= =
! )! (
!
,
the probability of occurrence
of the event r times in n
successive years
Where q = 1  P.
Frequency Analysis and Forecasting
Annual flood series.
Introduction
Flood frequency
Frequency distribution
Order No. m
Flood magnitude Q (m
3
/s)
T in years = 51/m
1 160 51.00
2 135 25.50
3 128 17.00
4 116 12.75
. . .
. . .
. . .
49 65 1.04
50 63 1.02
A plot of Q Vs T yields the probability distribution.
For interpolation/extrapolation of small return periods, a simple best
fitting curve through plotted points can be used as the probability
distribution.
However, when larger extrapolations of T are involved, theoretical
probability distributions (e.g. Gumbel extremevalue, LogPearson Type III,
and log normal distributions) have to be used.
example  a list of flood magnitudes of a river arranged in descending
order as shown below.
Frequency Analysis and Forecasting
Annual flood series.
In frequency analysis of floods the usual problem is to predict
extreme flood events. Towards this, specific extremevalue
distributions are assumed and the required statistical parameters
calculated from the available data. Using these, the flood
magnitude for a specific return period is estimated.
Chow (1951) has shown that most frequencydistribution
functions applicable in hydrologic studies can be expressed by the
following equation known as the general equation of hydrologic
frequency analysis:
Introduction
Flood frequency
Frequency distribution
x x K
T
= + o
Where x
T
= value of the variate X of a random hydrologic series with
a return period T, = mean of the variate, o = standard deviation of
the variate, K = frequency factor which depends upon the return
period, T and the assumed frequency distribution.
Frequency Analysis and Forecasting
Annual flood series.
the commonly used frequency distribution functions
for the prediction of extreme flood values are
Gumbels extreme value distribution
logPearson Type III distribution
logPearson Type II distribution
lognormal distribution
normal distribution
In order to select and recommend a certain frequency
distribution function (distribution model), the sample
statistics data distribution should be tested for goodness of
fit as a satisfactory basis for selection. (test of goodness, in
previous chapter)
Introduction
Flood frequency
Frequency distribution
Frequency Analysis and Forecasting
Gumbels distribution.
the most widely used probabilitydistribution
functions for extreme values in hydrologic and
meteorological studies for prediction of flood peaks,
maximum rainfalls, maximum wind speed, etc.
For annual series of peak flows, the probability of
occurrence of an event equal to or larger than a value
x
0
is
Introduction
Flood frequency
Frequency distribution
( )
P X x e
e
y
> =
0
1
( )
y
X X
X
=
+
128259
0577
.
.
o
In which y is a dimensionless variable given by
Frequency Analysis and Forecasting
Gumbels distribution.
Rearranging the dimensionless variables, y, equation in the form
of Chows general hydrologic frequency equation; the value of XT
with a return period T is:
In practice it is the value of X for a given P that is
required and the Gumbels equation is transposed as
y = 1n(1n(q)) = 1n(1n(1p))
Noting that the return period T = 1/P and designating;
y
T
= the value of y, commonly called the reduced
variate, for a given T
Introduction
Flood frequency
Frequency distribution
y
T
T
T
=

\

.


\

.

ln ln
1
x x K
x
T
= + o
( )
K
y
T
=
0577
12825
.
.
Where
Frequency Analysis and Forecasting
Gumbel's Equation for Practical Use.
The above Gumbel's equations are applicable to an infinite
sample size (i.e. N ), however, annual data series of extreme
hydrologic events have finite lengths of record. So that, Gumbels
equation is modified to account for finite N as given below for
practical use.
Introduction
Flood frequency
Frequency distribution
x x K
n
T
= +
o
1
Where o
n1
= standard deviation of the sample
( )
=
x x
N
2
1
K = frequency factor expressed as
K
y y
S
T n
n
=
y
T
= reduced variate, a function of T
y
n
= reduced mean, a function of sample size N and is given in
Gumbel distribution (Table 2.1 )
S
n
= reduced standard deviation, a function of sample size N (Table 2.2)
Frequency Analysis and Forecasting
Gumbel's Equation for Practical Use.
Introduction
Flood frequency
Frequency distribution
Table 2.1. Reduced mean in Gumbel's extreme value distribution, N = sample size
N 0 1 2 3 4 5 6 7 8 9
10 0.4952 0.4996 0.5035 0.5070 0.5100 0.5128 0.5157 0.5181 0.5202 0.5220
20 0.5236 0.5252 0.5268 0.5283 0.5296 0.5309 0.5320 0.5332 0.5343 0.5353
30 0.5362 0.5371 0.5380 0.5388 0.5396 0.5402 0.5410 0.5418 0.5424 0.5430
40 0.5436 0.5442 0.5448 0.5453 0.5458 0.5463 0.5468 0.5473 0.5477 0.5481
50 0.5485 0.5489 0.5493 0.5497 0.5501 0.5504 0.5508 0.5511 0.5515 0.5518
60 0.5521 0.5524 0.5527 0.5530 0.5533 0.5535 0.5538 0.5540 0.5543 0.5545
70 0.5548 0.5550 0.5552 0.5555 0.5557 0.5559 0.5561 0.5563 0.5565 0.5567
80 0.5569 0.5570 0.5572 0.5574 0.5576 0.5578 0.5580 0.5581 0.5583 0.5585
90 0.5586 0.5587 0.5589 0.5591 0.5592 0.5593 0.5595 0.5596 0.5598 0.5599
100 0.5600
n
y
Frequency Analysis and Forecasting
Gumbel's Equation for Practical Use.
Introduction
Flood frequency
Frequency distribution
Table 2.2. Reduced standard deviation S
n
in Gumbel's extreme value distribution, N = sample size
N 0 1 2 3 4 5 6 7 8 9
10 0.9496 0.9676 0.9833 0.9971 1.0095 1.0206 1.0316 1.0411 1.0493 1.0565
20 1.0628 1.0696 1.0754 1.0811 1.0864 1.0915 1.0961 1.1004 1.1047 1.1086
30 1.1124 1.1159 1.1193 1.1226 1.1255 1.1285 1.1313 1.1339 1.1363 1.1388
40 1.1413 1.1436 1.1458 1.1480 1.1499 1.1519 1.1538 1.1557 1.1574 1.1590
50 1.1607 1.1623 1.1638 1.1658 1.1667 1.1681 1.1696 1.1708 1.1721 1.1734
60 1.1747 1.1759 1.1770 1.1782 1.1793 1.1803 1.1814 1.1824 1.1834 1.1844
70 1.1854 1.1863 1.1873 1.1881 1.1890 1.1898 1.1906 1.1915 1.1923 1.1930
80 1.1938 1.1945 1.1953 1.1959 1.1967 1.1973 1.1980 1.1987 1.1994 1.2001
90 1.2007 1.2013 1.2020 1.2026 1.2032 1.2038 1.2044 1.2049 1.2055 1.2060
100 1.2065
Frequency Analysis and Forecasting
Application procedure Of Gumbel distribution to estimate the
flood magnitude for d/t return period.
Assemble the discharge data of record length N. Find mean and o
n1
for
the given data.
Using Tables 2.1 and 2.2 determine the reduced mean and standard
deviation for sample N.
Find the reduced variate, y
T
for a given T
Find K by
Determine the required x
T
by
Introduction
Flood frequency
Frequency distribution
y
T
T
T
=

\

.


\

.

ln ln
1
K
y y
S
T n
n
=
x x K
n
T
= +
o
1
Frequency Analysis and Forecasting
How to verify the annual flood series data follows the
assumes Gumbel distribution?
The value of x
T
for some return periods T<N are calculated by
using Gumbel's formula
plot x
T
Vs T on Gumbels probability paper, and this plot
should lie approximately on a straight line (property of
Gumbel propability plot paper).
Since Gumbel's distribution has the property which gives T =
2.33 years for the average of the annual series when N is very
large, the mean of N years flood series should be close to X
2.33
Introduction
Flood frequency
Frequency distribution
Therefore; For the given data, values of return periods (plotting positions) for various
recorded values, x of the variate are obtained by the relation T = (N+1)/m and plotted
on the graph described above. A good fit of observed data with the theoretical
variation line indicates the applicability of Gumbel's distribution to the given data
series.
Frequency Analysis and Forecasting
Example : Annual maximum recorded floods in a certain river, for
the period 1951 to 1977 is given below. Verify whether the
Gumbel extremevalue distribution fit the recorded values.
Estimate the flood discharge with return period of (i) 100 years
and (ii) 150 years by graphical extrapolation.
Introduction
Flood frequency
Frequency distribution
Year Max. flood (m
3
/s) Year Max. flood (m
3
/s) Year Max. flood (m
3
/s)
1951 2947 1960 4798 1969 6599
1952 3521 1961 4290 1970 3700
1953 2399 1962 4652 1971 4175
1954 4124 1963 5050 1972 2988
1955 3496 1964 6900 1973 2709
1956 2947 1965 4366 1974 3873
1957 5060 1966 3380 1975 4593
1958 4903 1967 7826 1976 6761
1959 3757 1968 3320 1977 1971
Frequency Analysis and Forecasting
SOLUTION: The flood discharge values are arranged in descending
order and the plotting position return period T
P
for each
discharge is obtained as
Introduction
Flood frequency
Frequency distribution
T
N
m m
P
=
+
=
1 28
Order
number
m
Flood
discharge
x (m
3
/s)
T
P
(years)
Order
number
m
Flood
discharge
x (m
3
/s)
T
P
(years)
1 7826 28.00 15 3873 1.87
2 6900 14.00 16 3757 1.75
3 6761 9.33 17 3700 1.65
4 6599 7.00 18 3521 1.56
5 5060 5.60 19 3496 1.47
6 5050 4.67 20 3380 1.40
7 4903 4.00 21 3320 1.33
8 4798 3.50 22 2988 1.27
9 4652 3.11 23 2947 
10 4593 2.80 24 2947 1.17
11 4366 2.55 25 2709 1.12
12 4290 2.33 26 2399 1.08
13 4175 2.15 27 1971 1.04
14 4124 2.00
N = 27 years, = 4263 m
3
/s, o
n1
= 1432.6 m
3
/s
x
Frequency Analysis and Forecasting
SOLUTION:
From Tables 2.1 and 2.2, for N = 27, y
n
= 0.5332 (reduced
mean)and S
n
= 1.1004 (reduced standard deviation) is calculated.
Y
T
and K for any choosen T is calculated by the formulas we
descussed
discharge x
T
for any chosen return interval is calculated by using
Gumbel's formulae, and the result for some T values are shown in
table below.
Introduction
Flood frequency
Frequency distribution
x
T
(years)
X
T
[obtained by Eq.(5.30)]
(m
3
/s)
5.0 5522
10.0 6499
20.0 7436
100.0 9558
150.0 10088
When these values are plotted on Gumbel probability paper, it is seen that these
points lie on a straight line according to the property of the Gumbel's extreme
probability paper.
Frequency Analysis and Forecasting
Confidence Limits of probability distribution estimates:
Since the value of the variate for a given return period, x
T
determined by Gumbel's method can have errors due to the
limited sample data used, an estimate of the confidence limits of
the estimate is desirable.
For a confidence probability c, the confidence interval of the
variate x
T
is bound by value x
1
and x
2
given by
x
1/2
= x
T
f (c)S
e
Where f(c) = function of the confidence probability c determined by using the
table of normal variate
Introduction
Flood frequency
Frequency distribution
x
c in per cent 50 68 80 90 95 99
f(c) 0.674 1.00 1.282 1.645 1.96 2.58
S
e
= probable error =
b
N
n
o
1
b K K = + + 1 13 11
2
. .
K = frequency factor given
o
n1
= standard deviation of the sample
N = sample size
Frequency Analysis and Forecasting
Confidence Limits of probability distribution estimates:
Example: Data covering a period of 92 years for a certain river
yielded the mean and standard deviation of the annual flood
series as 6437 and 2951 m
3
/s respectively. Using Gumbel's
method estimate the flood discharge with a return period of 500
years. What are the (a) 95% and (b) 80% confidence limits for
this estimate.
Solution
Introduction
Flood frequency
Frequency distribution
From Table 2.1 and 2.1 for N = 92 years,
y
n
= 0.5589, and S
n
= 1.2020.
Then y
50
= [ln((ln(500/499))] = 6.21361 , K
500
= (6.21361 
0.5589)/1.2020 = 4.7044 x
500
= 6437 + 4.7044*2951 = 20320m
3
/s.
For the 95% confidence probability f(c) = 1.96 and x
1/2
= 20320
(1.96*1726), which results in x
1
= 23703m
3
/s and x
2
= 16937m
3
/s.
Thus the estimated discharge of 20320m
3
/s has a 95% probability of
lying between 23700 and 16940m
3
/s.
Frequency Analysis and Forecasting
STANDARD FREQUENCY DISTRIBUTIONS
Several cumulative frequency distributions are commonly used
in the analysis of hydrologic data and, as a result, they have
been studied extensively and are now standardized.
The frequency distributions that have been found most useful in
hydrologic data analysis are
normal distribution,
lognormal distribution,
Gumbel extreme value distribution, and
logPearson Type III distribution.
Introduction
Flood frequency
Frequency distribution
Frequency Analysis and Forecasting
Normal Distribution
It is a classical mathematical
distribution commonly used in
the analysis of natural
phenomena.
It has a symmetrical,
unbounded, bellshaped curve
with the maximum value at the
central point and extending
from  to + .
Introduction
Flood frequency
Frequency distribution
Normal probability distribution
Standard normal distribution
Frequency Analysis and Forecasting
Normal Distribution characteristics
the maximum value occurs at the
mean.
Half of the flows will be below the
mean and half are above.
68.3% of the events fall between 1
standard deviation (S), 95% of the
events fall within 2S, and 99.7% fall
within 3S.
the coefficient of skew is zero
Introduction
Flood frequency
Frequency distribution
Normal probability distribution
Standard normal distribution
Frequency Analysis and Forecasting
Normal Distribution Frequency curve:
Following the characteristics of normal probability distribution,
frequency curve of normally distributed random variable X can be
developed by the ff procedure.
Compute the mean X and standard deviation S of the annual
flood series.
Plot two points on the probability paper: (a) X + S at an
exceedence probability of 0.159 (15.9%) and (b) X  S at an
exceedence probability of 0.841 (84.1%).
Draw a straight line through these two points; the accuracy of the
graphing can be checked by ensuring that the line passes through
the point defined by X at an exceedence probability of 0.50
(50%).
Introduction
Flood frequency
Frequency distribution
The straight line represents the assumed normal population. It can be used either
to make probability estimates for given values of X or to estimate values of X for
given exceedence probabilities.
Frequency Analysis and Forecasting
Normal Distribution Application:
Disadvantage: it is unbounded in the negative direction whereas most
hydrologic variables are bounded and can never be less than zero.
For this reason and the fact that many hydrologic variables exhibit a
pronounced skew, It usually has limited applications.
pdf for normal distribution
Introduction
Flood frequency
Frequency distribution
Hydrologic variables such as annual precipitation, annual average streamflow, or
annual average pollutant loadings follow normal distribution
2
2
1
2
1
) (

.

\

=
o
u
t o
x
X
e x f
u is the mean and o is the standard
deviation
Frequency Analysis and Forecasting
Standard Normal Distribution:
A standard normal distribution is a
normal distribution with mean (u) = 0
and standard deviation (o) = 1
Normal distribution is transformed to
standard normal distribution by using
the following formula:
Introduction
Flood frequency
Frequency distribution
z is called the standard normal variable
Frequency Analysis and Forecasting
Flow estimation by Normal Distribution.
Comparing the reduced variable (frequency factor) in the
normal distribution, and the standard normal variable in the
standard Normal distribution, we will have:
Introduction
Flood frequency
Frequency distribution
2
2
1
2
1
) (

.

\

=
o
u
t o
x
X
e x f
T
T
T
z
s
x x
K =
=
So the frequency factor for the Normal Distribution is the standard normal
variate, and the Estimated Value for some return period T is expressed as:
s z x s K x x
T T T
+ = + =
Frequency Analysis and Forecasting
Normal distribution estimation procedure:
To make a probability estimate p for a given magnitude, use the
following procedure:
compute the value of the standard normal deviate by
with the value of z and obtain the exceedence probability
from standard normal probability table or use excel function
NORMSDIST(z ).
To make estimates of the magnitude for a given exceedence
probability, use the following procedure:
with the exceedence probability obtain the corresponding
value of z from standard normal probability table or use excel
function NORMSINV(probability).
compute the magnitude X by:
Introduction
Flood frequency
Frequency distribution
s z x x
T T
+ =
Frequency Analysis and Forecasting
Log normal distribution:
It has the same characteristics as the normal distribution except
that the dependent variable, X, is replaced with its logarithm.
Introduction
Flood frequency
Frequency distribution
If the pdf of X is skewed, its not
normally distributed
If the pdf of Y = log (X) is normally
distributed, then X is said to be
lognormally distributed.
x log y and x
y
x
x f
y
y
= >


.

\

= , 0
2
) (
exp
2
1
) (
2
2
o
u
t o
Hydraulic conductivity, distribution of raindrop sizes in storm follow lognormal
distribution.
The characteristics of the lognormal distribution are that it is bounded on the left by
zero and it has a pronounced positive skew. These are both characteristics of many of
the frequency distributions that result from an analysis of hydrologic data.
Frequency Analysis and Forecasting
Log normal distribution probability plot:
Similar to normal distribution, If the logarithms of the peak
flows are normally distributed, the data will plot as a straight
line according to the equation:
Introduction
Flood frequency
Frequency distribution
procedure for developing the graph of the lognormal distribution:
1. Transform the values of the flood series X by taking logarithms: Y =
log X.
2. Compute the log mean (Y ) and log standard deviation (Sy) using
the logarithms.
3. Using Y and Sy, compute 10
Y + Sy
and 10
Y  Sy
. Using logarithmic
frequency paper, plot these two values at exceedence probabilities of
0.159 (15.9%) and 0.841 (84.1%), respectively.
4. Draw a straight line through the two points.
The data points can now be plotted on the logarithmic probability paper using the
same procedure as outlined for the normal distribution. Specifically, the flood
magnitudes are plotted against the probabilities from a plotting position formula
Frequency Analysis and Forecasting
Log normal distribution  Estimation:
Graphical estimates of either flood magnitudes or probabilities
can be taken directly from the line representing the assumed log
normal distribution. Values can also be computed using:
Introduction
Flood frequency
Frequency distribution
to obtain a probability for the logarithm of a given
magnitude (Y = log X)
to obtain a magnitude for a given probability.
Antilog transformation to obtain the true magnitude for
a given probability.
Two useful approximate relation of the logarithms b/n the original variables: