z

d. 9.a
Y O

vi
b N

4 J

* R R R R R R R R R R R R R R R R R R R F F N S *

A^l

N
N

U Z t6 fl E P E c? .q ,,I.t' o U F rid O N

o > 1 4 . < d

3 -f o
.rl t
: :!

3 EE E
' = e

Y a

dd
6 t s (q\

==
E E +HJ
El: Z Y = o ii@ E N -

O o

= - F F

z.

z. z. < =
^ ;i E tlJ 9

Y G EU < 6 J > oz @

5 *

z o z. u

O F

K 6

=
@

z.
u u r u
U

= z
z.

u F

v
q J

6 h N

4 @ N

L r N o N

< N

o
G

a eeeeeeeeeeeeEEeeeeeqeeeqeeEeeeeee
U

x U
q

z
E

F N o < 6 @ F @ o o F N O < o @ F @ O O - N O + D @ F O O F N O b b b h @ @ h b 6 @ @ @ @ @ @ @ @ F t s r F F F F F r t s @ o @ @

z.
=

= BBEBaaaBBaaaaBBBaaBEaaaaaBaaaaaaa
6 6 @ @ F F 6 6 O O O O F F N N O O O O F F N N m m + + 6 6 @ @ F F F N N N N N N N N O O O O O O O O O O O O O O O

=
<

z. z. z. z. z. z. z. z. z. z. z z z z z z z z z z z z z z z z z z z z z z z.
t a a a a -

N N N N N N N N N N N N N N N N N N N N N N N N N N N N N N N N N

d
^ E 5

ct d

o h N N o F O m O O @ @ O @ @ O @ F O @ q F F O @ O F N F o F @ O O O + N @ @ O + O @ 4 F F r + N F F F O 6 O N + h 6 O - O 6 @ + O N - -

ct ct et +

di qt ri

qt ct qt et ct -

<t F dt d

ci ct st qt -

ci et ct ri

ct ct d

e qqeqeaaErraaEEEaEEtaat + a E l F f rE q = q f e f i r u
x
U

= a

flE+ = = g

q q

o o o o o o o o o o o o o o o o o o o o o o o o o Q o o o o o o o

z.

O N F e T N - O A @ O N @ O O O o o o o o O o O O o O O O O O O O O - N N O 6 + O N F N - - O O O O o O O O O O O O O O O O O O O O

, , x

s=+ HEF
o
@ O O e N O + 6 @ F 6 O O - N O + 6 @ F O O F N O F E N N N N N N N N N N O O O O

o o a

= e
=
s
o o o o o o

^ ?.frtr
o <
=

z.
u

= z
z
a

5 = =

o o o e o e o o o o o o o o o o o o o o o o o o o o o o + + 6 6 @ @ F F @ 6 O O O O - - N N O o o N N N

e o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o F F N N O O + + 6 o @ N O O O O O O O O O O O O O

ll ! u ^ 2 '

aE+ =
-

=
< {
A

z z = z z z. z. z. z z. z. z. z. z. z. z. z. z. z. z. z. z. z. z. z. z. z. z. z, z z z z
a a a ? a a a 1 a a o o o o o o o o o o o o o m o o o m o o < < + + + + + + + + + + +

1 q= < ^
F V A

3 E
ts

N N N N N N N N N N N N N N N N N N N N N N N N N N N N N N N N N

= n a

j E
6

f;-

ci

ct ci

ci

ci

ct

ct d

ct ci

et d

c;

ci

cj

Q c;

BssEsaasaEa s E F F R R S $ Ss S A b E g
z

EEgEEEEEEEEEEEFE=EFFEEEgEE
7
>

z . z z z . z . z .- z -zj z z - z- . z . z z = - = - - -

- --- - ) i -

z 1 1 1 1 = z
- - - - -

z z z
- ' - ' -

z
-

- -

E N P K F R K R K R K RR R K R R K F R R FK R R K i
* * * * * * *
N N F

*
rt

*
ct r

*
rt

*
6i -

*
-

*
-

*
ci

*
ci

*
d ci

= ;
E

ct
F F

di il
C <

oi F

ct 6i

ct d

ct d

i a E s f f s 3 a b t s f i3 6 D 3 3 3 A b 3 E R = S R F F
z

EE E E E F E E E E E F F F F F R F F E E = E E E E
z.

7 z z z z .z z 1 z z z = = = = = = = = = = < = = = =
> = = = = - - - - - 5 - i i - - - - - - - - - - - -

k F
q

3 x s t * t t N N s * t F * t x * t S R R K F R K K

* * * * * * * *
u z,
o

=
u

nRRFEE$$EHR;$sFEil
K R R R S E S S S S A D E B E E $ S+
< 6 @ F A < < q 9 F = m

F 6 0 @ O F O

t
@ L

=
F a

q!ni!!

= o

=
z

EEESgEEEEEEEEEEEE=EFFEEEg ? s --Ki
F a
U < E = F x

=
r z, Z N

z z 2 z. z z z. z z 1 z z z- = = = = = = = z z z z z: = = = = = = = = = = = J
- - - = - - - - i - a i - - - - -

i s i
o 6 r

* t N * t s * * N * S F S t * N S S * S N t S N S
* * * * * * * * * * * * * * * + * * * * * * x *

5u?6

> =
@ I

k z
o

= ;>
u o +

-*RNEFBSEus'pRxf HexnsE neEfr F

o o - \ o + h @ F @ o o = S p = = g = P P R F N R t R
U ^

d 5 i i6 ;N \ NE
@ts u +

24 =

z.

E=
trt

z
v v

EFEEEEEFEEEEEFF=FRRFEEE=EE = -^
4 z,zz.z.z=1=zz==z.=z==4z zz <z=4 = = = i = = = - - = = = = - = - = = - 5 5 - - - - :ZO r
N

*
J

E * n R R R R R R R R R f l t R R R F R R R R t * t F t

MODELING 625 24.3 STORM SURGE FLOWS ANDAVERAGE PEAK OF TABLE24.4 RUNOFF SUMMARY SIMULATED 1963 STORM AT POINTS THROUGH FORTHEJUNE, 1 8
Peak Hydrographat Routed to Hydrographat 2 Combined Routed to Hydrographat 2 Combined Routedto Hydrographat 2 Combined Routed to Hydrograph at 2 Combined Routedto Hydrograph at 2 Combined Routedto Hydrographat 2 Combined Routed to Hydrographat 2 Combined
I

6-hr

24-hr

72-hl

Area

2 2 2 3 3 3
4 A A

5 5 5 6 6 6 7 7 7 8 8 8

34475. 25622. 17310. 30092. 26759. 15400. 27539. 25912. 3362. 25925. 23911. 7 100. 239tt. 23911. 3349. 23911. 22949. 4113. 22949. 22453. 1684. 22453.

24048. 20848. 11207. 26453. 2406r. 10503. 25828. 2455r. 2382. 24606. 22858. 5604. 22874. 22874. 2478. 22874. 22024. 2764. 22924. 21595. 868. 21595.

69'74. 6974. 3100. 10074. 10070. 2995. 13056. 13052. 689. 13702. 13523. 1816. 14717. 14717. 7s0. 15268. t4984.
I tJ,

3382. 3382. 1503. 4885. 4885. 1452. 6337. 8409. 441. 8850. 8875. 1162. 10038. 10039. 480. 10518. 10482. 495.

33.40 33.40 26.90 60.30 60.30

zt.3v

t5t12. 14994. 228. 14995.

rc977.
10908. 146. ' 11054.

87.60 87.60 9.20 96.80 96.80 28.30 r25.10 l25.to 28.00 153.10 1 5 3r.0 17.00 170.10 170.10 5.00 1 7 5l . 0

MODELING 24.3 STORM SURGE

stormsbut alsofrom storm floodsfrom single-event Coastalareasnot only experience Severalcomby (short-termchanges sealevel) normally caused hurricanes. in surges Most are surges.l2 storm hurriCane-produced puter models are availableto analyZe from the atmosof modelingthe physicalprocesses momentumtransfer deterministic, phereto the ocean.Parameters thesemodelscan be adjustedto allow analysisof in actual or hypothetical storms such as the probable maximum hurricane (PMH) or projecthurricane(SPH). standard private consultants,and universitieshave developedsurge Federal agencies, surgemodelsare listed inTable 24.5. open-coast models.Severalnonproprietary, User's manualsare availablefor all the modelslisted. Storm surgemodels simulatethe effects of wind momentum on ocean water and oceanography, wavehydrodyThis involvesprinciplesfrom meteorology, masses. geometry,and water-level about storm, namics,which-all operatefrom assumptions from surges damage conditions.Tide levelsare includedin mostof the modelsbecause peak surgelevel. entirely on concurrencewith the often depends The equationssolvedby the first three models in Table 24.5 are two-dimenof of sionalversions Eq. 13.51and the equatiotr continuity,Eq. 13.40.The fourth the model disregards continuity equation and solvesEq. 13.51 through a seriesof

626

MODELS SIMULATION 24 CHAPTER SINGLE.EVENT

MODELS SURGE TABLE24.5 STORM

Programname SPLASH

User/agencf
NWS

Data input radius to Atmosphericpressures, maximum wind, storm speed radius to Atmosphericpressures, maximum wind, storm speed radius'to Atmosphericpressures, maximum wind, storm speed, maximum wind speed,depth of shelf Wind field, pressuredifferences, radius to maximum wind, forward speed

equation Differential solutionmethod Finite-difference

coasts Applicable Mildly curved, Gulf and East coasts All coasts Gulf and Atlantic coasts

SSURGE FIA model

coE
FIA

Finite-difference Finite-difference

BATHYSTROPHIC

CERC

Finite-difference

All coasts

,NWS, National WeatherService;COE, U.S. Army Corps of Engineers; FIA, FederalInsuranceAdministration;CERC, Coastal Engineering ResearchCenter, Corps of Engrneers.

they treat time asa succession The assumptions. modelsarenot truly dynamicbecause water depthsat the end of eachtime stepusedin of steadystates.Output is a file of the simulation. Storms that can be simulatedinclude the SPH, PMH, or any prescribedwind field.

r summary
involvesthe useof singleBy far the largestnumberof hydrologicmodel applications in eventsimulationmodels,whetherthe generalversionsdescribed this chapteror the urban runoff models about to be presentedin Chapter 25. Data requirementsfor modelsare nominal-far lessthan for continuousmodelingstudies.In single-event the majority of cases,the data required for any subareaare easily obtained from readily availabletopographicand soils maps.Given the basin area,slope,soil types, of of land use,and location,estimates peak flow ratesand shapes runoff hydrographs can be obtainedfor given stormswithin a few ar severallocationsin the watershed hours' time usingthesemodels.They continueto be the primary tool usedby practicin ing engineers analysisand designof stormwaterhandling facilities.

PROBLEMS
for 24.1. Six numberedsubareas a river basin are as shownin the sketch.Preparea scheconnectrunoff components' matic diagramfor a modelstudyusingboxesas subbasin ing lines as channel routing links, circles as hydrograph combination nodes, and for the trianglesas reservoirrouting nodes.Then describe computationsequence this (Seesketchon nextpage). basinin the samemannershownin Fig.24.1. in a 24.2. Synthesize unit hydrographfor a watershed your locale using the HYMO model from Snyder'smethodand unit hydrographs Comparewith corresponding equations. the SCSmethodin Chapter12.

PROBLEMS

627

t-r'\t-;

Sketch for Problem 24.1 a to 24.3. Usethe HYMO modelequations synthesize unit hydrographfor the entire 258 sq. in mi Oak Creekwatershed Fig. 24.8. 24.4. A watershedexperiences l2-hr rainstormhaving a uniform intensity of 0.1 in./hr. a Using E : 0.7, Ko : 0.6, C : 3.0, and AK : 0.0, calculatethe hourly lossratesl, model.Determinethe total andpercent by asdetermined the HEC- 1 event-simulation lossesfor the storm. 24.5. RepeatProblem 24.4 wing CUML1 : 0.5 in. 13.7to the outlet of the 30-mi reachusingthe in 24,6. Routethe inflow hydrograph Pr-oblem pairs of flows two time incremethod by lagging averaged HEC- 1 straddle-stagger outflow rates. ments (12 hr). Comparethe routed and measured in 24.7, Routethe inflow hydrograph Problem 13.7through the reachby dividing the 30-mi and reach into three subreaches treat the outflow from eachas inflow to the next in and comparethe f,nal outflows line. Lag flows one time increment in eachsubreach values. with the measured 24,8. Study Table 24.3 andFig.24.8, and then definethe following terms from Table24.3: AMSKK,X, TAREA, NP, STORM, TP, CP, TC, R, RAIN, ANdEXCESS' COMP Q. 24.9, Search the HEC-I printout in Table 24.3 to determinevalues (give units) of the following: a. The time incrementusedin the model run. B. b. Snyder'sCe,Eq. 12.17input for Subarea unit hydrograph. Subarea-A c. The pbak flow rate for the synthesized A. d. The total runoff (in.) from Subarea A. e. The peak outflow rate from Subarea f. The peak-to-peaktime lag in routing the outflow hydrographfrom Point 1 to Point 2,Fig.24.8. g. The percent attenuationcausedby the reachbetweenPoints,l and 2. A peak outflow rate if Subarea is neglected. h. The Subarea-B peak outflow rate. i. The simulatedSubarea-B

628

CHAPTER24

MODELS SIMULATION SINGLE-EVENT *Refer questions: to the HEC-1 output in Tables24.3 and24.4to answerthe following 24.10. "PRECIPITATION PATTERN" actual rainfall depths' or a. Are the valueslabeled

Points 1 and 2 of Oak Creek. of Branched 24.11. Describehow, for a given storm,you wciulddeterminethe effectiveness enumerfloodingat Point 8. Answer^by "as oak reservoirat Point 9 in Fig. 24.8 to reduce "runs" You are alloweda maximumof two ating your computationlogic illustrated. may be usedas long as you describeyour wlttr ilBc- t, und uny nui-rb", of subareas subareas. 24.12.AhydrologistwishestomodelawatershedusingthesCScurvenumbermethodfo runoff hydrographthrough a reservoir determinenet rain, and to route the watershed usingthestorage-indicationmethod.BaseflowistobeincorporatedaSaconstant model or modelswould valuethroughoutthe storm duration.Which event-simulation the accomPlish task? a hydrographusing the 24.13. The following data were preparedfor a Yl card for routing is methodof UgC-t' The time increment 1'0 hr' straddle-stagger
Value

Field

Field
A

0
I

2 3

5 6 7

0.0 0.0 0.0

_ I

by the straddle-stagger Usethis information to route the following inflow hydrograph is zero. continue routing until outflow is zero again' method. The initial outflow
4

Time (hr) Inflow (cfs)

0 0

1 0 3 0

2 6

3 0

5 9 0

6 3

' 0

7 0

r20

8 0

REFERENCES
t. 2.
J.

4.

Chapter25

UrbanRunoffSirnulation Models

r Prologue
The purposeof this chaPteris to: for . Describenine of the most commonly usedcomputerpackages simulation of urban rainfall-runoff processes' . Show that the modelsdo more than simulaterunoff; they allow the user to systems' and designcompleteurban stormwatermanagement analyze precisionby comparingresultsob. Demonstrate the'models;capabilitiesand tainedwhensimulatingtheSamewatershedwithdifferentmodels. . provide a ,,shopper's giid"" to commercialandpublic domainurban stormwater software. of runoff and also the urbanization generallyhasthe effectof increasing volume of runoff. Early attemptsto apply the tendsto result in earlier ani greaterpeak rates and designwere successanalysis modelsof Chapter24 inurban system single-event total decreasing rainfall-runoff analogs' iutiy u""o*plished by seleciingmore intense zonesfrom the total perviousarea' 6y rainfaUabstiactions deductingthe impervious and channels, by and oi the increasing speed traveloverthe land surfaces in improved kinematic wave approximationsof adding componentsto the codes to allow for stormwater flow and hydrologicrouting of flow in storm sewersand urban overla=nd retention or detentionPonds. urban watershed betweenthesemodels'predictionsand observed Discrepancies and wholeclassof single-event continhaveresulteJin the dwelopmentof a response operating in -urbanizedsystems' processes uous streamflow models of the oniq.t" allow the userto but Thesemodelsnot only simulatethe raintitt-runoff process, also facilities or to management stormwater existingnetwork of interconnected analyzean detentionponds, storm sewers, (underground designnew componentsof the system ditches.streetinlet sizesand locations,etc')'

! = X ^ a

ts

E o

> E o - (!

U EY 8 :

cloq
G l \O

n
C.t

qq
O h ca +

iq\oqv'l

s t F L - q ' l <f, \n ca C..l t'-

=^ ' z
E E)N : = 5

ral c.l

0O<.\r)\Ocn

E - ^

h O

e { -

d . ? ) O N c n * i

o S E X s E dhd N
1.O
@ J

o > 5 = ^ ^
-

a.th

o cd \O

e{r) cJJ +:+

o\$cicio. r.ai-jio cn6.aa{$

o c) o

;O 6 C z
f - v

\r) ol

00 a.l

\o r)

c'{ \o e^l

o = t ^

5 =' e z

ho

o\

or)

cqNcqol$

o o o

l )

H"o
J

o a o tuq

d 8E E==tr=
H v r r A Y ^A r\ H v r ) n f ) n Y m 5 Y F l ; m FA m

R.9 9 { X

.999 .9 X X X ; i X

gs - 9 iBe
Q6' v . = - Y

a ul
T F IL

n\ n O *
oo ci t'-

cl
oo

o?n
* O i

\qqqq

F-it-oj*f i n C { *

QE(u v o o oi5-

- b R

Eo
.=^
v -

s 5s HSRHX
cn Y * N - v l O O ne{ Y - Y - i O \ F - * i cnr)'am\o

t--

l--

O\

o r!(E R(U F

>v \r

q \ m s

co o1

Eo
(u

q n
Vcn cn v

\ori
6\O i

OCIOl S re.lcnt N ca
a

ot--F-m\o

o t,rJ
[ l h

7aa - E tr !.! ulr

Fa1

-tr9 E o
';^
d]

F-O

+ n+ *
.t)

\o
O
d

N6 hr 61 6l

e..l\o N<.r 6lF-0Oe.l C.) cn O 0O cq

i i n

.o r

S Z

s< c lt)
o| IT

a o

tr
V)

638

CHAPTER25

MODELS URBAN RUNOFFSIMULATION * 6 1.2

, / ),^
a
R / 9a . I U .

n R

o a
^ a

/
a
o

,t

jlo 2

)- nd
9 . {o4

a 8 . / 3

t 4 'f1, ' Q t3

.s&.' a 7
t.2

2 4 peaks Observed (cfs)

(a)

a 1.5
E

;p0
O r n

^0.12

Y o.os
'F 0.04 ; z 3

r L ^ r /

t \

ttT,*'

10 11

Drive basin'Long Island'N9wYork' for Figure 25.3 RRL results Woodoak (After and (b) peaks, volumes, (c)thehydrographs' 19, of storm October 1966r(a) Stall andTerstrieP.e) ofthese Physicalunderstanding gutters,pipes,and openchannels' surfaces, presentunderstandingof the many ii flow phen-omena muctrgreaier than the complexphenomenagov"erningrunofffromruralareassuchasantecedent "infiltratilon, soil moisture movement, transpiration, moisture conditions, evaporation,and so forth. provide a function for 5. A modification of the RRL method that would grasSedareacontributionstorunoffcouldbedevelopedintoavaluable in This is believedto be possible spiteof the iesign tdol for urban drainage. flexibility could be ^offered by the *unly "o*plexities involvei. Further surfacestorage' addiiional irovision for routing surfacerunoff through RRL methodson an urban basin 6. The input data r"qoir"m"nt' foi o'" of the evaluationof a basin for storm drainage for are reasonable the engineering than the data daia are no more complexor_elaborate The necessary design. drainagedesign' ;;dlly compiled foi a traditional storm

25.1 URBANSTORMWATER SYSTEMMODELS

l.) 3 1

639

E zoo
o

E U

$ roo
a a

3! 1 z oo" a t 2 a 1l

1.0

i.'
o

ot<

, +

,,,' ,9r iir.t,'ol

' 1a " 8

'J - ;'i^':ls

r-'f;3:'r ,"ni
0.5 1.0 Observed runoff (in.) (b) 1.5

100

200

300

Observed peaks (cfs) (a)

150

o bo

k 100

2 Time (hr)
(c)

Figure25.4 RRLresults Echo for ParkAvenue basin, Angeles, Los California, storm ofApril 18,1965: peaks, volumes, thehydrogriphs. (a) (b) (c) (AfterStall andTerstriep.e) 7. It appears that rainfall occursin greateramountsin the United Statesthan in Great Britain. This may accountfor the fact that the RRL method is successfuland widely used in Great Britain and yet suffers the abovedescribed breakdownsfor someof the basinsstudiedin the united States. 8. Better urban rainfall and runoff data are required for the proper testingof all mathematicalmodels.Research basinsthat do not havehydraulic pioblems,suchasundersized drainsor inadequate inlets, shouldbe selected and instrumented.

lllinois UrbanDrainage Area Simulator, ILLUDAS

As mentioned, the RRL method only simulatesrunoff from paved areas of the basin that are directly connectedto the storm drainagesystem.Grassedareasand nonconnected pavedareasare excludedfrom consideration. The ILLUDAS10model

640

MODELS SIMULATION RUNOFF 25 CHAPTER URBAN ofthe RRL methodbut also pavedareatechnique incorporaiesthe direct$ connected grassedand nolconnectedpavedareas', ,""oggrir", and incorporatesrunoff from very similar to the areahydrogriphsfor the subbasins.is Computationof grassed usedto 25.5 showsthe samesubbasin Figuie hydrographs. approachfbr pavedar-ea shadedareais the contributinggrassed ihe illustrate pavedarearunofi inFig. ZS.Z. area,which is largely the front yards ofresidences.Rain falling on any not-directlygrassed to connectedpavedarea is assumed run off instantly onto the surrounding from back and sideyards often areahydrologytakes over. Runoff area,and grassed street'The drainsgraluaily to a commonbick lot line andthenlaterally to the nearest areasfrom consideration suchgrassed traveltime requiredfor this virtually eliminates duringre1ative$shortintenseStormSnorma1lyusedfordrainagedesign. areahasbeenidentified,the curvein Fig' 25'5 can After the contributinggrassed time of be constructed.Travel times acrossthe grass strips are equivalentto the equilibrium from Izzatd's equation,2
t": 0.033KLq"o'67

(2s.r)

97 reaches percentof4,, that is, which is the time when the overlandflow dischrage ()\ )\ 4" : 0.00002311L of where 4, = discharge overlandflow (cfs/ft of width) at equilibrium -i to : raii supplyrate (in./hr), assumed be 1'0 in ILLUDAS (ft) L : length of overlandflow and 6:(0.00071 *c)S-o33

(2s.3)

where S : surfaceslope (ftlft) turf' c : coefficienthaving a value of 0.046 for bluegrass illustrated in to The time-area curve is assumed be a straightline. The endpoint,as grassed the Fig. 25.5represents traveltime from the farthestpoint on the contributing storageis normally set at 0.20 in. but can be varied' In ILLUDAS, depression Infiltration is modeledusing Holtan's equation,lr atea,

f:a(s-F)'o+f"
where

(2s.4)

,f : infiltration rate at time / (in'/hr) turf a : avegetativefactor : 1'0 for bluegrass at : storige available the soil mantle(in.) (storage the soil porosity in S minus storageat the wilting point) of F ; water alreaJy storedin the soil at time /, in excess the wilting point(in.)(amountaccumulatedfrominfiltrationpriortotimet) s_F:Storagespaceremaininginthesoilmantleattime'(in') (in./hr), generallyequivalentto the f" : finallonstant infiltration rate hydraulic conductivity (in./hr) of the tightesthorizon saturated Presentin the soil Profile usedto computean If physicalpropertiesof the soil are known, the equationcan be_ general interrelation betweenthe various innttiation curve. Figure 25.6 showsthe infiftratisn rates-andstoragefactorsinvolved'

25j
(a) Subbasln map (contributing grassedarea shaded)

SYSTEMMODELS URBAN STORMWATER

(c) Rainfall

641

(d) Runoff from supplemental paved area

(e) Losses

(b) Time versus grassed area curve

n
(f) Grassed area supply mte

92
a

cAoce't
GA: GAz GA
0 1 2 3 4 5 6 7 Travel time to inlet (min)

al

3 4 s 6 7 8 9 1 0 Time (min) (g) Hydrograph Qr= GAt (GASRT) + Qz= GAz(GASR1) GAI(GASRI) + + Qz= GAz(GASRI) GA2(GASR) GA1(GASR3) + Qn=GAn(GASnr) "'+ GA1(GASR,)

,? o

u F
I

8 10121416

Time from start of rainfall (min)

(After Terstriepand hydrographs. of Figure 25.5 Elementsin the development grassed-area Stall.lo)

642

CHAPTER25

MODELS URBAN RUNOFFSIMULATION

Infiltrat lon curve ial foinilial infiltration rate

S = Area cross-hatchedbelow curve equals total storage in soil (in.)

!.1 E

s-r'

\%
Final, constant -; " infiltration rate

Available smragem soil (in.)

Time(hr)

Figure 25.6 Diagram of inflltration^relations used in ILLUDAS, Eq. 25.4. (After Terstriepand Stall.'u) _ TabIe 25.4 provides an example computation of an infiltration curve for bluegrass on a silt loam soil in which soil moisture S of 6.95 in. is available. The equation is

f:r(6.9s-F)t++0.50

r r 55 )

Standardinfiltration curveshavebeen devisedfor use in ILLUDAS for soils having SCS hydrologicgroupsA, B, C, and D (Chapter4). Thesecurveswere synthesized LOAM CURVE SILT OFINFILTRATION FOR 25,4 COMPUTATION TABLE
Available srorage, Water stored, A F F (in.) (in.) 0.9s "1.0 1.0 1.0 1.0 1.0 1.0 0 0.95 1.95 2.95 3.95 4.95 5.95 6.95
lnfiltrationrate

Time Lt' (hr) 0.07 0.09 0.11 0.16 0.24 0.43 r.43 t (h0 0 0.07 0.16 0.27 0.43 0.67 1.10 2.53

s-F
(in.) 6.95 6.00 5.0 4.0 3.0 2.0 1.0 0

(S- D'+
15.0 12.3 9.5 7.0 4.65 2.64 1.0 0

t (in./hr)
15.5 12.8 10.0 7.5 5 .l 5 3.r4 1.50 0.50

'avg

(in./hr)
t+.1 I 1.4

8.7 6.3
A 1

2.3 o.'l

"Incremental rime, Lt : A'F/f*". Source: Tersftiep and Stall.ro

MODELS 643 SYSTEM STORMWATER 25.1 URBAN

from the Horton equation where /o : ,f" : e: k : / : f: f"+ (fo- f,)e-k, initial infiltration rate (in'lhr) ultimate infiltration rate baseof natural logarithms as a shapefactor selected k : 2 time from start of rainfall

(2s.6)

The curves This equationis solvedin ILLUDAS by the Newton-Raphsontechnique. inFig.25.7. are shown To accountfor wet versus dry conditions, II-LUDAS divides the antecedent ranges'shownin Table 25'5'Each moisturecondition (AMC) into four user-selected 5-day precipitation prior to the storm day. Infiltration from is based on the total on F;q.25.6 is varied, depending the AMC value specified' the user to operatein two modes,analysisand design.For ILLUDAS allows pipe and hydrographs providesnominal stormsewer mode,the modelgenerates design to without surcharge, passthe peak flows. In analysis adequate, diameters thatare throughoutthe basinnodesandlinks andthen mode,the modelg"n".ul"t hydrographs input pipe diametersare too small' It also sumsthe volume of alerts the user if iny by runoff water backed up ai inlets becauseit could not be accommodated the sewers. storm undersized

Time(hr) turf on Figure 25.7 Standardinfiltration curvesfor bluegrass four SCS soil types used in ILLUDAS. (After Terstriep and Stall.lo)

644

MODELS RUNOFF SIMULATION 25 CHAPTER URBAN CONDITIONS MOISTURE TABLE25.F ANTECEDENT LAWNS FORBLUEGRASS
AMC number
I

Description
Bone dry Rather dry Rather wet Saturated

Total rainfallduring storm 5 days preceding (in.)

2
J A

0 0-0.s 0 . 5 -I Over1

Source: Terstriep and Stall. lo

The ILLUDAS model requires estimationof severalinput parameters.Other The the evaluated sensitivityof ILLUDAS to variationsin parameters. studies12'13,14 study in Ref. 13 concludes: as in L. The sensitivityof the peak flows to changes AMC increases the soil group changes from D to A. the 2. The rangesof sensitivityto the soil groupsand AMC are approximately 3. XT;;rr" in the AMC from2to 3 and a changein the soil group from B to C are critical for large designreturn periods, and from C to D for small designreturn periods. 4. The ti'me to peak for various combinationsof soil group and AMC, for different return periods,remainsthe same. from 1 to 5. The peak flow and runoff volume increaseas the AMC changes is 4. This increase particularly important betweenAMC 2 and 3 in general and betweenAMC 3 and 4 for soil group A. The peak flow and the runoff from A to D for constantAMC. volumeincreaseas the soil group changes This increaseis particularly important between soil groups B and C in generaland soil groupsC and D for AMC 1. markedly for time inerementslarger than 5 min, 6. The peak flows decrease significantly for time incrementslarger and the pipe diametersdecrease than 10 min. 7. The time increment should not substantiallyexceedthe paved area inlet time.

Ovefflow Runoff Model (STORM) Storage,Treatment,

STORM is the Corps of Engineerscontinuoussimulationmodel of the quantity and quality of urban Stormwater resulting from single eventsor continuousdaily rainfall.ls It also simulatesdry weatherflow from domestic,commercial, or industrial Wet weather hydrographs,simulated from intermittent or continuous discharges. rainfall, can be used for a variety of hydrologicstudy purposes. hourly (hydrographs that alsoprovidewaterquality characWet weatherpollutographs can be predicted for individual historical or synthetic eventsand used in teristics) consistof The pollutographs of assessments impactsof runoff on receiving streams. runoff rates,amountsof pollutants,and pollutant concentrations. hourly

MODELS 645 SYSTEM STORMWATER 25.1 URBAN

RainfalVsnowmelt

Treatment

Figure25.8ConceptualviewofurbansystemasusedinSTORM.(AfterU'S.Army of CorPs Engineers'") by the degreein The model is conceptualized Fig. 25.8. Snowmeltis simulated inlormation is output to aid in the selectionof day method (Chapter 14). Statistical desiredcontrol of storm sbrage "upu"iti., und treatment rates required to achieve annualerosion' averannualrunoff, avefage *ut". rurr&f. Statistics,suchas average annualpollutant overflowfrom and average ageannualoverflowvolumefrom storige, are storage, all Provided' (to signal the The model simulates interactionof precipitation,air temperature erosion, dry weather flow' snowfall), runoff, pollutant accumulation,land Jurface or treatment system'The storage,treatmentiates, and overflowsfrom the storage prograncomputescontinuousorsingle-.event.runofffromrainfall' rainfall and depresof nunotf is computedas a fractio-n the differencebetween dependson land use. Runoff in excessof the ,ion ,torug". The fiaction selected treatment'Runoff for capacityis divertedinto storage subsequent speciteOiTeatment becomesoverflow and is of in excess both the treatment rate and stofagecapacity diverted directly into the receiving waters'

No. SCSTechnicalRelease 55 GR'55)

for The SCS TR-55 Procedures an in areaswere described urbanized mat recommends theseProcedures severalvendorshave1 Procedures,

646

CHAPTER25

MODELS URBAN RUNOFFSIMULATION

from through a numberof outlets.A public domainversionis available them avai{able the U.S. National TechnicalInformation Service.l6 initial studrenditionsof TR-55 should-perform Usersof the vendor-developed TR-55 program would be one that to ies usinghand-checks verify the code.An ideal used,notifies statesall assumptions usesSCS sourcecode or has SCS endorsement, options for making adjustmentsto the user of range violations, and incorporates accountfor all or most of the following: 1. Changesin the 484 coefficient of Eq. 12.22 for steep, avefage,or flat watersheds. 2. Per centimperviousness. 3. Percentof channelthat is improved. 4. Pondingarea. range. length over width ratios that fall outsidethe assumed 5. Subarea 6. Slope. 7. Antecedentmoisture conditions' 8. Different storm distributions' 9. Proper lag time equation. that the duration for the derived 10. Recognitionof the SCS recommendation time of concentration' unit hydrographbe about 13 percentof the subarea abstractions,1,, greater than that have initial 11. Allowance for watersheds 20 percentof the potential maximum retention, S. Whether by manual or computer operations,the SCS cautions that TR-55 hydrograph methods should not be used to perform final design if an error of Their adviceis to useTR-20, after ZS peicettt,inpredictedvolumecannotbe tolerated. is urban wathershed very complex if -aking appropriateparameteradjustments, the regardingthe useof precautions Other of or if a higher degree accuracyis required.lT the graphical and tabular methodsare identified in Section 15.2' in among time parameters SCS Urban Time Relationships The relationships phenomreconciledwith observed SCShydrologicmethodshavenot beencompletely Several ena or time relationshipsin other modelsof urban rainfall-runoff processes. adjustmentsfor urban effects, are formulations for lag time, with miscellaneous mentioned in Section ll.7 and elsewherein SCS literature, and have substantial impact on the shape of the hydrographs.The rational formula (Chapter 15), to that the time of concentration,deflnedas the time for rain falling illustrate, assumes at the most remote location to reachthe outlet, equalsthe time to peak of the urban hydrograph.lzzwd found this to hold approximatelytrue in observingrunoff hydrographsfrom pavedareas.2 The greatestdiscrepancyfound when comparing SCS and known urban time methis relationships the prolongedtime basethat resultswhen SCSunit hydrograph runoff from is basedon observed shape The hydrograph odsin Chapter 12 areapplied. As rural watersheds, shownin Fig. 12.13,the time basefor the dimenundeveloped, unit hydrographis about 5.0 times the time to peak. It was shownin Chapsionless time must requiresthat the release of ter I}thatlinear superposition unit hydrographs is the time of concentration,/"' Recall equal the time baseof the IUH, which in turn

25j

SYSTEMMODELS URBAN STORMWATER

647

FOR A D-HR UNIT OF TABLE 25.6 * COMPARISON TIME RELATIONSHIPS BY HYDROGRAPH SCS AND URBAN RUNOFFMETHODS Rational/lzzardllUHModels of Urban Runoff Given: D: t" Time to peak: t, Releasetime : tc Solving: D:t" Time to peak = t Releasetime : tc Timebase:D+t,:2t,

Unit SCSDimensionless Hydrograph

D : O.2 X time to peak Time to peak : Iag time + D/2 Lag time : 0.6 t" (Mockus Equation) Time base = 5.0 X time to peak D : 0.133 t" Time of peak : 0.666 r. Time base: 3.33 t, time : Time base - D : 3.20 t, Release

that the excess-rainfallreleasetime, t,, was definedin Chapter 11 as the time from rain to end of direct runoff. As shownin the Table 25.6, time relationend of excess unit shipsfor the SCS dimensionless hydrographof Fig. 12.13 giveprolongedrunoff durationscomparedto other urban runoff models.Only the time to peak is approxiUrban runoff modelsbasedon SCSdimdnsionmately equivalentin this comparison. less unit Mrograph proceduresmay result in longer time basesand hydrograph than other methods. recessions

Model(DR3M) RoutingRainfall-Runoff USGSDistributed

The U.S. GeologicalSurvey simulationmodel for urban rainfall-runoff applications tnodel for small watersheds originatedin tgZg as a lumped parametersingle-event to was expanded distributedparameter (deicribed in Chapter 24) and subsequently status,intendedprimarily for urban applicability.lsAlso, a soil moistureroutine was simulation. addedallowing quasicontinuous miles from a few acresto severalsquare The modelian be appliedto watersheds or It in size (an upper limit ol10 mi2 is recommended). doesnot simulatesubsurface interflow coniributionsto streamflow,and thesemust be externally addedif considered imPortant to the simulation. Routing of rainfall to channelsis by unsteadyoverland flow hydraulics,and by routing hydrographsthrough channel reachesis accomplished kinematic-wave methods(."t"ito Ctrupter13). The differential routing equationsare solvedby one of The usermay specify an explicit or implicit finitethree optional numericalmethods. differencealgorithm, or the method of characteristics' Time may be discretizedby the user in as small as l-min increments.The smallesttime increment is used by the program during any dayshaving short-time interval rainfall, calledunit days. Otherdaysare simulatedas24-hr intervals.Movement of surfacewater is simulatedonly during unit days.For the rain-free intervals, daily rainfall is input and usedto modify the soil moisturebalanceleadinginto the The formatforrain datais compatiblewith that of the U.S. Geological nextunit day(s). Input data Survey systlm, WATSTORE(Water Data Storageand Retrieval System). can also be obtainedfrom any local National WeatherServiceoffice. practically any basincanbe studiedby breakingit into severalsetsof four types (must be approximately of model ,"g*"ntr. Theseinclude overland flow segments

648

25 CHAPTER

MODELS SIMULATION URBANRUNOFF

Legend: Overland flow boundary Stream channel

oF1

Overland flow segment I

1 Reservoir R-ESI 1 CH1 Channel

Segment

Lateral inflow Rahfall.excess Rainfall excess oF1, OFz

Upstream inflow

oF1 oF2 cHl RES oF3 oF4 CHz

cH1 Rainfall excess Rainfall excess oF3, OF4

RESl

Watemhed outlet

for of Figure25.9 Segmentation watershed DR3M' rectangular),detentionstoragefacilities, channels,and nodes.This is illustrated in in fig.2i.9. Eachsegment the flgure may haveinflow from either lateral or upstream (or both, as occurs for segmentCHz). sources Rainfall is uniformly distributedover the overlandflow rectangles'Each has a Laminar flow is assumed and given length, slope,roughness, percentimperviousness. 13'56arefound of bandmforEqs.13.28and Ioo".or&.rtheiesegrnents.Thevalues fromm: 3 and
n :K - -

'

88So u

(t\ 1\

where56is the slope,r: is the kinematic viscosityof water, equalto 0.0000141ftlsec (for 50"i water), and K is a coefficientrelating the Reynold's numberN. to Darcy's 24.Thevalue frictionfactorfbyK: fN,.FlowoverroughsurfacesislaminarifK) K is related to rainfall intensity by (25.8) K: Ko+ 101 where K6 is fouhd from Table 25.7, andl is the rainfall intensity (in./hr). Channelsin Fig. 25.9 representeither natural or artificial gutters or storm pipes are allowed). Inflow to channels sewers(either op"n .hunntls oi nonpressure Nodesare used overlandflow (aslateral inflow), or nodes. comesfrom otherchannels, or or reservoi-r, whenthe user contributeto a channel whenmore than three Segments wishesto specify an input or baseflow hydrograph. describedin Section 13'3' Channel routing is by kinematic-wavetechniques, parametersb and m. These arc Input is the channei length, slope, and routing )

25.1 URBAN STORMWATER SYSTEMMODELS TABLE 25.7 ROUGHNESS COEFFICIENTS FOR OVERLANDAND CHANNELSEGMENTS Sudacetype Concreteasphalt Bare sand Graveledsurface Bare clay-loam soil (eroded) Spatsevegetation Short grassprairie Bluegrass sod
Source: Alley and Smith.r8

649

Laminarflow Ko

Turbulentflow Manning's n

24-r08 30-120 90-400 100-500 1,000-4,000 3,00010,000 7,000-40,000

0.01-0.013 0.01-0.016 0.012-0.03 0.012-0.033 0.053-0.13 0.10-0.20 -0.48 0.17

developed from the Manning equation and slope of the channel,respectively.The equationsare m : 1.67 and
D : -

l.4gsto/z
n

(25.e)

where ruis obtainedfrom Table 25.7 or similar information. The model adjustsboth b and m for various shapes,including circular and triangular (seeTable 13.1 for severalru values). overbankflow is possible, secondsetof b andmparameters If a can be input for all flows in excess the channelcapacity. of Reservoirinflow hydrographs routed by either of two storage are routing methods. If a linear reservoir model is appropriate,the storagecoefflcient K, relating outflow to storage O : KS,is input. If the modifiedPuls methodis desired, table by a of outflows and correspondingstoragelevels must be input. The model assumes an initial reservoirlevel equal to that corresponding an outflow of 0.0 cfs. to Pondingbehind culverts can be modeledas a reservoir if an S-O relation is input. This should include data points correspondingto roadway overflow to allow simulation of this common phenomenon. Excess rainfall (runoff) from perviousareasis developed from the precipitation input, minus severalabstractions. Infiltration is simulatedby

r:r(' .{of)

(2s.10)

where K is the hydraulic conductivity, P is the averagesuction head acrossthe capillary zone, andmo andm are the soil moisturecontentsbeforeand after wetting. The term SMS i$the soil moisture storase.The rate of excess rain is found from

r :
and

ir1<s

Qs.rr)
(2s.1,2)

, : r - ti f 1 > , s

where1 is the rate of rain suppliedto infiltration.

650

CHAPTER25

MODELS URBAN RUNOFFSIMULATION

on Runoff from impervious areasdepends whetherthe areasare directly conto drainagesystem.Those not directly connectedare assumed flow nected to the perviousareas,wherethey are addedaslateral inflow. One third of immediatelyonto and the rain on direcdy connectedareasis abstracted, the rest is lateral inflow to the gutter or channel. Soil moisture is accountedfor in a two-layer hypothetical storagezone. The amount in storageaffects the infiltration rate and allows continuoussoil moisture rain input) During unit days(dayswith short-duration rain events. between accounting moisturefrom Eq. 25.10 is addbdto the upper storagezone. Between all infiltraied proportion of the daily rain is added to SMS. During unit days, a user-specified rates occursfrom SMS, usinginput pan evaporation evapotranspiration rainlessdays, until the next rhin event,at which a coefficient.ttris processcontinues multiplied by time infiltration is governedby the amount of soil moisture; acrossthe continentalU.S. and Applications of On:U havebeen documented Hawaii.leCalibration of computedand measuredrunoff for almost in Alaska and reveal a median error in peak flow estimatesand 400 storms over 37 watersheds with the best results obtainedfor highly 24 percent,respectively, volume of 21 and to Indicationsfrom verificationstudies datearethat the model imperviouswatershedi. the peak flow ratesfor simulatedfloodsfrom stormshavingmagnimay overestimate tudesin the designrangeof flood control facilities, and give better resultsfor smaller stormstypically used in runoff quality studies.

FHWAStorm SewerDesignModel (HYDRA)

of As part of a package integrateddesigncomputerprogramscalledHYDRAIN,2othe U.S. Federal Highway Administration developedthe HYDRA storm drain design The model is distributedundercontract model for useby federaland other engineers. with the FHWA through Milrans SoftwareCenterat the Civil EngineeringDepartHYDRA hasbeenlinked by commerment of the Universityof Florida at Gainesville. The program'sprimary useis analyzcial vendorsto an integratedCAD/GIS system. ing adequacyof existing storm drains or designingnew storm drains and inlets by the rational merhod dJscribedin Chapter 15 or by a modified rational method the which represents hydrographas a Eapezoidhaving a volume equal to the calculated net rain. by Cornmercialversionsof HYDRA allow design the modifiedrational method, SCS methods,or revised Santa Barbara hydrographmethods.2lHYDRA has one over other storm sewerdesignmodelsin that hydraulic gradelines through advantage can be checkedby hydraulic backwatercomputationsto determinetotal the systern Another or and losses whetherinlets,manholes, junction boxesare surcharged. system the inlet capacityofthe storm usefulfeatureis'that streetand gutter flowsthat exceed sewersare routed by HYDRA to the next downstreamlocation and added to the at hYdrograPh that Point'

Model (SWMM) Storm Water Management

and appliedstormrunoff simulationmodel wasjointly develA very widely accepted oped by Meicalf und Eddy, Inc., the University of Florida, and Water Resources for Engineers22 useby the U.S. EnvironmentalProtectionAgency (EPA). This model

SYSTEM MODELS 25.1 URBAN STORMWATER

651

rainfall is designei to simulatethe runoff of a drainagebasin for any predescribed pattern. The total watershedis broken into.a finite number of smaller units or subby that can readily be described their hydraulic or geometricproperties. catchments A flowchart for the processis shownin Fig. 25.10. The SWMM model hasthe capability of determining,for short-durationstorms of of given intensity,the locationsand magnitudes local floodsaswell asthe quantity and quality of storm water runoff at severallocationsboth in the systemand in the receiving waters. The original SWMM was an event-simulationmodel, and later versionsitkeep track of long-term water budgets. The fine detail in the designon the model allows the simulation of both water with urban runoff and combinedsewersysquantity and quality aspectsassociated here.Information obtainedfrom are the water quantity aspects described tems.Only for would be usedto designstorm sewersystems storm waterrunoff control, SWMM in model is limited to relatively small urban watersheds regions where Use of the documented. of in the quality aspects water are adequately differences seasonal

Subcatchments

Gutterflow 1. Overlandinput 2. Gutterinput 3. Flow (Manning's) 4. Depth(continuity)

Figure 25.10 Flowchart for SWMM Runoff Block hydrographiccompu-tation.({fter Metcalf and Eddy, Inc.22)

652

CHAPTER25

MODELS URBAN RUNiOFF SIMULATION

fhe simulation is facilitated by five main subroutineblocks. Each block has a devices specificfunction, and the resultsof eachblock are enteredonlworkingstorage part of the input to other blocks. to be usedas The main calling programof the modelis calledthe ExecutiveBlock. This block interfacingamongthe is the first and last to be usedand performs all the necessary other blocks. The Runoff Block usesManning's equationto route the uniform rainfall intensity over the overlandflow surfaces, through the small guttersand pipesof the sewer pipes,and out of the sewerpipesinto the receivingstreams. into the main sewer system pollutional graphs(pollutographs). This block also providestime-dependent the package subroutines, TransportBlock, determines quality and the A third of infiltration, and calculatesthe quantity of dry weather flow, calculatesthe system water quality of the flows in the system. is for A usefulpackage subroutines water quality determination containedin of specify or have the model the StorageBlock. The StorageBlock allows the user to treatmentfacility in selectsizesof severaltreatmentprocesses an optional wastewater percentage the peak flow. If used,this block simulates of that receivesa user-selected passes as of the changes the hydrographs in and pollutographs the sewage the sewage processes. sequence unit of through the selected The earlier version22 allowed simulationof any reservoirfor which the outflow could be approximated either a weir or orifice, or if the waterwaspumpedfrom the as reservoir. The newer,version23 allows input of 11 points of any storage-outflow relation and routes hydrographsthrough natural or artiflcial reservoirs,including backwaterareas behind culverts. Routing is by the modified Puls method, which assumes that the reservoiris small enoughthat the water surfaceis alwayslevel. Evaporationfrom reservoirsis simulatedby a monthly coefficient(suppliedby the user) multiplied by the surfacearea. completes hydraulic calculationsfor overlandflows, in the The Extran Block2a pipes and culverts.It solvesthe completehydrodynamicequations, channels, and in performsdynarnicrouting, and providesall the depth,velocity, surcharging, assesses gradeline information requested. and energy by areas,slopes, widths, and linkagesmustbe specified the user. Subcatchment be suppliedfor perviousand imperviousparts Manning's roughness coefficientscan of eachsubcatchment. model listed As indicated in Table 24.1, SWMM is the only event-simulation If infiltration losses. parameters that usesHorton's equationfor calculatingwatershed infiltration the for Horton's equationareunavailable, usercan specifyASCE standard for determined eachtime stepare compared capacitycurves.Infiltration amountsthus surfaceplus any amountsof water existingon the subcatchment with instantanEous if the infiltration lossis larger,it is set rainfall that occurredduring the time step,and of Input for Horton's equationconsists the maximal and equalto the amountavailable. k recession constant in Eq. 4.1. The Green-Ampt minimal infiltration ratesand the equationis also usedin SWMM. Urban storm drainagecomponentsare modeledusing,Manning'sequationand the continuity equation.The hydraulic radius of the trapezoidalguttersand circular if pipesis calculatedfrom componentdimensions and flow depths.A pipe surcharges

MODELS 653 SYSTEM 25j URBAN STORMWATER it is full;provided that the inflow is greaterthan the outflow capacity.In this case,the surcharged amountwill be computedand storedin the Runoff and TransportBlocks at the head end of the pipe. The pipe will remain full until the stored water is completelydrained.Alternatively,the Extran Block canbe usedto conducta dynamic conditions. simulation of the systemunder pressure-flow Necessaryinputs in the model are the surfacearea, width of subcatchment, ground slope,Manning's roughness coefficient,infiltration rate, and detentiondepth. coefficient,invert slope, Channel descriptionsare the length, Manning's roughness diameter for pipes, or cross-sectionaldimensions.General data requirementsare processaccounts all inflow, infiltration for in summarized Table25.8. A step-by-step areas,providing a calculateddischarge losses, and flow from upstreamsubcatchment hydrographat the drainagebasin outlet. The following descriptionof the simulation the processincorporatedin early versionsof SWMM will aid in understanding logic of the model.2s hyetograph: to 1. Rainfall is addedto the subcatchment'according the specified Dt: D,+ &Lt where D, : the water depth after rainfall at D, = the water depth of the subcatchment time / R, : the intensity of rainfall in time interval At 2. Infiltration 1, is computed by Horton's exponential function, I, = * f" + (fo f")e-o', and subtractedfrom the water depth existing on the subcatchment (2s.r4) Dz: D, - I, A,t in where f",fr, k : coefficients Horton's equation(Eq. 4.1) Dt: the intermediate water depth after accounting for infiltration
FOR STORMWATER TABLE 25.8 GENERALDATA REQUIREMENTS MANAGEMENT MODEL (SWMM) population distribution, censustract data, Item 7. Define the StudyArea. Land use, topography, aerial photos,and area boundaries. Plans of the collection systemto definebranching,sizes,and slopes; Item 2. Define the System. types and generallocationsof inlet structures. Item3. Define the System Specialties.Flow diversions,regulators,and storagebasins. Maintenance.Street sweeping(descriptionand frequency),catchbasin Item 4. Define the System cleaning;ffouble spots (flooding). facility operatingdata; Item 5. Define the BaseFIow (DWF). Measureddirectly or through sewerage (composited BOD hourly variation and weekdayversusweekend;the DWF characteristics quantities,and quality). and SS results);industrial flows (locations,average period (6 months or longer) Item 6. Define the Storm Flow. Duly rainfall totals over an extended continuousrunoff encompassing study events;continuousrainfall hyetographs, the (BOD and SS) for the study hydrographs, and combinedflow quality measurements fully when and how taken). events;discreteor compositedsamplesas available(describe

(25.1,3)

654

CHAPTER25

MODELS URBAN RUNOFFSIMULATION

D, 3. If the resultingwater depth of subcatchment is larger than the specified detentiondepthDr, an butflow rateis computedusinga modifiedManning's equation.
tf Y _

!Ep,
n

- Do),/,s,r,

(2s.rs)
(2s.16)

and

Q-:

vw(D2 - Dd)

where V:thevelocity n : Manning's coefficibnt S : the ground slope IV : the width Q. : the outflow rate 4. The continuity equationis solvedto determinewater depthof the subcatchments resulting from rainfall, infiltration, and outflow' Thus

D,.L,:Dz-ftLt
whereA is the surfaceareaof the subcatchment.

o..

(2s.r7)

are until computationsfor all subcatchments com5. Steps I-4 arerepeated pleted. 6. inflow (OJ to a gutter is computedas a summationof outflow from tribu(Q..,) and flow rate of immediate upstream gutters tary subcatchments (Qr,)

Q^:2Q-.,t2Qr.,

( 2s .18)

7. Theinflow is addedto raisethe existingwater depth of the gutter according to its geometry.Thus O',

Y.:Y,+7M

Q5.19)

where Yr, Y, : the water depth of the gutter A, : the mean water surfaceareabetweenY, and Y, 8. The outflow is calculatedfor the gutter using Manning's equation:
y :
1 ' A- O ' l '
fL

Pz/t51/z

(2s.20) (2s.2r)

and* where

Qr:

VA"

R : the hydraulic radius S; : the invert sloPe arcaaI Y1 A" : lhe cross-sectional

9. The continuity equationis solvedto determinethe waterdepth of the gutter resultine from the inflow and outflow. Thus

Yt+^t: Y, t (Qr^ oJf,

(25.22)
)

25,1 URBAN STORMWATER SYSTEMMODELS

655

t.0. Steps6-9 are repeateduntil all the gutters are finished. 11.The flows reachingthe point concernedare addedto producea hydrograph
coordinate alongthe time axis.

12. Theprocesses from 1 to 11 arerepeated succeeding periods in time until the

completehydrographis computed. Three general types of output are provided by SWMM. If waste treatment processes simulatedor proposed, capital,land, and operationand mainteare the nance costs are printed. Plots of water quality constituents versustime form the secondtype of output. Thesepollutographs producedfor severallocationsin the are systemand in the receiving waters.Quality constituents handledby SWMM include suspended solids,settleable solids,BOD, nitrogen,phosphorus, grease. and The third type of outputis hydrologic.Hydrographs any point, for example, end of a gutter at the or inlet, are printed for designated time periods. The StatisticsBlock will provide frequencyanalysisof storm eventsfrom a continuoussimulation.

University Cincinnati of UrbanRunoffModel(UCURM)

The University of cincinnati urban runoff model (ucuRM) was developed the by Division of WaterResources, Departmentof Civil Engineering,of the University the of Cincinnati.26 flowchartis reproduced Fig. 25.11.The programconsists three A in of sections:(l) MAlN-infiltration and depression storage,and two subroutines, (2) GUTFL-gutter flow, and (3) PIROU-pipe rouring.It is similar to the EpA model and divides the drainagebasin into subcatchments whose flows are routed overland into gutters and sewerpipes. The rainfall is read in as a hyetograph.The infiltration and depression storageare summedand subtractedfrom the rainfall to give overlandflow. This is routed through the gutter system storm water inlets and to the pipe network. Starting at the upstreaminlet, the flows are calculatedin successive segments the sewersystem,including discharges of from inlets, to producethe total outflow. The drainage areais divided into small subcatchments with closely matched characteristics. The rainfall data are introduced and the infiltration is computedfor eachsubcatchment. Principal elementsof the modelingprocessfollow: 1. It is assumedthat runoff begins wheneverthe rainfall rate equals the infiltration rate and the mass of precipitation balancesinfiltration. The equations representing theseconditionsare

t.I /:-krn1
and

I |

Qs.zt1

t*.'#0

- e k): mi(r) +

+ + r) - ,rrlr]-1 {;trr ffiliv

(2s.24)

MODELS SIMUI-ATION RUNOFF 25 CHAPTER URBAN

,/ '1. Read Data RainfallintensitY 2. Infiltration constant storage caPacitidr 5. rj"p.o.ion

Infiltration (Horton's equation) Depression, supply storage

t
/ |

/-----ffi

Data 1 Slop 2. Roughness J. Lenl ;th 4. Widrh

il Overlan flow 1. Detention 2. Dis:harge

ReadData 1. Layout 2. Length Gutter flow input 1. Overland put 2. Gutterin.. o 3. Discharg (continuitY) Gutters '@

1. SloPe 2. Roughness 3. Diarieterorheight 4. Length

1. 2. 3. 4.

Flow routlng Gutter input Average velocitY (Manning's) Time offset AddoffsethYdrograPh and new qutter input

Pipes

Print hydrograPh

--16A \ -,/

Figure 25.11 UCURM model flowchart' (ATterPapadakisand Preul'26)

25j

URBANSTORMWATER SYSTEMMODELS

657

* where mi(I) : i(1) : k : "fo: f": DT : t :

the massprecipitateduntil time r (in.) the ordinatesof rainfall intensity curve the decayrate of infiltration (units/min) the initial infiltration capacity(in./hr) the ultimateinfiltration capacity(in./hr) the time incrementof rainfall intensity curve the time to intersectionof rainfall curve and infiltration curve x : arrincreinent of DT

The infiltration curve is computedfrom the equationsand t, I, andx arc stored. 2. Surfaceretentionis relatedto depression storage an equationderivedby by Linsleyet al..2s (25.25) s : (t _ f)e-@-F)/sd where 57 : P : F : i : f : s: the total depression storage(in.) the accumulated rainfall in storage(in.) the accumulated infiltratioh (in.) the rainfall intensity (in./hr) the infiltration (in./hr) the surfaceretention (in./hr)

The infiltration and surfaceretentionare subtracted from the rainfall intensity to yield the runoff. 3. The hydrographof the overlandflow is derivedby solving

- T+?:
where

,/,Dinlt. *ry ou(#)'f'''

(2s.26)

(25.27) D" : (0.009'79n'o'6ro6lo'6)fso'3 Dr.z : the detention storageat the beginning and end of time interval I (in./unit area) : the overlandflow supplyat the beginningand end of time rr, rz interval t (cfs/min) n : Manning's coefficient l: the lengthofoverland flow s : the slope (ftlft) (in./hr per unit area) 4 : discharge

4. For the initial time increment, q1 : 0 and Dr : 0 are substituted, is D, calculated,and q, is found from

q: yfnl ,,,,p,,,1r * o.o(S)'l','

L \D"/ )

es.z8)

where the symbolsare as previously defined.The determinedD" and q, becomeDt and qt. The overlandflow hydrographis derivedby repeating this cycle.
I

658

MODELS SIMULATION RUNOFF 25 CHAPTER UEBAN 5.Thegutterflowiscomputedusingthecontinuityequation

60 *6!7: At 6x

q,

(2s.2e)

where Z is the width of the water surface' Thetelm(ayla|risneglectedbecausethechangeinthedepthofthe to gutterflow is very small with respect time. After integration,the equation becomes (25.30) Q: q"L * Qo where Qs : L : q": Q: upstreamgutter contributions the length of the gutter (ft) the overlandflow from the hydrograph the flow from the gutter system

METHODS OF TABLE 25.9 COMPARISON URBAN RUNOFFMODELSAND

Degree of sophistication of surface flow routing Low Degree of sophistication of pipe flow routing Low Accurate modelingof surcharging No Flexibilityof modeling of storm draln components Low

Model

Surfaoe routing
Peak flows only Yes Yes

Pipe flow routing Peak flows only No In combination with surface In combination with surface Yes No Ye$

Quality routing No

Ratiorlal method Chicago Unit hydrograph STORM RRL MIT EPA-SWMM Cincinnati HSPF ILLUDAS
Sour ce;,}/ialifred.sftef

No No

Moderate

NA Low

NA No

NA Low

:"*
Yes Low Low Low

Yes

Yes Yes Yes

No

Moderate High High

No Yes Yes Yes No

Lowmoderate NA Moderate

No NA Yes

Low NA High

Yes Yes

Yes Yes

High Moderate

Low Moderate

No No

Low Low

(ucuRM)

Yes

Modetate

High

No

Low

I,age13nd Smit!'-8

25.2 URBAN RUNOFFMODELSCOMPARED

659

25.2 URBANRUNOFF MODELS COMPARED

Severalquantitative comparisonsof the RRL, SWMM, UCURM, ILLUDAS, and STORM models have been reported in the literature. A qualitative comparisonof severalwas preparedby Lager and Smiths and is shownas Table 25.9. Table 25.10 provides a bullit matrix showing componentsof most of the same models. Other comparisonsinvolve quantitative analysisof results of the models when applied to actual guagedstorm events.One of the first was an applicationby Heepsand Mein6 of three modelsto two urban catchments Australia for a total of 20 storm events. in A similar statisticalcomparisonof the samethree modelsapplied to 12 stormsover eachof three urban watersheds performedby Marsaleket al.7Significantresults was Another quantitativecomfrom theseindependent evaluations summarizedhere. are parisonis providedby Huber.27 The Heepsand Mein conclusions model performanceare:6 of

1. The degree subdivision the catchment a significantinfluenceon the has of of

peak dischargepredicted by each of the models.The RRL and SWMM methodsgive lower peaks and the UCURM gives higher peaks for finer subdivision. The UCURM containsseveraldeficiencies. The major ones,the effectsof predictedhydrographs, storages arelhat depression which canbe seenin the

Explicit modeling of insystem Treatmenl modeling storage No

Receiving model available

Degree of calibration/ verification required

Simulation period Individual storms Individual storms Individual storms Long term

Availability Nonproprietary

Documentation Good

Data requirements Low

NA NA NA

No

Usually not verified

Moderate High

NA No

No No

Nonproprietary Nonproprietary

Fair Fair

Moderate Moderate

No

Yes

No

Low

Nonproprietary

Good

Moderate

No NA Yes

NA NA Yes

No No Yes

Moderate Moderate Moderate

No No

No No

No Yes

Moderate High

Moderate

Individual storms Individual storms Individual or conttnuous storms Individual storms Individual storms or long-term Individual storms

Nonproprietary Proprietary Nonproprietary Nonproprietary Nonproprietary Nonproprietary

Good Fair Good

Moderate Moderate Extensive

Fair Fair

Extensive Extensive

Good

Extensive

uer6o:d lelnduoC elqElrE^E

o
f q) c

a J

eun-lEeu lo4uoc
u6;seg suollelnduroc o a a a o a o a
a a a a a

o a o o
a a

o a o

o
al,

ollrl lBruolur osooqcuEc uorlelnurs snonurluoc

6urnreceg uonelnursI1r;enb ralenir 6utntecag uollElnurs /v\ollJlE/v\

^Uleng srxrolsuoo^iqeq acueleq

luouleoJl roleMelsEM
o 3 o) o

a a a a a
a a a

^lrlpno suollgEar
rnocs pue uorteluor1jrpes 6urlnor r{1r1eng ,{Ulenb Jolel rulols fllenb reqleerirfuO sa!iloolaA sluud e6els sluud

a
J

tJJ F

o
a a

6
n

o o z 9)
F

a a a o a a o
a

a a o

a a

o a

z u z
(L

e6eroig suorlels6urdr-un3

a
a

E
o 3 (l) U)

suolsroAro pue rvro;; ernsse.rd bur6reqcrng

pue loJluocMou uJEerlsu/v\op uerlson sloitlss ut Oupnorlrol3 a a
a

J UJ

o
I

a a a a
a

o o a o
o a

a o a o a o

o z
l (f,

suilols uoo/\laq acuelEqralEM sEale snotruoq uroJllJounH

a a a

a o

o a a

z
dl l t!

cc
c

uro,rl ssoJe snorrugour llounu

o E
o
G'

llau$ ous srlder6ole,{r.l lerolos}o }ndul

ill\Ol,lOt{}EO/tA ruO

o
z
U) (E

a o
a

a a
a o

o a
o

a a
f

a o
o a

o
!

sMollur lueuqclec eloruny\

o o
d
N

':
! E

u,l E

p
q

COMPARED 661 MODELS RUNOFF 25.2 URBAN are assignedfull when the rainfall intensity falls below the infiltration by are storages not depleted infiltration. The capacity,and that depression valuesof the rainfall intensity(difficult to obtain from useof instantaneous recordercharts)can causevolume errors. ' 3. The SWMM was the model with the best overall performancebut at the ' expense large computerstorageand time requirements. of 4. The RRL model predictedpoorly for stormsin which perviousrunoff was well for many other types of storms. significantbut performedreasonaf;ly in general,supportthoseof Stall and Terstriep'e The results, models is the prediction of 5 . A major problem with using noncontinuous by This problem is further aggravated use of the antecedeniconditions. is equationfor which prediction of the parameters virtuHorton infiltration ally impossible. usingthe samethree modelsfor three watersheds The Marsalek study results,T Maryland, indicated that the SWMM model performed in Illinois, Ontario, and slightly better than the RRL model and both thesemodelswere more accuratethan studied. the UCURM model for the small watersheds usedby Marsalekin of providesdescriptions the urban watersheds TabIe25.11 timesto and calculated of Typical comparisons observed the runoff modelevaluation. runoff volumesfor the three modelsare providedinFig.25.12 peak,peak flows,and in for the Calvin Park watershed Kingston, Ontario. to the information described developthe qualitative comMarsaleket al. used and, arrived at the following conclusions:7 parisonin Table 25.12 modelsfor urbanrunoff, suchasRRL, SWMM, 1. Uncalibrateddeterministic betweenthe simulatedand yielded a fairly good agreement and UCURM, of eventson typical urban catchments small size. runoff measured about 60 percentof the simulatedpeak flows were within 2. Onthe average, About the samescatterwasfound for the values. 20 percentof the measured betweenthe times to peak and runoff volumes.The agreement simulated and simulatedvaluescould be further improvedby modelcalibrameasured tion. 3. Wherr comparing the entire simulated and measuredhydrographsusing was the statisticalmeasuies, agreement found goodfor SWMM, goodto fair of RRL, and fat in the caseof UCURM' in the case TABLE25.11DESCR|PT|oNoFTESTURBANWATERSHEDSUSEDFoRTHE
OF EVALUATION URBAN RUNOFFMODELS

Testurban drainage basin

Oakdale Calvin Park Gray Haven

Location
Chicago, IL Kingston, Ontario Baltimore, MD

Area (acres) 13.0 89.4

zt,J

lmoerviousness
4f .6

Land use Residential Residentialand institutional Residential

27 52

PART SIX

STATIST]CAL METHODS

Chapter26

and Statistics Probability

r Prologue
The purposeof this chapteris to: . Introducethe basictenetsof probability theory as appliedto random,hydroon logic variables,with particular emphasis the relativefrequency definition foundationalto the frequencyanalysisproof probability-a conceptthat is in cedurespresented Chapter27 andthroughoutmany other chaptersof this text. . Describecommonprobability distributionsand showhow they are appliedto hydrologicphenomena. . Relate the fundamentalsof probabitity theory to hydrologic designcriteria in described Chapter16, Section16.3. . Acquaint the readerwith the theory behind linear regression and show how in hydrologyto predict how a study watershed is this powerful technique used to of by will respondto somechange examiningresponses the watershed past in of responses other similar watersheds inputs or by statisticallyscrutinizing predictive equationfor the subjectwatershed. order to developa ' Show how to transform many hydrologicvariablesthat havenonlinear relationships into new variatesthat can then be analyzedby performing linear on regression the transformedvariates. . Providethe theoreticaland practical foundationnecessary fully capitalize to in principlesdiscussed Chapter16 andthe time-series on the hydrologicdesign describedin Chapter22. modelingprocedures analysis*and commonly, the study of hydrologyis undertakenby readerswho lack the prerequisite backgroundin principles of statistics,probability theory, and frequency early in the review thesesubjects most hydrologycourses As analysis. a consequence, include chapterson ;tatistical methodsto Practically all hydrology texts schedule. summarizethe basic principlesof statistics,probability theory, probability distributime-seriesanalysis, tions, bivariate and multiple linear correlation and regression, the placementof this material at the end of this Thus,despite and frequencyanalysis. or that the readerhasthis background will studythe material text, the authorsassume

672

AND 26 CHAPTER PROBABILIry STATISTICS with an understanding in Part Sif, prior to beginninga study of Part Three. Readers analysis,and the basicsof probability distribution of statisticalmethods,regression functions may wish to turn directly to Chaptet27.

ANALYSIS AND STATISTICAL VARIABLES 26.1 RANDOM

that isn't sufflciently exA random variable is one that demonstrates.variability havethis tendency,apMany hydrologicphenomena plained by physicalprocesses. or pearing at times to be fully subjectto chancethemselves, driven by some other of closelyrelated factor. In practice,hydrologistsoften analyzeproblemsas systems For connectedrandom and deterministicprocesses. example,precipitation is often of statisticallyas a randomvariablebecause the complexityof understandevaluated processes are known to drive the precipitation that modelingthe atmospheric ing and system.Runoff that results from the precipitation, on the other hand, is viewed of that arethe nucleus the majority usingthe rainfall-runoff analogs deterministically, this textbook. of Hydrology relies heavily on principles from probability theory, statistics,and generaWhole texts on frequencyanalysismethods,stochastic information analysis. are and analysisof variance,and regional analyses available of data, regression tion are of containingthorough descriptions the principles.l'2Many hydrologicprocesses complexthat they can be interpretedand explainedonly in a probabilislic sense. so Hydrologic eventsappear as uncertaintiesof nature and are the result, it must be The inforwith randomor stochasticcomponents. of assumed, an underlyingprocess is mation to investigatetheseprocesses containedin recordsof hydrologicobservaobserved tions.Methods of statistical analysisprovidewaysto reduceand summarize data, to presentinformation in preciseand meaningfulform, to determinethe underphenomena, to makepredictionsconcerning and of lying characteristics the observed future behavior.Statisticalanalysisdealswith methodsfor drawinginferencesabout the population basedon examinationof samplevaluesfrom the population. These range,distributionwithin inferencesincludeinformation aboutthe central tendency, the range, variability around the central tendency,degreeof uncertainty, and frequencyof occurrenceof values. Statistical analysisinvolves two basic sets of problems, one descriptive, the otherinferentiat. The former is a straightforwardapplicationof statisticalmethods, little risk. The inferential problem,however, and requiringfew decisions representing of decisionsbearing some risk, and requires an understanding the methods entails and the dangersinvolvedin predicting and estimating.The most common employed when only the inferential problem is to describe whole classof possibleoccurrences The wholeclassis thepopulation andthe portion portion of themhasbeenobserved. a observedis the sample. understudyare continuousif they may take in The randomvariables the process on all values in the range of occurrence,including figures differing only by an infinitesimal amount; they are discreteif they are restrictedto specific,incremental values.Distribution of the variablesover the rangeof occurrenceis definedin terms of the frequencyor probabitity with which different valueshave occurredor might occur'"

26.2 coNcEPTSOF PROBABILITY 673

26.2 CONCEPTS PROBABILITY OF

The laws ofprobability underlieany study ofthe statisticalnature ofrepeatedobservations or trials. The probability of a singleevent, say Et, is definedas the relative number of occurrencesof the event after a long seriesof trials. Thus P(E ), the of probability of eventEr, is nrfN for n, occurrences the sameeventin N trials if N n, is sufficiently large. The number of occurrences is thefrequency. and nr/N the relatiw freQuenffobabilities and the rules governingtheir manipurationare known In intuitively or from experience. the familiar coin-tossingexperiment,P(heads) : P(tails) : |. Eachoutcomeof a singletoss(a trial) has a finite probability, and the is outcomes 1. Also, the outcomesaremutually sumof the probabilitiesof all possible exclusive;that is, if one occurs,saya head,then a tail cannotoccur.In two successive tests,there are four possibleoutcomes-HH, TT, HT, TH-each with a probability ofthe other one, probabilitiesfor eachtrial is independent of|. In this case,because trial) : L,x L : j. Again, the eachoutcomeare foundby P(first trial) X P(second sum of the probabilities of the possibleoutcomesis 1. Note that the probability of gettingexactlyoneheadand one tail duringthe experiment(without any regardto the is order) P(HT) + P(TH) : *. Summarizingthe rules of probability indicated by coin tossing,we find the following:

1. The probability of an event is nonnegative and never exceeds 1. 0<P(E)<1

( 26.r )

2. The sum of the probabilitiesof all possibleoutcomesin a singletrial is 1.

s
I

.1J

P(8,) : 1

(26.2)

and 3. The probability of a numberof independent mutually exclusiaeevents events. is the sum of the probabilitiesof the separate P ( 4 U E 2 ) : P ( E y )+ P ( E 2 )

(26.3)

The probability statement,P(Et U Et), signifies the probability of the "the probability of Et ot Er." union of two eventsand is read or eventsoccurring simultaneously in 4. The probability of two independent is succession the product of the individual probabilities. P(q ) E,) : P(E) x P(E) (26'4)

P(& a E ) is called probability of the intersectionof two eventsor ioint probability and is read "the probability of El and Er." or Considerthe following exampleof eventsthat are not independent mutually exclusive:An urban drainagecanal reachesflood stageeach summer with relative freqlrencyof 0.10; power failuresin industriesalongthe canal occur with probability

674

CHAPTER26

AND STATISTICS PROBABILITY

of showsthat whenthereis a flood the ch'ances a powerfailure for of 0.20; experience are to 0.40. The probability statements whateverreasonare raised P(power failure) : P(P) : 0'20 P ( f l o o d ) :P ( F ) : 0 ' 1 0 P(no powerfailure) : P(P) : 0'80 P(no flood) : P(F) : 0'90 P(power failure given that a flood occurs) : 0'40 is The last statement called a conditionalprobability. It signifiesthejoint occurrence of events and is usually written P(P I F).. Rules 3 and 4 no longer are strictly : 0.3. If the eventsreIf applicable. Rule 3 applied,P(F u P) : P(F) + P(P) theionditional probability P(P I F) would gqualthe marginal independent, maineO if probability f (p). T'nor the eventsare independent the probability of either is not i'conditionedby" or changed knowledgethat the other has occurred'For indepenby dent events,the joint probabilitieswould be P(F)P):0.1 x0.2:0'02 P(FaF):o.tXo.8:0.08 P(FnP):0.9X0.2:0.18 The probability of a flood or a power failure during the summerwould be the sum of the flrst three joint probabilitiesabove. P(FUP):P(F nP)+ P(F.P)+p(F lP):e'23 of however,from the statement conditionalprobabilThe eventsare dependent, : 0'1, a powerfailure will occurwith probability ity: When a flood o""u,. with P(F) : 0'1 x 0'4 : 0'04 : 0".4, and true joint probability is P(F') x P(P I F) : P(F) + P(P) P(F a P). The proUaUitityof the union is then P(F U P) P(F ) P) : 0.1 + 0.2 - 0.04 : 0.26' Note the contrast: P(F U P) : 0'30 for mutually exclusiveevents events P(r U P) : 0'28 for joint but independent : 0.26 otherwise P(F u P) The new, more generalrule for the union of probabilitiesis s. P(& u E,) : P(81) + P(E,) P(h . E2) and a sixth rule shouldbe addedfor conditional probabilities: 6_ P ( E , , l E P(Er- w "' r ) : n E2)

P(FaF):o.gxo.8:0.72

(26.s)

{26.6)

26.2 CONCEPTSOF PROBABILITY

675

be exceeded.Becausethe probability of any single, exact value of a continuous "occur" can also mean the level will be reachedor exceeded. the In variableis 0.0, oncein 10 years'lhus on exceeded the average or long run, the levelwould be reached return period* Z in yearsis definedas the average T : 1 P(Ft: (26.7)

and the following generalprobability relation.hold: in 1. The probability that F will be equalledor exceeded any year

P@):+
in 2. The probability that F will not be exceeded any year

(26.8)

P(F):l-P(F):l-+

(26.e)

in 3. The probability that F will not be equalledor exceeded any of n successlveyears / l\n

: p,(F)x &(Fl x . . . x P,(F) P(F)': ( t - ;l \ T l

(26.10)

at 4. The probability R, called risk, that F will be equalledor exceeded least years in n successive once

R : 1 - ( r - + l : 1- I P ( F ) F
WITH VARIOUSDEGREES TABLE 26,1 RETURNPERIODSASSOCIATED DESIGNLIFE OF RISKAND EXPECTED Expecteddesignlife (Years) Risk

(26.rr)

with variouslevelsof risk. Table 26.1showsreturn periods associated

e/.)
75 50 40 30 25 20 15 10 5 2
I

10 4.O2 2.00 7.74 3.43 10.3 4.44 14.5 6.12 7.46 " 17.9 22.9 9.47 31.3 12.8 48I . 19.5 98.0 39.5 248 99.5 498 198.4 6.69 t4.9 20.1 28.5 35.3 45.3 62.0 95.4 195.s 496 996

15

100 14.9 29.4 39.7 56.5 70.0 90.1 t23.6 190.3 390 990 1992 18.0 36.6 49.5 70.6 87.4 1r2.5 154.3 238 488 t238 2488 35.6 72.6 98.4 140.7 174.3 224.6 308 475 976 2475 : 4975 72.7 144.8 196.3 28r 348 449 616 950 1949 4950 9953

11.0 22,t 29.9 42.6 52.6 6'7.7 90.8 r42.9 292.9 743 1492

to * The terms return period arldrecurrenceinterval are usedinterchangeably denotethe reciprocal of-the.annualprobability of exceedence.

676

CHAPTER PRoBABILITANDSTATISTICS 26

EXAMPLE 26.1 If Zis the recurence interval for a flood with magnitudeQ*findtbe probability (risk) that the peak flow rate will equalor exceedQ" atleastoncein two consecutive years. Assumethe eventsare i4dependent. Solution. The solutionis easilyobtainedby substitution into Eq. 26.11.To assistin understapding equations,an alternativederivation follows. the The four possibleoutcomesfor the tyo years are: a.' nonexceedance both years in b.' exceedance the first year only in c.' exceedance the secondyear only in d: exceedance both years in Because thesefour representall possibleoutcomes, probability of the union the o f a , b , c , a n d d i s1 . 0 , o r f r o m E q . 2 6 . 2 , P ( a U U c U d ) : l . 0 . E x c e e d a n c e i n a t b least one year is satisfiedby b, c, or d, but not a. Thus the risk of at least one exceedance P(b U c U d), which is the total less the probability of a. From is Eqs.26.2and26.3,we find that ' z-yearris: P(b U c U d):1k P(a) From Eq. 26.3, we find that P(a) : P(Q < Q"inYear I) x P(Q I Q"inYear 2)

:(t-l-)ft-1-) r/ \ r/\
and

R i s k :r - P ( a ) - 1 - ( t - 1 - ) '
\

.r./

rr

EXAMPLE 26.2 What return period must a highway engineeruse in designinga critical underpass drain to accept only a 10 percent risk that flooding will occur in the next 5 years? Solution R : 1 - (' -;)" .10: 1-

Z : 48.1years

('-i)'
IT

26.3 PROBABILITY DISTRIBUTIONS

Randomvariables, either discreteor continuous, characteizedby the distribution are of probabilitiesattached the specificvaluesthat the variablemay assume. random to A variablethroughoutits rangeof occurrence generallydesignated a capital letter, is by and a specificvalueor outcomeof the randomprocessis designated a small letter. by

26.3 PROBABILIry DISTRIBUTIONS

677

^
b 0.2 5 ^ o.l

= P(0) 0.0s P(1= s.15 ) P(2)= 0.2s P(3)= g.2s

= P(4) s.15 = P(s) s.1s

P(6) = 9.63 P(7) = s.s2

Number of cloudY daYsPer week, x

per of distribution cloudydays week' Figure26.L Probability For example P(X : x,) is the probability that random variableX takeson the value , x,. A shoiter version is p(x,). Figur" 26.1 showsthe probability distribution of the the number of cloudy daysin a weet. ft is a discretedistributionbecause number of were taken, a day had which the relative frequencies daysis exact;in ihe rlcord from to-be describedas cloudy or not. Observethat eachof the seveneventshas a finite probability and the sum is 1; that is,

) r(-t,) = t
Another important property of random variablesis the cumulativedistribution in X is lessthan or equal function, CDF, definedur ttr" ptotubility that any outcome to a stated,limiting value x. The cumulative diitribution function is denotedF(;r). Thus F(x):P(X=x)

(26.12)

monotonicallyfrom a lower limit of zeroto an upperbound andthe function increases Figure 26.2 isthe CDF of the numberof cloudy daysin a week derivedfrom of unity. fig. Z6JUy taking cumulativeprobabilities.The function showsthat the probability there thai the numberof cloudy daysin the week will be 5 or less.Conversely, is gOqo This complementary is a 10 percentprobability that it will be cloudy for 6 or ! days. called G(x), where3 prolability is sometimes cumula-tive Q6'13) r(x) = P(X> x) G(x):1continuous variables present a slightly different picture. Figure 26.3 is the were grouped The r""oid of annualstreamflows. observations histogramof an 85-year from 0 to 900 cfs and the number falling in eachinterval into nine intervalsranging plot the was plotted as frequeniy on the left ordinate.A convenientalternativeis to record by the right ordinate.The cDF for the streamflow relative frequencyas shown the increase, continuousdistrinumberof observations is shownin Fig. 26.4.As the by bution will be developed reducingthe sizeof the intervals.In the limit, the broken curvesof Figs. 26.3 and'26.4 wilt appearas thosein Fig' 26'5' There ls a difference between the ordinates of Figs. 26'3 and 26.5a. Since the to with probability, it is convenient reconstitute relative frequencyis synonymous

.7'
CHAPTER 26 PROBABILITYAND STATISTICS

vl

3 o.a ^ -' .F o.o

II

P(x)

.l

0.4

U o.2

0 1 2 3 4 5 6 7

0.05 0.15 0.25 0.20 0.15 0.10 0.08 0.02

0.05 0.20 0.45 0.65 0.80 0.90 0.98 1.00

0 Figure 26.2

1 2 3 4 5 6 Numberof cloudy daysper week,.r

Cumulative distribution of cloudy days per week.

RIR

b
il lt a q

I o '5

o.top

3 4 5 6 7 Mean annualflow, -r (100 cfs)

Figure 26.3

Frequency distribution

of mean annual flows.

26.9 PROBABILITYDISTRIBUTIONS 679

'

Q . o' "r Y
o q

0.6

o I o > 04

(J

0.0 0

3 4 5 6 7 flow, x (100cfs) Meanannual

Figure 26.4 Cumulative frequency distribution of mean annual flows.

probabilitydisFigure 26.5, Csntinuous (a) function tributions: probabilitydensity function. distribution and(b) cumulative

680

CHAPTER26

PROBABILITY AND STATISTICS

histograrfrso that the area in each interval representsprobability; the total area containedis thus unity. To do this, the ordinatein eachinterval, sayn/N for relative frequencyor probability, is divided by the interval width, Ax.-The ratio nfN Lx rs the literally the probability per unit length in the interval and thereforerepresents on the densityof probability.The probability n/N inthe interval is fepresented average as CDF (beforethelimitingprocess) AF(x), or F(x + Lxlz) - F(x - L'xl2).We then can define
l\x) :

.. --;liln Ax-o Ax

AF(x)

dF(x)
dx

(26.r4)

which is called the probability densityfunction, PDF.3This function is the density (or intensity) of probability at any point; f(x) dx is describedas the differential probability. For continuous variables, f(r) > 0, sincenegativeprobabilitieshaveno meaning. Also, the function has the property that f(x) dx: 1
()61\\

which again is the requirement that the probabilities of all outcomessum to 1. Furthermore,the probability that x will fall betweenthe limits a and b is written

p(a-x-b):lu,r*,o*

(26.16)

Note that the probability that x takeson a particular value,saya, is zero;that is,
fo

ft-l dx : o

(26.17)

that nnite proUalilities are definedonly as areasunder the PDF which emphasizes betweenflnite limits. The CDF can now be deflnedin terms of the PDF as P(-* < X < x) : P(X= x) : F(x) : I f(u) du

t'

(26.18)

J_*

whereu is usedas a dummy variableto avoidconfusionwith the limit of integration. The area under the CDF has no meaning,only the ordinates,or the difference in to For example, P(*r3 X = x2),which is equivalent Eq. 26.16, can be ordinates. evaluated as.F(x) - F(xt). in that For discrete.distributions cannotbe summarized integral form, there are to corresponding the propertiesgiven in Eqs. 26.15, analogous arithmetic statements datetaken from a contin26.16, and26.18.In particular, the distributionof sampled uousdistributionis a specialcaseof discretedistributionsand canbe givenin the form Thus of arithmeticsummations.s

),f(',):
P(a-X<b):

t
x1= b

(26.re)
(26.20)

).f(r)
r 7 -

26.4 MOMENTS DISTRIBUTIONS OF

k

681

P(x<d:2fQ,)

(26.2r)

in For a finite numberof observations the sample,/(x) is the probability of xr for each outcomein the samplespaceand thereforeP(x,) : P(xr) : P(xt) : . . . : I /N. Hence/(x) can be replaced with P(x,) in Eqs.26.19,26.20,and26.21. 26.3 DXAMPLE Table B.1 containsthe areabeneatha "standardnormal" bell-shapedPDF.Because the distributionis symmetrical,areasare providedonly on one sideof the center.Use the distribution to determinethe valuesof 1 . P ( 0< z ' 2 ) . 2.P(-2=z=2). 3. P(z > 2\. 4. P(z< -1) Solution 1. P ( 0 = z < 2 ) : . 4 7 7 2 . ) F r o m s y m m e t r y ,P ( - 2 - z < 0 ) : Since P ( 03 2 3 2 ) : . 4 7 7 2 . P(-Z = z < 2) : P(-2 3 z 30) + P(0 3 z 3 2), thenP(=2 = 7 < 2) : .4772+ .4772: .9544. the 3. This is the areaunderthe curvein the right tail beyondz : 2.0. Because area right of center (z : 0) is .5000, P(z = 2) : P(z > 0) - P(0 = z = 2 ) ,o r P ( z > 2 ) : . 5 0 0 0 - . 4 7 7 2 : . 0 2 2 8 . 3 4 . F r o m t h e o l u t i o n t(o ) ,P ( z = - 1 ) : P ( z = 0 ) - P ( - l = z < 0 ) ' B y s = z s y m m e t r y , P ( - 1 z = 0 ) - P ( 0 3 z s ! ) : . 3 4 1 3 , a n d P (= - 1 ) : .5000- .3413: .1587. rr

16.4MOMENTS DISTRIBUTIONS OF
The propertiesof many random variablescan be definedin terms of the momentsof the distribution. The moments representparametersthat usually have physical or geometripalsignificance.Readersshould recognizethe analogybetweenstatistical momentsand the momentsof'areasstudiedin solid mechanics. The rth moment of a distribution about the oriein is definedaso

tri : p::

x'f(x) dx

(26.22) (26.23)

2*,rr,r:12,,,

The first moment aboutthe origin is the mean,or as it is commonly known, the the average. determines distancefrom the origin to the centroid ofthe distribution It frequencyfunction. The prime is normally usedto signify momentstaken about the origin, but the mean is often written as p insteadof y'.

682

CHAPTER26

AND STATISTICS PROBABILITY

Moments can be definedabout axesother than the origin; the axis usedextensively in deflninghigher momentsis lhe mean or, as given above,the first moment about the origin. Thus

p,:
or tr,:

I
I

(x - rd'f\xl dx
- p)'

(26.24\
(26.25)

1<
;)-G,

Wheneverp', or p', are defined for r =l 1, . , the distribution/(x) is comto pletd defined.It seldomis necessary computemore than the first three rnoments; ieveral important distributionsrequire only two. The momentsare usedto specifythe of parametersand descriptivecharacteristics distributions that follow in the next distributionsare describedby combinaof section.Becausevarious characteristics tions of the momentsaboutthe meanand origin, the following relationsare occasionally helpfull'a: Ft: 0 -p' Pz: FL
ttt : tJ"'z

- 3plr, * 2p'

(26.26) (26.27) (26.28)

CHARACTERISTICS 26.5 DISTRIBUTION

' , of by of Characteristics statisticaldistributionsare described the parameters probabilin terms of the moments.The principal ity functions, which in turn are expressed orprobability about the are characteristics centraltendency, groupingofobservations and skewof the variate or observations; a central value;variability, the dispersion The theoreticalfunctions shownin of ness,thedegree asymmetryof the distribution. Fig.26.6 exhibit approximatelythe samegrouping about a central value,but/2 has a much greatervariability thanfl, andf possesses pronouncedright-skewwhilefi is symmetrical.

Symbol Gonvention
of of In introducingthe parameters distributions,the usual sequence statisticalprobare lemswill be followed-that is, parameters derivedfrom the distributionof sample of of data and usedas estimates the parameters the population distribution' Summa-

Figure 26.6 Symmetrical and skewed probability distribution for continuousvariables.

CHARACTERISTICS 26.5 DISTRIBUTION

683

For example,the tion forms of integralsare usedto computemomentsfor samples. of it is designatedTand is usedasthe bestestimate the population meanof sampledata Greek letters are usedto denotepopulation parameters. mean.By convention,

CentralTendency
The familiar arithmetic average,the mean, is the most used measureof central tendency.It is the first moment about the origin and is designated

r:lf*,
n i=t

(26.2e)

The statisticx is the best estimateof the population mean p. 'Means other than the arithmetic mean-for example, the geometric mean : (x624 . . . x,)1/" harmonic Two addimeani : ,l> (Ilx,)-are alsoused. or T are the median,which is the middle value of the of tional measures central tendency observeddata and divides the distribution into equal areas,and the mode, which in discretevariablesis the value occurring most frequently and in continuousvariables is the peak value of probability density.A11three are illustrated in Fig. 26.6.

Variability
deviation by Dispersioncan be represented the total rangeof valuesor by the average of aboutthe mean;however,the parameter statisticalimportanceis the meansquared deviation as measuredby the secondmoment about the mean. The parameteris by termed the varianceand is designated

c r : rn f

i=l

G , -p ) '

(26.30)

to is But the population mean /-{, not known precisely and therefore it is necessary computeinstead
n

s-:

Z\xt-xf

(26.3r)

- 1 in place of n in As the best estimateof o2, the quantity s2 is found using n of the lossof a degree freedom involves for Eq. 26.30.The reasoning this substitution scopeof this text. by using 7 insteadof p, but a proof is beyondthe The squareroot of the varianceis a statisticknown as the standard deviation in (o or s), in which form variability is measured the sameunits asthe variateand the to mean,and henceis easier interpretand manipulate.The coefficientof variation C", useful in comparingrelative variability. definedas cf p, or sfi, is an expression

Skewness
A fully symmetricaldistribution would exhibit the property that all odd moments weightto either side would haveexcessive distribution,however, equalLero.A skewed of the centerand the odd momentswould exist. The third moment a is

o : ! n2 @ , - p ) '
i:l

(26.32)

684

CHAPTER26

AND STATISTICS PROBABILITY

The best-estimate the third moment is computedby of

u -

(n-r)(n-2)

Z-J

\xi-

x)'

(26.33)

The cofficient of skewness the ratio afc3 and is estimatedby is

n ---; (-- -- Q
J

(26.34)

For syfnmetricaldistributions,the third moment is zero and C" : 0; for right skewness(i.e.,thelongtailtotherightside)C">0,andforleftskewnessC"<0.ThePDF is forfi shownin Fig. 26.6has a right or positive skew.The property of skewness of questionable from lessthan 50 sampledata statisticalvaluewhenit mustbe estimatdd points. EXAMPLE 26.4 Determinethe distributionparameters and comparethe distributionsof annualrainfall for the records shown inTable 26.2.

TABLE 26,2 ANNUAL RAINFALLFOR SELECTEDCITIES Annual rainfall(in.)

Year

Anniston, AL 48 49 )) 98 43 53 56 47 69 57 61 64 99 54 40. 47 58 44 44 64 44 51

LosAngeles, CA
9
lo

Richinond, VA.

t928 r927 t926 t925 r924 1923 1922 t92l 1920 1919 1918 t9t7 1916 1915 t914 t913 t9t2 191 I 1910 1909 1908 1907 1906

t9 9 8 6 15 20 11 o 18 8 23 t7 23 17 10 18 5
)4

43 44 38 31 47 49 52 31 51 40 4l 43
3 t

36
J+

38 36 37 43 34
JJ

l9 l5 21

49 47

FUNCTIONS DISTRIBUTION PROBABILITY 26.7 CONTINUOUS

685

Solution
Parameter
Mean, ;r Standarddeviation,s of Coefficient variation,C, : s/V C": a/s3 Coefficient of skewness,

Anniston
57.2 in. 15.5in. 0.27 1.69

LosAngeles
14.9 in. 5.9 in. 0.40 -0.16

Richmond

41.5 in. 6.7in. 0.16 0.16

Comments.(1) Anniston's record shows a high annual averageand a fair$ large right skew,caused variability. In particular,Anniston'sdistributionhas a pronounced valuesin this short period of record. (2) Los principally by two very large observed but Angeleshasa small annualaverage a very largevariability and a slightly negative (3) Richmondhasthe mostuniform distribution:a relativelysmall variabilskewness. ity and only a slight positive skewness. I I

FUNCTIONS DISTRIBUTION 26.6 TYPESOF PROBABILITY

hydroMany standardtheoreticalprobability distributionshavebeenusedto describe that any theoretical distribution is not an processes. should be emphasized It logic of exactrepresentation the natural processbut only a descriptionthat approximates phenomenon and has proved usefui in describingthe observeddata. the under$ing giving the PDF,mean,andvariance summarizes commondistributions, Table26.3 {he functions. The distributions presentedin the table have experiencedwide of the In in applicationand are derivedand discussed many standardtextbookson statistics. of material to follow, only aspects the most usbd distributionsare given. the discreteprobability distributionsinTabIe26.3 The usesof binomial and Poisson generallyto thoserandomeventsin which the outcomecanbe described arerestricted and trials are independent or either as a success failure. Furthermore,the successive probability of success remains constant from trial to trial.3'a In a sense,the the common discretedistributionsare counting or enumeratingtechniques' The binomial distributionis frequentlyusedto approximateother distributions, and vice versa.For example,with discretevalues,when n is large andp small (such the thal np ( 5 preferably),the binomial approaches Poissondistribution.This is a distribution (i : np) and is very useful in describringarrivals in single-parameter queueingtheory. When p approaches and n grows large, the binomial becomes I in from the normal distribution described the next section. indistinguishable

FUNCTIONS DISTRIBUTION PROBABILIry 26.7 CONTINUOUS

and to Most hydrologicvariablesare assumed be continuousrandom processes, the as distributionsare usedto fit historical sequences, in frequency continuous common for analysis, example(Chapter 27). Other applicationsare also important for contindistributions.The elementaryuniform distribution is the basis for computing uous

7
686 CHAPTER 26 PROBABILITYAND STATISTICS

^ ,-" ,
i l {
N H

+
'\-..7
a

,
C)N

n d l
d

, L t ,
I

.gb
> o
(UN

g b

G
,

B l
I

q ^

r
d

i
t v

B -'i.l
co. bl tl
H

.i l \ F t.x

+
x -:

Nl'\a T

,
tl
h

x el

a f f
tx o o

r-

dlJa , +' f

N I N t i - l xrnl -il

+ x J CO+

sr

\, {

B
l
a

+
l s l
-

-{.

o J o
I

Ir o z o ut a
U)

tr

(!

s \/ " x
rir

: " vr
x

" ' \/t :'

\/r "'

o x

8 vr '' R
\,r vt

8 vl ^
Vl

-\? .

^r
k

o ,? ,9,
vl x "' ' ' \/l

;
/\r
.,

M
d

u ;
6 ^

-r

l\t

IL

tr l
dl r

o o
o o

fr

a o
z o
E U'

o o
lt

o
E E

0)

o tu J o

(!

s
F

IL

a? (o
6|

UJ @

EE E ;F
=E . . '

E E af 6 .

= i ' 5E

E "

E !3 E

EiE 6

F[ E S EZ B i ' oO e F E

FUNCTIONS PROBABILIW DISTRIBUTION 26.7 CONTINUOUS

687

The wholebody of materialin the randomnumberssoimportant in simulationstudies. area of reliability and estimatingdependson derived distributionslike Student'sl, that follow concernthe more chi-squared,and the F distribution. The explanations The common distributionsappliedin fitting hydrologicsequences. readeris referred to standardtexts for more detailedtreatment.3-6

NormalDistribution
frequencyfunction, alsoknown be,ll-shaped The normal distributionis a symmetrical, many processes distributionor the natural law of errors. It describes as the Gaussian variations.The whole basisfor a large that are subjectto random and independent body of statisticsinvolving testing and quality control is the normal distribution. of Although it often does not perfectly fit sequences hydrologic data, it has wide for example, in dealing with transformed data that do follow the application, normal distributionand in estimatingsamplereliability by virture of the central limit theorem. The normal distribution has two parameters,the mean p, and the standard By deviationa, for which 7 and s, derivedfrom sampledata, aresubstituted. a simple function only. the distribution can be written as a single-parameter transformation, Definingz : \x - t-i/o, dx : o dz,the PDF becomes

(26.3s)
and the CDF becomes F\z) :

-r*l-

e-"2/2du

(26.36)

The variablez is called the standardnormal variate; it is normally distributed with zero mean and unit standard deviation. Tables of areas under the standard normal curve, as given in Appendix B, TableB.1, serveall normal distributionsafter Given a cumulativeprobability,the deviatez is found of standardization the variables. in the table of areasand x is found from the inversetransform: x:p+za EXAMPLE 26.5 Assumethat the Richmond,Virginia, annualrainfall in Table 26.2 follows a noimal distribution.Usethe standardnormal transformationto find the rain depththat would havea recurrqnceinterval of 100 years. Solution. From example26.4,the meanis 41.5 in. and the standarddeviation is 6.7 in. This gives x: 4L5 + z(6.7) or x:7lzs

(26.37)

Equation 26.18 showsthat the areaunder the PDF to the right of z is the exceedence probabilprobability of the event.For the 100-yrevent ,F;q.26.7 givesthe exceedence Table 8.1 in From the figure accompanying tlI00:0.01. ity Pk): llT,:

688

CHAPTER26

PROBABILIry AND STATISTICS

Appendix B, F(z) : 0.5 - p(z) : 0.49, and z : 2.326 by interpolating the table' The expected100-yr rain depth is therefore x : 4t.5 + O.32O X 6.7 :57.1 in: The 100-yr event for a normal distribution is 2.326 standarddeviationsabove the mean. rt

Log-NormalDistribution
partly dueto the influence Many hydrologicvariablesexhibit a markedright skewness, phenomena havingvaluesgreaterthan zero, or someother lower limit, and of natural will frequencies theoreiically,in the upper range.In suchcases, beingunconstrained, normal distribution,but their logarithmsmay follow a normal distribunot follow the y : tion.7The PDF shownin Table 26.3 for the log-normal comesfrom substituting ln x in the normal. With p, andcy as the mean and standarddeviation,respectively, of the following relations have been found to hold betweenthe characteristics the variatex and the transformedvariatey:r'7 untransformid

p:exp(p'y+412) oz:p,2lexp(d)_11 a : lexp(3fi) - 3 exp(fi) + 2lC3 C,:lexp(dr) - r1t'' C,:3C" * Cl

(26.38) (26.3e) (26.40) (26.4r) (26.42)

Also p, : lfl M, whereM is the median value and the geometricmean of the x's. opensthe extenthe,transformation usefulbecause The log-normal is especially oI theoreticaland applied usesof the normal distribution. Since both the sive body only to comdistributions,it is necessary normal and log-normal are two-parameter variatex and solveEqs. 26.38 and varianceof the untransformed mean-and pute the or Information on three-parameter truncatedlog-normal dis26.39 simultaneously. can be found in the literature.r'7 tributions

lll) TYPe Gamma(and Pearson

The gammadistributionhaswide applicationin mathematicalstatisticsandhas been usedincreasinglyin hydrologicstudiesnow that computingfacilities make_iteasyto evaluate the gimma functioi insteadof relying on the painstakingmethod of using tablesof the incompletegammafunction that lead to the CDF, P(X < x). In greater useis a specialcaseof gamma: tbePearsonType/1L This distributionhasbeenwidely adoptedas the standardmethodfor flood frequencyanalysisin a form known as the log-pearson /11 in which the transform y :1og x is used to reduce skewness.8-r0 Aithough all three momentsare requiredto fit the distribution,it is extremelyflexible in that a zeroskewwill reducethe log-PearsonIII distribution to a log-normal and the pearsonType III to a normal. Tablesof the cumulativefunction are availableand variates A will be explainedin a later section.lo'11 very important property of gamm-a normals)is that the sumof two such aswell asnormal variates(includingtransformed variablesretains the samedistribution. This feature is important in generatingsyn-thefie hy-drologic sequences.l''''

26J

DISTRIBUTION FUNCTIONS PROBABILITY CONTINUOUS

689

Distribution Gumbel'sExtremal
The theory of extreme valuesconsidersthe distribution of -the largest(qr smallest) The distribution of the nt observations occurringin eachgroup of repeatedsamples. extreme values taken from n, samples,with each samplehaving n2 observations, Gumbel was the first to dependson the distribution of the nrn, total observations. frequencies.la Chow has demonemploy extremevalue theory for analysisof flood log-normal with constantskewstratedthat the Gumbel distribution is essentiallya ness.ts The CDF of the densityfunction given in Table 26.3 is

P(X - x) : F(x) : exp{-expl-o(, - u)l}

(26.43)

a a convenient form to evaluate function. Parameters andu are given asfunctions the of the meanand standarddeviationin Table26.3.Tablesof the doubleexponentialare usually in terms of the reducedvariate,y - a(x - u).tuGumbel also has proposed (minimumannual) to anotherextremevaluedistributionthat appears fit instantaneous droughtflows.17'18

CDFsin Hydrology
Normal and Pearson distributionscan often be usedto describehydrologicvariables if the variableis the sum or mean of severalother random variables.The sum of a random variablesis approximatelynormally distributed.For numberof independent example, annualrainfall is the sumof the daily rain totals,eachof which is viewed the as a random variable. Other examplesinclude annual lake evaporation, annual pumpagefrom a well, annual flow in a stream,and mean monthly temperature. usedin approximatingthe distribuThe log-normal CDF hasbeen successfully tion of variablesthat are the product of powersof many other randomvariables.The the normally distributedbecause logarithm logarithm of the variableis approximately of productsis a sum of transformedvariables. Examples variablesthat havebeenknown to follow a log-normal distribution of include:

1. Annual seriesof peak flow rates. 2. Daily precipitationdepthsand stremflowvolumes(alsomonthly, seasonal,

and annual). rates. 3. Daily peak discharge 4. Annual precipitation and runoff (primarily in the westernUnited States). Earthquakemagnitudgs, 6. Intervalsbetweenearthquakes. 7. Yield stressin steel. 8. Sediment sizes in streamswhere fracturing and breakageof larger into smaller sizesis involved.
5.

The PearsonType III (a form of gamma) has been applied to a number of variablessuch as precipitation depths in the easternUnited Statesand cumulative watershed runoff at any point in time during a given storm event.The transformed the Type III is most usedto approximate CDF for annualflood peaks.If log-Pearson the skew coefficient C" of the variable is zero, the CDF reverts to a log-normal.

690

AND 26 CHAPTER PROBABILIry STATISTICS of usedwith monthly precipitationdepthandyield strengths concrete It hasalso.been members. A useful CDF for values of annual extremeis the Gurnbel or extreme value probability of distribution.The meanof the distributionhas a theoreticalexceedance interval T of 2.33 years.Flood peaksin natural streamshave 0.43 and a recurrence exhibited strong conformanceto this disffibution, including means with 2.33-year recurrenceintervals. Graph paper that producesa ptraight-linefit for Gumbel variables is a availableand useful for graphical tests of annual extremes.A sampleis rates,peak shownin Fig. 27.2. The CDF has been applie! to peak annual discharge magnitudesand intervals, maximum rainfall intensitiesof wind velocities,drought events. given durations,and other hydrologicextremesthat are independent

AND REGRESSION CORRELATION LINEAR 26.8 BIVARIATE

proceduresare widely used in hydrology and other sciCorrelation and regression The premiseof the methodsis that one variableis often conditionedby the ences.te value of another,or of severalothers,or the distribution of one may be conditioned by the value of another. Just as there are probability density functions (PDFs) for evaluatingthemarginal probability of a variable(seeSection26.2), so also are there PDFsforlhe conditional probabilities (also describedin Section 26.2) of variables. The conceptis illustrated in Fig. 26.7. For two variables,the bivariatedensityfuncfor tion,/(y li,), ptottedin the vertical on the frgure,changes eachvalueof x' The one shownappliesonly to variationsin y whenx : xr. Different distributionsmight occur for other valuesof x. A measureof the degreeof linear correlationbetweentwo variablesx and y is thelinear correlation cofficient, P*,y. Avalue of p',, : 0.0 indicatesa lack of linear

line Pylr regression

with conditional probability Figure 26.7 Bivariate regression function.

AND REGRESSION CORRELATION691 LINEAR 26.8 BIVARIATE correlation andp,,, : + 1.0 meansperfect correlation.The correlationcoefficientis found from
u - , , : -

cov(x, y)
(l*ay

cr.v
aroy

(26.44)

(seeEq. 26.30), and where o, ando, are the variancesof eachvariable,respectively, definedas cov(x, y) is the covariancesharedby the two variables,

y) cov(.r, : c,.,:

The samplecorrelationcoefficient,r : s,.rfs*s* is usedto estimatep',r. The sample covarianceis found from the squareroot of

f _f _Q'-

dv p)(y - p,)f(x,v) dx

(26'4s)

: s?,,

(26.46)

line shown on Fig. 26.7 is derived to passthrough the mean The regression of the distributions,so that for any given value of x, the mean value of y I x values line. The standarderror of the by (read"y givenx ") can be estimated the regression of y I x is depictedby the line drawnthrough the conditional distributionsat estimate a distanceof one standarddeviationfrom the mean.If the conditionaldistributionsat of are all x-values normal,it canbe shownthat the meanvalue,&y1,, the conditional distribution is related to the meansof .r and H or
a & , 1 , : l t ', * ( p 2,( * and the variance is tr,)

t26.47)

4t.
where

: #['"

, @- p)'1

(26.48)

a?: 40 - p')

(26.4e)

which is the variance of the residualsof the regression.Just as the mean of the distribution requiressubstitutionof the given value of x into Eq. 26.47, so also does to thevariance ,Eq.26.48. Whenthe valueof x in Fig. 26.7is setiequal A the standard error of the meanis
O-t=: ae -----7

(26.s0)

VN

betweeny and .r as the Equation 25.47 is linear and expresses linear dependence value of y can be computedfor fixed valuesof x' Also, slrownin Fig. 26J.The mean if the correlationbetweenthem is significant,one can predict the valuesof y with less error than the marginal distribution of y alone.In fact, from Eq. 26.49, the fraction is by of the original varianceexplainedor accounted the regression

o":t-*

(26.5r)

692

CHAPTER26

PROBABILIry AND STATISTICS

line is It can be seenalso from Eq. (26.47) that the slopeof the regression
cy

PA:

tl,yl,

lLy

r - rr^

(26.s2)

then p itself is the slope,where or, ifx andy are standardized, - p,r)/a, }rnt. (26.s3) p = (x - t*)/o, to The bivariatecasecan be expanded coverhigher-order,multivariatedistributions.

EQUATIONS REGRESSION 26.9 FITTING

ing value of x. The line to be fitted is -l Bx; Y,: a

(26.s4)

N 6 I

, ? ; xbo
J . F

Year
A1 /1

Jackson River

Cowpasture River

3[i
e o

43 44 46 47 48 49 50 51 52 r40 120 100 80 Lowest annualflow for 1 day (cfs) Virginia, 1941-1952 River at FallingSprings, Jackson 60
Mean = Standard deviation =

F r
; h

61 92 65 72 82 67 74 t 18 t24 r08 65 8884.7 21.7

58 81 70 63 68 58 74 105 134 108 93 85 83.1 23.2

B U

line: Iz : 4.94 + 0.923X; r : 0.86. Figure 26.8 Cross-correlationof low floWs.Regression

EQUATIONS 693 REGRESSION 26.9 FITTING The best estimates of o and B are sought. Thus to minimize

) 0, - f), : ) ly, (o + px,)1,

(26.ss)

valuesfrom Eq. 26.54,take values and!,ate the estimated wherey, are the observed follows: partial derivativesas

- (q+ n')r} *{> b, tv, (o* B';l'} #{>

- n d- B ) " , : o

(26.s6) (26.s7)

resultin a and two After carrying out the differentiationsand summations, equations B, callednormal equations.

)y,

(26.s8) (26.se)

2*,y,-")xt-F2*?:o
yields SolvingEqs. 26,58 and 26.59 simultaneously e:-' 2 v , F 2 ^i : y - B T n n

(26.60) (26.6r)

P: =7 - (>;y[
Recall the slopeis p(arf o), or as estimatedfrom sampledata B: rt;

(26.62)

equationis variancein the regression Also, the unexplained

4:4Q-p')

(26.63)

and is the squareroot of which is the standarddeviationof residuals(seeFig' 26'8) cailed the standard error of estimate.Thesecan be estimatedfrom

*-: 4 n -l ( r - r , ) s z
s2": 2(y,-il'

(26.64) (26.6s)

wherey, and i, are as definedpreviously(seeEq' 26'55)' ivtanynyarotogicvariablesare linearly related,and after estimatingthe regresrangeof sion coeffici"ntr, p."di"tion of y can be madefor any value of x within the is often performedbut shouldbe x observed values.Extrapolationoutsidethe range

694

CHAPTER26

PROBABILIry AND STATISTICS

done wfth caution.Equation 26.48 showsthat the variancein the estimateof y for a givenx valuebecomes largewhenx is severalstandarddeviationsaboveor belowthe mean. EXAMPLE 26.6 Rivers are and Cowpasture The lowestannualflows for a l2-yr period on the Jackson tabulatedin Fig. 26.8. The stationsare upstramof the confluenceof the two rivers equationand the correlationbetween that form the JamesRiver. Find the regression low flows. Solution are2 x : 1016;) y : 997;2 x2 : 9 1 , 2 1 6 ; 2 : The basicstatistics l" 88,777;and2 xy : 89,209. For the two-variableregressiona and B are found from Eqs.26;60and 26.61,.

_ [(89,20e) (l0t6xee7)/(12)]: Qo? o (9r,216) (tor6)' 102)

o:

ee7 (0.e23x1016)' 9 r : 4 i

is The regression y : 4.91 + 0.923x. 3. The correlation coefficientfrom Eq.26.62 is

(0.e23)(2r.7) :
23.2

0:86

s,, error of estimate, is 11.7,which is plotted 4. From F;q.26.64the standard line regression in Fig. 26.8. rl as limits aroundthe

for Coefficientof Determination the Regression

BeA regression equationreplaces(and extends)the data used in its development. loss of some causeit cannot reproduceall the basedata, the processresults in the information. This not only includeslossof information aboutparticular pairs of data, of but alsoaboutthe variability of the data.The variancesf is a statisticalmeasure the the spread valuesof y. The greaterthe valueof sl, the wider variability of the measured of of points aroundthe mean.The percentage information aboutthe variancein y that equationis called the cofficient of deteris retained,or explainedby, the regrdssion (differences between its mination, Cr. To determine value,the residualsor departures the unacy actual and estimated values)haveknown variance4, which represents equation.The explainedvariance would be the countedvariance in the regression - o2, and the percentageretained (coefficient of determination)is difference, 4

(26.66)

EQUATIONS 26.9 FITTINGREGRESSION

695

Comparisonwith Eq. 26.49 revealsthat

Co= Pz

(26.67)

of ofthe correlationcoefficientp is the percentage d, explained Thus the square regression.For any sample of data the coefficient of determinationr2 is by the equationto the as estimated sl,rlslsl . A large r2 indicatesa goodfit of the regression of for the equationaccounts or is able to explaina large percentage the databecause in the data. variation
EXAMPLE 26.7

in Determinethe coefficientof determinationfor the regression Example26.6. r2, Solution, From Eq. 26.67,the coefficientof determination, is 0.7396. "accountsfor" about 74 explainsor equationadequately Thus, the regression pefcent of the original information abouty containedin the raw data. Twentysix percentof the information is lost. I r For to can be extended multiple linear r'bgressions. examThe bivariateexample variable andx1 andx2 ple, the linear model in three variables,with y the dependent variables,has the form the independent y:a*F$t*Fzxz The normal equationsare ) y : an-r Br) t' * FrZ *, (26.68)

(26.6e) (26.70) (26.7r)

)y"':

")

xtt FrZ*?+ Fr2*,*,

2 yr r : * ) xzI Fr 2 *r *, + F"2 *7
The squareof the standarderror of estimateis

s7:

2(y,-r,)'

(26.72)

where y, are the observedvaluesand y-,are predictedby Eq. (26.68). The multiple correlation coefficient is

n : ( t in LinearTransformations Hydrology

+)
^2\ | /2

s;/

(26.73)

Strongnonlinearbivariateand multivariatecorrelationsare also commonin hydrology, ind various mathematical models have been used to describethe relations. Piiabolic, exponential, hyperbolic, power, and other forms have provided better

696

CHAPTER26

AND STATISTICS PROBABILITY

graphicalfits than straight lines. Becauseof difficulties in the derivation of normal equationsusing least squaresfor thesemodels,many can be transformedto linear forms. The most familiar transformationis a linearizationof mtiltiplicative nonlinear relationsby using logarithms.For example,the equation
Y : g,xft1$z

(26.74) (26.7s)

becomes linear when logarithms are taken, or

log y : log * + Bllo$ x1 -f B"Iog x2

The log transformationprocedureresultsin a linear form when the logarithms in are of one or both setsof measurements substituted Eqs. 26.60 and 26.61.For by example,if a bivariateparabolicform I : qXb is suggested the data, logarithms allow use of the linear form log Y : log a -t b log X. The normal equationscan be usedby redefiningy : log Y,x : logX, e : log a, andF : b,thereby transforming can the equationto y : a * Bx. The regression now be performedon the logarithms, and the estimateof a is found by taking the antilog of valuesof a and B determined, a. This transformationis possiblefor severalother nonlinearmodels,someof which are shown in Table 26.4. The variablesx and y must be nonnegative,with values preferably greaterthan 1.0 to avoid problemswith the log transformation.
OF TABLE 26,4 LINEARTRANSFORMATIONS NONLINEARFORMS

Equation Y=A+BX
Y = BeAx Y:AXB Y:ABx

Abscissa
X log X X

Ordinate
I

Eouationin linearform

log Y log Y log Y

lY): A + B[x] e)[x] tbc rl : locB + A(1og X] tbc rl : bc A + B[toc B)[X] A lloCrl : loC + (loC

Note.'Variables in brackets are the regressionvariates.

EXAMPLE 26.8 the by In the following exhibit (Table 26.5) prepared Beard,20 regionalcorrelationis deviationof flow logarithmswith the logarithmsof the drainage soughtof the standard areasize'andthe numberof rainy daysper year;X, is setequalto ( 1 t log s) to avoid equationand the multiple correlationcoefflcient. values.Find the regression negative Solution are 1. From Eqs.26.69,26.70, and26.7l, the parameters -0.49 a : 1.34; Fr : -0.013; Fz: equationis and the regression Xt : 1.34- 0.0I3X2- 0.49X3 log s : 0.34 - 0.013log(DA) - 0.49 log(days) or 2. The multiple correlation coefficient from Eqs' 26.72 and 26.73 is R : 0.56. ll

APPLICATIONS AND CORRELATION 26.10 REGRESSION DATA FOR 50 GAUGINGSTATIONS TABLE 26.5 LOGARITHMIC Xr:1 + logs Station number (1)
I

697

Xz: logDA

Per & : log numberof rainY-daYs Year

X2 (2)

x3 (3) 2.11 2.12 2.ll 2.04 2.07 2.04 2.09 1.99

x1 (4) 0.29 0.18 o.l7 0.44 0.38 0.3'7 0.30 0.35 0.16 0.11 0.32 0.34 0.25 0.43 0.2'l 0.25 0.52 0.18 0.39 0.40 0.25 0.23 0.54 0.51 0.45 0.63 0.45 0.59 0.46 0.32 0.96 0.12

Station numDer (5) 33 34 35 36 37 38 39 40 4I 42 43 44 46 4'7 48 49 50

x2 (6)
1.94
z.tJ

xs
(7\ 1.87 t.36 1.81 1.58 1.48 1.89

x1 (8) 0.20 0.58 0.64 0.37 o.27 0.54 0.63 0.78 0.46 0.44 0.24 -0.03 0.30 0.17 0.14 0.10 0.27 0.18 17.89 0.358 51.1527 52.7934 -r.640' 7 33.2598 34.4347 -r.r749 8.1635 6.4010 L7625

2 3 4 5
o

7 8
q

1.61 2.89 4.38 3.20 3.92 1.61 3.2r 3.65

3.23 z.rs

t0 11 t2
I J

t4 15 16 17 18 19 20 21 22
LJ

24 25 26 27 28 29 30 3l 32

2.08 4.33 1.60 2.09 2.00 2.82 2.00 2.40 2.09 3.69 2.18 2.19 2.09 2.17 |.91 4.48 1.9s 4.95 2.21. r.97 2.08 3.4r 4.82 l 88 r.93 r.78 L.74 4.39 3.23 2.01 3.58 2.04 1.64 1.78 1.7 6 4.58 3.26 1.93 1.81 4.29 1.23 1.89 1.48 3.44 |.97 2.lt

3.63 1.91 2.26 2.97 0.70 0.30 3.38 2.87 2.42 4.53 3.04 4.13 |.49 5.37 1.36 2.31 r4'7.55 2.951 503.7779 435.4200 68.3579

r.32 1.54
1.62 2.03 2.26 1.93 1.78 2.00 2.Or 1.95 2.tl 2.23 96.24 1.925 285.5627 284.0042 1.5585 r8'7.59r2 185.2428 2.3484

>X x
2 XX, 2 X2 X2/n 2 XX, 2 X2 X,ln

1.5585

2X2X'ln 2 xx1

-1.6407

Note: x = X - X. (AfterBeard.2o)

2 6 ' I o R E GR E S S |o N A N D C o RRELAT|oNAPPL|CAT|oNS

698

26 AND CHAPTER PROBABILIry STATISTICS desiredJtatisticalparameteras dependent variable,and the appropriatephysicaland The climatic variableswithin the basinor regionas the independenlvariables. proceand dures are signiflcantly better than using relatively short historical sequences point-frequency analysis. Not only doesthe methodreducethe inherently large sampling errors but it furnishesa meansto estimateparametersat ungauged locations. There are limitations to the techniques Section 26.9. First, the analyst asof sumesthe form of the model that can expressonly linear, or logarithmically linear, analdependence. Second,the independent variablesto be includedin the regression ysis are selected. variablesare And, third, the theory assumes that the independent indeedindependent are observed determined or without error. Advancedstatistiand cal methodsthat are beyondthe scopeof this text offer meansto overcomesomeof to theselimitations but in practiceit may be impossible satisfy them. Therefore,care must be exercised selectingthe model and in interpretingresults. in Accidental or casualcorrelationmay existbetweenvariablesthat are not funcbetweenhydrotionally correlated.For this reason,correlationshouldbe determined Because the natural of logic variablesonly when a physicalrelation can be presumed. dependence between many factors treated as independentvariables in hydrologic variableand eachof the independent studies,the correlationbetweenthe dependent variableswhen variablesis different from the relative effect of the sameindependent togetherin a multivariatemodel. One way to guard againstthis effect is by analyzed screeningthe variablesinitially by graphical methods.Another is to examine the results of the final regression equationto determinephysicalrelevance. significantvariaid Alternatively,regression techniques themselves in screening a ables. When electroniccomputationis available, procedurecanbe followedin which model, and the successive independent variableS addedto the multiple regression are in relative effect of eachis judged by the increase the multiple correlationcoefficient. it Although statisticaltestscan be employedto judge significance, is useful otherwise to specify that any variable remain in the regressionequation if it contributesor explains,say,1 or 5 percentofthe total variance,or ofR2. A frequentlyusedrqethod is to computethe partial correlation cofficients for each variable.This statistic represents relative decrease the varianceremaining( 1 - R') by the addition of the in the variablein question.If the varianceremainingwith the variableincludedin the regression is (I - Rz) : pz and the variance remaining after removal is (l - R'') : D'', then the partial regression correlation coefficient is Most PC spreadsheet software packageshave statistical routines for all the described here and many more. Most are extremelyflexible,requiringminianalyses mal instructions-and input data other than raw data. Specialmanipulationscan effect variables,bring one variableat a time an interchange dependent ol and independent into the regression equation, rearrange the independent variables in order of significance, and perform various statisticaltests.

\D'' - D')lD''.

Hydrologic Records Extending

frequently can be used to extend short records if significant Regression techniques correlation existsbetweenthe station of short record and a nearby station with a - lorigerreeord.Iq Example26.6,if the Jackson River recordswerecompletefromI94l

PROBLEMS 699 recordswere incompleteafter 1952,the cross-correlation to datebut the Cowpasture equationfor I could be usedto estimatethe missingyearsby solvingthe regression The reliability of suchmethodsdepends from 1953 on usingthe X flows as observed. on the correlationcoefficientand the length ofthe concurrentrecords.Ifthe concurrent record is too short or the correlation weak, the standarderror of the parameter to be estimatedcan be increasedand nothing is gained.The limiting value of crosscorrelation for estimatingmeansis approximatelyp : l/\/ n, where n is the length Thus any correlation above0.3 would improve the Cowof the concurrentrecord.21 pasturerecords.Estirnatesof other parameterswith larger standarderrors require for highercross-correlation significantimprovement.Extendingor filling in deficient for recordsoften is necessary regional studiesin which every record usedshould be adjustedto the samelength.

Hydrologic Variables Regionalized Predicting.

studiedvariousmethodsof regional flood analysisand found the Cruff and Rantzz2 techniquea better predictor than either the index-floodmethod multiple regression (Chapter27) or the fitting of theoreticalfrequencydistributionsto individual historitechniques extendall recordsto a common to cal records.They flrst usedregression base length. Next they extrapolatedby various methods to estimate the 50- and 100-yearflood events and with multiple correlation examined several dependent variables including the drainage areaA, the basin-shapefactor (the ratio of the parallel to the main diameterof a circle of sizeA to the length of the basin measured channel)Sa,channelslopeS, the annualprecipitationP, and others.They found only A and S, to be significant, which resulted in prediction equations of the form The Q, : cAS!,. These equationswere superior to those of the other techniques. multiple correlation coefficientwas as high as 0.954.It is interestingthat regression regionalvaluesof the techniques wereemployedin still a third way,that is, to estimate mean and standarddeviation after adjustingthe record length. Example26.8 illusanalysisto regionalizethe standarddeviationof trated the applicationof regression annual maximum flow logarithms as a function of the drainage area size and the number of rainy dayseachyear.

'

r summary
Statisticsis a diversesubject,and the treatmentin this chapterhasbeennothing more and practitionersmust return againand again than an introduction. Seriousstudents They will find that evaluatingnew developments to the theory in standardworks.23 of must claim a large shareof their time. Only certain aspects, statisand techniques principally the common distributionsand the tical hydrologyhave been presented, methodsfor analyzingfrequency of eventsobservedat a single point. In the next to chapterthis information is extended common applicationsin hydrology.

PROBLEMS
26.1. The probabilitiesof eventsE1 andE2 arc each.3. What is the probability that E1 or but not mutually exclusive,and E2 will occur when (a) the eventsare independent (b) whenthe probabilityof Et, given E2is .l? -

700

CHAPTER26

AND PROBABILITY STATISTICS

26.2. EventsA and B are independenteventshaving marginal probabilities of.4 and .5,
respectively.Determine for a single trial (a) the probability that both A and B will and (b) the probability that neither occurs. occur simultaneously, 26.3. The conditional probability, P(E, I E,r),of a power failure (given that a flood occurs) is .9, and the conditionalprobability,P(Ez I E), of a flood (given that a powerfailure occurs)is .2. If the joint probability, P (\ andE), of a power failure and a flood is .1, determinethe marginal probabilities,P(E) and P(E). (c) 26.4. Describetwo random eventsthat are (a) mutually exclusive,(b) dependent, both nor and (d)"neithermutually exclusive dependent. mutually exclusiveand dependent, cofferdamis to be built to protect the 5-yearconstructionactivity for a 26.5. A temporar;1 to major crossvalley dam. If the cofferdam is designed withstand the 20-yearflood, what is the probability that the structurewill be overtopped(a) in the flrst year, (b) in the third year exactly,(c) at leastonce in the 5-yearconstrucfionperiod, and (d) not at all during the 5-yearperiod? The to 26.6. A 33-yearrecord of peak annualflow rateswas subjected a frequencyanalysis. median value is defined as the midvalue in the table of rank-ordered magnitudes. Estimatethe following probabilities. the , a. The probability that the annualpeak will equalor exceed medianvaluein any singleyear. retlrrn period of the median value. b. The average c. The probability that the annual peak in 1993 will equal or exceedthe median value. d. The probability that the peak flow rate next year will be less than the median value. yearswill e. The probability that the peak flow rate in all of the next 10 successive value. be lessthan the median f. The probability that the peak flow rate will equal or exceedthe median value at years. least oncein l0 successive g. The probability that the peak flow rates in both of two consecutiveyears will equal or exceedthe median value. h. The probability that, for a2-yearperiod, the peak flow rate will equal or exceed the median value in the secondyear but not in the first' 26.7. What return period must an engineeruse in his or her designof a bridge openingif there is to be only a 50 percent risk that flooding will occur at least once in two years?Repeatfor a risk of 100 percent. successive 26.8. A temporary flood wall has been constructedto protect several homes in the floodplain. The wall was built to withstand any dischargeup to the 20-year flood magnitude.The,wall will be removed at the end of the 3-year period after all the homeshavebeen relocated.Determinethe probabilities of the following events: a. The wall will be overtoppedin any year. b. The wall will not be overtoppedduring the relocation operation. c. The wall will be overtoppedat leastonce before all the homesare relocated. d. The wall will be overtoppedexactly once before all the homesare relocated. for e. The wall will be adequate the flrst 2 years and then overtoppedin the third year. return periods(shownon the next page)are known 26.9. Waveheightsand their respective campsitewill accepta 25 percent Ownersof a downstream for a 40-mi long reservoir. risk that a proiective wall will be overtoppedby wavesat least once in a 2}-yeat period. Determinethe minimum height of the protective wall.

PROBLEMS 701 Waveheight (ft) 10.0 8.5 7.4 5.0 3.5 period Return (years) 100 50 30 10 5

26.10. Assumethat the channel capacityof 12,000cfs near a private home was equaledor exceeded 3 of the past 60 years.Find the following: in a. The frequencyof the 12,000-cfsvalue. b. The probability that the home will be floodednext year. c. The return period of the 12,000-cfsvalue. d. The probability that the home will not be floodednext year. e. The probability of two consecutive, safeyears. f. The probability of a flood at leastonce in the next 20 years. g. The probability of a flood in the second, not the first, of two consecutive years. but h. The 20-yearflood risk. 26,11, The distribution of mean annual rainfall at 35 stations in the JamesRiver Basin, Virginia, is given in the following summary:

Interval(2-in. groupings) Numberof observations Interval (2-in. groupings) Numberofobservations

36 or 37 in. Z 44 or 45 in. 5

38 or 39 in. 4 46 or 47 in. 4

40 or 41 in. j 48 or 49 in. 2

42 or 43 in.

50 or 51 in. 2

(seeChapter27) andplot the frequencydistribution Computethe relative frequencies andthe cumulativedistribution.Estimatethe probability that the meanannualrainfall (a) will exceed40 in., (b) will exceed50 in., and (c) will be betweenthesevalues. 26.12. Write a simpleprogram to READ N data points and compute the mean, standard in deviation,and skewness coefficient. 26.13. A normally distributedrandom variablehas a mean of 4.0 and a standarddeviation of 2.0. Determinethe value of

"8

dx I f(x)

f@

26.14. For a standardnormal densitv function. use Table B.1 to determinethe value of '
fr+'o

J*-ro 26.15. A normal variableX has a meanof 5.0 and a standarddeviationof 1.0.Determinethe value of X that has a cumulativeprobability of 0.330. 26.16. If the mode of a PDF is considerablylarger than the median, would the skew most likely be positive or negative?

fG) dx

702

ANDSTATISTICS 26 CHAPTER PROBABILIW about the properties of a PDF by 26.17. tomplete the following mathematicalstatements insertingin the boxeson the left the correct item numberfrom the right. Assumethat X is a seriesof annual occuffencesfrom a normal distributibn. I r.Zerc a. I f(x) dx: J t " b.

: dx f'_ro, r
dx: '34 l-.o 'o'

2.Unity eachyear 3. Valuewith 5 percentchanceofexceedance 4. 0.68 5. Valueexpectedevery 50 r"urc on the average 6. P(X < mr) ' 7 . P ( m 1X ' * r )
8 . P ( m 1- X = m z )

d.

f,o"

dx:r

dx: .5 f-to,
f(x) dx : Z

I rt, dx: .02

l.

9. Median

10. Standarddeviation 26.18, The mean monthly temperaturefor Septemberat a weather station is found to be normally distributed.The mean is 65.5" F, the varianceis 39.3'F2, and the record is completefor 63 years.With the aid of TableB.1, find (a) the midrangewithin which two thirds of all future mean monthly valuesare expectedto fall, (b) the midrange within which 95 percentof all future valuesare expected,(c) the limit below which to and (d) the valuesthat are expected be 80 percentof all future valuesare expected, exceeded with a frequencyof oncein l0 yearsand oncein 100years.Verify the results by plotting the cumulativedistribution on normal probability paper. to basinis determined be approximately 26.L9. The total annualrunoff from a small drainage normal with a mean of 14.0 in. and a varianceof 9.0 in.2.Determinethe probability that the annualrunoff from the basinwill be lessthan I 1.0 in. in all three of the next years. three consecutive of 26.20. In the past60 years,a discharge 30,000cfs at a streamgaugingstationwasequaled return period (years) of this or exceeded only three times. Determine the average value. and havemarginal probabilitiesof .4 and .5, respec26.21. EventsA and B are independent tively. Determine the following for a single trial: a. The probability that both A and B occur. b. The probability that neither occurs. c. The probability that B, but notA, occurs. Existingrecordsrevealthe following information aboutEventsA and 4 whereA = a ?6il,

IongMerch'warmspellandB 1q

:!94!$99!.

PROBLEMS Year A : warm March? B : April flood? No Yes No No Yes No No Yes Yes Yes
No Yes

703
1n

Yes No

No Yes

Yes Yes

No No

On the basisof the 10-yearrecord, answerthe following: Prove. a. Are variablesA and B independent? Prove. b. Are variablesA andB mutually exolusive? c. Determinethe marginal probability of an April flood. d. Determinethe probability of having a cold March next year. e. Determinethe probability (onevalue)of havingboth a cold March and a floodfree April next year. f. If a long March warm spell hasjust endedtoday,what is the best estimateof the probability of a flood in April? 26.23. Two dependent eventsarc A : a flood will occur in Omaha next year and B : an ice-jam will form near Omahain the Missouri River next year. Useyour judgment to rank from largestto smallestthe following probabilities:P(A), P(A andB), P (A or B), P(A I B). 26.24. The probability of having a specifiedreturn period, [, is definedas: or P(annual valuewill be equaled exceeded /, - I \'-' : exactly oncein a periodof r : I years) T,f \' Also, p (annualvalue will be equaledor exceeded _ exactly r times in a period of n years) pn_r(l _ p)r

the in a. According to the descriptions parentheses, secondprobability shouldequal the first when n and r are equal to what values? resultin the sameprobability for an annualvaluewhose b. Showthat both equations years.Discusss. frequencyis 33{ percentand the return period is Z.: /:3 below, find (a) the number b that will make the function 26.25. For the function described of a probability densityfunction, and (b) the probability that a singlemeasurement x will be lessthanl.
-

.l \^)

{i",,'

forx ( 0 for0<x=b forxlb

depth of precipitation in July. Betweenvaluesof 26.26. The random variabler represents -r : 0 and x : 30, the probability densityfunction has the equation/(-r) : x/40p',. In the past, the average July precipitation p,, was 30 in. a. Determine the probability that next July's precipitation will not exceed20 in. b. Determinethe singleprobability that the July precipitatjon will equal or exceed years. 30 in. in all of five consecutive depth of precipitation in July. Betweenvaluesof 26.27. The random variable-r represents x : 0 and .r : 30, the probability densityfunction has the equation/(.x) : x/1200. a. Determine the probability that next July's precipitation will not exceed20 in. 30 b. Determinethe probability that next July's precipitationwill equalor exceed in.

l---

704

AND 26 CHAPTER PROBABILITY STATISTICS 26.28.

26.29.

26.30. 26.31.. 26.32.

26.33. 26.34.

discharge Measured (1000 acre-ft) 12.2 10.4 10.6 1.2.6 14.2 13.0 14.0 t2.0 10.4 tl.4

recharge Estimated (1000acre-ft) t2.o 9.8 I 1.0

t3.z

14.6 14.0 14.0

I z,+

10.4 11.6

26.35. F i t a r e g r e s s i o n e q u a t i o n t o t h e d a t a i n P r o b l e m 2 6 ' 3 4 , t r e a t i l g d i s c h a r g e a s t h e
dependentvariable.computethestandarderrorofestimate.Estimatetheexpeoted if of discharge no when recharg"ir 13 Kti what would be the estimate discharge is the relative improvementprovided what information were availatie on recharge? estimate? by the regression The program linear regression' 26.36. prepare a computerprogramfor simple,two-variable, X' (b) computethe means'variances' should(a) read in N pain oioUt"tuuiions, Y and constants'the (c) find the regression and standarddeviationsoiUott' Y andX, and with the data in coefficient. verify standarderror of estimal-, u"J,ir" correlation Problem26.34. of the mean annual rainfall with the 26.37. From the following observations of variation How iinea, predi.tion equationfor the catchment. altitude of the gauge,d";;;;; well correlatedare rainfall and altitude?

PROBLEMS Gauge number Mean annual rainfall(in.) Altitudeof gauge (1000ft)

A A

705

1 2 3
4

22 28 25 31
JZ J I

4.4
1 < < A

5 6 7 8 9 10 1l 12

36 35 36 46 4l 4I

5.6 5.6 5.8 6.0 6.6 6.6 6.8 7.0

26.38. Estimatethe expectedrainfall in Problem 26.31 for a gaugeto be installed at an

altitude of 5500 ft. equationY : A + estimates A and B in the bivariateregression of 26.39. The least-squares : 3.0,whereYis atransformation as defined logro)andXis BXarcA: 2.0 andB the Ify as 1o916;r. and r are relatedby y : axb, determine valuesof a and b. defined (Z), deflnedasthe time for a streamto rise from 26.40. The time of rise of flood hydrographs low waterto maximum depthfollowing a storm,is relatedto the streamlength (L) and in slope(S). From the information given below for 11 watersheds Texas, the average New Mexico, and Oklahoma, derive a functional relation of the form T,: aLbS'.

Watershed number
I

Ir

(min)

(1000 ft) 18.5 14.2 25.3 tt.7 9.7 8.1 21.'7 3.9 1.2
J.J J.)

(fv10oo ft)
7.93 19.0 t2.a 13.3 11.0 15.0 16.7 146.0 20.0 64.0 33.0

2 3
A

5 6 7 8 9 l0 11

150 90 60 60 100 75 90 30 30 45 50

26.4r, Repeat the exercise in Problem 26'40 by fitting the relation T,: dF", whete
F : L/{S

with t in mi and S in ftimi. Plot the results on log-log paper. 26.42. The squareof the linear correlationcoefficientis calledthe proportion of the variance Describethe meaningof this phraseby evaluatthat is "explainedby the regression." and what doesthe term givenin the text.What varianceis explained, ing the equations "explained" mean? pairs of valuesof normally distributedvariablesX and Y are ana26.43. Twenty measured lyzed, yielding valuesof X : 3O,V : 20, s' = 20, ands, : 0. Determinethe values

706

CHAPTER26

PROBABILITY STATISTICS AND

d a andb and the standarddeviationof residuals a least-squares usingthe linear for fit equationY:a+bX. 26.44. The least-squares estimates A and B in the bivariateregres*sion of equationy : A + BXareA:2.OandB : I.0, whereyis atransformation defined log1eLIf Iand as X are relatedby Y : a(.b)',determine valuesof a and b. the

26.45. Given a table of ten valuesof mean annual floods and corresponding drainageareas
for a numberofdrainagebasins,statehow linear regression techniques would be used to determinethe coefficientand exponent(p and 4) in the equation Qzzz : pAq.

26.46. What choiceof transformedvariablesI and X would provide a linear transformation

for y : a/(x3 + b)? Also, if a regressionon these transformed variables yields I : 100 + 10X determinethe corresponding valuesof a and b. Would the linear transformationbe applicableto all possiblepairs and valuesofx and y? 26.47. Which measure variation in a regression : a * bX is generallylargerin magniof Y tude, the standarddeviation of I or the standarddeviation of residuals? For what condition would the two valuesbe equal?

REFERENCES
1. Ven T. Chow, "Statisticaland ProbabilityAnalysisof HydrologicData," Sec. 8-1, in Handbookof Applied Hydrology (V. T. Chow, ed.). New York: McGraw-Hill, 1964. 2. M. B. Fiering, "Information Analysis," in Water Supply and Waste Water Removal (G. M. Fair, J. C. Geyer,and D. A. Okun, eds.).New York: Wiley, 1966,Chap.4. 3. J. R. Benjamin and C. Cornell, Probability, Statisticsand Decisionfor Civil Engineers. New York: McGraw-Hill. 1969. 4. A. M. Mood and F. A. Graybill, Introduction to the Theory of Statistics, 2nd ed. New York: McGraw-Hill. 1963. 5. A. J. Duncan,Quality Control and Statistics.Homewood, RichardD. Irwin, Inc., IL: 1959. 6. P. G. HoeI, Introduction to MathematicalStatistics,3rd ed. New York: Wiley, 1962. 7. J. Aitchison and J. A. C. Brown, The Log-Normal Distributlon. New York: Cambridge UniversityPress,1957. 8. H. A. Foster,"Theoretical Frequency Curves,"Trans.ASCE 87,142-203(1924). 9. L. R. Beatd, Statistical Methods in Hydrology, Civil Works Investigations, U.S. Army Corps of Engineers,Sacramento District, 1962. 10. "A Uniform Techniquefor DeterminingFlood Flow Frequencies," Bull. No. 178, U.S. Geological Survey,1989. 11. "New Tablesof Percentage Points of the PearsonType III Distribution," Tech. Release No. 38, CentralTechnical of Unit, U.S. Department Agriculture,1968. 12. M. B. Fiering, StreamflowSynthesis. Cambridge,MA: Harvard University Press,1967. 13. F. E. Perkins,SimulationLecture Notes,SummerInstitute, "Applied MathematicalProgramming in WaterResources," University of Nebraska,1970. 14. E. J. Gumbel, "The Return Period of Flood Flows," Ann. Math. Statist. l2(2), L63190(June 1941). 15. Ven T. Chow, "The Log-Probability and Its EngineeringApplication," Proc. ASCE 80, 1-25(Nov. 1954). 16. "Probability Tables and Other Analysis of Extreme Value Data," Series22, National Bureauof Standards Applied Mathematics, 1953. 17. E. J. Gumbel, Statisticsof Extremes.New York: ColumbiaUniversity Press,1958.

Chapter27

Frequency Analysis

r Prologue
The purposeof this chapteris to: . Presentmethodsused in hydrology to evaluatethe recurrenceof particular magnitudes and durationsof random hydrologicvariables. . Elaborateon the definitionsof freqUency, reiurrence interval, return period, and risk analysis introduced Chapter10, Section10.4. in . Illustrate the diverseapplicationsof frequencyanalysisin hydrology. . Teachseveral methodsfor conductingfrequencyanalyses, includingthe useof that allow estimationof recurrenceintervals for variables frequencyfactors that follow conventionalprobability distribution functions. . Introduce methods of point and regional frequency analysis and describe regional USGSregression equationsthat have been 4doptedthroughoutthe U.S. for estimating flood flows for use in structure design and floodplain analysis. . Establishhow to estimatethe reliability of estimatesderived from point or regional frequencyanalyses. Explainthe widelyusedBulletin No. 17BLog-Pearson Type III procedures for performing uniform flood flow frequencyanalyses. Describehow variousfederal agencies apply frequencymethodsin designor analysisof water resources systems. In Chapter 26, probability and statistical characteristics random variables of were introduced,alongwith common distributionfunctions and principlesof regression and correlatiqn.The presentchapterprovidesapplicationsof theseprinciplesto common hydrologicvariables.

27.1 FREQUENCY ANALYSIS

presented Chapter26 areusedmostfrequentlyin describing in The statisticalmethods hydrologicdata suchas rainfall depthsand intensities,peak annual discharge, flood flows, low-flow durations,and the like. Frequencyanalysiswas introduced in Sec-

ANALYSIS 709 FREQUENCY 27.2 GRAPHICAL tion 10.4 and is defined as the inve or recurrencA probabilitiesof magnit wise, the frequencYof a hYdrologic discretevariablewill occur or some' in exceeded anY given Year'The lat freq-uenc probability or exceedance ir"qo"n"Y is a ProbabilitYand has as frequencY, shownbYEc ceedancs Two methods of frequencYa plotting techniqueto obtain the cum iu"to.t. The cumulativedistribution the probability of an eventequal to is used to obtain recurrencerntervz tioned whenworking with recordssl ofexpectedhydrologiceventsgreaterthantwicetherecordlength.

ANALYSIS FREQUENCY 27.2 GRAPHICAL

The frequencYof an event can be When annualmaximum valuesare imated as the meantime in Years'\ The onceon the average' t exceeded be shownto be x: where 7 = N = n : m: tuf,lj
m

(21.r)

the mean number of exceedances the number of future trials -oi the number of values values,with largestequal to I descending therunt t : 1' N : T' and If the mean number of exceedances
4 I _ -

n ) l
m

(21.2)

indicatingthattherecurfenceintervalisequaltothenumberofyearsofrecordplus 1, divided bY the rank of the event' They give different results as Severalpictting p*liion formulas are available'l rntervalsoutainedfor 10 yearsof record .1, notedin Table2'7' .The rangein recurrence plotting position formulasdo not account is illustratedin the right-hani column.Most formula ihat doesaccountfor samplesize for the samplesizeor length of record. One generalform wa, giuen by Gringorten' and has the
.r1 -

r -

n * l - 2 a
m - a

(27.3)

710

CHAPTER2TFREQUENCYANALYSIS
TABLE 27.1 PLOTTINGPOSITIONFORMULAS

Form:1 andn=10
Method California Hazen Beard Weibull Chegadayev Blom Tukey Solve for P (X > x\ m n 2m-l P

.10
.05 .067 .091 .06'l .061 .065

l0 20 14.9 11 14.9 t6.4 15.5

1 - (0.5)'/'
fn n * l

m-0.3 n+0.4

,m-i n+i
3m-l 3n*l

where n : the number of yearsof record the rank a parameterdependingon n as follows:

n a n a

10 0.448 60 0.440

20 0.443 70 0.440

30 0.442 80 0.440

40 0.441 90 0.439

50 0.440 100 0.439

in In general,a : 0.4 is recommended the Gringortenequation.If the distribution is approximatelynormal, , : fi is used.A value of a : 0.44 is usedif the data follows a Gumbel distribution. order or in is The technique all cases to arrangethe datain increasing decreasing of magnitudeand to assignorder number m to the ranked values.The most efficient formula for computingplotting positionsfor unspecifieddistributions,l and the one now commonly usedfor most sampledata, is the Weibull equation P _ m n-fI (27.4)

Whenm is rankedfrom lowestto highest,P is an estimateof the probability of values being equalto or lessthan the rankedvalue,that is, P(X < x); whenthe rank is from in the highestto lowest,P is P(X > x). For probabilitiesexpressed percentages, value is IA\ml@ + 1). The probability that X: .x is zero for any continuousvariable.

ANALYSIS 27.4 REGIONALFREQUENCY

721

regionalvaluesof skewness regiond'lstudies. Methodsof "smoothing" and averaging havealsobeenproposed.ro'15 did Many techniques used in the past for generalizingregi6nal characteristics methodof extending The so-calledstation-year not rely on statisticalconsiderations. statisticalvalidity, especially rainfall recordshasprovedhelpful but has questionable areas.The method seriesor to stationsin nonhomogeneous if applied to dependent has been used to combine, say,two 25-yearrecords to obtain a single 50-year sequence. In practice, the analyst may have to use imagination and ingenuity to while remaining awareof actual and theoretical summarizeregional characteristics, considerations, lndex Flood Method The index-floodmethodusedin the pastby the U.S. GeologicalSurveyis an example The successfully.t''tu method usesstatistical of summarizingregional characteristics and It them in graphicalsummaries. canbe supplemented generally databut combines techimproved by using statisticalmethods,employing,for example,the regression niquesexplainedin Chapter26.The index method,as illustrated in Fig. 27.4, canbe outlined as follows. 1. Preparesingle-stationflood-frequencycurves for each station within the homogeneous region (Fig. 27.4a). taken from the curves at various 2. Compute the ratio of flood discharges frequencies the mean annual flood for the samestation. to 3. Compileratios for all stationsand find the medianratio for eachfrequency (Fi5.27.4b). 4. Plot the median ratios againstrecurrenceinterval to produce a regional curve (Fig. 27.4c). frequency involved are (1) a homogeneitytest to justify Two statistical considerations shortrecordsto placeall stations definitionof a region, and(2) a methodfor extending by on the samebaseperiod. A somewhatsimilar techniquewas developed Potterfor It the Bureau of Public Roads.rT relies on the graphical correlation of floods with physical and climatic variables and is thus a techniquethat refers in part to the in discussion Chapter26 (seealso Chapter16).

Regression Equations Regional U.S.G.S.

survey institutedaprocessof correlatingflood Early in the 1950s, U.S. Geological the Setsof regreswith drainagebasin characteristics. flow magnitudqs and frequencies floodshavebeendeveloped for sionequations the 2-,5-, l0-,25-,50-, and 100-year regionin every state.The work was for practically everyhydrologicallyhomogeneous largely inauguratedto developmethodsfor estimatingpeak flow rates for designof highway structuresat ungaugedbasins.Data from gaugedsites was evaluatedby modelsto the data. regional analysisto provide the best fit of regression gaugedata were consultedto Continuouswater stagerecordersand crest-stage Given the frequencycurves,a developfrequencycurvesfor all gaugedwatersheds.

6000 5000
o

4000
a

o po

3000 2000 1000 0

1.01 1.1

I ts
c") c..l
I

i'
2 5 1 0 2 (Yr) interval Recurrence
(al Recurrenceintervals (Yr)

50

100

station
1

1.1
^ ALt

1.5
n ?<

5
1 AA

10
1 q?

20
).55

50
3.03

(b)

3.0
2.8
I

2.2
6 o

2.0 1.8
l.o
' : I

po

7.4 t.2 1.0 0.8 0.6 0.4

1.01 1.1

t.52

10 (Yr) interval Recurrence

(cl

50

100

Figure27.4 Index-floodmethod of regional flood frequency flood frequencycurve, (b) ratios Q' (a) an-alysis: single-station and (c) regional flood frequencycurve' to Qr.rrfor 15 stations,

ANALYSIS 723 FREQUENCY 27.4 REGIONAL to numberof correlationtestswere madeusingmultiple linear regression predict the parameterssuch as drainage peak flows from various easily obtained independent duringthe snowmelt meantemperature aspect,elevation, area,basinslope,watershed and hundredsof other variables' season, Each study was reportedby state.The open file or water resourceinvestigation of reports are availablefrom the USGS and include discussions the equations,comcopies 1n"ntron rangeof applicability,information on the reliability of the equations, of of all the gaugedbasin frequencycurves,and.sets equationsfor estimatingfloods Equationsfor all stateshavebeen compiledby the U.S. in ungaugid watersheds. Ceotogicit Survey into a PC software packagecalled NFF (National Flood FreHYDRAIN)' qu"n"y;, availablefrom tJreusGS (or FHWA as part of their package, was Regionalregression Figure 27.5 showsthe six regionsfor the stateof Texas. gaugedstationdata.As in many of by independently regionusingavailable conducted variableswere sethe reports, the Texasmanual revealsthat different independent fgr The equationsdeveloped Region2 are: lectedfor eachregion.18
Q, Qs Qro : : : 216 Aos74So'12s 184 322 Ao62oso

389 Ao6a6So'2ta 236 = 485 Ao 668510 Qr, : 555 Ao 6825'0'250 Qro : 628 Ao 6e4s0261 Qrco

(27.r2) (27.r3) (27.r4) (27.rs)

' (27.16)

(27.r7)

for where g : peak discharge given frequency,cfs A : drainagearea,squaremiles betweenpoints 10 and 85 percentof the slopeof the streambed S : average distancealons the main streamchannelfrom the mouth to the basin divide, feet per mile

EXAMPLE 27.6
Develop estimates of flood peaks for a 200-square-mile rural watershed near Dallas. The mean slope between the 10 and 85 percent points is 3.4 ft per mile. Solution. Dallas is in Region 2. Equations 2712-27.17 2 1 6 A o s 1 a s o l 2:s 5 , 2 7 0 c f s : c Os:322406205018o 10,770 fs : 1 a 5 ' 4 9 0c f s Q r o : 3 8 9A o ' 6 a 6 s o ' 2 1 :22'300 cfs Q r r : 4 8 5A 0 6 6 8 5 0 2 3 6 : Q r o : 5 5 5 A o 6 8 2 s o 2 5 o 2 7 , 8 0 0c f s : 34,170 cfs ll Qrco : 628 Ao6e4so261 Qr: give:

724

CHAPTER2T FREQUENCYANALYSTS

UNDEFINED

Pacoa

,'7I

N D R E

Figure 27.5 Hydrologicregionsin Texasfor 1976USGSregional regression equations.(From Ref. 18.).

ANALYSIS 27.4 REGIONALFREQUENCY

725

726

CHAPTER2TFREQUENCYANALYSIS

(NFD Program NationalFlood Frequency

floodlike equations Eqs.27.I2 through27:t7 for estimating Since1973,regression beenpublished,at leastonce, have watersheds peak discharges rural, unregulated for for every stateand the Commonwealthof PuertoRico. In 1993the USGS,in cooperManageation with the FederalHighwayAdministration and the FederalEmergency and metropolitanarearegression ment Agency,compiled all of the current statewide equationsinto a microcomputerprogramtitled the National Flood Frequency(NFF) equationsfor estimatingflood-peak regression This program summarizes Program.re for techniques estimatinga typical flood It for discharges all52 states. also addresses probability peak discharge hydrographfor a given recurrenceinterval or exceedence prograr4lists statewideregression The for unregulatedrural and urban watersheds. and providesmuch of the referenceinformation and equationsfor rural watersheds equationsfor estimating input data neededto run the computerprogram.Regression for urban flood-peak discharges severalmetropolitan areasin at least 13 statesare also available. and Information on computerspecifications the computerprogramare given.rT Instructionsfor installing NFF on a personalcomputerand a descriptionof the NFF data base of regressionstatistics are also given. The program and the associated program is available as part of the Federal Highway Administration package, HYDRAIN, or by itself. Thoughthe USGSand FHWA do not distributeor servicethe software,information about vendorswho provide softwaresalesand servicecan be obtainedby contactingthe agencies.

from ChannelGeometry Flood Frequency

Stream channelsin alluvial systemsdeveloptheir width, depth, slope, and other that flow through hydrographs from the composite hydraulicgeometrycharacteristics that the shapeof some streamchannels,if their valleys. It has been demonstrated can be correlatedwith the mean annual properly evaluated trained hydrologists, by disflow, peak annual flow, bank-full flow, and the dominant, or channel=forming, similar to Eqs. 27.12-27.I7, havebeen successfully equations, charge.Regression standarderrors of estimate. with very reasonable derivedfor many perennialstreams for Measurements these studiesare normally obtained during low flow. The channelof interestis that channelbeing maintainedby the current flow regime.It is by characterized the activechannel,limited laterally by the point barsand mostrecent the geologicfloodplaindeposits. is felt that theserepresent most recentdepositions, It and are thereforeindicative of the width neededby the current flow and sediment the transport regipe. Figure 27.6 rll:ustrates principle in determining the active, the floodplain-buildingchannel,established'for exampleas the width A-A . Colorado,and in havebeenconducted Nevadd,California, Kansas, Suchstudies A elsewhere. USGSinvestigationof 53 gaugedstreamsin mountainregionsof Colorado resultedin the following equations.2o
Qz = 0.666 Wr'eoaD*o Q, : 1.53 W1.682D-o2stAo'017
Qr6 : 2.38 W1 2se t43 53o A.o D-o

(27.r8) (27.r9) (27.20)

ANALYSIS FREQUENCY 27.4 REGIONAL

Referenceline Low-flow water level

727

Distance, in feet

Figure27.6 Ref. 20.)

illustrating active channeldimensions'(From Typical streamcross-section,

Q25 Qro

: :

3.70 Wr'372D-o263Ao2rs 4.93 W127aD-02s6Ao'257

(27.2r) (27.22)

where Q W D A

: = : :

peakflow for the given frequency,cfs iop width of streamat bank-full condition, ft mean depth for bank-full flow, ft cross-sectionareaat bank-full flow, sq ft

The multiple correlation coefficientsfor theseequationsranged from 0-80 for the 42' ranged_from 1 errors,respectively, 50-yrflow to 0.89 for the2-yr event.Standard offer yet anothertool for usein Thesetypes of investigations percentto 32.2percent. estimatingp"uk flo*r, and allow the hydrologistto evaluatefloods by site-specific conditionsversusmore uncertain regionalparameters'

RainfallCharacteristics Regional
The variation of rainfall frequencieswith duration was introduced in Chapter 2' can be usedto I analysis Regression thoseshownin ChaPter15, and the Many formulashavebeenusedin t a form with intensity(i) inverselypt of the form i : AIQ + B) to fit featuresof both the rainfall region A constants andB thereforeserveascharacteristic and the frequencyof occurrencein eacharea'

728

CHAPTER27

ANALYSIS FREQUENCY

EXAMPLE 27.7

intensity-duration-frequency Fit the following rainfall datato determinethe 10-year curve.

r : duration (min) i : intensity (in./hr) Ui

r0
5.9 o.l7

7.r o.t4

15 5.1 0.20

30 3.8 0.26

60
z-)

t20
1 A

0.43

0.71,

Solution in 1.. A model of the form i : AIG * B) can be expressed linear form as ,, of The regression l.li versust yields l/i : 0.005t + 0'12, from which A : 2 0 0 a n d B: 2 4 . 3. Thus the rainfall formula is i : 200/(t + 24). The correlation coefficient is -0.997. ll

lli:tlA+BlA.

Maximum averagerainfall depths have been publishedby the U.S. Weather for Bureau22 durationsbetween30 min and24hr and for recurrenceintervalsbetween

depthrelation showninFig.27.7.
c
o F o

100

'a

9u

N
\\
3hr
mln

24ft

6hr

t 8 0
F

e 'ro
q

lhr

* 6 0 b - "

50

150

350

400

(mi2) Area Area-depth curves for use with duration frequency Figare 27,7 (U.S. Weather Bureau.) values.

FREQUENCY ANALYSIS 729 27.4 REGIONAL hedvily on the densityand location The accuracyof arearainfall data depends of of gauges throughoutthe areaconsidered. The simpleaveraging the accumulation givesno consideration the effectiveareaaround eachgaugeor to the in all gauges of of in stormpattern.Two methodsare available calculatingthe weightedaverage gauge records, the Thiessen polygon method and the isohyetal method. The Thiessen Perpendicmethodassumes linear variation of rainfall betweeneachpair of gauges. a ular bisectorsof the connectinglines form polygonsaround each gauge(or partial polygonswithin the areaboundary).If a sufficientnumberof gauges availableto are the constructcontoursof rainfall depth (isohy.ets), weightingprocesscan be carried out by using the average depth betweenisohyetsand the area includedbetweenthe Figure 27.8 showsboth schemes. isohyetsand the areaboundaries. An exampleof the effect of gaugelocation and density is shown inFig.27.9. stormrainFigure 27.9ashowsthe increase variability betweenThiessen-weighted in from the watershed falls and rainfall at a singlegaugeas the distanceof singlegauges Figure 27.9b showsthe effect of gaugedensityand total areaon the centerincreases. of standarderror of the mean.Completestudies precipitationpatternsoverlargeareas requiredetailedanalysis depth-area-duration datathat dependon the masscurves of in from a network of gauges. The methodis described detail in other of accumulation Figure 27.I0 depictsthe depth-area relation for the 24-hr storm references.23-25. taken at variousdurationsand the shownin Fig. 27.8.It also required observations successive determinationof average depthsby the isohyetalmethod.

(a)

(b)

(a) Figure 27.8 Methods of determiningrainfall averages: Thiessennetwork (24-hr total; precipitation = 2.54 in.) and (b) isohyetal map (24-hr total; averagebasin averagebasin precipitation = 2.50 in.), The arithmetic average over the basin = 39.10/15 * 2.61 in.

730

OHAPTER27

ANALYSIS FREQUENCY
H U.+

>.= O F

(n n

Distance from rain gauge to watershed center (mi) (a)

H 6
6 6 o

b g

9 R 6

500 (mi2) per Area gauge (b) Figure27.9 (a) Relationbetweenthe standarddeviationof the watershed

a4tlt,ililliT,i',l:##-:"::: :1111i:ifli'ffi ill"lH"rffi

cipitation as a function of the network density and drainageare for the Bureau') basin.(U.S. Weather Muskingum

STUDIES OF 27.5 RELIABILIW FREQUENCY

of A significantdevelopment theoretical statisticsis the central limit theorem.As a .on*-rqu"n"" of the law of large numbers,the central limit theorem statesthat for a populationwith finite varianceoz anda meanp, the distributionof samplemeansitrut ir, a numberof equally good meansfrom repeatedsamples-will be distributed as themselves a normal disiribution with meanp, anda varianceequalto a2fn, where or is the population standard deviation. This theorem does not limit the type of under$ing population distributionbut saysthat the distribution of the sampl" *9-uts The statisticalY n will approacha normal distribution as the samplesizeincreases. is the siandarddeviationof the distributionof meansand is called lhe standarderror

STUDIES 27.5 RELIABILIWOF FREQUENCY

731

9
o 00

<

500

1000
Area (mi2)

1500

2000 Figure27.10 Depth-area-duration curves for24-hr stormofFig.27'8. (FromRef' 23')

their of the mean. Listed in Table 27.6 are severalparametersof distributionsand and therefore reliability, are standard errors. It is apparent that standard errors, almost completelya function of the samplesize'

Limits Confidence
based of It is possibleto placeconfidencelimits on the measurement a samplemean of the underlyingpopulation' on the normal distribution of all meansand regardless variate of two As mentionedearlier,approximately thirds of the observations a normal - 1 standarddeviation'Therefore,two thirds shouldfall betweentheiimits of + i and +olfi. The 95 percent of all sample means should occur between the limits confidencelimits for the mean are ment requires knowledgeof the u only s2insteadof o2 is known and meat limits for a samPle confidence For more use of samplingdistributionsthat are beyond the scope of this text. the testingand hypothesis information in ine neta of inferential statistics-in particular, statisticaldecisiontheory-the readermust turn to other sources. Approximateerror limits or control cufvescan be placedon freqlency curves. curve by metdd proposed Beardsinvolvesplacinglines aboveand below the fitted 27.6 TABLE
Measure
Mean Standarddeviation Coefficientof variation Coefficient of skewness

error Standard

"/{!_
a/\/2n

c,\/r + zcl/vzn
x6t"

- D/tn + l)(n - 2\(n + 3)

732

CHAPTER2T FREQUENCYANALYSIS TABLE27.7 ERROR LIMITSFOR FLOOD FREQUENCY CURVES Exceedancefrequency (7o,at 5ololevel) 99.9 99

Yearsof record (n) 5 10

I.)

50 0.76 0.57 0.48 0.42 0.35 0.31 0.28 0.24 o.2l 10 0.95 0.58 0.46 0.39 0.31 0.27 0.24 0.20 0,17 50

10 2.12 1.07 0.79

Q.64

0.1

20 30 40 50 '70 100

r.22 o.94 0.80 0.71 0.60 0.53 0.49 0.42 0.37 0.1

1.00 0.76 0.65 0.58 0.49 0.43 0.39 0.34 0.29

3.4r
1.65 1.19 0.97 0.74 0.61 0.54 0.44 0.36
99

0.50 0.42 0.36 0.30 0.25 90

4.41 2.11 1.52 t.23 0.93 0.77 0.67 0.s5 0.45

99.9

Exceedancefrequency (%, at 95% level)

Nate: Tabular valuesare multiples ofthe standarddeviation of the variate. Five percent error limits are addedto the flood value from the fitted curve at the sameexceedance frequency and the sum plotted. Ninety-five percent limits are subtractedfrom the flood value at the sameexceedance frequency. Log values are added or subtractedbefore antilogging and plotting.

to form a reliability band. Table27.7 showsthe factorsby which the standarddeviations of the variatemust be multiplied to mark off a 90 percentreliability band above and below the frequencycurve. The 5 percentlevel, for example,meansthat only 5 percentof future valuesshouldfall higherthan the limit, and,similarly,only 5 percent shouldfall under the 95 percentlimit. Nine of ten should fall within the band.
EXAMPLE 27.8

The maximum annual instantaneous flows from the Maury River near Lexington, Virginia, for a26-yearperiod are listed in Table 27.8. Plot the log-PearsonIII curve of best fit and determinethe magnitudeof the flood to be equaledor exceeded once in 5, 10, 50, and 100 years.Using Table 27.7, also plot the upper and lower confidencelimits.

Water (year) 1926 1927 1928 1,929 1930 t93l 1932 1933 1934

Discharge (cfs) 6,730. 9,150 6,310 10,000 15,000 2,950 8,650 I 1,100 6,360 1935 t936 1937 1938 1939 1940 t941 1942 t943 13,800 40,000 10,200 13,400 8,950 11,900 5,840 20:t00 12,300

Water (year) 1944 1945 1946 t947 1948 1949 1950 1951

Discharge (cfs) 6,680 6,540 5,560 7,700 8,630 14,500 23,700 15,100

STUDIES 733 OF 27.5 RELIABILITYFREQUENCY Solution 1. The statisticalcalculationsare summarizedas follows:

Arithmetic Mean 7 Variance s2 Skew coefficient C"

Log

11,606
53.87 x 106 " 2.4

4.001
0.051b 0.38

2. After forming an array and computingplotting positions,the data are plotted in Fig. 27.11. from TableB.2' III 3. Plottingdatafor log-Pearson (Table27.9) aredeveloped Confidencelimits are plotted in Fig. 27.II usingTable 27.7. r I
0.01 0.05 0.1 0.2 0.5 1 2 5 F t 0 g r o
d

Log-Pearson IIIFit

V
',;/

I bo 30

"

:l
t.,

? + o
d '"

5r)

360
9" 70
d

6 a o
o . 9 0 95 98 99 99.8 99.9 99.99
o

t
a

/a,

1 0 2 0 4 0 6 (1000 maximum discharge cfs) Annual

Figure 27.11 Maximum instantaneous annual flows, Maury River, Lexington,Virginia.

734

CHAPTER2T FREQUENCYANALYSIS TABLE 27,9

Chance

('/")
99 95 90 80 50 20 l0 4 2
I

/ (vr)
1.01 1.05 1.11 1.25 2 5 l0 25 50 100 200

(c" = 0.38) K
-2.044 - 1.530
\.zJ+

(t: 4.001) (sy= 0'227) y + Ksy: logQ 3.537 3.653 3.721 3.760 3.987 4.ft1 4.300 4.426 4.512 4.591, 4.666
1 44?

4,498
< )6n

0.5

-0.855 -0,062 0.818 1.315 t.874 2.251, 2.60r 2.930

9,705 15,380 19,950 26,690 32,5t0 39,030 46,360

SERIES DURATION ANALYSIS PARTIAL OF 27.6 FREQUENCY

In earlier examplesof frequency analysis,only the seriesof annual maximum or Theseextremes in minimum occurrences the hydrologicrecord havebeen described. with frequencyanalysisand the manipis consistent constitutean annual series thal ulation of annualprobabilitiesof occurrence.All the observeddata-say, all floods or all the daily streamflows-would constitutea completeseries.Any subsetof the the completeseriesis a partial series.In selecting maximumannualeventsfrom a the eventin one year exceeds annual that the secondgreatest record, it often happens year. Analysis of the annual seriesneglectssuch events. maximum in some other Although they generally contain the same number of events,the extreme values is without regardfor the period (i.e., year)of occurrence, usuallytermedthe analyzed partial duration series. In Table27.10 themaximumrainfall depthsthat occurredfor any 30-min period in are Maryland,1.945-1954, shown the order rainfallsat Baltimore, duringexcessive The 11 maximum a represent completeseries, The 65 observations of occurrence. annual eventsare underlinedand representthe annual series.the greatest11 events throughoutthe record are identifiedby an asteriskand representthe partial duration series. The larger numbersoccur in both series,and hencerecurrenceintervalsfor the eventsare the same.The theoreticaldifferencesin recurrenceintervals less-frequent based on annual and partial duration series of the same length are shown in TabIe27.Il. The differencefor intervalsgreaterthan 10 years is negligible.The following exampleis illustrative. EXAMPLE 27.9 Performa frequencyanalysisof the 30-min Baltimore rainfall data in Table27.L0 as an annual and apafiial duration seriesand plot the results. Solution. SeeTable27.I2. The data are plottedin Fig' 27'1'2. r r

DURATION SERIES ANALYSIS PARTIAL OF 27.6 FREQUENCY MD, BALTIMORE, 1945_1954 RAINFALL DEPTHS, TABLE i7.1O MMIMUM3O-MIN Storm num0er 1
z J
I

735

Year 1945

RF depth (in.) 0.38 0.47 0.39 0.76 0.56 0.35 0.43 0.40 0.36 0.62 0.55 0.88 0.47 0.36 1.15* 0.75 1.53* 0.51 0.88 2.04* 0.76 0,97 0.71 1.07* 0.94 r.20* 1.33* 0.65 0.47 0.84 0.68 0.63 0.47 0,52 0.49

Year 1953

Storm number 1
z
J A

RF depth (in.) 0.40 0.45 0.53 2.50* r.03 0.75 0.70 1.00*

5 6 7 8 9 1 2
J

5 6 7 8

5 6 7 8 9

0.55 0.63 0.69 t.27* 1.10* 0.88 0.97 0.59 0.46 0.50 0.55 t952 I 2 3 4 5 6 7 8 , 0.47 1.20* 0.93 0.70 0.57 0.46 0.48 1.30*

1 2 3 4 1955 1 2 3 4 5 6 'l 8 9 0 I 2

0.42 0.70 0.85 0-30 0.70 0.95 t.o2 0.50 0.65 0.55 0.52 0.45 0.54 0.60 0.80 0.95

1947

I z
J A

5 6 7 8

Nole: Underlineditems are the annual series.Asterisks identify the partial duration series.

BETWEEN THE TABLE27.1'I RELATION PARTIAL DURATION SERIES SERIES ANDTHEANNUAL

interval(yr) Recurrence Partialdurationseries

Annual series t.2 1.6 2.0 2.5 5.5 10.5

0.5 1.0 1.5 2.0 5.0 10.0

736

CHAPTER2T FREQUENCYANALYSIS TABLE27.12

Depth(in.)
Annual series
1 z
J i

Partial series

Recurrence interval (n + 1)/m 12 6

A J A A

5 6 7 8 9 10 l1

2.50 2.04 1.53 1.33 1.30 1.27 1.02 0.97 0.85 0.76 0.52

2.50 2.04 1.53 1.33 1.30 1.27 r.20 t.20 1.15 1.10 t.o7

1.7 1.5 1.3 1.2 1.1

The preceding example leads to considerationof the frequency analysis of rainfall depth or iniensity for various durations of rainfall. Design problems often require the estimationof expectedintensitiesfor a critical time period. Frequency of analysis the rainfall record for.periodsother than the 30-min duration-for example, ihe maximum 5-, 10-, 20-, and 60-min occurrences-would yield a family of thesedata is to .oru"r similar to thoseof Fig. 27.10. The usual methodof presenting aize thedatain intensityconveftdepthin inchesto an intensityin in. /hr andto summ duration-frequency curves as shown in Fig. 27.I3. Thesecurves are typical of the that point analysis rainfall data.It shouldbe emphasized the frequencycurves.join -o""u.t"tt""t of from the samestorm; that is, they do not reprethat are not necessarily

2.5

/^

.i E 0) E t <

_x--

>-P

t
d

1.0

zAnnta

series

1.01

1.r1.21.1.5 3

2,

5 678910

20

interval (Yr) Recunence Figune 27.12 Difference in annual and partial duration series 1 1-year record of maximum 30-min durations' Baltimore, Maryland'

PROBLEMS 741 27.t Expand the computerprogram of Problem 26.12 to include the computationof the coefflcientofthe logarithmsofthe input data. mean,standarddeviation;and skewness order and Also, include a routine to sort the data by placing them in descending plotting positions.verify, usiiig the datain Ptoblem27.4. computethe corresponding 27.g, perform a completefrequencyanalysison one of the three 33-yearrecordsgiven in III type III or log-Pearson and comparewith the normal the tablebelow.Fit a Pearson or log-normal ofbest fit. Plot the dataand placecontrol curvesaroundthe theoretical 27.7. curve of best flt usins.Table

Year t928 t929 1930 r93l 1932 1933 1934 r935 1936 193'7 r938 1939 1940 t94l 1942 1943 1944 1945 t946 1947 1948 t949 1950 1951 1952 1953 1954 1955 1956 1957' 1958 1959 1960

River Trempeuleau Wl Dodge, (DA : 643 mi') Qp..r.(cfs) 3,700 1,700 3,360 1,650 3,600 11,000 2,5'70 4,490 7,180 1,780 3,170 6,400 3,120 2,890 5,680 5,060 2,040 8,120 4,570 5,4r0 4,840 1920 3,600 4,840 6,950 4,040 5,710 10,400 r7,400 713

Blow River Banff, Alberta, Canada (DA = 858 mi' ) (cfs) Opua,

James River VA Scottsville, (DA : 4570 mi2) Qp""r (cls)

10,200 7,590 9,280 6,610 9,850 11,000 9,490 6,940 7,'720 5,210 7;770 6,270 7,220 4,450 5,850 7,380 5,590 4,450 7,2r0 5,880 r0,320
4 )qo

75,600 44,700 45,800

2t,roo
31,400
5q 500

38,800 93,400 126,000 62,200 87,400 68,400 130,000 27,100 80,600 95,200 133,000 57,000 41,200 33,200 59,600 94,200 64900 54 snn 67,000 62,900 70,000 20,400 64,200 44,500 ,o ?no 64.200

10,080 8,570 5,460 9,180 10,120 8,680 9,060

5 160

r,r40
8,000 1,480

6,730 7,480 6,440

742

CHAPTER FREQUENCY 27 ANALYSIS

27.10. Compare results of Problem 27.9 wfih estimatesby Gumbel's extreme-valuedistribu'

tion for the 50- and 100-year events.

27.11. The pan-evaporation data (in.) for the'month of July at a site4nMissouri are
9.7 11.2 9.3 9.8 lt.7 8.8 9.2 8.7 It.2 tt.4 9.3 I1.5
11.3 I 1.8 9.3 10.9 l 1.5 8.9 10.4 10.2

Determinethe mean, standarddeviation,and coefficient of variation. What are the standarderrors of thesestatistics? Establishthe approximate95 percentconfldence limits of the mean. 27.12. On which type of plotting paper(probability,log-probability, rectangularcoordinate, log-log, semilog,extreme-value, none)would eachof the following plot as a straight line? a. Normal frequencydistribution. b. Gumbel frequencydistribution. c.Y=3X+4 d. Log-normal frequencydistribution. e. PearsonType III with a skew of zero. t' Q: 43Ao1' g. Log-PearsonType III with a skew of logarithmsequal to zero. h. PearsonType III with a skew of 3.0. 27.13. Determine the 50-yearpeak (cfs) for a log-PearsonType III distribution of annual peaksfor a major river if the skew coefficient of logarithms (base 10) is -0.1, the mean'oflogarithms(base10) is 3.0, and the standarddeviationof basel0logarithms is 1.0. 27.14. A 4Oaear record of rainfall indicatesthat the mean monthly precipitation during April is 3.85 in. with a standarddeviationof 0.92. The distribution is normal. With 95 percentconfidence, estimatethe limits within which (a) next April's precipitation is expectedto fall, and (b) the mean April precipitation for the next 40 years is expectedto fall. 27.15. Given a table of valuesof mean annual floods and corresponding drainageareasfor a number of basinsin a region, describehow regressionanalysiscould be used to determinethe coefficientsp and q in the relation Qz.tz: pAq. gaugelocationhasa range 27.16. The 80-yearrecord of annualprecipitationat a midwestern between14 in. in 1936and 42 in. in 1965.The recordha5a meanof 27.6 in. anda standarddeviationof 6.06 in. Assuminga normal distribution, (a) plot the frequency curve on probability paper,(b) determinethe probability of a droughtworsethan the 1936value,and (c) determine recurrence interval.ofthe 1965maximumdepthand the comparei&ith the apparentrecurrenceinterval. 27.17. A reservoirin the localeof Problem27.16 wlll overfill when the annualprecipitation exceeds in. Determinethe probability that the reservoirwill overfill (a) next year, 30 (b) at leastonce in three successive ybars. years,and (c) in each of three successive

PROBLEMS

743

27.18. *Using Eqs. 26.38 and 26.39 and log-probability paper, solve Problems27.16 and that the annual precipitation is log-normal. 27.17 assuming the determine return 27.19. Given the following valuesof peak flow ratesfor a small stre.am, period (years)for a flood of 100 cfs by flrst using annual peaksfor an annual series and then using all the data for a partial duration series.

Year

Peak (cfs) June I Aug.3 June7 July 2 May 18 June 3 July 4

Year 1968 1969 r970 t97r r972

Date
May 11 June 8 Sept.4 Aug. 8 May 9 Sept.8 May 4

Peak (cfs)

1963 t964 1965 1966 1967

90 300 60 80 r00 90 40

800 700 90 400 30 700 80

27,20. Recordedmaximum depths (in.) of precipitation for a 30-min duration at a single station are:

Year
1963 1964 1965 1966 1967

Date May 3 June 3 June7 June 2 June 1 Aug. 3 July 4

(in.) Depth 2.0 1.0 1.0 1.0 1.0 3.0 1.0

Year 1968 1969

Date
Aug.8 May 6 June 8 Sept.4 May 4 Sept.8 May 9

Depth (in.)

4.O 6.0 5.0 1.0 1.0 5.0 1.0

a. Determine the return period (years) for a depth of 2.0 in. using the california method with an annual series. b. RepeatPart (a) using a partial duration serles. to c. Determinefrom the partial duration seriesthe depth of 30-min rain expectbd be 8 years. (on once every equaledor exceeded the average) 27.21. For a 60-yearrecord of precipitation intensitiesand durations,a 30-min intensity of a 2.50 in.lhr was equaledor exceeded total of 85 times. Al1 but 5 of the 60 years experiencedone or more 30-min intensitiesequaling or exceedingthe 2.50-in./hr value.Use the Weibull formula to determinethe return period of this intensity using (a) a paltial seriesand (b) an annual series.

744

ANALYSIS 27 CHAPTER FREQUENCY 27,22. from the data given in the accompanyingtable of low flows, prepare a set of low-flow frequency curves for the daily, weekly, and monthly durations. (cfs) MEANDISCHARGE FORTHE LOWEST NUMBER CONSECUTIVE OF FOLLOWING VISTA, NEAR BUENA MAURY RIVER DAYS, VIRGINIA
Year 1-day 7-day 3O-day

1939 1940 194l 1942 1943 t944 1945 1946 1947 t948 1949 1950 195 I 1952 1953 1954 1955 t956 1957 1958 1959 1960 1961 1962 1,963 1964 1965

100.0 167.0 22.0 101.0 86.0 62.0 78.0 76.0 97.0 r54.0 136.0 113.0 95.0 115.0 85.0 68.0 83.0 64.0 62.0 88.0 76.0 83.0 99.0 90.0 60.0 51.0 64.0

103.0 171.0 59.4 127.0 93.9 65.9 80.7 78.6 102.0 176.0 138.0 125.0 95.3 116.0 86.1 70.0 96.1 66.3 64.1 92.6 80.9 91.7 103.0 95.0 60.6 54.1 68.7

r25.0 t94.0 69.1 1'73.0 103.0 77.4 90.3 87.1 123.0 215.0 163.0 139.0 101.0 r20.0 90.8 81.7 99.9 7r.7 75.8 107.0 117.0 103.0 152.0 105.0 70.8 62.0 76.2

27.23. For the 7-day low flows at Buena Vista given in Problem 27.22, attempt to fit a probability paper' prostraight-linefrequencycurve on log-normalor extreme-value ceedingas follows:From the original plot of the data, estimatethe lowestflow (say;4) flows (Q - q : at the high recurrenceintervals;subtractthis flow from all observed The Qr); and rcplot Qr versusthe original recurrenceintervals.Repeatif necessary. frequencydistribution. best fittiBg curve will be a three-parameter

PROBLEMS 745 the 27.24. The following table surirmarizes number of occurrencesof intensitiesof various for durationsfor a34-yearrecordof rainfall. Maximum intensities the given durations storms and a count made of the exceedances' were determinedfor all excessive on expected a 5-yearfrequencyand numberof exceedances Interpolatefor the average plot the 5-yearintensity-duration-frequency curve'
were equaledor exceeded Numbei of times stated ihtensities Intensity(in./hr) Duration (min)

2.0
68

3.0 '73 51
LJ

4.0 48 26 11 3 1

5.0 2I l1
I

6.0
o

7.0 2
I

5 10 15 30 60 120

3
I

72 29
1 8 5

3s l7
6 l

7 2

analysisof over 200 flood recordsin Virginia led 27.25. The resultsof a multiple regression to the following regional flood frequencyequations: . Qt'z'v': 9'l3AeoeS2e3 Qzzt-v':20'84861S30e 3oo : Qs'v, 38'1A83oS 283 Qn', : 63'0A8o2S : Qzs-v, 104A71e5266 : Qso'v' l18A7esS21e for whefe the flood discharge the given frequencyis in cfs, A is the drainageareain betweenthe points that are 10 and mi2, and S is the channeislopein ftlmi (measured g5 percent of the total rivei miles upstreamof the gauging station to the drainage diviAe;. Devise a method for graphically portraying theseregional flood frequency relations.(Note that there are four factors,Q, T, A' and S') .25, flnd the predicted floods for the Using the regressionequationsin Problem 2'1 River, at cootes Store.Drainage atea : 215 mi2and chanNorth Fork, shenandoa-h : 44.3 ftlmi. nel slope equationsin Problem 27 '25 with the Comparethe predictions from the regression : by valuesestimated the frequencyanafsis inqxample27.8. Drainageatea 487 fii2 : 2I.l ftltrl'l. and channelsloPe storm rainfall over the city of Baltimore Referringto Fig. 2.6b, comparethe average on SeptemberI0, Ig57, computedby the isohyetalmethod, with the simpleaverage within the city. Neglectthe areato the south of total accumulationat the riin gauges 1.0-in.isohYet' of the Fit the forrnula i : AIQ + B) to the data derived in Problem 21.24 for the 5-year intensity-duration-frequency curve.

27.26.

27.27.

21.28.

27,29.

746

27 CHAPTER FREQUENCYANALYSIS 27.30. Deielop a regional flood index curve for the RappahannockRiver basin from the flood frequency data given in the following table:
(ft3/sec) FOR STATIONSlN THE PEAK FLOOD FREQUENCYDISCHARGES RIVER BASIN RAPPAHANNOCK

Station River Rappahannock near Warrenton, VA Rush River at VA Washington, Thornton River near Laurel Mills, VA Hazel River at fuxeyville, VA River Rappahannock at Remington,VA River Rappahannock at Kellys Ford, VA Mountain Run near Culpeper,VA Rapidan River near Ruckerwille, VA RobinsonRiver near Locust Dale, VA Rapidan Rivet near Culpeper,VA RappahannockRiver near Fredericksburg, VA

Drainage Typeof area(mi2) series 192 15.2 142 286 616 641 14.'7 111 180 456 | sqg
Annual Partial Annual Partial Annual Partial Annual Partial Annual Partial Annual Partial Annual Partial Annual Partial Annual Partial Annual Partial Annual Partial

2.33 (mean).
4,150 4,600 530 610 5,900 7,200 7300 8,300 11,000 12,000 12,300 14,000 '750 950 3,950 4,700 4,600 5,150 9,100 10,800 26,000 29.300

period years in Return 10 8,350 8,650 860 900 11,500 12,500 11,800 t2,400 14,500 15,200 19,000 20,000 t,750 1,900 7,100 7,700 7,000 7,300 16,400 17,600 39,900 42,000 9,000 9,20A r,290 1,310 19,900 20,500 17,200 18,000 18,100 18,900 26,800 27,500 3,350 3,550 11,600 12,000 9,800 10,100 26,900 27,600 55,000 57,500 25 14,000 14,000 2,100 2,100 34,000 34,000 25,000 25,500 24,500 25,000 42,000 42,000 6,000 6,000 21,000 21,000 !5,400 15,800 50,000 50,000 85,000 85,000 19,250 19,250 3,000 3,000 48,000 48,000 41,000 41,000 31,000 31,000 57,500 57,500 10,000 10,000 34,000 34,000 2r,5oo 21,500 78,000 78,000 117,000 117,000

27.31. From the information given in Problem 27.30, find the relation betweenthe mean shouldbe of annual flow and the drainagearea.(Note that the functional expression : the form Qz.zz rA".) the 27.32. Using the resultsof Problems27.30 and27.3l, estima{e 30-yearflood for an with a drainagearcaof 540 mi2. watershed ungauged 27.33. Annual flood recordsfor a lO-yearperiod are given by:

Year Flood

1 300

2 700

3 200

4 400

5 1000

6 900

7 800

500

100

10 600

Mean : 550 cfs, median : 550 cfs, standarddeviation : 300 cfs. Use an annual seriesand the definition offrequency in a frequencyanalysisto determinethe magnitude of the 4-yearflood. Comparethis historicalvaiuewith the analytical4-yearflood obtained assuminsfloods follow a normal distribution.

PROBLEMS 747 a and intensities durations, 30-min intensityof 2.50 recordof precipitation 27.34. For a 60-year a in.i hr wasequaled exceeded total of 85 times.All but 5 of the 60 yearsexperienced or the one or more 30-min intensitiesequalingor exceeding 2.50-in./hr value. Use the Kimball formula to determine the return period of this intensity using (a) a partial duration seriesand (b) an annualseries. to basinis determined be approximately 27.35. The total annualrunoff from a small drainage normal with a mean of 14.0 in. and a standarddeviation of 3 in. Determine the probabilitythat the annualrunofffrom the basinwill be lessthan 8.0 in. in the second yeafs. year only of the next three consecutive 27.36. Six yearsof peakrunoff ratesare given below.Assumethat the floodsfollow exactlya peak. the normal distributionand determine magnitudeof the 5O-year

Year Runoff (cfs)

1 200

2 800

3 500

4 600

5 400

6 500

21.37. Annual floodsfor a streamare normally distributedwith a mean of 30,000 cfs and a flood returnperiodT,of a32,000-cfs the of Determine average variance I x 106cfs2. in the stream. 27.38. Annual floodsfor a streamhavea normal frequencydistribution.The 2-yearflood is 40,000cfs and the l0-year flood is 52,820cfs.Determinethe magnitudeof the 25-year flood. 27.39. The 80-yearrecord of annual precipitation at Linclon, Nebraska,yields a range of valuesbetween10 and 50 in. with a meanannualvalueof 25.00 in. and a standard a annualprecipitationrepresents sum of many random deviationof 5.30 in. Because that annual variables(i.e., depth of precipitation for eachday of the year), assume precipitation is normally distributed. a. In 1936 the precipitation at Lincoln was a mere 14 in. Determinethe probability that the annual precipitation will be 14 in. or lessnext year. this amountwould be equaledor b. In 1965Lincoln received42 in. On the average, exceeded once in how many years? c. Comparethe theoreticaland apparentreturn periods of the record-highvalue of 50.00in. 27.40. Annual floods (cfs) at a particular site on a river follow a zero-skew log-Pearson Type III distributions.If the mean of logarithms(base 10) of annual floods is 2.946 and the standarddeviationof base-10 logarithmsis 1.000,determinethe magnitude of the 50-yearflood. 27,41,. Annual floods (cfs) at a particular site on a river follow a zero-skew log-Pearson 1.733and annualfloodsis Type Illdistribution. Ifthemeanoflogarithms(base10)of the standarddeviationof base-l0logarithms is L.420,determinetfe magnitudeof the 10O-year flood. of flood magnitudes record for a drainagebasin gives 10- and 5O-year 27.42 The 100-year 12,500 and22,000 cfs. Determinethe magnitudeof the mean annual flood if (a) the flood peaksfollow the index-flood curve of Fig. 27.4c atd (b) the flood peaksfollow distribution. a Gumbel extreme-value

748

CHAPTER27

FREQUENCYANALYSIS *The following parameters were computed for a stream: 27.43.

Period ofrecord : 1960-1984, inelusive. Mean annual flood : 7000 cfs Standarddeviationof annual floods : 1000 cfs Skew coefficientof annual floods : 2.0 Mean qf logarithms(base l0) of annual floods : 3.52 Standarddeviationof logarithms : 0.50 Coefficientof skew of logarithms : -2I Determine the magnitudeof the 25-yearflood by assumingthat the peaks follow a (d) log-PearsonType III distribution, (b) Gumbel distribution, and (c) log-normal distribution. 27.44. Peak annual dischargerates in the Elkhorn river at Waterloo,Nebraska,yield the following statistics: Period of record : 1930-1969, inclusive Mean flood : 16,900cfs Standarddeviation : 17,600cfs Skew of annual floods : 0.8 Mean of logarithms(base 10) : 4.0923 Standarddeviationof logarithms : 0.3045 Skew of loearithms : 2.5 Determinethe 100-yearflood magnitudeusing the uniform techniqueadoptedby Council for all federal evaluations. the U.S. Water Resources b. Determinethe'100-yearflood magnitudeassumingthat the floods follow a twoparametergamma distribution. Type III variableX has a mean of 4.0, a standarddeviationof 2.0, and a 27.45. A Pearson of coefficientofskewof 0.0.Determinethevalue(foursignificantfigures)I:f.(X)dX. . 27.46, A timber railroad bridge in Hydrologic Region2 of Texasat Milepost 738.04 on the railroad systemshownin the sketchis to be replacedwith a new concretestructure. The 50- and 100-yearflood magnitudesare neededto establishthe low chord and Determinethe designflow ratesusingthe USGS respectively. elevations, embankment Regression Equations.The drainagearea is 0.43 sq. mi, and the streambedslopeis 62 ftoer mi.

REFERENCES
t . M. A. Benson,"Plotting Positionsand Economicsof EngineeringPlanning,"Proc. ASCE
J. Hyd. Div. 88057-71(Nov. 1962). Res'68(3), L I. Gringorten,"A Plotting Rule for ExtremeProbability Paper,"J. Geoplrys. 1 8 1 3 - 8 1 4 ( F e b .9 6 3 ) .

ucENoftp.*

l l * *l* --)
.J$
^ " : 1- ' .

,
i

t \

I
I

Slrip

\l Sketch for Problem 27.46

750

CHAPTER2TFREQUENCYANALYSIS 3. VerfT. Chow, "A General Formula for Hydrologic FrequencyAnalysis," Trans. Am. Geophy Union 32, 231-237 (1951). s. 4. Water ResourcesCouncil, Hydrology Committee, "Guidelines:for Determining Flood Frequency,"Bulletin 17B, (Revised)U.S. Water Resources Council, Washington, D.C., Sept.,1981. 5. L. R. Beard, Statistical Methods in Hydrology, Civil Works Investigations, U.S. Army Corps of Engineers,Sacramento District, 1962. 6. A. Hazen, Flood Flows. New York: Wiley, 1930. 7. V. T. Chow, "Statistical and Probability Analyqisof Hydrologic Data," in Handbookof Applied Hydrology.New York: McGraw-Hill, 1964. 8. P. Victorov, "Effect of Period of Record on Flood Prediction," Proc. ASCE J. Hyd. Div. 97(Nov.1971). 9. L. R. Beard, Statistical Methodsin Hydrology, Civil Works Investigations; Sacramento District, U.S. Army Corps of Engineers,1962. 10. M. A. Benson and N.C. Matalas, "Synthetic Hydrology Based on Regional Statistical " Parameters,Wat r Re ources Re 3(4)(1967). e s s. 11. N. C. Matalas,"MathematicalAssessment SyntheticHydrology," WaterResources of Res. 12. Yet T. Chow, "Statistical and Probability Analysis of Hydrologic Data," Sec. 8-I, in Handbookof Applied Hydrology (V. T. Chow, ed.). New York: McGraw-Hill, 1964. 13. T. Dalrymple, "Flood-Frequency Analysis,"Manual of Hydrology,Part 3, U.S. Geological Paper1543=A. Washington, D.C.: U.S. Government Printing Offlce, $urveyWater-Supply 1960. 14. G. M. Fair,J. C. Geyer,andD. A. Okun,Water andWasteWater Engineerlng. New York: Wiley, 1966. 15. "Monthly Stream Simulation," Hydrologic EngineeringCenter, ComputerProgram23C-L267, Sacramento District, U.S. Army Corps of Engineers,July 1967. 16. R.W.CruffandS.E.Rantz,"AComparisonofMethodsUsedinFloodFrequencyStudies for CoastalBasinsin California," Flood Hydrology,U.S.G.S.Water Supply Paper 1580: Washington, D.C.: U.S. GovernmentPrinting Office, 1965. 17. W. D. Potter, "Peak Ratesof Runoff from Small Watersheds," Hydraulic Design Series No. 2, Bureau of Public Roads,Washington,D.C.: U.S. GovernmentPrinting Offlce, Apr. 1961. 18. U.S. GeologicalSurvey, "Technique for Estimating the Magnitude and Frequencyof Floods in Texas," WaterResources Investigations Report 77 -110, 7977. t9. M. E. Jennings,W. O. Thomas, Jr., and H. C. Riggs, "Nationwide Summary of U.S. GeologicalSurvey's RegionalRegression Equationsfor Estimating Magnitude and Frequencyof Floodsat Ungauged Sites,"U.S.G.S. WRI 93-1, Reston, VA, 1993. 20. U.S. Geological Survey, "Selected Streamflow Characteristicsas Related to Channel Geometry of Perennial Streamsin Colorado," Open-File Report 12-160, Water ResourcesDivision, Lakewood,Colorado,May 1972. 2 r . E. W. Steel,l{ater Supplyand Sewerage,4th ed. New York: McGraw-Hill, 1960. 22. D. M. Hershfield,"Rainfall FrequencyAtlas of the United States,"Tech.PaperNo. 40, U.S. Weather Bureau.1961. 23. HydrologyHandbook, ASCE Manual of Practice,No. 28, 1949. 24. W. G. Knisel, Jr.,and R. W Baird, in Al?SPrecipitation Facilities and RelatedStudies. Washington, D.C.: U.S. Departmentof Agriculture, Agricultural Research Service,1971, Chap.14, , \ R. K. Linsley, Jr., M. A, Kohler, and J. L. H. Paulhus,Applied llydrology. New York: McGraw-Hill, 1949.

3(4)(re67).

Appendixes
A APPENDIX:
FACTORS AND PROPERTIES. CONSTANTS. CONVERSION TABLEA.1 WATER
atm Heat of vaporizationof water at 1-.0 Gasconstants(R) 540 callg: 970 Btu/Ib R : 0.0821(atm)(liter)/(g-mol)(K) R : 1.987g-ca1/(g-mol)(K) R : 1.987Btu/(lb-mo1)('R) Specificheat of air Accelerationof gravity (standard) g : 32.17ftlsecz:980.6 cm/sec2 Cp : 0.238call(g)("C) Density of dry air at OoCand 760 mm Hg 0.001293g/cm3 Heat of fusion of water 19.7 callg: l44BtulIb Conversionfactors per I second-foot-day squaremile : 0.03719 inch

per *t" 1inch runofr square of

: 33:3 :::::r1;?t-0"t,
: 2,323,200cubic feet

e = 2.71828 log e = 0.43429 ln 10 : 2.30259 Metricequivalents foot : 0.3048 meter mile : 1.609 kilometers acre : 0.4047 hectare 4047 squaremeters I squaremile (mi') : 259 hectares 2.59 squarekilometers (km2) 1 acre foot (acre-ft) : 1233 cubic meters I million cubic feet (mcf ) : 28,320 cubic meters I cubic foot per second(cfs) : 6.6rta, cubic metersper second 1.699 cubic metersPer minute I acre-in. per hour : 1.008cubic feet per second(cfs) " 1 second-foolday (cfsd) = 2447 cubic meters I million gallons (mg) = 3785 cubic meters 3.785 million liters 1 million gallons per day (mgd) = 694.4 gallons per minute (gpm) 2,629 cubic meters per :ninute 3785 cubic metersper day

I cubic per root second t3jL":'"i'ff1":* n'"' : ? I horsepower : 3Jr-f,f*:llld,p".,..ond

752

A APPENDIX OF TABLE A,2* PROPERTIES WATER

Traditional Units U.S. Temperature Specific gravity ('F)

iz

Unit weight 0b/ft3)

Heat of vaporization (Btu/lb)

Kinematic viscositY (ft'lsec)

Vaporpressure
psl in.Hg

40 50 60 70 80 90 100

0.99987 0.99999 0.99975 099907 0.99802 0.99669 0.99510 0.99318

62.416 62.423 62.408 62.366 62.300 62.217 62.11,8 61.998

t073 1066 1059 r054 r049 l044 1039 1033

1,.93x L67 x - 1.41 X 1.21x 1.06 x 0.929x x 0.828 X 0.741

10*5 10-s 10-5 105 10-s l}-s l0-5 10-5

6.11 0.09 8.36 0.12 12.19 0.18 17.51 0.26 24.79 0.36 34.61 0.51 47.68 0.70 64.88 0.95

0.18 0.25 0.36 0.52 0.74 1.03 1.42 r.94

Sl Units Temperature Specific Density gravity (o/cm1 fC) 0 5 10 15 20 25

JU

Heat of vaporization (cal/g)

Kinematic viscosity (cs)

Vaporpressure (mmHg) 4.58 6.54 9.20 12.78 17.53 23.76 31.83 42.18 55.34 92.56 149.46 233.79 355.28 525.89 760.00 (mb)
(g/cm')

35 40 50 60 70 80 90 100

0.99987 0.99999 0.99973 0.99913 0.99824 0.99708 0.99568 0.99407 0.99225 0.98807 0.98323 0.97780 0.97182 0.96534 0.95839

0.99984 0.99996 0.99970 0.99910 0.9982r 0.99705 0.99565 0.99404 0.99222 0.98804 0.98320 0.97777 0.97179 0.96531 0.95836

597.3 594.5 591.7 588.9 586.0 583.2 580.4 577.6 574.7 569.0 563.2 551.4 545.3 539.1

1..790 .t.520 1.310 1.140 1.000 0.893 0.801 0.723 0.658 0.554 0,4'14 0.4r3 0.365 0.326 0.294

6.23 6.11 8.89 8.72 L:2.27 r2.s1 17.38 17.04 23.83 23.37 32.30 31.67 43.2.7 42.43 5'7.34 56.24 75.23 73.78 123.40 125.83 199.26 203.19 311.69 317.84 473.67 483.01 70r.13 714.95 1013.25 1033.23

B APPENDIX

753

B APPENDIX:
TABLE 8.1 AREAS UNDERTHE NORMALCURVE

=1, ,<,t f:*"*'"t'a,

01 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 r.0 1.1 r.2 r.3 r.4 1.5 r.6 r.7 1.8 r.9 2.0 2.1 2.2
z-5

.02 .0080 .0478 .0871 .1255 .1628 .1985 .2324 .2642 .2939 .3212 .3461 .3686 .3888 .4066 .4222 .4357 .44't4 .4573 .4656 .4' 126 .4783 .4830 .4868 .4898 .4922 .4941 .4956 .4967 .4976 .4983 .4987 .4991 .4994 .4996 .4997

.03 .0120 .0517 .0910 .1293 .1664 .2019 .2357 .2673 .2967 .3238 .3485 .3708 .390'7 .4082 .4236 .4370 .4485 .4582 .4664 .4732 .4788 .4834 .487r .490r .4925 .4943 .4957 .4968 .49'77 .4983 .4988 .499r .4994 .4996 .4997

.04 .0159 .0557 .0948 .1331 .1700 .2054 .2389 .2704 .2995 .3264 .3508 .3729 .3925 .4099 .4251 .4382 .4495 .459r .467r .4738 .4793 .4838 .4875 .4904 .4927 .4945 .4959. .4969 .4977 .4984 .4988 .4992 .4994 .4996 .4997

.05 .0199 .0596 .0987 .1368 .1736 .2088 .2422 .2734 .3023 .3289 .3531 .3749 .3944 .4115 .4265 .4394 .4505 .4599 .4678 .4744 .4798 .4842 .48"18 .4906 .4929 .4946 .4960 .4970 .49'78 .4984 .4989 .4992 .4994 .4996 .4997 .0239 .0636 .1026 .1406 .r772 .2123 .2454 .2'764 .3051 .33t5 .3s54 .3770 .3962 .4t3t .4279 .4406 .4515 .4608 .4686 .4750 .4803 .4846 .4881 .4909 .4931 .4948 .4961 .4971 .4979 .4985 .4989 .4992 .4994 .4996 .4997

.o7
.0279 .0675 .1064 .1443 .1808 .21,57 .2486 .2794 .3078 .3340 .3577 .3790 .3980 .4147 .4292 .4418 .4525 .4616 .4693 .4756 .4808 .4850 .4884 .4911 .4932 .4949 .4962 .49' 72 .4980 .4985 .4989 .4992 .4995 .4996 .4997

08 .0319 .07t4 .1103 .1480 .1844 .2190 .2518 .2823 .3106 .3365 .3599 .3810 .3997 .4162 .4306 .4430 .4535 .4625 .4699 .4762 .4812 .4854 .4887 .4913 .4934 .4951 .4963 .4973 .4980 .4986 .4990 .4993 .4995 .4996 .4998

.09 .0359 .0753 .rr4r .1517 .1879 .2224 .2549 .2852 .3133 .3389 .3621 .3830 .4015 .4177 .4319 .444r .4545 .4633 .4606 .4767 .4817 .4857 .4890 .4916 .4936 .4952 .4964 .4974 .498r .4986 .4990 .4993 .4995 .4997 .4998

2.4 2.5 2.6 2.7 2.8 2.9 3.0 3.1

J.Z 5.5

3.4

.0000 .0398 .0793 .1179 .t554 .1915 .2257 .2580 .2881 .3159 .3413 .3643 .3849 .4032 .4192 .4332 .4452 .4554 .4641 .4713 .4772 .4821 .486r .4893 .4918 .4938 .4953 .4965 .4974 .498t .4987 .4990 .4993 .4995 .4997

.0040 .0438 .0832 .1217 .1591 .1950 .2291, .2611 .2910 .3186 .3438 .3665 .3869 .4049 .4207 .4345 .4463 .4564 .4649 .4719 .4778 .4826 .4865 .4896 .4920 .4940 .4955 .4966 .49:15 .4982 .4987 .4991 .499? .4995 .499't

4.0

.499968

London: CambridgeUniversity Press,1957 (for Mathematical Statistics. Source:AfrerC. E. Weatherburn, z : 0 to z : 3,1); C. H. Richardson,An Intoduction to Statistical Analysis. Orlando, FL: Harcourt Brace Jovanovich.1994(for z = 3.2to z: 3.4); A. H. Bowker and G. J. Lieberman,EngineeringStatistics. EaglwooGCliffs,Nf: Prentice-Hall, 1959 (for z : 4.0 and 5.0)'

u?

o F o\ t -

o\ 6 $ o J

a.. < +

cn xt F *

F$

c{ r) \o S

d h +

r) J n +

$ Ri + r+

N F a. st

co cr, o sf

cn N N $

F. $ i *

o\ \c) c) <

o 6 o\ o

o oo ; N o\ 6 m,cn

h + r o

-l \O \o o

r) r n co

0\ 6 + o

n O + co

c..l o

co c{ o.l o

ar 6 cn n

n !. Q cn I

q\ $ o c{ h

\o q) al

!! \O \ al

o r-\o d

\o F. n cn

- + n rO n a\ + q e q oi * * * \ O q q "? n i q i + cq o s o o o o cn .n ci o o o

cO al O\ r) (j

O cn h ri

\O rn cO cr\ S cO O : t rr oo o F. -i
cn ca o

o\ N n ! S l-.. Q N + Sp N q q -t q q q oq oq \ I 9 n =q n cl
o cQ o N ol N ot ol N N al N e.l

O\ F- N F- rr $ \a

1t E.l Q \O

o
lf,

C.t $ ri o J J

st ; *

ci oi -

r O

+ O

6 d O

r o\ 6

di ri d ;
Fq cN Ft-q N rl,ot F- t-q q ol at

;
t-. \o (\l 6l

.d ;
c.l \o c.l cl \O r) ol 6l

6i 6i 6i 6i 6i c.i 6i 6i ..i $i 6i ci 6i Gi oi 6i ..i t..i c.i c.i 6i 6i .'i c'i

O + cl N O ci q al o\ d q N FO 6! ol ca o\ o.l O\ F.. -': 6l o \o -': ol \O S -1 al c-t -: N 0o O .': c.l l= \9.o !9 $ ! c q q q a] ol N e.:l !1 t> O, O Q O\ 0o r) q\ I q] n S - q q q q oc q oq \ i i : : * * # r * \n \ :

V) O c) N c-.1 i oo + J F + : O\ O\ O\ oO oO cO t--

s + F-

\Q \Q \O n or O \o e{ oo * t-- \O \O r) rn

o\ rf

ci h .+

ro +

O\ * r) d \o ca cO N

Ci

O. t' :f, n o'!a) r i O

oo $ cl d

o o c

O n O + co O .i + + + c.l O - oi = e.l 6 v) \o F.- co o\ o ; * - i cl cj q q c\ cl el cl Q c.l q

: * * * i * d i i i i i * * i * d * ; i d * H *

S O\ m F o\ O -' O o\ @ cn oQ cQ tt- q\ n Cl a'l N N co 6 o d S + o cn cn ei cl - Q Q o\ e1 rl q q q q crl c.l q q dl q c.l c'l 6) a] cl q

0)
G

e
o
N

N d a'-o\ o* 6l +\DF 6* cn rt F- o\o n!-{ !! A:t n s n i n n v? v-1 n v? 9 9 9 9 9 q \ \ \ \ \ \ \ \ t t ? T I T oq

O O O O\ O\

t'.. r) $

o] O\ Ft o

O r) O r) \o

O\ N u) oo q, a a Q 9p \O $ o sf r \o 0oo\QQ

Q \O N

z
F

o o
O (L

c) J (t,

\O O * 5 d\

) -

d) tf

F \O ) r) * - - O O. \O n + c{ - O\ \O $ m C9 I !o !! - - n "1 { { { "1 I q - q .t i c] ct q al -1 -1 -.: - -'1 -l q c q ? I c r l r r l r r l l l l r l l l l l l l l l l l l l l l

i \O n \E \o \O : U :U U * oo \o : U \O : 5 O tr: U U :U r) O\ .i c.i F tr: U 9: 9 al ri a:9 r) \o F: F F t*-l tr-: 0\ 6\ a-: 6 tt- h N O ;i oo @ -1,: -i N cn 0a -l @ cn oo -i + < -i ,ai + r) 0o oo -i -i * A oo ^ \O !A 0o -l t.A j F!Q i \O A 0o A \o !? 0o A !! !! oo A Q !A oo A

\O oo O -

O O. t-- t

c{ oo $

O \a O !a O +

oo N \O g\ 6 - q\

O .i

F-

l
N t m A

U)

o (E

C)

o (L

3
lu TL r

ro (\l

\O 6 \o : U

\O $ A

I o i

l 6l $

| r- -

| o

| *

I -

| -{ !.

| I

| F

l m F

| \o o

I cl

z a

CE lrJ (L (f
IL

F- 0Q ; + \o or' cat st t- O\ S \O O a.l S \O 0o O - c.l S G oo c.l rrq q O \ + \ O. "9 "q "e * q q q q q C q I -: - - - - ci c'l oi q cj cl q' i \ r | | r I r I r | | | | | | | I | | | l ' l I I I | | | | |

!n o\ s

o\ n

r) o

Q \o -

r) oo o vl

(o (\l ro

r) oo ro o6 6.O a o I I

* * rcj |

\O i< o |

cl \o f.o I

O 5 F o I

o\ O ol V O\ ;,h dj * + oo od - O\ O\ o o o d o I I I t l

+ oo O\ o I

O c-t c) -.j I

\O n O .i I

6 o\ O -j |

- A \9 c! co \o o d i - c.l a{ .j .i -i .j | I I I

O oo c.l -j I

l-i o i I 6

cn rl m .j |

6 cq .j |

ci e'I $ ..i |

oo !D t .i |

! q\ + i |

{ oI n ..j l q

!Q I \o rl \n oo i S n n O -'i -.i..i .j , l I $ 9o or \O O t'- v) cl

IJJ l

tt* O + \O 6 .i

v
c! o lrl
J C > o

O C O O C O C C O O O -

$ co O !a C.l + \o o\ aj b O + 6 o oo + O, r} q q \ i \ \ "c * q q q q g -1 -1 c'l dt dt n n n \? \ oq og q q a 1 q n - - c ^ l c l o l c { o l

O o\ O\ ol F- r) \O O at- (-'. O t\

\O 6 co q, * H

Q 6 \Q

\Q O h oo

t t

l l

l l l

: V ( /E) o

6 .E...

d 6 i 6 i 6 i o l c t c . l a l e { 6 . . { q . . t -

q q oq \ I n =q q ol - q q o9 \ 9 !.1 n (.l q -1 q q T \ 9 ! 1 I I : ? O O O C O O O C O O

ro

gi qR EE A A $ E q E X B $ $ F n $5 3 5i E 3 5 E E ' '
X:.i :l:l " . iX -':'j: .'-i -':'j -j -i -: -l J a : 1 . . 1 . ' lX : : - -j: ': a -' -j -':r - 'r - 'r 6

R = i t n E : 3E f iF I 3 E E i 5 E E E g E E E g E F *:
; f : :': : : -' -.r' : -i' r i - - .-i j - j .

|o

EtA$[Fg3$$*ilx:s:ssi5353$5S5
s a q E E r s: s fE:l'iF ri l R: R -sd E c; c; d 3 5on $ g a s F R r i : 8 o ci ci d -' Y o = :
: : : : : ! i : i
* l * * i * i

r.c)
A,I

RKlITE=:::==9===Iii35:A35ii=
g.
(d

c)

o C) J 0) o O (L

s & ;d U R # dtF h H obloq "e r + n e \s \I E \A\ F I: p A $ :F: = B i l # H cb dt "e i i l "C e oq s \ I : : X ;

d i o o i o c i c - 3 3 3 3 ; i i c d d b l c o o d o o i c

o) o c)
N

z
ro

0)

: E E E E a : l = g E 5 : A = n Fn qX HoEfi * 8 f l R c ig o :c di d $ : 33 33: ; =; = = d:j:j : J ;i J J J J o o c;

r
o

(r

(\

s: s $3; dse s e d FR RcfH+re +Eti Eo q -qI qcIqI [| !-| r r i $ 9 - { S F : bP br ; io +r t ' -i T Tit t o cI c e b i; 3 : 3

o

r I r | | | | I r , ' T T T T T T

R g R = g g 3 3 3, X d +i 3 '-s ,-E ;I_di -i i x ni ' ;' . -;i i. i i i - s n : , ::'i::',

c - 6 r 66n eq e ,
I I | |

i J f | I I I | | I | | I I r r T l ' r r I | | I I I I I |

H X X R Rn N= E E , 'X
-i-

e {

(o
lo

(\

,: $ ie ET T*TE T ia a iE T T Tq q T ET T T T T T? ? i = F Ta Tii T$A TT l T l5:ne T ii TT

s
O q o ul

O)

? $ * : F s s h $ 8 3$i qn 8 f i A E r l S ; 1 l !eEEf
F<

+TtTTtTtTiiTiTTiiiTTTTTTTTTTTT

o
F

=.9
j o v

c)

=t : : : : : : 3 : I : : l : : : * t f ri l I ? ? ? ? ? ? ? ? T T T Tr r I | | I | | I | | I I I I | |

: 3 R . .ir x : i

3
q

755

759 Open channelflow, 131 Orographicprecipitation, 19 Overland flow, 561 Pan, classA, 129 Parshallflume, 131 Partial duration series,734 Partially penetratingwells, 473 Peakflow,3lI-343 Penmanmethod"97 Permeability,438 Piezometrichead,435 Point precipitation, 27 101 Potentialevapotranspiration, Potentialtheory, 463 Precipitablewater, 16 7, Precipitation, 15-39, 164 Probability, 67I-699 Probablemaximum flood. 373 Probablemaximum precipitation (PMP), 34,392 Rainfall. 281.727 127 Rain gauges, Rationalmethod,312 Recession, 574 equation,692 Regression Relativehumidity, 16 135 Remotesensing, 245 Reservoirs, Reynoldsnumber,437 Laboratory (RRL) Road Research Method,631 Routing,234-257 Runoff, 153-169, 302 Runoff curve number,73 Saltwaterintrusion. 474 Service SCS:seeSoil Conservation SCSAtt-Kin Tr-20 Method,254 SCSmethod,73 SCSTR-55 Method, 320,445 SCSTP-149Method,331 198-201 S-hydrograph, Simulationmodels,548-59 1, 594-628,630 Snow,265-305 Snowmelt,271-284 Snyder'smethod,207 Service(SCS) Soil Conservation number,73 runoff curve method,211 unit hydrograph 55, 165 Soil moisture,25, Specificyrelds,432 Standarddeviation,683 Standardproject storm (SPS),395 Model IV Stanford Watershed

(swM-IV),550
67 Statisticalanalysis, | -699, 7O8-739 Steadyflow routing, 252 methods,508 Stochastic. 402 Storm drainage, Storm Water ManagementModel (swMM), 604,650 1 Streamflow,11-118, l7 l, 574 and Reservoir StreamflowSynthesis Model (SSARR), Regulation 565 446 Streamlines, 448 Surfaceof seepage, runoff. 177 Surface

swMM,650

Syntheticunit hydrographs,205-227, 335 Temperatureindexes,291 467 Theis' nonequilibrium approach, Thiessenmethod. 30 20 Thunderstorms, 182-185 Time of concentration, 535-544 Time series, 444 Transmissivity, 95 Transpiration, - 100 Unconfined aquifer, 443, 461 Uniform flow field,463 188-227 Unit hydrographs, U.S. GeologicalSurvey Index-Flood Method.336 Survey U,S. Geological Rainfall-Runoff Model, 597

University of Cincinnati Urban Runoff Model (UCURM), 655-658 flow,442,467 Unsteady flow routing, 252 Unsteady 309 Urban drainage, Urban runoff models.309-351 1 Velocity measurement, 14 Velocity potential, 440

Water budget,86, 293, 5'l.I Water equivalent,268 : Water t'apor, 16 WatershedHydrology Model

(usDAHL),563

t53,570 Watersheds, Weirs,131 Wellfields,.465 Well function, 468 Wells,460-473