SeismicRefra Burger92

EXPLORATION G EOPHYSICS
OF THE
SHALLOW SUBSURFACE
H. Robert Burger
Smith College
Accompanying Macintosh Computer Software

by
Douglas C. Burger
and
H. Robeft Burger
--J'E
=E
---#
Prentice Hall
Englewood Cliffs, New Jersey 07632
Seismic E*ploration:
The Refraction Method
Refraction seismology has a long and distinguishcd history. Although seismographs

were
developed in the latter part of the nineteenth century, it was not unl.il the early
twentieth
century that major discoveries resulted frorn refractio, techniques.
In 1909 A- Mohorovibii determined that a velocity discontinuity exists at the base of
I'he crust based on his interpretation ol'wave arrivals fl'onr a nearby
eafthquake. Then, i,
l9l3 B- Gutenbelg determined the depth to the Earth's core. Thus, refraction seisrlology
was responsible fbr our early knowledge of the basic structure ol the Earth.
By thc 1920s refraction techniques were being vigorously applied to oil exploration
and were very successful. As seisnric techniques and equipnrent irnproved, reflectiol
eveptually replaced refiaction as the dorninant method applied in oil explor-a1ion. However,
refraction continues to be the most l'requently applied seisr.nic technique fbr shallow
subsurIt is rnainly for this reason that we consider it firsr in this text.
face investigations.
HOMOGENEOUS SUBSURFACE
Before beginning to examine the refi'action rnethod, let's examinc once again
waves spreading throughout a homogeneous subsurface. As the hernispherical wavefront passes
by a
string ofequally spaced geophones, each records the ground displacement due
[o this wave.
The time of transit of this wavel'ront fronr the energy source (shot point) to each geophone
can be detern.rined
frorn a field seismogram. From the geophone spacing (geophone inter-
57
Seismic Exploration: The Refraction
58
Method
Chap' 3
A Single
Subsurface
This means that if we

mine the slope (rise/n
are in meters and mil
determined from the t
tr
the subsurlace is minir

l{orizontal distance lrorr shot point
face because of its lor,r

of useful inlormation r
--}
Figure 3-l Generalized diagram illustrating

ray paths in a uniform matcrirl with no discontinuitjes Time-distance relationships for
the po(ion of the wave traveling directly
from energy source to receiver are shown in
the lravel-time curve
(shot offset), we can conval) and the distance from the shot point to the first geophone
(Fig' 3-l)'
struct a graph (travel-time curve) that plots time against distance
the geophones are equally
and
velocity
a
constant
a[
Because the wave is traveling
somewhat trivial in this
spaced, this time-clistance plot wili Ue a straight line. Although
throughout most o[ the rest of
case, it is instructive to begin an analysis that we will follow
for this line' The form
our seismic studies. trt's ask ourselves i[ we can write an equation
of the equation will be
Path
veloctLY
(3- 1)
A SINGLE SUBSURFACE
IN
You realize, of course,

to have several interfar
interfaces will producr
will limit our analysis

retuming to the surfac
tionship that leads to r
tious time to review Sr

Figure 3-3 illustr
ically refracted ray foll
the interface between

interface at the critica
head wave is generater
from N to G. Several tr
(3-2)
v1
can determine the velocity

For an equation this simple in form, it is fairly obvious that we
lt's useful to review
values'
distance
and
time
the
of the homogeneous rnat,erial if we know
this equation with
of
derivative
first
the
Taking
(or present) , ult or calculus at this point.
respcct Lo -r gives us the slope of the line' Thus,
dt=l
dx
(3-3)
Vl
and
slope
I
v,
(3-4)
Figure
The
ve
3-2
Tra'
locity is 2[
A Single Subsurface lnterface
59
r/
'' -
(3--5)
rrropa
if we have a travel-time curve similar to that in Figure 3-1, we can determine the slope (rise/run), take the inverse, and multiply by 1000 if the units for our graph
are in meters and milliseconds. Following this procedure, the 'r,elocity of the direct wave
determined from the travel-time curve in Figure 3-2 is 286 m/s. Although this insight into
the subsurface is minimal, it does give us some information about the material near the surface because of its low velocity (see Table 2-2). However, as we soon shall see. the amount
of useful information we can extract from typical travel-time curves is much greater.
This means that
SINGLE SUBSURFACE INTERFACE

You realiz.e, of course, that the subsurface usually is not homogencous. We normally expect
to have several interfaces present. Based on our discussion in chapter 2, it is clear that such
will produce reflections, refractions, and wave conversions. At the nroment we

will limit our analysis to refractions and, in particular, to refractions that result in energy
retuming to the surface where it may be sensed and recorded You should recall the relainterlaces
tionship that leads to critical refraction and produces a head wave. This rnight be a propitious time to review Snell's law, the critical angle, and critical refraction.
Figure 3-3 illustrates the path fronr energy source to geophorre G that one such critically refrac(cd ray follows. A comprcssional wave generatcd at E and traveling at V, strikes
the intefacc between materials with different velocities, V, and V,. The ray that strikes this
interface at the critical angle,0,,, is refracted parallel to the interlace and travels at %. A
head wave is generated as energy is retumed to the surface along rays such as that traveling
from N to G. Several possible questions arise from an analysis of Figure 3-3. Can we detect
210.O
,l68
Vclocity = 286 m/s
126 0
o
E
840
420
00
Distance (nt)
Figure 3-2 'Il-avcl tirne curvc (tirne-distance graph) illustrating only direct wave arrivals
Thc velocity is 286 m/s as deternrined from the inverse ol'thc slope
Seismic Exploration: The Refraction Method
60
ll
Chap.
A Single
Subsurfac,
so
tr
I
G
Equation 3-8 is the ba

at a single interface. Il
ht=
thickness
distance
of layer I
r lrorr energ'
at this point, but Eq.
care ofthis task bcfort
'
,'
"'
n.",
V:'vr
Figure
3-3
refiacted
Next, we utilize the re
Diagrant illustrating syrntrols uscd in dcrivation ol tirnc of travel lbr critically
raY.
to change Eq. 3-9 to
thc arrivals ofthis refracted energy? How long does it take for the encrgy generated at E to
arrive at G? How would the head wave arrivals plot on a travel-time curve? Can we extract
any information about the subsurface from the head u,ave assunting we can identify it on
seisrnogranrs'?
Consolidating, we obt,
Derivation of Travel-llme Equation

Using the familiar
A logical first step is to follow a procedure such as we undenook for the direct wave analysis in the preceding section. After we derive a travel-time equation for the refracted wave,
we can analyze the equation and sce what is possibtc. ln deriving an cquation of form similar to
rela
and the identities notei
Eq 3-1, wc will
use the relationships and quantities in Figure 3-3.

Thc first step is a simple one. The total timc of travcl nrust be
tlme
EM MN
vt v2
-+-+-
NG
(3-6)
vl
Sincc
COS
l',
EM
ancl
EM=NG
then
With the llnal fomr ol

(3-1)
EM
A1so,
EA=BG=
MN=x-2hrtan9,,
san'rc procedurc as w
tion with rcspect to r
tA Single Subsurface lnterface
61
SO
h,
x-2httan9i.
cosO,.
V2
I
(3-8)
Equation 3-8 is the basic formula for determining time of travel for a critically refracted ray
at a single interface. If we know hr, V. and Vr, we then can compute the travel time for any
disl"ance -r from energy source to receiver. We could deduce several interesting relationships
at this point, but Eq. 3-8 is more useful if we recast it in a slightly different form. Let's take
care of this task belore proceeding to an analysis. First, we rearrange terms to produce
lihad--I-
2h,
2h,tan9.. x
Vz
V)
I{ cos 0,,
(3-e)
Next, we utilize the relationships
"
sin0
and sin0
,, =
cos 0,,.
V'
V2
to change Eq. 3-9 to
i,,h6-J-rrra-
h,
h,sin)
0,
-r
( cos0,. I cosO,.
(3-
l0)
V2
Consolidating, we obtarn
tlme =
zht(.1
sin2 9,, )
V,cos9,,
+_v2x
(3- l 1)
Using the familiar relationship

sin2
and the
0,. +
cos2
0,, =
identities noted previously, we can substitute to obtain

Z h,cos9,- x
.
llme=''+vt
V2
.
2h,(t - (v,'lvl\\'t2
.vl
tilne =
2ht(v.)2
(3-t2)
(3-
l3)
v2
- v,')tt' +-x
(3- r 4)
v)
VrVt
With the final form of our travel-tirne equation (3-14) directly in front of us, let's follow the
same procedure as we did fbr the direct wave and take the first derivative of the equation with respect to -r. This provides a perhaps surprisingly sirnple lesult:
dtl
dr
(3- l 5)
V,
Seismic ExPloration: The Refraction
62
Method
the critically refracted ray is an

Therelbre, we can say that the travel-time equation for
V. a sloPe of 1lV, must be less
:
equation of a straight line with a slope of 1/V. Since Vr>
than a slope of
A Single Subsurfa
ChaP' 3
TABLE
3-1
Arrival Times
fr
llV,.
0
Analysis of Arrival'llmes
ray can be utilized in

This is sufficient intbrmation to decide if the critically refracted
geolrletry in Figure 3-4 and
rcsolving subsurfhce geology. In our analysis we will use the
equations for the direct and
the values in Table 3-1. W! can place the derived travel-time
times for various geophone
criticatly refracted rays in a spreadsheet and calculate travel
of geophone intervals and
placements. Table 3-l shows ihe results for one combination
geology.
wave arrives first at

we scan the arrival times in this table, we see that the direct
refracted wave arrives before the direcl- wave
<Jistances up to 27 m fiorn the shot but that the
we had a geophone at each distance
at all geophones at distances greater than2l m. Thus, if
geophones l-10 would be
positiln in Table 3-1, the first event on a field seismogratrt for
I 1-24 would be the refracted wave
thc direct wavc arrival, and the flrst event for geophones
If we examine the subarrival. what is a simple physical explanation of this relationship?
frorn left to right at 1400 m/s'
surface modcl in Figure 3-4, we see that the direct ray travels
rlVs' is refracted and travels paralThe critically refracted ray travels I'o the interface at 1400
at 1400 m/s' At significant
lel ro the interface at 4500 m/s, and then returns to the surlace
If
t
5
E
a
lntcrcePt ttlnc =
1r
No refraction
arrivals in
this area
{.
Slope
l2
t5
IC
l8
l2
21
t5
24
t1
21
l9
30
2l
-tl
23
l6
25
39
27
1)
30
45
32
48
34
5t
36
5.1
3rl
-57
40
60
42
6l
45
66
41
69
49
= l/V,
distances fiom the en

velocity more than thr
the refracted wave mr
Slope = I /V,
I
Distancc
(nr) -->
V, = 4500 m/s
velocity.
Is this informat
refracted wave? lf we
source (geophone por
Figure 3-4. The direct
through the origin (asr
times plot along a stra
would be the case whr

two-layer case (single
compute VrandVr.T
of the
subsurface mat
However. the travel-ti

must be some way to
3-1 Arrival
TABLE
Times for Direct and Refracted Waves

Arrival Times (ms)
Distance fronr
Direct Wave
Shot
63
Refracted Wave
0.00
t3.58
214
14.24
4.79
14.91
643
15
l2
16.24
t5
l0.l
16 9t
t8
I2 tt6
t7.5u
ffi
21
t-5
u)
18.24
o.n,',
24
l7
t4
t8 9r
-57
-5u
21
19.29
t9.5tt
30
2t
)0
33
23 51
20_9
36
75.1
2t
43
21
t
5tt
39
21
.86
22.24
42
30.u)
22 9t
45
32.11
23.
48
34.29
74 24
-16.43
24 9t
-5
54
51
60
63
66
69
57
40 71
47.86
45 m
41 t4
49 29
t8
-5
Vckrci(v 2-P
tnr,s)
ffi
I
lo
I
CcoPlronc intcrval trn
-l
I
ffi
I
45fi)
i,,,,
intcrvrls hcti,rc
rrn.t
tr.55
|
25.58
2624
26.9t
27
-5u
28.24
28 9t
distances frorn the energy source, the greatest portion of the refract-ed wave's path is at a
velocity rnore thdn three times as great as the direct wave. At sorne surface point, therefore,
the refracted wave l)tust arrive before the direct wave, which is traveling at a much slower
velocity.
Is this information sufficieut for us to reach a decision on the utility of the critically
refiactcd wave? If we plot the arrival times in Table 3-l against distance from the energy
source (geophonc posil-ions), we produce a graph very similar to the travel-time curve in
Figure 3-4. Thc direct wave times plot along a straight line with slope
= 1lV, that must pass
tkough the origin (assuming the energy source [o be at the surface), an<l the refracted wave
times plot along a straight line with slope l/Vr.If we had only these times an<I distances (as
would be the case when analyzing field data), we could deduce that we were dealing with a
two-layer case (single interface) because we have two travel-time segments, and we could
compute V, and Vr. The velocity values would provide us with some insight into the nature
of the subsurface materials, but we still would not know the thickness of the upper layer.
However, the travel-time equation we derived contains a layer-thickness telm h,, so there
must be some way to solve for /2, if we know V, and Vr.
Chap. 3
A Single
Subsurfar
sect at a point. The I

distance. Crossover r
to the interface. At -r
equal, and, therefore,
Note that Table 3-'1 also is a dynamic table and that values in
bold may be changed. Experiment with the velocity and thickness values to convince yourself that the refracted wave ulti-
mately will arrive before the direct wave. lf you select velocities that are similar (remember the condition that Vr>V,,), or if
you select a very large depth to the interface, the times at the
24 distance positions in the table may not illustrate this relationship. The reason is that the most distant position is not
far enough from the shot point for the refracted wave to
arrive first. lf this happens, increase the geophone interval
and
shortly.)
If
until you are satisfied that the relationship holds. (Actuatly, if

you have this difficulty, you have discovered one of the limitations of the refraction method, which will be discussed
we rearange temls,
and
Determining Thickness
Exantinc Figure 3-4 once again. Thc straight line passing througlt thc arrival tinles lbr the
critically refracted ray can be extendcd until it intersccts thc tirne axis. This time is termed
the interceltt time t.. Rccall that it has no real physical signilicance, bccause no refractions
arrive at the energy source (x = 0). Howevel, trt,r = 0 our travel-tirne Eq. 3-14 reduces to
time=t,=Zht
(v.? -
v,1 Ttlt
The VrV, terms
cance
plify Eq. 3-20 to
(3- r 6)
V.V,
It is mainly
and, l"heretbre,
(3- r 7)
Thus, fbr a single horizontal interface, i1 we can detennine times 1br direct and retiactcd
arrivals fiom a field seismogram, we can calculate the thickncss of the material above the
interface and the velocities of the materials above and below the interface.
Crossover Distance
Before proceeding further, it is perhaps a good time to exatninc two other matters of interest. First, although we now have all the information necessary to solve problerns related l"o a
single horizontal interface, there is an additional approach that occasionally is used, and we
now develop it for sake of completeness. If you examine the travel-time curye in Figure
3-4, you see that the straight lines for the direct wave times and refr-acted wave tirnes inter-
a matter ol
thickness. However, t
whereas t- requires on
computer program Rel
Critical Distance
The second item of in
You should recall that,
tance from the energ
received. This distahcr
3-5. Please note that
ir
tables used in this text,

These times obviously
ming a spreadsheet to
refraction, especially u
65
sect at a point. The horizontal coordinate of this point, x.,, is refered to as the crossover
disttmce. Crossover distance can be used instead of /- to develop a solution for h,, the depth
to the interface. A[ x.o the times of travel for the direct wave and the refracted wave
are
equal, and, therefore,
IlIe
,ri."", *ou"
X,,,
(see 3-2)
vl
tllnertlractcd wa"e =
2 ht
(V22
- V,')'l'
V,V,
X,,,
(see 3-14)
v2
and
xru
Y
If we rearrange terms,
2ht(v21
vr',)',t'
x,.,,
v)
(3- 1 8)
u,e arnve at
h,
(vr2 -
v,1 1tt)
(3- I e)
V,V,
and
,,,=\(+)t.r+,^]
Thc V,
/,
terrns cancel, and remembering that
(3-20)
(Vl- Vl) = (Vz- V) (Vz+ V,), we can sim-
plily Eq. 3-20 to
x'' (" ' = 2\Vr+V,) ''\t''
lr,
(3-2r)
It is rnainly a matl-erof convcnicnce whether we use Eq. 3-2 I or Eq.3-17 to determine layer
thickncss. However, because determination of x,.,, involves fitting rruo Iines to field data,
whercas l- requires only one line, t. usually can be detennined with morc accuracy. ln the
computer program RefractSolvc that accornpanies this book we utilize t. (Eq. 3-17).
Critical Distance
The second itern of interest is one you have seen several tirnes belbre (see Problem 2-18).
You should recall that, based on the geometry of the critical refraction, there is a finite distance from the energy source to the first point at which this refracted energy can be
receivcd. This distance is ref'erred to as the critical distance and is equal to x..,, in Figure
3-5. Please note that in many refiaction modeling programs and in the tables and dynamic
tables used in this text, reliaction times are given for distances less than the critical distance.
These tirnes obviously cannot be valid but are given to avoid the complexity of programming a spreadsheet to exclude them. You rnust remember, however, the geometry of critical
refraction, especially when planning and interpreting field suryeys.
-Y
66
L{, r.tl2
-rcrit
Method
ChaP. 3
TABLE
+l
A Single Subsurfa
3-2
Critica
Thickness (nr)
7
4
6
VzrVt
3-5
This diagram illustratcs the relationships lbr calculating critical disttnce
r0
Critical distancc is thc nlinirnum distance
t2
fronr tlle encrgy source at which thc first criLical rel-raction can be reccivcd
t4
Figure
t6
t8
Dcrivatiorr of an cquation to calculate critical distance is straightfbrward (u'hich you

already know ifyou solved Problem 2-18).
.\,,,
tan
0,, =
,/
20
rl-)?r
-f
Try a va
co ntrast
2-2in or
and, since
sin
then
0."v)
= A.
I /y)l
'I|=- '"'/
tunlsrtt'l
\ v- ,/.1
(3-23)
t,,
and
x.,u
21,,
trn[.in
t# )]
(3-24)
Equation 3-24 is quite sufficient for our purposes in this day of inexpensive hand calculatois which handle trigonontetric functions with ease. Howcver, it is possible to use the
trigonopet-ric identitics we employed when deriving the travel-tinle cquation for head
wavcs to arrive at a dill-crent lbrm for Eq.3-24:
You rnay have

surface is the same as
reflected and retiact(
demonstrate, all other
fere with our analysis
Constructing
Now that you
Table 3-2 presents some critical-distance values fbr a range of layer thicknesses' The
velocities used Lo calculate these values are similar to thosc for unsaturated sands overlying
saturated sands (see Table 2-2). Note that the percentage of critical distance relative to
thickness is substantial. In a case such as this the positioning of geophones is iruporlant if
one is to gather the maxitnunr alnount of intbrmation in the most economical fashion-
have
and thickness for a sir
ine a field
(3-2s)
a Tra
scisrnogra
straightforward exam
Recall that we are int
In Figure 3-6 this is i
marked by arrows to I
Construct a tabl
at each geophone. Th
Figure 3-7(a). Use a s

slopes and the interc
times, it's a good ide

TABLE
3-2
Thickncss
67
Critical-Distance Values for Various Thicknesses
(nr)
Critical Distance (m)
r53
1.06
6
IJ
t0
459
Vclocity l-P
612
165
Velocity 2-P
1.1
9 Ilt
t0 7t
t6
t)
t8
t3 16
20
12
(nr/s)
(nr/s)
500
1400
2.1
-5.29
Try a variety of velocity values to investigate the role velocity

contrast has in affecting the critical distance. Examine Table
2-2 in order to use realistic velocities.
You rnay have noliced that thc path of the first critically retiacted ray to arrivc at the
sur{acc is thc sarne as a ret-lection which strikes the interlace at the critical angle . Thus, the
reflected and relracted wavcs arrive at the sarne time at r..,,. However, as we will later
dcmonstrate, all other reflcctions an'ive later than the critical retiacl.ion and so will not interfere with our analysis, which depends on the first wave arriving at each recciver.
Constructing a Trave!--l'ime Curve from a Field Seismogram

Now that you havc a good grasp of the relationships that enable us to deternrine velocities
and thickness fbr a single horizclntal interlrrce separating two differing materials, let's exarrine a ficld seismogranr liom a sitc with this subsurtace geology. Figure 3-6 illustratcs a
straightfonvard cxarnple. Perlinent information is contained in the caption for this trgure.
Recall that we are interested in the time at which the first energy arrives at each geophone.
In Figure 3-6 this is indicated by the first sharp downtum in each geophone's output and is
rnarked by arrows to help you recognize this first break.
Construct a table sirnilar to Table 3-3 in which you record the time of each first break
at each geophone. Then graph this daLa so as to create a time-distance plot similar to that in
Figurc 3-7(a). Usc a sufliciently large sheet of cross-ruled paper (at least 8Z X 1 I in.) so that
slopes and the intercept time can be determined accurately. When measuring firslbreak
timcs, it's a good idea to construct a ternplate on a transparent sheel- on which you have
68
A Single Subsurfa
Chap. 3
(a)
80
?20
64.O
560
3 l-^=
48.0
itr +oo
E szo
4 i--.*
5
240
6,0
80
00
l0
lll-
t2
i
I
l<nts
-5
0ms
80 nrs
12.0
3-6
Ficld scisntograrn from thc Connecticut Vallcy, Massachusctts CeoPhone

(races are labclcd l-l 2. The first geophone is lcated 5 m fronr thc energy source, The geophone interval is 3 m First breaks for each trace re indicatcd by a downward directcd
arow Tirning lines are at 5 nts inlervals. The record encoll'lpasses l(xl ms This scisnroFigurc
640
grarn exhibiLs a classic two layer pa[tem
a
F
to read
drawn lines at 0.5- or l-ms i1tervals. Place this sheet on the seisrnogram in order
values accurately and consistently.
Your ncxt task is to interpret the data points. The first eight data points dcfine a
pass through
straight line reasonably well (Fig. 3-7(b)). The line for the direct wave lnusf
that criterisatisfies
the origin if the energy source was located on the surface, and this line
case
like
a
farniliar
on. The last four data points also lie along a straight line. This plot looks
in which we receive the direct wave a[ the closest geophones and then the head wave at the
commost distant geophones. Next, measure the slopes of these lines and from the slopes
(46
the
ms),
and
pute the velocities. The intercept time can be read from the veftical axis
subsurface structure computed fiom Eq.3-17. Our computations suggest a 9'2-rn-thick
layer with a velocity of 385 m/s overlying material with a velocity of 1300 m/s.
Now that you have leamed some of the basics of the refraction method, it seerns like
an
the computer prograrns included on the disk that

appropriate time to introduce
this textbook. Instructions for using these programs are in Appendix A'
two of
"cio*pori"s
3-3
Cieophone
Dr
stance ( nr)
Tinre (nrs)
560
480
400
320
240
160
80
00
Figure
1l
I
l
l
t.r
-j-7
through drt:r p
RefractModel
As such, it is an extr
them in the pages th
Using RefractModel and RefractSolve
TABLE
800
(b)
travel-time curves b
with curves represen

be. So use RefractV
data as if energy sou
common. buf as we
Time-Distance Values for Seismogram in Figure 3-6
123
-5 8 ll
135 21.-5 285
4561tt9
t4 11 20 23 76
370 435 I -5 s90 625
-5
29
6tt-5
t0'l12
32 35
710 730
data points represent

3u
750
RefractSolve
RefractModel. You t
69
800
(a)
t20
640
E
E
t:
560
480
400
520
240
l6L-t
80
00
Distancc (nr)
800
(b)
120
640
560
E
E
4to
t: 320
240
,/,/
160
,l
./
OU
00
11 14 1? 20 21 26 29 32 35
38
Distancc (nr)
3-7 (r) Tinrc distitncc drta

throuqh dlttil poirrts in (a)
Figure
llont thc scisnlrgnrrrr in Figurc 3-6. (b) Lines drawn
RefractModcl produccs tirnc-distance data for subsurf'ace models that you designate.
As such, it is an extrernely useful tool fbr investigating various relationships as we discuss
thent in the pages tlrirt lbllow. Because the goal of this chapter is to enable you to interpret
travel-time curves bascd on the arrival of refracted waves, the more experience you have
with curves representing difl'erent subsur{ace conditions, the better your interpretations will
be So use RefractModcl fieely and often. A final note: RefractModel plots time-distance
data as ifenergy soLrrccs were placcd at both ends ofa stilng ofgeophones. Such practice is
comlrton, but as rr"'c have not yet discussed the reasons for this procedure, study only the
data points representcd by the same symbols as in Figure 3-7.
RefractSolve plots time-distance data that 1,61u enter or from data files created by
RefractModel. YrLr thcn fit straight lines to the data by drawing with a rnouse. The power of
70
Method
Chap'
subsurthis program is in its ease of use. Because you can draw lines quickly and because
when
face structures are computed quickly, you easily can attempt several interpretations
interpretastraight lines do not perfectly pass through all field-data points. Such multiple
your data'
tions are encouraged in order to gain some notion of the range of models that fit
From time to time in the following pages, suggestions are offered for exercises utilizsince
ing these computer prograrns. Try to spend some time investigating these suggestions'
by doing so yoLrr interpretative abilities wilI improve substantially'
It
Two Horizontal
lnter
200
I
a
i'=
is possible to save RcfractModel data as a data file, transfer directly to
RefractSolvc, and then analyze the RefractModel clata. This capability is a useful learning
aid, and you are urged to try this as well.
0
Epie cn
Use RefractModel to investigate a number of subsurface variations for the single interface case. First, hold layer thickness
Distant:c lron
te r'
try a variety of velocity contrasts to galn an

veloclties and related slopes' Remember
of
appreciation
always that we assume V, , V.,. Next hold velocities con-
constant and
stant and vary the thickness from very small to very large'
Note the manner in which the travel-time curves change' Can
you formulate a number of situations in the field for which
you might not recognize that two layers (one interface) are
present?
^^'l
the velocity discontinu

achievement at such an
If you examine
z.ontal inl-erface). The d

tlrrough the origin, becr
from the focus to the etr
The Mohorovidic DiscontinuitY

Virtually everyonc who has taken a physical geology course is familiar with the velocity
discontinuity that separates the eallh's crust tiom the mantle. Most retrernber this interlace
.'Moho" or M-cliscontinuity, a few recall that this tenninology incorporates (to some
as the
extenr) thc name of its discoverer, Andrija Mohorovidii (1857 1936), and f'ewer still are
aware of the evidence advanced by Mohorovidii to demonstrate the existence of the
i[ i t t rl i s c o n t i n u i t1'.
Following a strong earthquake in October 1909 that affected his place of work at the
Zagreb Meteorological Observatory, Mohorovidii plotted the main deflections on seistnogrz,.s recordedby 29 seismological observatories located at distances from very near the
epicenter out to 2400 km from the epicenter (Bonini and Bonini, 1979). Considering only
the p-wave arrivals, Mohorovidii noted that on seismograms recorded at stations close to
the cpicenl.er only one pulse appeared, but on seismograms from more distant stations there
were two clearly distinguishable pulses. This same paU.em held for S-wave arrivals (see
Figure 3-8). Mohorovitii reasoned that the early pulses recorded at the more distant sta-
M,
ho
rov
tions must be due to waves that traveled at a greater velocity over a considerable portion of
their path distance. He cornputeil velocities fiom the travel-time curves and calculated that
waves are shown (S-rar

waves are rneasured in
'ulficicnt time i
down before the arrival
work).
arrivals. Although this

stand the essential elem
Depending on th
spread, you often will
subsurlhce geology is n
the next step up in corn;
TWO HORIZONTAL
INTERF/
It is not difficult to enl

interfaces (often refen
sequence of sands and
nom.rally will be above
Two Horizontal lnterfaces
71
t
z
a
Distance
l(XX)
liorr cpiccntcr (krr)
--)
Earthquake
focus
-5
6 krn/s
V-J.J ktrls
Irigure 3-8 'lavel-timc curves sirnplilicd

ll-onr Mohorovili6's original plots. I,( is thc
clircct cornprcssi<tntl u,avc, and l',, is tlrc
refr-rctcd corrrprcssionll wavc Tcnrrinology
firr shear wrves S is similar Thc scclion
bclorv the trlvel-tinre curve illustralcs

Mohorovi[i6's intcrprelation ol thc truveltilrc cur\/cs
the velocity discontinuity was located at a depth o1'50 km. This was quite a remarkable
achievement at such an early date in the history of seisntology.
If you examine Figure 3-8, you should recognize our classic two-layer case (one horizontal interface). The direct wave curves for P- and S-waves (labeled P, and .f,) do not pass
through the origin, because the ea(hquake occurred not on the surface but at depth. Travel
tiom the focus to the epicenter took a tlnite amount of time. Also, curves for both P- and Swaves are shown (S-ray paths are not shown). Becausc thc distances traversed by P- and Swaves are measured in kilometers in this case (rather than in meters as in most exploration
work), sufficient time is present fbr the disturbance due to the arrival of the P-wave to die
down before thc arrival of the S-wave. Therefore, it usually is possible to recognize S-wave
an'ivals. Allhough this presentation is simplified somewhat, you should be able to understand the essential elements of the evidcnce for the Molrorovi6ii discontinuity.
Depending on the objectives of your field study and the length of your geophone
spread, you often will encounter a two-layer case. As you are well aware, howeveq most
subsurtace geology is more complex than a single horizontal interface. Let's take a look at
the next step up in cornplexity.
TWO HORIZONTAL TNTERFACES
It is not dilllcult to cnvisage a subsurface conf iguration in which there are two horizontal
intcrfaces (often reI'erred to as a thrcc-laycr case). For instance, consider an alluvial
scquence of sands and gravcls ovcrlying bcdrock. The sands and gravels near the surface
nornrally will be abovc thc water table, dry, nnd possess a relatively low velociLy. The water
72
Method
Chap'
table forms the llrst interface, as the saturaled sands and gravels will have a substantially
forms the
higher velocity than their dry counterparts Finally, the top olthe bedrock surface
atternpting
to
In
sediments'
saturated
than
higher
are
velocities
second intedace. as bcdrock
manner
the
same
as
precisely
in
procced
we
situation,
fbr
this
derive a travel-time equation
for a single interlirce.
Two Horizontal
lnte
sln
Everything we now d
plete travel-time equz
Derivation of a Travel-f ime Equation

lravel fi'otrl E to G
Rel'erring 1o Figurc 3-9 wc scc that firr this geology,a direct rvave rvill
to thc situation of
identical
is
This
path
EI'QG.
thc
lr'avcrsc
and that u,avc energy also will
Assurlling V, >
interluce'
V"-V.,
the
will
strike
also
encrgy
Wavc
thc single horizontal interlace.
EPRSQG'
follows
paLh
that
wavc
a
itr
rr)l-crested
arc
wc
so
occur,
Vr, criiical reliactiort will
procecdbelbre
answcr
g,
we
llltls[
quesliorl
rnajor
The
I/'l
V.i
=
Wi: already know l-hat sin
Figure 3-9 is
ing is how to deterDtinc thc'g, rctluired for ra1' /'R to be criticalll' reliactcd
carefully
drawing
tlie
Study
qucstion'
to
this
ansrvcr
thc
to
c1ruLwn tc'r call yclur attcntioll
As before,
E
Also,
L],
so that
belbre cclntinuing
We know fron Eq. 2-25 (Sncll's law) that
sin0,
sin 0,
and, since 0,
RJ
Vl
These rclationships
V.
becornes
0,., then
[ilnc
r-l
It is wo(h\\'hile
to pr
similar to thosc used

idcntities that wc user
t
G
tllll!4,!
th ic kness
ol Jaycr
vl
Vi
th
ickncss
of layer
[:me =
h,=
Irnrc =
2
**iaiai-
trne =
ttnte
and, finally
v|>v2>
v)
Diagram illustrating symbols used in derivation ol time of travel for ray critically refracted along the second interface io a tkee-layer case'
Figure
3-9
Once again wt
observed by now, if v
,-r

sin
0. =
73
sin
0,, =
and,
therefore,
sin
e, =
(3-26)
Everything we now derive follows the same procedure as for the single interface. The complete travel-time equation is
EP PR RS SQ,QC
vt v2 v3 v2 vl
(3-21)
tltTtc=T-T-TAs before,
l" and
cos 0,
PR
= SO
- =
= httan9i and
PC
= DQ -
L,l' = OG =
h'cos 0,,
Also.
EA
BG
l4tan9,,
so that
RS=x
2h,tan9,-2h"tan9l
(3-28)
These relationships can be substituted into Eq. 3-27, so that our travel-time equation
becomes
.
(
2h,
x
Vrcos9,,
2h,
cos 0,
2h,ran9,-2h,tan9,..
(3-2e)
Vt
It is worthwhile to present the major steps in simplifying this equation, but as the steps are
sirnilar to those used to arrive at 84. 3-14, they are presented without comment. The same
identities that we used previously are adequate.
ume,.. =
2h,,tan
VT
V\
2h,
0,
2hrtan
0,,
(3-30)
VJ
2h,
*
time
= *
-2h,sitt20' -2h'sin20''
"'^'"-%'{cos0,'Vrcos0,.(cosO,V,cosg,.
r 2h, - 2h,sin2 0 2h - 2l4sin2
Vrcos9,
V,cos9,
V
.
x 2h,cos9,
'
+ 2l4cos0,,
ume=-+
v2
vt
v3
.
ItttlL
and,
(3-3 1 )
0,,
(3-32)
finally
tlme =
r *2h,(V.2 -v,2)tt2
v3
v.v'
2h2(%2
- Vr')'l'
(3-33)
V.V,
Once again we finish with an equation for a straight line. As you probably have
observed by now,
if we once again take a derivative-voila!
Chap.
74
Two Horizontal
V.
dt _l
dx Vt
(3-34)
same type of
Although it is beginning to sound a little repetitious, we can continue with the
you
convince
should
3-10
analysii as we diJ for the single interface. Examination of Figure
eventually
interface
that at some point the ray that has a part of its path along the second
will be thc first to arive at one of the disLant geophones. We see that the velocity of
line passthc third layer carr be deterrnined by taking thc inversc of the slope of the straight
tJre
second
along
relracted
ing through the a.rival timcs fiom the ray that was critically
i
tcrtace.
Vr, and V,
Thc line corresponding to refract.ions ll'onr the second interlace can he extended back
1,r' If we use
to Lhe verlical axis (Fig. 3-10) and is referred to as the sccttttl intcrceltt litrrc,or
Eq. 3-33 and solve for /r-, at -r'= 0. wc urrive e(
r
1,,.
l-15l
are con
travel-time curve in Fi
84. 3-L1 . The second
Based on our di
Figure 3-10 as represr

each with a lower slo
are available front the
depths to the two inte
the local geology, we
the exploration site.
In the
n, =
lnte
precedinl
thickness of the secon

in the case of the singl
solutions use intercep
This derivation is left

field work, however, i
Critical Distance
The critical distance r

interface is EG in Figt
toEA+PC+CQ+Bt
A
sin 9. = V,/V, and sin
c
E
F
lntcrcept titne =
Slope
= l/V,
Slope
= l/V,
Intcrcept tirre = /i
Distancc
(nr) -->
, , , , , , , , , , t=, ,,t,t ,t I
/
Vr= 400 m/s
1500 m/s
= 4500 nr/s
v|>v2>vl
Figure 3-10 Gencralized diagram illustrating ray paths in a rnalerial with two horiz'ontai
are
interfaces Time-distance relationships Ior the direct and two critically retirtcted rays
shown in thc
trrrel titnc curvc-
Two Horizontal
lnterfaces
75
V,, Vr,and V., are computcd from thc inverse of the slopes of the three line segrnents on the
travel-time curve in Figure 3-10. The thickness of the first layer h, is computed using r-, and
F4.3-11 . The second layer thickness is then calculated using r,. and Eq. 3-35.
Based on our discussions to this point, we interpret a traiel-time curve such as that in
Figtrre 3-10 as represenLing a three-layer case, because it has three straightline segnrents,
each with a lower slope than the one closer to the energy source. Vclocities for each layer
are available frortr the curve as are the thicknesses of the frrst two layers (and, of coLrrse, the
deplhs to l"hc two inter{aces). From this information alone and sonre general knowlcdge of
the local gcology, we can arive at a reasonably good rnodel fbr the subsurfnce gcology at
thc cxploration site.
Irt the preceding analysis we used only intercept tirncs to devclop an equation l'or
thickness of the sccond layer. We could develop an equation utilizing crossover distances as
in the case oI thc single intcrlace, but fbr reasons cited previously and as mosI cornputerized
solutions usc intercept tin.res, it does not seem necessary to produce yet ur()rc equations.
This derivation is lelt lbr those interested in such an exercise. Because of implications lor
f reld work, however, it docs seern wise to once again tum to the topic of critical distance.
Critical Distance
The critical distance x..,, lbr the head wave produced by critical refraction along the second
inlertace is EG in Figure 3-l l. The ray path for this geometry is EPRQG. Thus, r. , is equal
to EA + PC + CQ+ BG. Because EA = BG = h,tan 0 and PC = CQ= hrtan 0,,, and because
sin 0 = VrlV.and sin 0, = V,lV.,then
thickness
of lirycr
ht=
ickness
of layer 2
th
Figure 3-ll Diagrarn ilJusLrating s_vmbols

uscd in dcrivation ol critical distance for rav
v3>v2> vl
criticalJy relracted aJong the second intellace

in a three-layer case
Chap. 3
76
- -')
sin
_,*
[, (,
-1
erf
*,
"'
sin
g,.
,* r5'r'
Two Horizontal
(3-36)
The
[,-
depth
magn
-1
V'/
/V.
lt
and ir
ues
which
(u/,.)')'
3-4 rel
pa rrso
sig
ifi
and
vl
(3-38)
('.' ',')"'
In orclcr to investigate the inrplications of this fbrmula fbr critical distance, let's creaLe
a table sirr.rilar to Table 3-2.
The velocities in Table 3-4 are fairly representative of what rnight be expected in an
the magniarea wiLh unsaturated and saturated sediments lying abovc bedrock' Although
point
where
the
[o
source
energy
the
frorn
(the
distance
tucle of thc second critical distance
secthe
to
depth
lralfthe
than
ntore
is
received)
be
can
retiactions from the second interface
who
examined
to
anyonc
ullexpected
be
not
should
this
ond interlace (surn of hrand h-r),
Figure 3-l I closclY.
3-4
TABLE
Although crit
the first arrival at e
traveling path EPR
along the first inted
siderable distance a
head wave generat

RefractModel. Figu
layer thicknesses ft,
for such values is l,
before the head war

interface.
Critical
Critical Distances for Two Horizontal lnterfaces
distar
arrivals on a fleld
First X,.,, (nr)
(nr)
800
5l
120
640
560
t2
076
r53
229
l6
306
il37
?(l
3U2
t4 2l
24
4.59
t7
28
535
19 90
3)
612
36
rJlt
25 5ll
24
t0
40
765
78 42
160
li
Sccond
tt4
-5 6u
tt
0-5
480
E
400
320
t:
t)
80
00
Figure 3-12
Table 3-4 with
face to be a fir
distance
77
The magnitude of the second critical distance relative to the

depth to the second interface varies according to the relative
magnitudes of V,,V, and V.. Use Table 3-4 as a dynamic table
and insert realistic velocity values for a case in which the values are reasonably similar in magnitude and for a case in
which the velocities are very different. Let the values in Table
3-4 represent the standard or general case for means of comparison. Do velocity ratios affect the critical distances to any
significant extent?
Although critical distance is irnportant, remernber that the relraction rxcthod utilizes
the first arrival at each geophone. Refer to Figure 3-ll. lt should be obvious that a wave
traveling path EPRQG will take longer than a wave traveling iiom E to G that is refracted
along the flrst interface. A wave refracted parallel to the second interface must travel a considerable distance at V., before overcoming its longer path distance and arriving prior to the
head wave generated by the frrst interlace. We can verify this impression by utilizing
RefractModel. Figure 3-12 illustrates a plot using the same velocities as in Table 3-4 and
layer thicknesscs /r,- 5 m and h.=)11 rn. Table 3-4 tells us that the second critical distance
for such valucs is 14.21 m, whereas Figure 3-12 denronstrates that it is approximately 60 rn
belore the hcad wave from the second interface arrives prior to the head wave from the first
interface.
Critical disl"ance is of primary practical impofiance when trying to recognize later

arrivals on a field seisntogram. Il'we have a reasonable rnodel for the subsurf'ace, we can
800
120
640
560
Second
criticrl
distrnce
4E0
E
E
i:
Hcld wave arrival Irorl
4DO
second intcrfacc
320
240
t60
80
00
45
55
65
75
Distance (m)
Figure 3-12 Travel-Lirle curvc based on a RefractModel plor using velocity values
in
Table 3-4 with lrr = 5 rn and 14=21], ,.. Note that thc first refraction from the second interface [o be a first arrit'al is located at a distance considerabJy greater tl)an the second critica]
di stance
78
Method
Chap'
Multiple lnterfaces
less than
calculatc critical distance and compare waveforms from geophones at distances
to recable
be
might
and great-er than thc critical distance. If conditions are very 8ood, we
would
then
points
ognii. events in addition to the ftrst arrivals. These additional data
cnablc us to refine our interpretation.
Analyzing a Second Field Seismogram
500
4s0
400
350
soo
200
150
let's interthat wc have increased our knowledge of the refl'acLion nlethod sourewhat,
approxinrately
pret another field seisn.rogranl. The seismogram in Figure 3-13 was rccorded
that the ellergy
tl.S t,o norlh of the lield location for thc seismogram in Figure 3-6' Note
l3
3 is reasonably
source gcophonc otliet and the geophone inten'al arc different. Figure
distinct as in Figure
stririghfforward to intcrPret except that the first breaks rrc tttr[ qtrite as
background noise
3 6. Also, lurplif ier gains on channels ll and 12 were quite high, so l-hat
Proceed in l-he same matlis evidcnt and ntakes it a little mclre difficult to Iocate first breaks.
ner as belore to gather your tirne-distance data'
your data'
Whcn you are rcady, either use RefractSolve or work by hand to interpret
a
was by hatnmer
Figure 3-1zl illustrates our prefered interpretal'ion. Errergy in this case
the graph' As the
blow on the surfhce, so the direct wave should pass through the origin of
points
2,3, and 4
ilata point from geophone I does not fit well on the line dctenlrined by
2, 3, and 4 are head
and lies below the line, we interpret it to be a direct wave arrival' Points
change in
wave arrivals f,rom a shallow inter-face. Points 5-12 define a line with a clear
do noL
points
case, the
slope and represent arrivals fiom a second interface. As is usually Lhe
maxito
the
fit any straight linc, so we visually fit a line that passes tlrrough or close
N.w
100
50
00
Figure 3'1u
mum number of poir

points for which first
Our analysis
re
layer, a thickness of I
of 3871 m/s for the tl
overlie Triassic arkosr
and the second interfa
Another remint
e*actty
order here. Fitting lin

Therefore, you shoul,
much variation is pre
This takes only a few
you with a good senst
2
)
adopt.
5
6
1
MULTIPLE INTERFACES
l0
l1
t2
--->l nls
-5
80 ms
Although the tkee-la

might ask what a trat
two interfaces and vel
those just discussed il
we reproduce Eq.
notice
Figure 3-13 Field seisrnogram from the Conneclicut Vlrlley' Massachusetts Geophone
,.-". o," labeled I I2 Thc first geophone is located 3 nt fiom the cnergy sourcc Thc geo
directctl
phone intcrval is l0 tr First breaks lor each trace re indicated by a downwud
arrow Timing lines ire at 5 ms intervals The record enconlpasses
3-
ICX)
nls'
progressive
-r
Multiple lnterfaces
79
500
450
400
350
300
I 250
200
100
50
00
13
23
33
43
53
63
73
83
93
103
113
Distance (m)
Figure3-14
Prct'cred intcrpretation of time-distancedatataken liorn Figure3-I3.
mum number of points. In fitting this line we also could disregard or weight less heavily
points for which first breaks are not especially clear (e.g., trace ll).
Our analysis results in a thickness of 1.4 m and a velocity of 387 m./s for the first
layer, a thickness of 10.0 m and a velocity of 1820 m/s forthe second layer, and a velocity
of 387 I m./s for the third layer. The field site is in an area where glacial sands and gravels
overlie Triassic arkoses. We therefore interpret the first interface to represent the water table
and the second interface to represent the contact between sands and gravels and bedrock.
Another reminder about the utility of computer programs such as Refractsolve is in
orderhere. Fitting lines to the field data from Figure 3-13 is notcompletely unambiguous.
Therefore, you should attempt several different interpretations in order to determine how
much veriation is present in the subsurface configuration computed for each intelpretation.
This takes only a few more minutes with RefractSolve (or a similar program) but provides
you with a good sense of how much unceftainty is present in the final geologic solution you
adopt.
MULTIPLE INTERFACES
Although the three-layer case represents a configuration with multiple interfaces (two), we
might ask what a travel-time curve looks like for a situation in which there are more than
fwo interfaces and velocity increases tbr each layer. Must we proceed through analyses like
ttrosejust discussed in detail for a four-layer case, a five-layer case, and an n-layer case? If
we reproduce Eq. 3-14 and 3-33 with the terms slightly rearranged, you can't help but
notice a progressive similarity.
l-irne
- v2r
2lt,(Vr2 - V,')'t'
V,V,
(3- r 4)
T
L
80
trtne
x
= -vl *
2 h,
(vt2
v,' )tl'
Chap. 3
Dipping lnterfaces
(3-33)
V.V,
(a)
different form using a sut.mnation symbol and

We can rewrite these equations in a slightly
(thus, the trumber of interfaces is n - l)'

letting the letter n represent the nurnber of layers
for a wave that is criticatly retiacted at the n - I
The equation represents the time o[ travel
interface.
t * 1-t n
- V,,
V,, Z,
a
tlme,,
nl
(3-3e)
[-
(b)
'
20m
3-33 and therclore represents the cquetlon

Equation 3-39 has the satne form as Eq.3-14 and
ofastraightlincwithanintercept,,,,,"ndwithslope=llV,,.Basedontlrisinforrllation,we
witl a greater
geology with n layers' each

can say that- a travel-titne curvc for'h'subsurlace
segments witlr eaclr progressivc segment
line
ll
straight
velocitl, than the otle above, has
3-39 can be uscd to develop an equahaving a lesscr slope than thc preceding one. Equation
for thickness /t,,,. Velocities are detertion that in.o,po,"t", intercept time /,, and solves
are calculaLed using intercept tirnes
mined from the inverses of the n slopes,land thicknesses
and beginning with layer 1 and proceedin
Equations for thickness are not dev
already has been presented' Also' tbr mo
especially commoll for morc than tlrrec I
t\
(c)
E
E
t:
,"oron. for this, some of which wc will
velocitiesofsaturatedsedimentsandthevelocityContrastbel-weetrtherrrandbcdrock.This
and
the preceding exercise: the watcr table

results in two obvious interlaces that we saw in
the sedirlrent-bedrock contact'
itself very nicely to colnputer
Fortunately, the iterative nature of the process lends
RefractSolve'
by
utilized
one
the
is
solution. The process just describcd
Figure
DIPPING INTEBFACES
the subsur{ace consisted only of horizontal interfaces
vety interesting or challenging' In our prevtbetwee-n rocks and sedimenis, it would not bc
horizontal inter{aces' We all can visuous discussion of the refraction nlethod, we assumed
inclined or are irregular' First' let's
are
interfaces
al.izc numerous natural situations in which
Although
it
would be sirr.rpler
3-15
(a)
the interfrce ol 5(
depLh to that in (a
Travel time curve
if
examinethecaseofasirrgledippinginter{aceandseehowwemightdealwiththisfirst
dcparture lrotn our idealized world'
Analyzing the Problem

Tobegin,studyFigure3-l5.Thegeologyin(a)and(b)isidenticalexceptthattheinterface
travel-tirne
conducted at (a) produces a

in (b) dips at 4" down to the left. A refraction survey
the same type of plot that a survey
exactly
is
curve with two straight-line segrnents, which
to distinguish an inclined i.terway
no
co,ducted at (b) would produ.J. There is absolutely
face from a horizonta

time curves. How thet
lem, part (c) does cont
IlV, as you alread'

less than the prior val
We refer to this as an,
Our clue lies in
subsurface configurat
each geophone is earl
to
interface. The reason
''...'.....-
Dipping lnterfaces
(c)
81
I 60
140
100
of,n
tr
i:
60
2a
40
60
80
100
120
Distance (m)
l-igure 3-15 (a) Horizontal intcrface at a depth ol 20 rn with vel<rities above and below
thc interlace ol 5(X) rrVs and l-5fi) m./s, rcspectively (b) A dipping interface with identical
depLh to that in (a) at thc sitc o1'thc hlmmcr intpact and with identical velocities to (a)- (c)
Travel-tinre curves [or (a) and (b)
face from a horizontal interface based on this type of field survey and the resulting travel-
time curves. How then should we proceed? Although Figure 3-15 does not solve our problem, part (c) docs contain a clue. The'slope ofthe line Iabeled "horizontal interface" is equal
to llV, as you already know. Howeveq the slope of the lined labeled "dipping interface" is
Iess than the prior value and therefore gives a velocity that is greater than the true velocity.
We refer to this as an apparent velocity.
Our clue lies in the reason for the different arrival times for the head wave in the two
subsurface configurations. Notice in Figure 3-15(c) that the arrival of the refracted wavc a[
each geophone is early relative to similar geophone positions in the case of the horizontal
interface. The reason for this is evident when comparing (a) to (b). All rays in (a) retuming
82
Method
Chap'
all are longer

to the surface from the interface have identical path distances to traverse, and
ray retumeach
right,
to
the
upward
than those in (b). Because the surface in (b) is inclined
to its left'
immediately
ray
the
than
ing to the surlace through V, traverses a shorter distance
and
the
horizontal
from
Thus, the gap between critically refracted wave trat'el-tirue,s
(c)'
course,
Of
in
slopes
different
the
inclined interlhces increases up-dip. This accounts for
would a travel-time
this observation <joes not solve our problem, but you might ask, What
the lcft edge ol the
from
sourcc
curve look like if we reversed the location of the energy
difl-crerlt fron'r
will
appear
it
that
predict
diagrarn to t5e right eclge? You shoLrld be able to
for thc dipcurve
to
thc
cotnpared
the horizonLal inter{ace casc. How will it appear when
ping interfacc that wc alrcady exanrined?
energy
Before attacking tliis qucstion, rve flrst need to compare thc results ol- an
such a
illustratcs
3-16
Figure
source position switch lbr the case of the horizontal interlhce.
placed
as
is
sotrrcc
energy
an
and
situation. Geophones arc distributed along the surface,
position,
the
this
fiolr
arc
recorclcd
shown at the lcft edge o1'the dia-srant. After travel titlcs
Dipping lnterfaces
energy source ls mov

of the diagram, and I
verse because it is tl
because the energy st

pose of Figure 3-16 it
cal from either energ'
face distance from
thr
salrre as the surface d

tances for both refract
in travel-time curves I
It is customary
Figure 3-16. When tl
other. Note, however,

lower lett ol the grap
that because the curvt
segrnents are the same
IR
There is one
ac
critical relationship to
it takes fbr wave ener
the energy source for
reverse traverse as
Io
the lorward energy so

the same. This equiva
evident because the u
the reverse and forwe
with the energy sourc(
Now we can rel
look tike i[ we reverst

to thc right edge? Fi1
reverse traverse is ad
160
140
reverse traverse are n(
120
data for both a tbrwt

inclined interface is
100
;-,
opposed to Figure 3-l

The reason thal
,60
Geophoncs lFand
as are geophones 2F
40
refracLed wave. If we
greater than a similar
shown on the travel-ti
20
o
Distancc
(rl)
Figure 3-16 Conclation o1'a travel titre cutve with rvave paths to geophones at equal
disLanccs
li
ft
otn an cnergl' sottrcc for a forward and a re\ crse Iraversc
the reverse traverse I

between the forward
increases. Whai is tht

reciprocity. Exarnine
-J
Dipping lnterfaces
83
energy source is moved to the other end of the geophone string as shown at the right edge
of the diagram, and times are again recorded. The first position is called the forward traverse because it is the initial position. The second position is termed a reverse traverse
because the energy source position is reversed relative to the first traverse. The main purpose of Figure 3-16 is to demonstral-e that path distances and, hence, travel times are identical from either energy source position to a given geophone interval. For instance, the surface distance frorn the energy source forthe forward traverse to geophone lF(40 m) is the
same as the surface distance from the reverse energy source to geophone lR. The path distances forboth refl'acted waves are identical, as are their [ravel times. Such geometry results
in travel-time curves for each traverse that are identical.
I1 is custornary to plot data fror.n forward and reverse traverses in the manner of
Figure 3-16. When the curves are ploLtcd in this rnanner, they are mirror irnages ol'each
other. Notc, however, that if the data frorn both traverses were plotted with the origin at the
lowcr lclt of the graph, l-he curves would fit precisely on top of each other. Also observc
that because the curves are idcntical, the slopes of the respective direct and refracted wave
segnlents are the sarne, as are the inl"ercept tintes.
There is one additional inrportant feature o[ this travel-time curve that illuslrates a
critical relationship to rcmcrnber when we invcstigate dipping interfaces. Refer to the tintc
it takcs lbr wave encrgy that is refractcd along the interface to travel frorn the position of
the energy sourcc lbr the forward traverse to the position o1' the energy source lbr the
reverse traverse asTo.To is the travel tinre fronr the reverse enelgy source to the position of
thc forward energy sourcc. T,.and l* must be equal because the patlrs ol'travel are prccisely
thc sarrre. This equivalcnce is known as tcciprcc'ih,. Thatreciprocity holds in Figurc 3-16 is
cvidcnl because l-hc upperr.nost horizorrtal line on thc travel-l.ime curve indicatcs that both
the revcrsc and fbnvard hcad wave curvc scgments intersect the vcrtical axes (coincident
with thc energy sourcc locations) at 160 ms.
Now we can return to onc of our original qucsl-ions: What would a travcl-tinte curve
look Iikc if we reversed the location ol'thc encrgy source lrom the left edge of the diagran'r
to thc right edge? Figure 3-17 illustrates the sante situation as Figure 3-15 exccpt that a
revcrse lravcrse is added. As you likely guesscd by not, the curves for a forward and
reverse traverse are no[ thc sanre when an intcrlhce is dipping. Therefore, weulvuays collcct
lor both a lbrward and reversc lraverse, as this is thc only way to detenlinc il- an
inclined intcrlacc is presenl. l[ the cr-u-ves are not syrnntetrical, as in Figurc 3-17 (as
data
opposed to Figure 3-16), thcn the interface is inclined.
The reason that the curves arc not symrretrical should be clear frorn Figure 3-17.
Geophones li? and 1R are the same distances from their respective encrgy sources (40 rn)
as arc gcophones 2Fand 2R (60 m). Exarnine the path traveled by thc irrcident and critically
refracled wave. lf we rneasure the distance traveled in the V, ]ayer to geophone lF, it is
greater than a sinrilar measurcrnent fbr geophonc 1R. This produces a grcaLer travel tirnc, as
shown on Lhe travel-time curve. This greatcr travel time for thc lorward fravel'se rclat-ive to
the rcverse travcrse holds fbr all source-geophone positions except one. The dilfbrential
betweett the forward and reverse times decrcases as the sourcc to geophone distance
incrcascs. What is thc one case for which the times are eqLral? Recall the requirernent of
rcciprocity. Exarnine Figure 3-17 and deterrline to your satistaction that I. and I*, when
Dipping lnterfaces
Chap.

2F 2R
t
20
rl
Forward
'll
160
140
^.
120
1OO
rrl
:, -1
,ol
Rcverse
Rcciprocal times
--l --
lr^ _y
E
'a
'td
80
Intercept time greater lor

traverse with encrgy
60.
source a( down-dip (r1)

portion ol interface
40
Figure 3-18
that rcciproca
pornts to a ge
l:rbclcd lfi)F
Distance (m)
Figure3-lTConelationol.atrrr,cltitnccurvcwithgcophoncpositionsabor,easinglc
rlipping intcriacc. Thc purpose
ol
thrs rJiagram is to denlonslratc thc diffcrcnt path dis-
:l;::lranivaltimcslorgeophoncslocatedx(irlenticalollsctsloraforrvardandret'erse
again these particular
defined as in our initial discussion ol-reciprocity, must be equal. once
the vefiical axes at
intersect
segl]lents
path differences are the same, and head wave curve
142 ms fbr both forward and reverse plots.
2
In our discussion of Figure 3-15 we pointed out thal- the apparent velocity for layer
segtraverse
reverse
the
from
a
velocity
was greater than tl're true velocity. If we calculate
if dips
ment, we discover that the apparent velocity is /ess than the l-rue velocity' Actually,
true
the
to
equal
nearly
very
is
are low (<10') the average of the two apparent velocities
velocity.
interBefore we begin a derivation of a travel-tirne equation for the single dipping

in
are
sununarized
in
mind
to
keep
face, pcrhaps a brief review is in order. The main points
Forward and reverse travcl-time curves are not symmetrical when a dipping
inlrfac" is present. Velocity V, for the layer above the interface is determined in the stanReciprocal times must
dard way. V, values fbr the forward and reverse traverses must agree.
The average of V2F
give
Vr.
be eqrai. lnverses of slopes for the head wave arrivals do not
to rn, and nt, fol'
equated
unO izr^ approximates V, when the dip is low' Slope values are
Rigurl:-ti.
lbrward (F) ar
lowing the convention

slope. We can determi
less for the traverse w

vertical distance to thr
everything is importan
ln order t
type, entr
into
Refr
when yot
layer situ
rules you
lar relatio
ventron
represent
atrve ang
the right.
Dipping lnterfaces
85
Slopc
= I /"vzx
Reciprocal tirnes
nrust be equal
-r1
G
l()(lF
t(nR
120
/"v2F'= nt
tr
tl
e0
,t), ----7
60.
40
20
0
020
./
tl\
Slopc = l/V,
60
lroru,l rd
I)islarrcc (nr)
lntcrccpr tirrc grealcr Ibr
traver se with encrgy
sourcc at down-dip (r/)
porlion ol interlucc
<-
Cro
120
Rcversc
lntcrccpt tirrc lcss lirr

trrvcrsc with cncrgy
sourcc ilt up-dip (ir)
porlion ol'intcrlilcc
Figurc 3-ltl IlllPonant I'cattttcs <11'rt travcl tinrc cur vc lirr a srnglc dipping
interlrce- Notc
that recillrttirl tillles ntttsl bc cqLrrl []owcvcr, thc tiltcs l-ror1 forrvard and
rcvcrsc sho(
lx)ltts to a leol)l)onc at a givcn distancc arc r)ot cqurl This is illustrated b), thc .lt_ows
labclcd I (X)F and I 00R, which tlcsignltc tt avcl (inrcs to qcoplrorrcs locir(cd i (X)
nr I r 6rrr thc
Iblward ([r.) rnd rcvcrsc (R)shots
lowing the convention for thc equation o['a straight line: r,= nt-r + b, u,hcrc,1
designates
slopc' Wc can detellnillc thc dircction ol clip ol-thc intcrlace bccause the intcrcept
tirlc is
less lbr the traverse with its enL.rgy s()Lrrcc .rt the up dip porlion
o1-thc iltcllncc (wherc thc
vcflical distlrrlcc to the intcrlace is lcas(). Although this secrns Iikc a lot to kecp
straight,
evcrlthing is irnportant whcn interpreting curves rclatcd to dipping intcrtrccs.
ln order to become more adept at working with curves of this

type, enter several configurations of velocity, depth, and dip
into RefractModel. At first work only with two layers, but
when you feel at ease with these curves, attempt some three_
layer situations with both interfaces dipping. Following the
rules you already know, you should be able to work out simi_
lar relationships for this more compricated situation. The con
vention used in this text for dip angles is: a positive angle
represents a surface inclined upward toward the right; a neg
ative angle represents a surface incrined downward toward
the right.
86
Method
Chap'
Dipping lnterfact
3
In examininl
Derivation of a Travel-Time Equation

interface essentially is the
The derivation of a travel-time equation for the single dipping
a few more terms' and
has
equation
initial
same as for the single horizontal interface. The
other-wise everyeffort'
more
bit
a
requires
reducing the equation to a more compact form
thing is farniliar.
interface. First, we
tion (we are "shoot
ThesymbolsemployedinthefollowingderivationareidentifiedinFigure3-19.The
j, ate
equation time, rePresen
position. The intercePt
quantities h,, and
and sincc
frorn the derivative of

located at the up-dip portion of thc interface,
of that used in most exploin /r,,. You snouta bc aware that this convenLion is the opposite
is easier to keep
terminology
this
that
textbooks. However, we belicve
ration gcophysics
source'
straight if 1'ou relate the subscript to the position of the energy
uses ft in the same manThis
inter{ace'
the
Note also that ft is the vertical distance to
(as well as many
RefractModel
and
ner as in the case ol'horizontal intelfaces. Refractsolve
thickness in the
is
This
thickness.
as
other computer progrants) refer to this quantity
_not
explorationists'
field
by
used
sense
the
stratigraphic sense but is usually thickness in
when the
computed
bc
can
always
Thickuess as rneasured perpendicular to layer boundaries
*
Up
PG =
.rcos0,
This givcs us the bl
tirnc,
" -
J't
1', cosO,,
Wc citn rcduce this

in our prcviotrs der
dip is known.
:!
Since MN = PG -
rrip
Down-diP
Finally, we arrive
Our next step is to

dip. The equation r
again, we have
These two equatior

horizontal single-ir
a ritual, is to deter
the derivatives of
vt'Y,
F-igure3-lgDiagrarnillustratingsl,mbolsusedinderivittiotloftraveltilneequal.ionlora
singlc dipping interface
1(tirnc
tlr
Dipping
lnterfaces
A7
In examining Figure 3-i9, recall that 0,, is measured frorn the perpendicular to the
interlace. First, we tackle the case when the energy source is located in the down-dip position (we are "shooting" up-dip).
.EMMN
+-NG
tlnte, =
"vtv)
EM= lr
cos
and
since
lu =
(3-40)
vl
NG
cos
0,,
0,,
rsinB,
-EP-i,ti,, - xstnB
NC _
J,t
cos 0,,
Since MN =
PG
PG-
-rcosD
AM
NC, we determine the relevalrt' qtrantities
AM - j,, tttnl,,, and NC = J, t,.,n0,, - (,t,
rsirrB)tan0,,
This givcs us thc basic travel-tittle cquation
tirttc,
" -
rcosp
J'r
V, cos
- j,, tanl,, (;,, -
rsinB)
tan0,,
* i,r I(
V.
9,,
.rsin0
cos0,,
1-t_41.)
Wc can recluce this equation to a sintplcr lornr by using thc Lrigonomctric itlerltiLies we tlscd
in our prcvior-rs derivatittns and by relnctnbering
sin(0,, B) :
sin(6,, + 13) -
sin0,,
cosB
cos0,, sinB
cos6,, sinB
sino,, cosp
Finalty, we eLrrive a1
tirne,,
UtrL+f
sin(o,,
(3-42)
P)
Our next step is lo tnove thc energy source to the up-dip pa( of thc interlace to shoot clowndip. The equarion we arrive at looks remarkably similar to Eq. 3-41. Aficr sirnplilying once
again, wc have
rirne,,
- rtrr+f
sin(e,,nB)
(3-43)
Thcsc two equutions are not only vcry sin.rilar to each other but are not as dill-crcnt lrorll the
horiz-6ntal si6gle-interlacc casc as you rnighL havc predictcd. Our nexL step, by tlow almost
a ritLral, is 1o cleterrnine thc slopcs of the straight lincs that thesc equations def inc. Taking
the derivatives ofEq. 3-42 and 3-43 results in
d(time,,)
dx
sin(0,,
-0) and r/(time,,) dx
Vt
sin(e,,
Vl
+0)
(3-44)
88
Method
Chap'
Multiple Dipping
3
and intercept times, we

Following the temrinology established in Figure 3-18 tbr slopes
Noting from Figure
single dipping inter
begin by rewriting E,q. 3-44 in the forrn
nt,,
sin(o,,
-\j1
in(o^+0)
and ttt, = -:i--
-B)
(3-4s)
V, if we know 0''' Of
Renrembering that sin 0n = VtlV, we see that we can deten't-titlc
+ and - signs in Eq' 3thc
Noting
interface.
the
of
Oip
B
course. we also want [o tro* it',e
p
0,,- Therefbre' we put
and
both
1br
equations
slope
4-5. wc see that we can solve these
these cquations in tl're lbnn
si, (v'n'") (3-46)
ei, 0 = sin '(v,nr,,) and 0" + 0 =
Becausc
tl1,,,
'
and
we solve for 0,..:

tl7,1,atld V, are determined directly from the lravcl-tinrc Curve,
0,, = sin '(v,nr,,) + p and 0" = sitr-' (v""")

20,, - sin-' (V,rn,i ) + sin '(V,"',)
{)
MULTIPLE DIPP!NG INTI
and, finally,
(3-41)
ei,
We now turn to th(
one interlace is pret

technique follows tl
Similarly, wc dcterrnine the relationship fbr B using
F-0,, -si1 '(v,"r)
and
B--o''+sin
plex. You would gz

ply presenl" the resu
(q''")
details.
to arrtvc itt
2tJ
= sin-'
Travel-f ime Equi
(V,rrr,,)- sin'(V,rrr,)
A clezr
and
o
sin' (t\nr,,)
- sin' (U,r,,,)
(3-48)
exposition
faces is presented L
changed to be cons
from the perpendic
and refraction. An1
an cquation that provides a
Ijou,cvcr, we are nol yel- quite I'inishcd. we still need to dcrive
portions of thc interlhce' At this
solurion lbr thickness, aL both Lhe up-dip and down-dip
we can express Eqs'
point we know the critical angle ancl- the velocity of the first layer, so
3-42 and 3-43 in terms of intercept times:
other subscripts fol

The cquation
(3-4e)
j,,
cos 0,,
vl
(3-s0)
wheren=thenum
tion of the procedr
lntercept times
are
-Multiple Dipping
lnterfaces
89
Noting from Figure 3-19 that cos B=irllt* and cos F= ju/h,,, we complete our task lor the
single dipping interfacc
t,,,v,
2 cos
(3-s I )
0,.
t,,,V,
,
J'
- 2.r.
(3-s2)
,,
'
11
.t,t
l-5 l)
cosp
-
and
"
J,,
(3-54)
cosB
MULTIPLE DIPPING INTERFACES

We now turn to thc development of travel-time equations for a situation where more than
one interface is prescnt and at least one interface is inclined to the horizontal. Thc geornetric
technique lbllows the sarre pattem we utilized several times alrcady but is nruch rnore complex. You wcluld gain relatively little new knowledgc by pursuing this in dctail, so we simply prcscnt the results and cr.rcourage you Lo consult original sourccs if you are interested in
details.
Travel-llme Equation
A clcar cxposition o1'thc developrrent of travel-tirnc
equal-ions fbr several dipping interfaces is preserrtcd by Adachi (1954). Figure 3-20 is based on his work, but the synrbols are
changcd to be coltsistcnt with those used in this tcxt. Note that angles 0, and @- arc mcasured
from thc pcrpendicular to t-he various interfaces and therefore represent angles of incidence
and refiaction. Anglcs u and d. arc measured frorn the vertical. Thesc synrbols as well as
othcr subscripts follow l-he sanre usage as in the case of thc single dipping interlace.
Thc cquations developcd by Adachi are
I
n-l
h,t
-lcosd
Vi'
L(.n.a
V,'
+ cosuIt
(3-55)
+ cosa
(3-56)
where n = thc number of interf aces. Solutions to thcse equations are obtained by a combina-
tion of the procedures for the single dipping interface and multiple horizontal intertaccs.
Intercept timcs are determined for each straight line on the travel-time curve. The same
Chap. 3
90
Multiple Dipping
veying. In the major

are sufhcient for our
\rrl'{
Analyzing Field S
You now have the
order to review procr

ine a representative s
Figure 3-21 ill
h,,
fron-r a foru,ard trave

ters are noted in the l
channel capability, lc
phone segments. Her
These records
D,+
are fairly typical. Fir
soufce are reasonab

times very difficult
ularly efficient
RefractSolve require
ignored when drawir
you know which firs
during your interpret
time curves is only a
dal"a to determine to
line fits.
v4>v1>\/">vl
Figure3.20DiagrarrtiIlustratingsymbolsusedintraveltirnccquationlbrrnultiplcdip
Consult the tra

and lR should have
ping intcrfaccs (altcr Adachi' 1954)
critical angle. These quantitics

techniques are applied to determine slopes, angle of dip, and
related to the shallowest interthen arc used to solve for vT ltr,,and fir,. once all quanl-ities
and the process begins again'
face arc k,own, Eqs. 3-55 ina i SO a.e'e*panded with i = 2,
task, and RefractSolve is
computer
appropriate
especially
As bcfore, this technique is an
slightly diflerent but

curve is dcfined onl'
Here is a case where

layer thickness are lil
malion probably is
interpretations due tc
blscd on lhesc cquatiorrs.
ln order to demonstrate the applicability of Eqs' 3-55 and 3-56
for all cases, rewrite them for the case of B = 0 and n =
t<
enerl
1'
and compare the result to Eq. 3-2'
or two dipln practice Adachi's equations usually need to be expancled for only one
of the refracping intcfaces. The rnany complexities of the subsurface and the limjtations
to more
approach
this
apply
to
iion method co,bine to make it generally irnpracticable
surrefraction
of
shallow
goals
cornplicated situations. On the positive side are the typical
You
me
Figu re
are in tl
Do not
arrival
fact, yo
Multiple Dipping lnterfaces
91
veying. In the rnajority of cases, the objectives of the survey are such that these equations
are sufficient for our purposes.
Analyzing Field Seismograms

You now have the knowledge to interpret the majority of routine refraction surveys. In
order to review procedures ar.rd to practice with more complicatcd seismograms, let's exatnrne a representatrve sun,ey.
Figurc 3-21 illustrates four ficld seismograrns of the wiggle-trace variety. Two are
from a frrrwald (raverse, and two arc from alr accompanying reverse lraverse. Field paratretcrs are noted in l-l-rc figure capticln. Bccause the instn-rment used in this survey has only 12channcl capability, long traverses requiring 24 or 36 geophones must be recordcd in l 2-geophonc segrnents. Heltce, tltc need fbr twcl records for the forward and reversc travcrses.
Thcsc rccords are instructive for a var-iety of reasons. First, conditions at the field site
are fairly typical. First brcaks on the geophones near thc Bulfalo-gun ("sholgLrn") cncrgy
source are reasonably clear, bul- first breaks on the morc distant geophones arc sonrctinres vcry difllcult to identify. This is not at all unusual at localities that do not havc parlicularly ellicicnt energy-coupling chzuacteristics (dry sand). Conrpul-cr progranrs such as
Refi'actSolve require first-break tin.res for every geophone, but questionablc l"irtrcs can be
ignored when drawing travel-tinlc curves. You therefore should kcep careful rccords so that
you know which first-break tilnes you have nrore cont'idence in. Thcse should guicic you
during your intcrprc(ation. Once again, remenrber (hat obtaining a solution frorn the lraveltimc curves is only a lnaltcr of a fcw nrinutes, so be surc to try several possiblc (its to your
da(a to deterruine to what extcut your computed subsurlace nrodel varies depending upon
line Ilts.
Consult the travel [irne curve based on or-rr first-break picks (Fig. 3 22). Segurents lF
and lR should havc thc sarnc slopes if the trrst layer is horlogencous. Thcir slopcs are
slightly different but lead to only very slightly diff'erent vclocity values. Unlbrtunatcly, cach
curve is dclined only by one point, so it is difficult to know which line is r.ttore rcliablc.
Hcre is a case where experirncnting with interpretations could provide useful inlbrrrtatiort. [1.
laycr thickness are little afl'ectcd by thc possible extrerlcs in thc V, values, additional inlbrntation probably is not warranted. Howevcr, if you cannot tolcratc the eror in thickness
inlcrpretations duc to t-his uncerlainty, the obvious solution is to place a string ol-gcophoncs
You may wish to analyze the seismograms from

Figure 3-2'l independently. Our first-break picks
are in the RefractSolve file:
ffi
K826 8802 1-24
Do not hesitate to substitute your own interpretations of

arrival times or to draw different fits to the data points. ln
fact, you are encouraged to do so.
Method
Multiple Dipping lr
Chap' 3
(b)
(a)
I
1
l0
t0
ll
ll
1Z
l2
lOnrt
---l
Il0nrs ---l s t-
5 l-<-rns
ms
l3
t3
l4
t4
t5
t5
t6
l6
t7
t'7
l8
l9
l8
l9
20
20
2t
21
22
22
23
23
24
24
--->l
l<-
[-igure3-21(.r)FicltlseisllogramlronltheConnecticutValley,Massachusctts.Forward
thc
arc labeled 1-24 The first geophone is located 3 nl from
travcrse Geophone trtrces
energySoulCe..I'hegcclphoneintervalisl0m.Firstbreakslbrcachtracearcindicatedbya
*-r*rrddirectcd-arrowT'irninglinesareat-5msintervalsTherecordcncompassesl8-5
in order to record more direct wave

with a smaller inteNal at the beginning of each traverse
tolerances'
arrivals and hence to determine velocities to closer
AlthoughSomescatteringofpointsexists,segments2Fand2Rarereasonablywell
first interface is essentially horizontal'

defined. Because ry'r and ,,., i almost equal, the
of reciprocity, even though the
Note that these two ,"g-[}l,, must satisfy the principle
I 85ms--->l
ms
Figurc3-21 (
[raverse
intersections of their
ing segments 3f
in Figure 3-22.
and
Thest
zontal dotted line joir

less than 1.,.,. This in<
Iowest pafi-is
locater
however. that if
a thir
-Y
Multiple Dipping lnterfaces
93
(b)
I
2
,
4
5
6
7
8
l0
ll
t2
I|0rns ---l
-5
IS
l<--
I| 80
nrs
l3
t4
l5
I6
t7
l8
l9
20
2t
22
23
24
I 85
|
rns
---> -5 <IS
IlflIJr,
t ,i,i'
@ontinued) (b) Reverse traverse All paramerers the sanre as for thc lbrward
intersections of their extensions with the veftical axes are not shown. The data points def rning segments 3F and 3R have considerable scatter, but the majority fit on the lines as drawn
in Figure 3-22. These two lines satisfy reciprocity, as indicated by R, and R, and the horizontal dottecl line joining these points. However,3Fand 3R have different slopes, and /,r, is
less than 1.,r. This indicates that the second intcrface is dipping and that the up-dip or sliallowest part'is locatcd at the energy source position used for thc forward traverse. Note.
however, that if a third interfacc is present and is dipping, we cannot use this simple check
94
Chap.
The Nonideal Sub
THE NONIDEAL SUBSUI
',\
p\
120
108
96,
'
/o'
3"*"t'*
*.i.r..r..l-
t. . - . ),
I
I
i:i4 0
E
i:
ti,.4
?2
.<-
After the previous k
sume that we have c

situation. Fortunatel
cially if you
are not
information such as
6,0 ll
ly common and siml
4AO
One that come

date have assumed I
360
,,,,
.
24.O
it, ------->
'
1Z O
.<-
the one immediatell

horizontal or incline
00
123
Iror
tvrrLl
l''igrrrc
--r>
I)istancc
3-22 Trrltcl tillrc cttrvc
(rll)
hascd tltt delr
183
of velocity increasin
low subsurface invet
2u3
which an interface
<-
Rcvcrsc
velocities can result

tions from a shallow
ltottl []rgurc 3 2l
layer, and quantities
afl'ects Lhc intcrcept

to detcrrnine its dirccti6n g1-dip lrccause the dip o[thc scconcl intcrlacc
timcs 1br the third itttcrl'acc
(Fig.
If you exarninc thc structllr.c contputcd li-onr this positioning ol'curvc segnlents
studl'
ablc
to
arc
3 23), you will scc thirt ouf analysis is corrcct. we hopc that b)' llow i'ou
This is a valuable
curves-of rhis contplexity arrcl tr.t arrivc at thc satrle gcnc|al conclusions.
wltile still
ability, for it enablcs ),ou Lo plot rluta and tltakc a vcrl gencralized intcrpretation
to gatharld/or
in thc fleld. This is thc n]ost appropt'latc trrrc [o try arrd restl]ve uncertllintics
er dal-a 1br a lllore ctltllplctc lllalysis.
00
Hidden Zones: Th
In all our derivations
iayer imrnediately at
many routine shallor
likely you can
creat
layer above. In
such
interface is present.
Three-layer sit
160
case that a surface lr

through the higher v
deeper layers. The b
:
o
space, these are the
tion does not occur, t

direct wave, the tral
107
21
fident you can think

or in combination, r
Lo(
will go undetected
3
Simjlarly, the deepet
This situation produ,
48O
occurs at the second

interface will be con
Let's now con
higher velocities. Ba
=1
42
l-()I WlIal
---------->
<-
Revet-se
l'igurc 3-23 Struclute computcd by Relrlctsolve ftolt curve seglnellts illusLrated in

Figurc 3-22
ity layer in the midd
The Nonideal Subsurface
95
THE NONIDEAL SUBSURFACE
After the previous lengthy introduction to the seismic refraction

method, you may well pre-
sume that we have covered most contingencies.

Unfortunately, that is nowhere near the true
situation. Fortunatery, what you have leamed is suffrcient,o
-r". a great many cases, espe_
cially if you are not concemed with detailed subsurface structure
bulneed only one item of
information such as depth to the water table r r bedrock
surface. Can you think of some fairly common and simple subsurface configurations that we
might add tL our repertoire?
One that cornes to mind immediately
models to
date have assumed lateral homogeneity. An
of velocity increasing witn aepttr. Atthough
low subsurface investigations, it often is possi
ower than
the one immediately above it. All our derivations
also assume planar interfaces, whether
horizontal or inclined- What happen
lted? Can we deal with cases in
which an interface has undulations?
nations of layer thicknesses and
velocities can result in critical refract
tions from a shalow rayer. rf this happens, no dara p.i",,
layer, and quantities computed from such a travel-time
curve will contain errors. we,re confident you can think of other situations we should
discuss, but those mentioned, either alone
or in combination, cover much common subsurface geology.
For the sake of tirne and
space, [hese are the ones we analyze in the followrng
pages.
th
;;iT:ffi
;ll["lli,h?T"tfi:J[:ffi;
Hidden Zones: The Low-Velocity Layer

In all our derivations to this point we assurncd that each
layer had a higher velocity than the
laycr immediately above it. Although we mentioned
that this often is G" .or", espccially in
many routine shallow investigations, there is no guarantee
that this assumption holds. Most
likely you can create reasonable models for which a
layer has a lower velocity than the
case
car refracduc to the

ce that an
reria
_time
Three-layer situations provide more combinations
for analysis. In the sornewhat rare

case thal a surface layer.has a higher-velocity
than those below, the direct wave traveling
through the higher velocity medium always
will arrive before refractions coming fronr thc

will not "see" l.hese layers, and their presence
information is available (e.g., from weit togs;.
deeper layers' The basic refraction method
revealed in a configuration in which V, . Vr,

V.,.
-time curve for V, and Vr,but no critical refiaction
h thc third layer is undetected, the dcpth to thc first
Let's now cot.tsider the effect of a low velocity layer

sandwiched between layers with
highcr velocities' Based on the preceding discussion,
we can understand why the Iow-velocity la-vcr in the middle of the stack will go undetected.
But what is the appearance of a trav-
96
Method
Chap'
are introduced into our analyel-time curve for this arrangement, and what types of errors
refraction occurs at the first
sis? The ray paths in Figure 3-24 illustrate that no critical
than the angle of incidence
interface, because the angle of refraction must be smaller
so a plot of time-dis(Snell,s law). critical refraction does occur at the second interface,
(Fig' 3-24)' The
segments
two
hnce data for this model produces a travel-time curve with
the first interat
refracted
is
first segment arises from iirect wave arrivals. As wave energy
for critical
angle
the appropriate
face, some energy will arrive at the second interface at
segment
line
second
the
slope of
refraction. equaiion 3-33 still is valid, and therefore the
information
only
the
was
to 11v1. If the travel-time curve in Figure 3-24
must be
low"qrut
the person would have no evidence for the presellce of the
interpreter,
an
available to
layer
the
v,
of
top
the
to
a depth
velocity layer and would compute a two-layer model with
The Nonideal Sut
of 25.5 m. This dep

the computation. B
small?
Note t
ments
segme
a curv(
WAVE
the
t/r-
segme
velocit'
120
100
InLercept tinre = ri
RefractModel
to create a travel-tin
E
essary information.
AN
SfoP" =
E
'a:
\__ -
Slopc
= l/1
Figure 3-24. The de

Thus, the depth base
the Vr-layer. You c<
dynamic table. Entr
RefractSolve, use Rt
'lU'
20
20
40
6,0
80
100
120
140
Dislrncc (nr)
TABLE
3-5
Refrac
Layer
l0
t0
l0
.10
-50
60
10
It0
90
l(x)
V7>V1 <V1
Il0
[0
Figurc3-24AplotofLime-disl.ancedatafrornTable35conelatedrvilhadiagranlol.the
y'
wave geller:Itcd at the yl
to calculate thc data' Note that therc is no hcad

ol the V' - Iayer on lhc travel time
interface, since V, < V, Therefore, therc is no indication
,r.,od"l
curvc-
,."d
97
of 25.5 m. This depth obviously is in error, as the presence of the Vr-layer did not enter into
the computation. But what is the magnitude of the error, and is the depth too large or too
small?
Note the convention for referring to travel-time curve segments in this and the following section. We refer to these
segments relative to the existing geologic subsurface. Thus, if
a curve has two segments, but one segment is due to head
wave arrivals from the top of the Vr-layer, we refer to this as
the V.-segment. Normally, however, if a curve has only two
segments, we refer to the one with the lower slope (higher
velocity) as the Y2-segment.
RefractModel assumes increasing velocity with depth, so we cannot use this prograln
to create a travel-time curve to analyze the problem. However, Table 3-5 provides the necessary information. Table 3-5 contains the model and the time-distance data illustrated in
Figurc 3-24. The depth to the top of the V.,-layer is the sum of /2, and 11, which is 15 m.
Thus, the depth based on the two-segment curve is greater than the true depth to the top o1'
the V.,-layer. You could continue to investigal-e Lhis problem by utilizing Table 3-5 as a
dynamic table. EnLer a model in Table 3-5, type the resulting time-distance data into
RefractSolve, use RefractSolve to compute a rnodel, and compare 0re result with the origiTABLE
3-5
Refraction Times for a Three-Layer Case with a Low-Velocity lntermediate

Layer
Distance
(nr)
l0
20
30
40
50
fi)
10
tto
90
l(x)
ilo
tzo
Dircct Wavc
(nrs)
Intcrllrcc
Rcl'raction (ms)
r00
200
30 0
400
500
600
700
800
900
100 0
il00
r20 0
-5t.9
544
56.9
669
V, (rn/s)
V. (nr/s)
V, (m/s)
/rr (rn)
694
fi.
119
lntcrval
594
6t9
644
14.4
169
19 4
Intcrccpt tirne
(nrs)
49 1
(m)
(rn)
1000
-500
4000
5
lo
l0
98
Method
Chap'
find that the

nal model you entered into Table 3-5. In each case you investigate, you will
sense of the error
incorrect depth aiways is greater than the actual depth. The reason for the
The vertical
is straightfor-ward. A two-segment travel-time cur.re warrants use of Eq. 3-17.
travel
time of the
thickneis of the first layer, hI is directly proportionai to /,. However, the
critically refracted wave is slowed by passage through V, and these arrival times are later
time
than they would be if the entire distance was through V, material. Hence, the intercept
the
actual
is greater, and it follows that the calculated depth to the interlace is greater than
depth.
Unless some additional evidence
is available,
the low-velocity layer will
remain
undetected. lf, however, strealn cuts, road cuts, well logs, or other geophysical methods
possible to
dcmonstrate the presence of a low-velocity layer in the geologic section, it is
approximate the depth to the second interface using a l-ravel-tirne curve such as the one in
Figure 3-24. This travel{irle cunr'e provides valucs for V' and V,. We dcmonstrated that
eq. :-:: rentains valid in the case o1 the "sandwichcd" low-velocity layer. Tlris substanti-
Eq' 335'
ates that thc intercepL tinte t, in Figure 3-24 is correct and is equivalent to ,i2 in
^If
Eq' 3solve
o-nly
we
need
V,
/i,
and
tbr
good
estinlate
a
providcs
evidcllcc
independent
our
35 to detennile h,. Table 3-6 provides a ureans for testing this asserlion. It is a dynamic
same
table and uses the values just discussed to calculate /2,. Values used in the table are the
as those used
in Table 3-5 and Figure 3-24.
The Nonideal
Su
Because the
should be made to
you are working. 1
of its major shortc<
Hidden Zones:I
Unfbrlunately, a lo
ing delcction by se
contrast with a la'
occurs and what

RetiactModel are
RefractModel or ir
and h.,
15 m. R
solution to this datt

Now change the va
a three-layer modeJ
First study 1
interlace arrive latt
identily only the fi
identitled. Hence,
tion I[ we use the

Explore the sensitivity of h, to the magnitudes of V, and h','
Note that V,, Vy and the intercept time are interrelated and
should not be changed independently. lf you want to investigate additional values, use Table 3-5 as a dynamic table to
determine values for a new model. Then insert these into
Table 3-6.
the computed vertir
value of l5 rr.
The hidden-l
depth, whereas
TABLE
3-7
Disttnce
(m)
TABLE
3-6
Computation of Thickness for Low

VelocltY Second LaYer
t0
l-5
V, (nr/s)
V, (m/s) (rssunrcd)
V, (nr/s)
h, (rn)
(assunredl
l(XX)
20
5(X)
25
4000
30
-5
35
40
45
50
55
60
Lhe
Reft
99
Because the presence of a low-velocity layer is a common possibility, every eftbr-t

should be made to obtain independent information about the subsurface in the area in which
you are working. The inability of the refraction method to detect low-velocity layers is one
of its major shortcomings.
Hidden Zones: The Thin Layer

Unlbrtunatcly, a lclw-velocity layer is not the only circurrstance that rcsull-s in a layer escaping dctcction by seisrnic rcliaction rlethods. Thin layers clr layers that havc:r low-velocily
contrast with a laycr below also will cscapc dctection. Let's try 1o deternrinc why tlris
occurs and what type error results. A dynanric tablc (Tablc 3-7), RcfraclSolve, and
RclractModel are very useful tbr our purposes. First euter thc lbllowing rnodel into
Rel}actModcl or inlo Tablc 3-7: Vr = 500 m/s, V, = 1400 trt/s, V. = 4-500 tn/s, fi, = I0 tu,
and h, = l5 nr. RefractModel produces a classic l"hrcc-scgrncnl" tl'avel-tin1e curve, and a
solLrtion to this data using RefractSolve yields values thaL zrre idcntical to the stafling model.
Now change the value of fi, to 5 m. This time the result is a two-segnrcnl curvc cvcn though
a three-layer model was entered.
First study Table 3-7 and tl.ren inspect Figure 3-25. AII head wavcs liorn the flrst
intcrlhcc arrive later than the head waves from the second interlhce. Usually wc art: able to
identify only the first energy to arive a1 a geophone, so thc later-arriving waves will not be
identillcd. Ilence, wc observe a two-layer case when we have a thrce-layer gcologic section. lf we use the tinre-distance data from our sccond nrodel in RefractSolve. we find Lhat
thc conrpul"ed verlical thickness to the top of the V.,-layer is I1.7 m. This is less than the truc
valuc of l-5 rn.
Thc hidden-layer problern due to a thin laycr yiclds a dcpth value less than the truc
dcpth, whcrcas thc lorv-velocity layer resulted in a depth value grcatcr than l"tre lrue value.
TABLE
3-7
l)istancc
(nr)
.5
t0
r-5
20
2s
30
3.5
40
45
50
5-5
60
Refraction Times for a Thin lntermediate Layer

l)ircct Wrve
(rns)
l0
20
30
40
50
60
70
80
90
100
1r0
120
lntcrluce I
Rcl-raction
50
5t6-5
58 79
65 93
73 08
tto 22
87 36
94 50
r0r 65
rou 79
tr5 93
r23 08
44
lntcrlacc
(rrrs) Rcliaction
(rns)
.17 6-5
4E76
.19 tt7
50 99
52 I0
53 2r
54 32
5-5 43
-56 54
57
(rn/s)
(rn/s)
(nr/s)
h, (nr)
/r. (nr)
V,
V,
V,
6_5
-sti 76
-59 u7
lnterval (nr)
500
1,100
4500
l0
5
100
Chap.
The Nonideal Sul
120
Slope = l/V,
100
>\.
t
E
tr
\-l
Slope
l/v,
\-_-_
Slole =
l/V.,
T
10
20
30
50
I)istance (m)
Figure 3-2(
depth to the
calculate a uaximur
vides some limits. I
head wave arrivals
another layer, V} is
and therefore can dr
illustrated in Figure
V2= 1400 m/s
hz=5m
^^^'
Y1 = 4500 m/s
assuming that the V

were recorded). Thi
first interface.
You can use t
ing thc V., line,

v1> v2>vl
Figure 3-2-5 A rhrn intermediaLe laycr that is not dcLccted by rcl'raction The time-dis
tance data fronr Tablc 3-7 are corelated with a diagranr of thc rrtodcl used to calculate thc
data. No hcad wave arivals for the first interface amve bcfbre the direct w:rve or the head
wave fiom the second interfacc due to thc V. -V. velocity contrast atld the thinncss of the
V,
IaYer-
Why? Once again, a two-segment travel-time curve warrants use of equation 3- 17. The vertical thickness of the first layeq ft,, is directly proportional to /,. However, the travel tirne of
the critically refracted wave (second interface) is speeded by passage through V, and these
arrival times are earlier than they would be if the entire distance was through V,-material.
Hence, l-he intercept time is less, and it follows that the calculated depth to the interface is
beca
the slope ofthe V, li
Laterally Varying
I ateral variations ir
rather than the exce;
significantly, and ca
sense of the eifect
<
RefiactSolve.
In this sectior
(Fig. 3-27). The tra
less than the actual dePth-
the direction of the

horizontal interiace
no indepcndent infbrmation that the Vr-layer may be present, then the layer will be missed,
and the computed depth will be accepted as accurate when it actually is too shallow. If independent information indicates that we should be dealing with a three-layer situation, we can
slopes for the first s

do. However', it ma1
Do we have any recourse as we did in the case of the low-velocity layer? [f we have
the first segment hi
101

'120
100
o80
a.
F
60
i,
->
40
\
0
10
stor"=
20
ttu,
30
40
Distance (rn)
l.'igure 3-26 Hou,to utilizc (hc travel tinre curvc in Figurc 3
dcpth to thc second intcrlicc Prtxedurc cxplained in thc text.
2-5
to calcu)ate a nraxitnunt
calculatc a ntaximutr depth. As we already have a rnininrum depth, this process at leastprovides sorle limits. We have values for V,, V.,, and the intercept time for the line through
heacl wave arrivals tiom the second interface. If independent inlormation indicates that
y2
anolher layer, y,, is present, we likety have a reasonably good constraint on the value of
and thereforc can draw a line with the correct slope on our curve. This line is adjusted as
illustrated in Figure 3-26 until it just touches the intersection of the V, and V'. lincs (we are
assurning that thc V, arrivals were just a bit too slow to arrive before the V. arrivals that
were recorded). This gives us the lowesl- possible intercept tinte for the head wave from the
lrrst inlerfacc.
You can usc this approach in RcfractSolve. Be sure to draw the V, line befbre drawing thc V, linc, becausc lines arc assigncd to layers in the order drawn. You thcn can adjust
the slopc ol-1he y2 line untit thc velocity recorded in a solul-ion is corect.
Laterally Varying Velocity

Lateral varialions in tnaterials at fhe surface, at least in a miltor way, may well be the rule
rathcr than thc exception. lf these variations are not extrenrc, sutiacc velocities will not vary
signilicant"ly, and calculated dcpths to intcrfaces will have only sntall etrors. You can gain a
sense of the clTect o1' small valiations by changing the slope of thc direct wave arrivals in
RelraclSol ve.
ln this section we analyze the efl'ect of an abrupt vertical change in surface rlatcrials
(Fig.3-27). Thc Lravel-tirnc curve in Figure 3-27 illustrates that a licld survey shooting in
the direction of the higher-velocity unit sees a curye analogous to that produced by a single
horizontal intcr{ace. Howcver, a reverse traverse produces a very unusual curve, becausc
the first segment has a lower slope than the second segnrent. lt should be clear why the
slopes fbr the first segments of the forward and reverse traverses have the valucs that they
do. However, it may not be as evident in the case of the second segrnents.
102
220
Chap. 3
The Nonideal Sub
I
I
200
't80
Slope = 1/V,
't
120
100
Slope
la
model
to dete
Slope
face.
/V,
60
40
Slope = l/\.j
20
0
ond
= I/V,
80
Tz
layer d
40
tr
160
Use
160
02040
case of the vertical

by the sense of the
In this case a threeOnce again, the slo
the n.rodel. The gen
the second and thir<
80
120
100
Distance (rn)
Forward
---)
{--
Revcrse
nuity posihon is
betwecn these point
I
I
1?O
100
E
.'60
E
i:'40
Figurc3-2TTheefl.cctofaverticalvelocitydiscontinuil'yonforwardandrevcrsetrar,el
directiontimc curvcs Note the distinct paftem of slope reversals depending on traverse
020
The posirion o1'the vertical discontinuity is marked by the crossover Pornt
Once again, we can understand the result more easily
if
we develop a travel-time
Forrvnrd
---)
cquation and determine the slope (refer to Fig.3-27):
time-I+vl
dtl
dr
x-)
(for -r 2 y)
(3-s7)
v2
(3-s8)
V2
discontinuity
Thcrefore, in such a case you can recognize the presence of a vertical surfacc
materials'
of
the
velocities
the
determine
curves,
travel-time
the
of
by thc special pattern
of the forward and
and locate the position of the discontinuity by the crossover distances
reverse curve segments.
is buried beneath a surin the

you examine Figure, 3-i8, you will conclude that the answer is yes' As
Will such a unique

face layer?
lf
Pattern be produced
if the discontinuity
v.
Figure 3-28
Slope rcvcrsa
103
Use Table 3-7 andlor RefractModel
to investigate a
hidden
layer due to a low velocity-contrast between the first and second layers in a three-layer sequence. Once you obtain a
model that produces a two-segment curve, use RefractSolve
to determine the error in the depth value to the second interface.
a rcversal in thc slope paltcnls is prtlduced

by the scnse of thc chlngc in velocitics between a lbrward traversc and a rcvcrsc (raversc.
In this casc a thrce-scgt't'tcnt curve arises duc to the la1'gt'above thc vcrlical cliscontinuity'
Oncc ageirr, thc slclpes of thc l-ravel-tinre curve segrrenLs yicld all vclocitics illLrstrrtcd in
case
of the verlicul discontinuity at the surface,
the model. The gcncral position ol the discontinuity is indicttcd by the change in slope o1
the second and third sc_qntents o1- each [raverse dir-ecl-ion. Notc, howe ver, that lhc discontinuity position is only brackctcd b),the slope changcs ancl also that it is not rnidway
betwccn these points.
1F
120
100
.a
S opc
6,0
-O
- l/V,
80
Slopc
/a
= l/V.,
i:'40
<
o20
I
Forward
Slopc=l/V,
40
---)
Slope=llV, .'
leb
I)istancc (m)
120
100
{-
Rcversc
I
I
I
l'igure 3-2ti 'fravcl lirnc curvcs lil a vcrlical

S
discontinLrit-r, Lrcncath
a surficittl laycr:
Iopc r cver sal s arc prLrsenI lirr heud wavc arrivals f:onr tlre horizonral interlace
104
Method
Chap.
The Nonideal
Sub
lf you iue interested in substantiating the values of the slopes of the travel-time curye
segrnenti or in calculating the positions of the crossover points, examine Eq' 3-8 and
."*.,t" it for the case of the vertical discontinuity. Don't wory about simplifying the equation beyond the fom of Eq. 3-8.
Interface Discontinuities
(1) a sharp
In our discussion of irtter{ace discontinuities, we confine ourselves to lwo cases:
than a flat
rather
surface
(2)
an
irregular
and
verlical offsct in a horiz.ontal discontinuity,
arl erorepresent
might
which
offset,
vertical
the
sudace. First. let's exarninc the case of
:iorral slcP or fault scaqr.
Figure 3-29 illustrates ray paths for forward and reversc U'avefses ovcf a vertical step
vertical
and rcsulting travel-tilne curves. The area of interest is the region surrounding the
fi-om
distances
at
significant
situal-ion
step, for wc essentially are dealing with a two-layer
consist
iilrivals
Eiu-ly
tfaverse.
lorward
the
on
the step in eithcr direction. First concentrate
of the direct and head waves in the classic pattem producing a two-segment travel-l-irne
curve. The last ray path to follow the normal state of atlhirs is QA. The position of A on the
geosurface is con[rolled by /r, and the critical angle, but it is extrernely unlikely that a
head
to
record
geophone
last
the
3-29
phone will be located at this exact spot. In Figure
wave energy from the more shallow portion of the interlace is imnediately to the left of A.
Some wave energy fbllows the path PR, encounters the bottom of the step, produces
the
difTractions, and travels along RS. That energy traveling along the dceper ponion of
at
the
will
return
also
which
SG,
R.B
and
interface generates waves fbllowing such paths as
angle
critical
the
by
is
conLrolled
critical ungl". Or." again the position of B on the surface
plot
and ft, + .. A, ,o-" point to the nght of B this encrgy arives first. These arrival times
time
later
in
displaced
rnust
be
line
along a straight line with slope l/V, . Note that this
an inlerbecause ray paths such as SG are longer than ones such as BA. This later line has
cept time I,,. whercas the earlier line has an intercept tlme /it'
l[ ui conrparc QA and R8. we see (hat the drl'l-crencc in interccpt lllnes' /-2 - /ir. ls
and
due to the extra distance, RD, traveled by RB at Vi. RD is related to the critical ahglc
Forwrrd
the height of the step r:
4p =
1cos0,,
l-59)
V.-
so that t1e intercept time difference can be computed and z determined by the following:
zcos 6,
'it
'it
Figurt
3-29
ovcr a vcrtrc,
(3-60)
Vl
and
(3-61)
cos 0i.
If
the reason fc
to the step can be enr

140
105
Points do not, fall on drawn segments
12lD
100
E
'a
ao
60
40
t; -V
20
o
Slopc
o20
Forward
,60
----)
- lA{ -/
'to0
80
120
<-
Disrancc (m)
Rcvcrsc
I
I
I
I
Vr= 500 nr/s
t)h
+:
4=
I'l(X) m/s
Vr= 500 m/s
Vr= l4(X)
Figure
ovcr
nr/s
3-29 Ray paths and rcsulting travcl tinre curves for forwiud and rcvcrsc Irnvcrscs
vcrtical step A dctailcd explanrlion is given in the text
(t,,
t,)vrv,
(r,' - ,,')"'
(3-62)
If thc reason for this computation is not yet obvious, note that the time deLay at B due
to the step can be envisagcd as lollows. lf the step was not present, the timc delay difference
!4 l"I
1rr,
106
beLwecn positions
Method
and B would be due to travel along QC at Vr. Because
Chap'
of the step
the
is equal to
time tlifibrent:e is due to RC at V, . However, the time to travel along QC at V,
V''
at
RD
travel
to
is
the
time
travel along DC at V ,, so that the delay at B
The Delay-Time
reverse traverses. I
presence of the
ste
step. The precedin
point located abovt
In all
the prr
face geometries, p
data to which strail
vt
is
deduced. Even i
adjustment to this r
to question how on
irregular surface w
are considered in th
Before proce
Due prirnarily to sJ
cussed here are n<
engage in analyses
result from a dippi
when an interface <
Show that the time to travel OC at V, is equal to the time to

travel DC at V,.
surface structures
al
Belore proceeding to an analysis of the reverse traverse, we need to point out that the
point 0.
actual arrival pattern is a bit more complex than presentcd. Ditfraction occurs at
will
they
but
that
o1'ray
than
time
QA,
These clilliacted waves itrrive everywhere at a latcr
yr
of
B,
right
point
to
the
At
sotlrc
.
at
arrive earlier at lJ Lhan ray RR, as both are traveling
difliacted
the
before
will
arrive
waves rcll-acted liorr the deeper po(ion ol' the intcrface
waYcs. This distance is determincd by the velocity ratio and :. Although in Figure 3-29
only ope ditlr-actcd arrival rr.ridway belweeu the early and late refraction scglllents is illustrated, in practicc sevcral such points may be present and may well make it morc dilficult
to discern thc deccptivcly sirnple pattem of Figure 3-29- Herc we havc a clcar-cut case
wherc dilll.acted waves play a key role in the fbrm of the travel-titne curve, albeil- not a
THE DELAY-TIME METH(
claritying one.
Nalurally, information l.rorn a reverse traverse is collectcd. In the case illustrated in
to
Figure 3-29, it irppcars Lhat the step geometry yields two curve segments with slope equal
head
to
is
due
l/V"
scgments
tli. in addition to the direct wave segment. The first of the
wave arrivals fl-om the dcepcr po(ion of the inteface. The last strch path received at the
ence in time it take

tance between sour<
the concept of delal
Delay time ci
the V,-layer (the slr
surlacc is eC. Wave energy will travcl up toward the surface frorn point R along paths such
interlace,
as RDB and RpA. This cnergy is refiacted upward from the shallower part of the
such as
the
sl"ep,
to
but the angles of refract-ion do not equal the critical angle. Rays close
with a
but
path
length
RDIJ. will arrivc earlicr than BC, as they follow a sonrewhat similar
left of
to
thc
port-ion of travel at V". Rays such as RPA that travel to a considerable distance
refiacof
anglc
R have a portion of tieir paft (rRP) at a shallow angle to the interface. The
tion of such rays begins to approximate the critical angle, so that a plot of their
tinres versus distance approximates 1/Vr.
tLrrival
The oll'set of the lN, segments ior the reverse traverse is the opposite of the oft\et
fbr the forward traverse. Thus, the critical criteria for recognizing the presence of the step or
and
far.rlt scarp are the ofTset 1/V, segments, with the offset sensc switched for forward
curves is presented
In this section
we
irregular refractor
one. However. if wr
the refractor surlac
the projection of tt
1913, p. 10). For thr
where /^ is the total

ty). In more advar
Hence, we use the
uppercase I is used
times that can be mr
The Delay-Time Method
107
reverse traverses. Ren-rembering these relationships is important. not only for detecting the
presence of the step, but for selecting the corect traverse for computing the height of the
step. The preceding suggests that under good field conditions the traverse with the shot
point located above the shallower portion of the interface is likely to yield superior data.
In all the preceding discussions of the refraction method applied to varioLrs subsurface geometries, planar interlltces were assumed. These interlaces produce tirre-distancc
data to which straight lines are fit, and from which pararreters are dctermined and geology
is deduced. Even in thc casc of a latcral velocity change or an olfsct intcrlacc, thc only
adjustrrenl to this approach is a single vcrtical contact. By this (irre you nray bc bcginning
to qucstion how one copes with situations when an inLerlhce is not planar but possesses arr
in'egular-surlhcc with highs and lows a rnore realisl"ic geologic sccnaricl. Such situations
are considered in the lbllowing scction.
Beforc proceeding to the next section, we call your attention to a valuable resourcc.
Due prirrarily to spacc considerations, nrany variations on the subsurlace gconrctries dis
cussed here are noL includcd. With sornc carcful consideration you should bc ablc to
engagc in analyses o1'nrore cornplicated geonretries. What Lype o1'travel-tirne curvc would
result from a dipping inter{ace wilh a vertical step? How are timc-distance data all'ected
when an interface changes inclination lrom horizontal to dipping? A good varicty ol' sLrbsurface strucLures and theirassociated travel-tirne curves as well as analyses ofa nunrberol'
curves is presented in Mooney (1911 ,p. l5-1 to l5-18 and l6-l to l6-9).
THE DELAY-TIME METHOD
In this section we take a slightll,dilterent approach that ultimal-ely will allow us to nrap
irregular refractor surt'aces. Study Figure 3-30(a). Ol course, thc path EBCG is a l'arniliar
one. However, if wc concentrate oul attention on thc intcrface belween V, and V, , which is
the refiactor surface we arc int-crcsted in mapping, we might think in tems of the diffcrence in tirne it takes a wave to travcl EBCG as opposed to the tirne to travel the sarne distance between sourcc and receiver y in the Vr-laycr (path AD). Such a considcration leads to
the concept of deLay time.
Delay tirnc can be detlned as thc l-irnc spent by a wave to travcl up or dor'vrt through
the V,-layer (the slant path) compared to the timc the wave would spend if travcling along
the projection of the slanl- path on the refractor (AB or CD in Figure 3-30(a)) (Rcdpath,
1973, p. 10). For the illustration in Figurc 3-30(a) Lhe totaldelay time is
fu"
- 'n - t
(3-63)
where t^ is the total travel time alotg EBCG or GCBE (recall the requirernent of reciprocity). In rrore advanccd analyses this total travel tirne is rclerred Lo as recipt'ocoL tinre.
Hence, we usc the symbol rft to refer to this quantity. In the developrnent that follows
uppercase I is used to represent delay tirnes, whereas lowercase I is uscd to refer to travel
timcs thal" czrn be measured fror.n time-distance plots.
Method
108
Chap. 3
The Delay-Time
t\
(a)
Also, by definition,
Using simple trigon
can demonstrate tha
and, finally,
Thus, we can calcu
determine the delay

determine Io if we r
Once again, I
geophone is locater
rewrite Eq. 3-65 as
Following our earli
measured on a held
ofEq.3-69, we see
Figurc
3-30
tinrc rclaLionships (b) Schematic

(a) Symbols used in derivarion of delay
ol
and
ficldsetupfu'a"t"''"iningdclaytirresanddepthstorclracLors(c)Mappingincgular
rcfractors with the delay-timc approltch
Since, bY dellnition,
T''"=To+To
(3-64)
If
we assume that tl
wecanderiveanexpressionforthedelaytimeatageophone:
Tu+To=t^-i
and
and
Ir,
109
Tc =
tn
v-
(3-65)
Also, by definition,
"(V)
CG
CD
(3-66)
Using simple trigonometry and the samc rclationships wc used when deriving Eq.3-14, we
can dcmonstrate that
lro
I{
hr. tan 0,,
cos 0,,
v2
- , r/l
(v,'' - Vr')
Ir..
'
(3-61)
V,V.
and, finally,
(3-68)
Thus, we can calculate the depth of the 7,-V, interface beneath the geophone if we can
determine the dclay time at the geophone (and, of course, know V, and V).How can we
deternrine Io if we can'( measure it directly?
Once again, the reverse traverse comes [o our rescue. Consider Figure 3-30(b). A
geophone is located at G, a forward source is a1 Eo, and a reverse source is at ER. We
rewrite Eq. 3-65 as an cxprcssion for reciprocal tirne:
tn-Tc+7,,+l
(3-6e)
Following our earlier convention, Zono represents the travel time frorn Eo to G which is
measured on a field seismograrn and plotted on a tinre-distance plot. Using the basic fbrm
of Eq. 3-69, we see that
r
tu," = Tu, + 1..
"v.+ _
(3-70)
and
Itrc=Tro+7"^+Y;.*
ur
If
(3-71)
V2
we assume that the refractor surface is planar beneath G, then
7", =
T.^
and
t,:rc *
te
nc = TEu + TEr + zTc
x
+-+
v1
y-x
V?
(3-12)
110
TABLE
Because of the requirement of reciprocity,
and. therefore,
F,..c "
ttc
'
LRa
3-8
+L
Distance (m
Forward source
Ceophone
l0
Ceophone 2
20
30
Gcophone 4
40
Ceophone
50
Ceophone
v)
'ERG
ald
(3-73)
entirely o[ quantities [hat are

Since wc now have au expressioll fbr T" that is colrposed
3-68 to calculate /rr' This salne proceobtainable from lleld observations, *. .- use Eq.
spread, so that, in theory at least' we
dure can be repeal-ed for each geophone along a typical
lnany geopholles as required data
can map the form of a refiacloris su.face by using as
points (Fig. 3-30(c)).
'
the delay-time method that
Sere.al assumptions are included in the devclopment of
surface irnmediately
limit its applicability, namely (1) that there was a planar refiactor
expressions for delay tirnes and
beneath the geophorre and the energy source when dcriving
r and y' Thc first
(2) that distances along lhe refrocto, were equivalerlt to surface distances
surfaces
refractor
urru-ption effectively linlits the application of this n.rethod to incgular
depth
refiactor
to the average
only if the relief on tl.re refractor is imall in amplitude relative
the
lirrrits
assumption
(Kearey and Brooks, 1984, p. 124; Redpath, |g13,p. 26). The second
method,sapplicationtorefractordipsollessthanl0.(KeirreyandBrooks'1984,p.125).
clelay-time rnethod is to work
Perhaps the best way to visralire Lhe application of the
and reverse tirnes for an
with a specific example. fuUt" :-S (a dynarnic table) lists for-ward
Gcophone 6
60
Geophono
10
Geophone
80
Ceophone 9
90
Geophone l0
r00
Geophone I I
il0
Ccophonc l2
t20
Reveme sourcc
t30
V, (m/s)
V, (nvs)
Ceophonc
TABLE
3.9
Ceophone
Ceophone 2
Geophone 3
geophonetimesin Eq.3-l3.i.""ip.o."ltimeisthetiurefrolntheforwardsourcepositionto
into Tablc 3-9 (also a
source position. These tirnes and V, and V, are entered
Geophone 6
dynamictable)whichusesEq.3-T3tocomputedclirytimesandEq.3-68tocalculate
Ceophone
i"r..r.
it to perform depth calculadlpths. Because Table 3-9 is a dynamic table, you also can use
in visullizing the steps in
aid
3-9
and
3-8
Tables
tions lbr acrual field cxarnples. Although
dept-h computations,
curv(].
you should appty tne delay-time approach to an actual travel-time
in Table 3-8, but only

Figure 3-3 l(a) is a time-distance plot of the tirnes calculated
where depths
positions
only
Thc
plotted'
are
arrivals
the times which would appear as first
posigeophorre
only
the
are
these
as
outline,
can be detennined are enclosed in a dashed
few
These
sources'
reverse
and
forward
the
both
tions where travel tilnes are available from
addi
obtain
can
we
Usually,
refractor.
the
about
points lirnit the amount of infonnation
Refrat
Position
Each set of lbrward and

irregutar refractor similar to Lhat itlustrated in Figure 3-30(c).
to the two energy source-toreverse times for a specific geophone distance is equivalent
ih"
l!
Head Wave Arr
Position
v
to=Tun+Tu^ +L
The Delay-Time
Chap. 3
Geophone 4
Geophone 5
Geophone 7
8
Geophone 9
Geophone l0
Geophone I I
Ceophone l2

TABLE
3-8
111
Head Wave Arrival Times for an lrregular Refractor

Dis(ance (m)
Position
Forward source
Ceophone I
Ceophone 2
(m)
Direct Wave (ms)
Forward Time (ms)
Reverse Time (ms)
l5
0.00
l0
t4
l3
714
2t.91
52.t4
t4 29
23.62
49 09
?o
55 .20
30
t2
2t
25.33
46.01
Gcophonc 4
40
78 51
28
31t
44
33
Geophone
50
35.7
30 76
4t
94
Gcophone 6
60
l3
l3
t4
42 86
33, ti2
10 24
Geophonc 7
70
t5
50.00
36.87
3U 53
Geophonc
80
l6
57 l4
39 97
36.82
90
Geophonc
Geophone 9
43
lorc
t7
t9
64 29
Ccophone I0
Geophone I
llt)
20
18.51
Gcophone I2
t2t)
2t
r3-5
r30
2t
92 86
Reversc source
V'
7.
(rn/s)
(nVs)
3.9
43
42 98
35 It
16 1t
49 76
32
-52 82
30 67
34 08
55.20
1400
(rn)
I0
Refractor D"p!h. Computed Using Delay Times

Forward Time (nrs)
Rcverse'l'irne (nrs)
21.9
52.1
Ceophone 2
23.6
49.r
94
8u
Posi
tion
Ceophone
3l
4200
Gcophonc intcrval
TABLE
]t
Dclay Tirne
(rns)
Depth (m)
t4
t3
Ccophonc 3
25.3
46-0
tJ.
t2
Gcophone 4
?4.4
u.3
ulJ
l3
Ceophone
30.8
42.0
rJ8
t3
Geophone 6
33.8
40.2
9.4
t4
Geophone 7
36.9
38.5
t0
l-5
Geophone 8
36.8
t0
r6
Ceophone 9
39.9
43.0
35. r
I 1.4
t7
Geophone l0
46.7
34.
Ceophone I
49.8
32.4
t2.ti
t3 5
20
Geophone I 2
52.8
30.7
t4.t
2I
V'
V,
(m/s)
(rrVs)
Rcciprocal timc
1400
4200
(rns)
55.20
t9
112
Chap. 3
Other Methods
(a)
/a
<-tn
a
"*"
tr
.r-
a
a
--------->
a
^40
.l
930
a
a
[-igure 3from eacl
.=
posltron
| , .-r;
value for I/r, we cl
l0
Eqs. 3-70 and 3-71

6,0
20
Forward
----)
100
e0
<_
120
Revcrse
t,'
r,,
Distance (rn)
and
(b)
Equation 3-75 demr

with a slope of 77.
E
o*o
10
Table 3-8. Using thr

The delay-rin
refraction data and

refracting interface
delay-time method
-'t0
9
-2O
\slope =o4-16=)lv.
refraction methods,
In the next s
100
20
Although these ma
120
cussed, they are mor
Distance (nl)
to provide an idea ol
Figure 3-3I
(a) A plot of the dircct wavc and lirrwlrd and revcrsc head wave times listcd
in Table 3-8 (b) A graph of the dilfercnce bctwcen forward rnd revcrse travcl times for
cach geophone plottcd agrrinst distancc The slope of thc linc llt to these points is used to
OTHER METHODS
calculate V..
tional points by increasing the number of geophones, but we must be careful nof to
use
arrivals from a deeper refractor. Strategies fbr dealing with thcse problems are covered
clearly and in detail in Redpath (1913).
Recall that in the delay-time method the quantity ir" is defined as pelpendicular to
the refiactor surface. Therefore, once depths beneath each geophone are detemrined, the
refractor can be mapped by drawing arcs with a radius equal to the calculated depths and
constructing a smooth curve that is tangent to these arcs (Fig. 3-32).
If you have entertained a nagging suspicion that something is slightly amiss, congratulations! Heacl wave times from an irregular refractor will not lie along a straight line in a
tinte-distance plot and, therefore, V, cannol- be determined accurately. Lacking a good
Wavefront Methr
The wavefront metl

struction to locate p
of horizontal or dipl
tor
surfaces. Howe,
of the tin
replaced by other m
Because
tioned later.
Nevertheless,
Other Methods
113
3-32 A refractor surface is construcred from delay-tinte depths by drawing an arc

frorn each gcophone location Thc alc's radius is equal to thc calculatcd dcpth at (hat
F-igure
posl tron-
value for 7,, we cannot calculate ho. How can we circunlvent this problent?
Eqs. 3-70 and 3-7l, we cau arrive at the form
t-t-Tt'l,trG
tt-kc
'Lt
, ,c,
+_v2r - 7,. t,R
T..
t,P
_)-r
If
we rewnte
(3-74)
v2
and
trrc-tr,rc=Tr,r-To*
+_2x _v
v2
(3-1s)
v2
Equation 3-75 demonstrates that travel-time differences plotted against distance yield a line
with a slope of 2/V2. Figure 3-3 1(b) is a graph of this type using the times and distances in
Table 3-8. Using the calculated slope of 0.476, we arrive at a value for Vrof 4202 mls.
The delay-time method represents a somewhat different approach to using typical
refraction data and under good conditions provides a much more detailed picture of a
refracting interface than the basic approach discussed first in this chapter. However, the
delay-time method is fraught with the same difficulties as most corrunonly employed
refraction methods, the most obvious of which is the hidden-layer problem.
In the next section we briefly mention other more advanced refraction methods.
Although these may circumvent some of the limitations of the approaches already discussed, they are more complex in design and implementation. They are presented primarily
to provide an idea of additional possibilities.
OTHER METHODS
Wavefront Method
The wavefront method was developed by Thornburgh (1930) and utilizes wavefront construction to locate points on refracting surfaces. The method is simple to apply in the case
ofhorizontal or dipping planar interfaces but is most useful for delineating irregular refractor surfaces. Howeveq its application in the case of more complex surlaces is tedious.
Because of the time involved in constructing these wavefronts, the method has been
replaced by other more powerful methods such as the generalized reciprocal method mentioned later.
Nevertheless,
it is instructive to illustrate how the wavefront method is applied in a
114
Method
Chap' 3
arrival times and to work with

simple case, as this illustrates yet an additional way to view
in Figure 3-33(a)'
illustrated
interlace
horizontal
planar
Other Methods
(a)
seismic data. Consider the
RefractedenergyarrivingatthegeophoneGtravelsapathE.ABG.Theotherendofthe
point P. By using the definition

planar wavefroni u..iuini at G is locatea on the refractor at
to show that the time for a
straightforward
is
it
of the critical angle and simple geometry,
travel from B to P at Vr"
to
wave
a
for
time
the
to
wave [o travel l'rorn B to G ui V, It equal
sin
0,.
vl
v2
(3-16)
-BG
BP
and
.BGBP
:
time,,,,
tirne,,,,
i -
(3-11)
(b)
500
450
400
at E^ defines
A similar wavelionl- associatcd with point P resulting frotl] a reverse shot
point ly' on thc surface. As in the previous derivation'
time.,,
time."
(3-78)
,n the time to travel

must be
The time to travel Er-ABP plus Lhe time to travel E*DCP
:qu.o]
in E'qs' 3-77
is reciprocal time /^. From this reiationship and the identities
EFADERwhich
and 3-78, it becomes aPparent that
E
E
aF
350
500
250
200
150
tLrc
'
t-t
'Etll
(3-1e)
-'R
position with a head

Knowing V, and Vr, we calculate 0,,' If we have a geophone
as the one at 40 m in Figure 3wave arrival (G), we can'construct a planaiwavefront such
to frnd the unknown time, and hence
33(b). Given a reciprocal time, we need only subtract
lrl
00
Dr
10
then construct a second wavefront which
poriiio, H, relative to the energy source a[ %. w"

locatespointP(Fig.3-33(b)).rnisprocedurecanbeduplicatedforeverygeophonethat
this would be a waste of effort for

records an arrival from the refractor of interest. Alttrough
dipping refractor,
a horizontal surlace, two or more polnts would delimit a planar,
and
\',-
In thesc latter cases the wavetnxny points would map in sonte detail an irregular refractor.
warranted in light of other
front construction becornes morc complicated and may not be
superior results with equivalent
t1erhods which producc similar results wil-h less eflbrr or
eI'lbrr.
Ray Tracing and the Generalized Reciprocal Method

material in this chapl"er, the
Although corrputer capabilities were applied for rnuch of the
of ray tracing
pro..rring capabilities of courputers barely were challerrged. The approaches
utilize com1980)
(Pahner,
nrethod
reciprocal
generalized
(ceweny et;t:., 1974) and the
more cornplicated subsurputer capabilitics morc lully in Jealing with refraction data from
text'
this
in
face situations than have been treated
ln applying thc ray-tracing method an investigator proposes a subsurface tnodel
surface geology, well data, and
basecl on whatever inlormation is available including local
by computer and travel
calculated
are
then
model
this
cven refraction data. Ray paths for
Figurc 3-33
Relationships
CUTVE
times determined. Th
with the goal of impn
new tlmes are compt
conforms to krown g
correspondence to me
The generalizer
ward, approach It ut
Other Methods
115
(a)
9,.
Vr= 1400m'/s
(b)
500
450
400
350
tr 300
250
't:tr 200
lR----------->
Ir?
t50
l0 0
00
I)7 lo
Distancc (m)
Figure 3-33 Dingrarr illustrrting thc basic prrrcdure li.tr thc wavcliont rncthod (1r)
(b) Locating point on rcfracLor li'oln travcl timc
Relationships used to dcrivc mcthod.

curve
timcs dcLerrnined. These times are compared with observcd times, the rnodel is adjustcd
with the goal cl1 irnproving the corespondence between observed and conrputed tirres, arrd
new times are conrputed. This process is continued until a reillistic model is in hand that
contbnns to known geotogic information and that produces travel tirnes which have a close
correspondence to measured timcs.
Thc gcneralized leciprocal method (CRM) is a powerfirl, yet relatively sn'aightforwald, approach. It Lrtilizes reciprocal tirne and thc arrival timcs at two geophones as does
116
Method
Chap'
thedelay-timemethod.Howevcr,theextensionsinanalysisthatcharacterizetheGRM
irregular re[X",'."^?
enable it ,o a""r'"f"'iallv well with
n'J::*T::t;t*l::1lr':].H
thin laYer), which Provides
an
o calculate dePths to intedaces

critical Parameters can be achi
al method as comPared to "cl

, 1990).
FIELD PROCEDURES
Field Procedures
is of similar or grer
retum on a nonwin
some
of
the
inhe
Obtaining reliable
Equipment Cons
In the majority of v
are less than 100 m
ence englneerrng ar
objectives and covel
sulting finns. All th
including energy
Site Selection and Planning Considerations
ri
sor
At this point
dealing with energl
you will
Once you identify a locality that
you
that
are a number of considerations
m
as
likely sources in order to determine
seismograph, a ham
not exceed 50 m anc
ditions and for great
face and subsur{ace, beftire undertaking

ma
seem self-evident, it is amazing how
c
data
ordcr to gct started with seismic
ry. Explosive chargr

tions often rule them
sulted after geophysical data is in
ha
Premeasurement PrcParatlon'
Assume for a minute that You are

If
covered by glacial sands and gravels'
cable with 3-m, 5-m' or l0-m take-outs
bedrock depths vary from surface expos
with the greater depths, you might plan
quick
at dePths of from 50 to 70 m' A
(500
m/s
unsaturated sands and gravels
reve
and a bedrock velocity of 4500 m/s
you
m from the encrgy source' Becausc
2
tions, your line length must be at least
10-rn take-ouLs' A good rule of thumb
lengths at least thrce tirnes the target's
is intended to illustrate the importance
a
and funds are limited You tnust work
data'
useful
collecting
are
you
sible that
Other considerations enter into th
As demonstrat
be at least three tin
attenuatron occurs, r
lengths also result ir
geophones are empl
more rounded, makir
only increase the
net
data.
If test records
returning to the site i
as frequently in refra
is a problem, filters
seismographs have a
of ground motion.
choice. Another optir

quencies from a test
frorr.r signais of intere
Geophone-Sprea<
coupled with the ground' Firm placemet

ble of augering
rn deep holes' filling
placing rhc gcophone: irr lhc- holes yi

wo(hwhilc if at alt possihle' la1' cable
made (see fbllowing section)'

various elevations, correctious must be
Windandcontinualtrafflcmustbeavoided.Bothcausegroundl-notion,whichoften
The standard geopho

3-34(a). The energy
cable and usually is
take-out inter.val. Thr
Field
Procedures
is of similar or greater antplitLrdc tltarr tlrc rvur.

rerum on a nonwindy day or duri,g ,r,. ,,igr.,,.tfn,i,:X
1fi
ffi,1'lil'"1"r'ffit;rT'ltJ;:[J:l
some of the inherent uncc(rtirrtics in lt'll'lcrion

surr/cys, reliable data are essential.
obtaining reliabre data under trre prccctrirrg c.rrtrili..s
i. n"^t tn impossible.
Equipment Considerations
In the majority of what oftcn arc tcrrttctl .r'/rrrliorl rcll'lction
surveys, target depths typically
are less than 100 m' Many. survcys lllttsr ()pcrirtc
rrr populatcd areas. Results usually influence engineering and'/or planning clccisions. conrracts
tcnd to be for specific, short-term
objectives and cover a limited arca ol'irrvcstigution.
rrrLrs lavoring ,ai";ara and small consulting firms' All these factors dircctly irrtlrrtlncc tlrr.
t1,pc of su,irey that can be conducted,
including energy sources' equiprnc.t. crl.rprrrirrs rrrciritics,
and tirne.
At this point it would be a gotld itlcu to rcatl oncc

again the matcrial in Chapter 2
dealing with energy sources arrd scisttric cquipurcnl.
when using a signal-enhancemenf
seismograph, a hamnrer sourcc is :.tr.lcqua[. lirr
ruilnt, rcliacl.ion surveys if target depths do
not exceed -50 m and saturatcd scditnettts art closc
to thc surlace. Under more difllcult conditions and for SreaLer dcpths, a tnittiururrr ol'rr slrotgun
or wcight-drop .orra" is mandatory. Explosive charges likely wourd bc Prcrcrr.ctr. t.,ut r,-.gar, ."r:.ty,
onj cxpense considera_
tions ol'tcn rulc lhem out.
As dcmonstrated previously irr this sccti()n. linc
lcngLhs for retiaction surveys must
be at least thrce tirnes thc targct's tlcpth. Thrs
nrcuus trrat ray paths are long, that much
attenuation occurs, and that llrst-brcilk unrplitudcs
ar.r.. likely to be smau. These long path
lengths also result in the.disappcrlrurlcc ot'highcr
li'..qucncies. Even though low-frequency
geophoncs are employed in rcl'rrtctiorr sru'\cvs.
lhc. rvavelbrrn tends to be snroother and
more rounded' making the llrst-brclrk,,,,.",
nro't- dilllcult to recognize. These factors
",",,sclcction,
only increase the need for carcl'ul plrrrtrting. sitc
and patience in gathering quality
data.
If
test records
contain otr.icctiorurblc n()lsL.. anal

the problem cannot be solved by
returning to l-he sil-e at a more 1'rt'opititrus tinrr'. l'iltcr.s
,l.,ry *^r" the day.-Filters are not used
as frequently in rel'raction survcys as thcy arc
in rcrrccli.r, *u*.yr, but at sitcs where noisc
is a problem' filters may reducc thc cl'lcct ol'tlrc
rrtrisc to sor-nc extent. Most rnodcrn field
seismographs havc an orcilloscopc tlisPlll rvith
un trption pcrrnitting real-timc observation
of ground nrotion. viewing thc crisplrry *tlilc switchine
|ilters ,roy"h"rp in narowing the
choice' Another option is 1() lllcilsLlrc tlotrinant rrorsc
s,avclengths and detemine noise tiequencies from a test record' o['cotrrsc. utilizing
l-ilte.s alscl ,",,".,or.. frequency cornpouents
frorn signals of interest, which uriry lrc.rr.rc trclc.rcr-itrus (ha.
noise intertcrcnce.
Geophone-Spread Geometries and placements

The standard geophone sprcacl is rclcrrctl to us 1l1g
irt-lirtt, ,s:ltrcurl and is illustrated in Figure
3-34(a)' Thc energy source, clt'"shot"
Point. is plilccd:rlong the same line as the geophone
cable and usually is placcd u[ u rlistiurcc lrtrontl
t5c l'irsL gcophone that is similar to the
take-out intcr'"al' The tevc|sc sll.1 i5
Plrtectl at :rrr ct;uirrrrent position beyond thc other cnd
118
a-
(c)
)..
! --
tlrrrlr
llll
(b)
face form with anor

domes ir earlier, de
low exploration. A
This is followed by
r r r r r I ' I X<:-E'nergysource
)iirlrr
(a)
Field Procedures
ChaP. 3
The Refraction Method

Seismic Exploration:
tances from the shc

geophone lines can
In addition tc
cially in complex
By
A
Illrrlll
I..
CeoPhone
allll'r'ltll
as
it is being
collect
the subsurface struc

would reveal the lil
(d)
ture might be possil
I
I
gram), taking a shot

In the last sec
make this conectio
face layer is essenti
geophone intervals.
I
I
t
have objectives anr

flexible.
In some
spread arrangements
Figure 3-34 Ceophonc
shoottng'
cenlcr shot' (d) Fan
(c) ln(b) Offset sprerd

(a) In-line spread
this chapter, some
iinl *itrt
and shots
['i"l'"i':"!
tl
rs avar
providcd sufllcient enerBv
travel times' Reciprocal

tant geoPhones'
not give exact total
3-34-do
Figure
in
(c)
Note that (a) and
the offset
timescanbeesumatec]fromtheseu,-ung.*"n.ibutincases..suchasthedelay-ume
iin']' o 1]'tecommends
i, i;;r;;'i"'.ur.ururion.. *.lr",n
method. where r*
3-34(b))';;;;er viabre "'T.{*""#lr}i:;'.:-r::il:'i"T:'l:Pii:

producsi11ntml:Il;:,'ff;ri"* *", 1 at rhat position and
an<t rhe energy source
r-
500
o=
n --].
ooo -]
''
oJil:[:3l*,:;:tr,:"."1"*mfi ::""'ffi:i'l]1ffi ;'o*henrs'[raver

;[':::'i:i"it*[:*.T-T"*:':f [tT::l'i[:
arongtherineor'"'"'il'^i"""ffi
;i;;il6'
time curve in Figun

Imagine you
ofthespread.ThisisthearrangementusedinmanydiagramslnthisChapterandin
the target
to record arr vals from
i.tracttt'loatl and RcfractSolve'
length is not sufflcient
*r"
*r-ra
a
ln cases where
i^l] H,:'
n'[e
r res'l'1he gc ome
il*'
'
at
taken
Pos:
shot Points
2' and 3 ""t" ;;;;-'"n' ^ n1"" 11i1,'if;:::'::tffi";
taken at 0"";;;'
{}'i,i:l* [ i;;Jtiill.;i
caset
spread Iengths requ

depending on the sr
o
soo
ss
^?l
--l
lr
: 2501
t
Fl
-..1
150
-.1
,oo
-.]
20
=o
-.]
oo
l/
spread (Fig.
used
brow
then
ILtl o" shooting
"",?"rn'"."
x
rpie :-:Xtall
x in FiBure3-34(?
)ol
;|S;,T;,*"J3
ar rhar
position.
u*'",1xJt"':;ffi1';:l;fi::'Fli:lirf.
Such
an ;?#; tJ;"J'*n'"
searching ror a subsur-
Figure
3-3
bcdrock wl
thrt
a bedrc
Field Procedures
119
facc form with anomalous velocity (the stippted oval in Fig. 3-3a(d)). Exarnples include salt
domes in earlier, deeper refraction studies (Nettleton, 1940) or buried bedrock hills in shallow exploration. A normal refraction line is shot to determine local velocities and depths.
This is followed by one or more shots with geophones arranged in a fan shape at equal distances from the shot point. Anomalous increases or decreases in velocity along some shotgeophone lines can aid in defining the general area of the target.
In addition to thc geophone spread chosen, placernent of the line is important, especially in complex regions. Also, one always should undertake preliminary analysis of data
as it is being collected. In order to illustrate the importance of both of these points, consider
the subsurface structure diagrammed in Figure 3-29. Plotting time-disl-ance data in the field
would reveal the likely presence of a fault or step. A more complete analysis of the strtrcture might be possible by renroving the spread to the left of its present position (in the diagram), taking a shot, and then repeating the process by rnoving to the right of the diagram.
In the last section of this chapter the "weathering" correction is mentioned. In order to
make this corection reliably, good control of thickncss and velocity variations in the surfacc layer is esscntial. This can be achieved only by several traverses with relatively small
gcophone intervals. The main advice here is to plan carefully, continually nronitor data,
have objectives and data requirernents carefully worked out and clearly in mind, and be
llexiblc.
In solnc cases energy constraints. sitc access, or other factors may precludc the long
spread lcngths required to receive refractions fiorn thc target. This may not spell disastcr
dcpending on thc survey requirenrents. Il-you have a good grasp ol'the mcthods dctailed in
this chapter, sonte inlbnnation always can be salvaged. For example, consider the travcltimc curve in Figurc 3--15.
Imagine you obtained this curve in an investigation where your goal was to dcter-
500
450
Assunrc bedrock rcl'raction arri vcs

just after this position
Slope for velocity
ol -5000 m/s
400
500
E
250
200
i-
150
100
50
00
Distancc (rn)
Figurc 3-35 Ttitvcl tilrc curve illustrating lrrcthod to dctcrntinc ntininturn depth to
bcdrock u,hcn no bcdrock rcfractions lrc rcccivcd at lcophoncs This lpproach itssunrcs
that a bcdrock refraction is prcscntjust bcl,ond tlrc ltrst position sartrplecl in tlte survey
120
Method
Chap'
Field Procedures
3
minebedrockdepth.AnalysisoftheCurvesuggestsatwo-Iayercasewithvelocitiesof800
which to
are present' and you hul"^n: time in
m/s and I500 m/s. No bcirock refractions
ZE
runalongerspread.Canyoudeduceanythingaboutbedrockdepth?Theanswertsyes.
have a reasonably good value
you can obtain a minimum figure for bedrock depth if
you
for bedrock velocitY.
Assumeabedrockrefractionimmediatelyafterthelastgeophoneinthespreadin
bare,lop" io. bedrock velocity such that the line
Figure 3-35. Draw a line with the correct
lytouchesthelastpointonthecurve.Repeatforthereversetraverse'Yoursolutionnow
hu'" a depth which is the minimum depth at
will be for a three-layer case, and you *ill
of Figure 3-35 the minimum depth is 24 m'
bedrock could be present' In the case
which
if your assignment was to determine if any
This information could well be all you require
15 m'
bedrock in the area is shallower than
two-layer case' Transfer

Use RefractModel to create data for a
set
this data
RefractSolve, saving this data set' Pretend
Assumassignment'
consulting
a
fiom
r."pr"."na. field data
the minimum
determine
m/s'
4500
of
velocity
inl
"-u"Oro"k
at which bedrock could be present'
Oeptf,
to
Figurc 3-36
BecauseComputerprogramssuchasRefractsolveprovideeasydatainput,linefitting,andspeedycornputations,usethemcreativelytogleanasmuchinformationasthedata
supports.
which can be reduct
Corrections to Data
If a geophone spread cannot be anang

portion of the obsewed variations in
at the same elevation' a

due to these elevation differhow corrections for this situaation in topography along the
ors
ences and not to subsurface structure'

tion are derived, consult Figure 3-36' Beca
a-re
The same relationsl
Thus. the total eleva
spread,theenergySourceandgeophonesareprojectedtoliealol8.acomtTlondatum.For
below the surface, and
that E is located at A, avertical distance of Z.
example, we assume
thatGisatD.Theactualwavepalh,EQRG,istransformedtoAPSD'Traveltimesinthe
(denoted by the minus sign in Figure 3-36)
surface layer are reduced by Eaiv, and CGlv,
we
by the plus sign)' As AB=PQ and CD=RS'
and increased by PQlVrun[ nSl% (denoted
EAB and GCD'
can confine our derivation to triangles
source ls
energy
the
at
time
The correction
t''-
Z,
Zr
tan9,
-Yfoq
Y
to
By the critical-angle equation, Eq' 3-80 simplifies
If
a buried shot is u
phone is below the
negative value.
If the surface
(3-80)
unacceptable time v
such cases a colTect
ering correclion).
surface layer thickn
Field Procedures
121
ZE
Vz>
Vr
+B
Figure
3-36
Sy
l)al rr nr
mbols and gcottrctry uscd in thc
lc\t
t() (i!'r i\ c I he
d,. - sinr
'' = L'y+'.tr(l
c le
r rrt
iol
c( )t-l
ccli()tr
(.]-ri
I)
which can be reduced further to

7
-L,
%u
3_ri2)
The same relationship holds at the geophone positiorl c\ccpt tlurt Z,. substitutcs lirr 2,.
Thus. the tolal elevation correction ls
/"r"urri,rn
: (2,' + Zrr)
(%'
- q';'"
I/
(l-tt3)
a buried shot is used, its depth simply is subtlacted ti'ttrn Z,-. ll'thc cnergy soul-cc or gcophone is below the datum, the relationship ren.rains thr'slnlc cxccpt that Z,.or-2,- takcs ir^ a
negative value.
If the surlace layer possesses significant variittiorts irr thickncss, this ca1 irrtr.sducc
If
unacceptable time variations due to the fact that this lavcr's vclocity is typically vcry 1rw. I,
such cases a correction must be rnade (often, though ntislcrttlintll,, rcl'crrccl to as thc wt:rtllt-
ering correction). Before such a correction can bc introtlttcctl. lrowcvcr, good crlpt16l ..
surlace layer thickness and velocity tnust be irl hancl. As (his rctluir-os a nunrbcr 6l'cl.scly
122
Method
Chap'
sPaccdgcophonespreads.suchanexpcnditureo[ellortismadeonlywhenoneisintcrested
provide an
Dobrin and Savit (1988, p' 494)
in detailcd nrapping ()1.a refracror irrfac".
explanaLion tbr this comectlon
Applications
Using
nation was understa

ed in state funding fr
The study wa
contracted some of I
to the extent of the
was mandatory. Fort
REFRACTION
APPLICATIONS USING SEISMIC
available. These
lo6
layer sandwiched b
WhatelY, Massachusetts
co
glacial till and./or arl

rier between the dee
south. Over a Period of several mon

covered in the wells of many homeown
aquifer. The consulti

confined aquifer if a
geophysical investi5
which might be a pri
is a rural
The town of WhatelY

Although a number of its PoPulatio
truck farming for their livelihood'
contamrpenetrate an unconflned aquifer' the

munity is derived from shaliow wells which
sup
method was able to

but both were able ti
physical survey delir
gravels to an extent
source now is devel,
due to the impermeal
A typical tral
Although this time-c
Fine sand and silt
l0
Seisrnic refrac
1984 and were
Finc sand and sotne graY claY
20
I
)
3
Gray clay with occasional silt lenses
;30
o
-5
8
9
40
IO
ll
t2
-50
I|
Mcdium to co:trsc sand atld f-ine gravcl
{)
nrs
--*l'o
Figurc
llcdr ock
(r0
from
Figure 3-37 Represeniative well 1og
Whately, Massachusetts.
Irls
3-38
labclcd I-12
interval is l0
Trming lincs
Applications Using Seismic Refraction

nation was understandably viewecl with grcrrt
ed in state funding for a sl-udy ol-thc
123
alrLr_nt.
l-lrc lcve I ol.c()nccnl ultimately rcsult_
c<lnta,rinati., pr.blc.r ,,,r.1 p,,..iii" renredies.

The study was undertakcn by a Boston-buscd cngirrccrin!
l'irrrr, which in turn subcontracted some of the subsurface invcstigations to
r gcophysic.l c..sulting company. Duc
to the extent of the conlamination, it soon bccanrc
otr\ious
was mandatory' Forlunately, the local gettlogic
data brrsc
tlrut un lrltcnrate water source

,o.re well logs were
*,,. *u.*i
available' Thesc logs (see Figurc 3-37) clcrrronst*rtc-rl

",r.,
thc pr.csc.c..
.l- a sand and gravel
laycr sandwiched between a thick scqucncc ol-glacial-rakr..l,,,.-"[ruu"
and compacted
glacial till and/or arkosic bedrock bencath. r'nc
rntr cl,ry lu1,cr i.;,;.,,;';,-, impermeable bar_
rier betwecn the deeper sands and gravcls artd thc surlircc
slirtrs tlurt lor,r the contaminated
aquifer' The consulting firms rcasoned lhat thc decper
suncrs u.tr
*;,,r;i. .right constir-ute a
confined aquifer if areas of suff cienr thickncss could
hc disc.r,crritl. So,rc prlor, but spotty,
geophysical investigations suggestcd ll.tc prcscnce
ol' an.n.r.lorr.lf ,r""p
bedrock areas
which might be a prime location for thc aqui|cr trrickncsscs
,.",1.,i,..,t.'
Seismic refraction and clectrical rcsistivity invcstigations
rr,crc initiated in october
1984 and were supplcmcntcd by carcfulrl, sclecre-cl
.r,ir[,g .,,.... N"i,h"r geophysicar
method was ablc to differcntiaLc bctwccrr thc clay
laycr and u,.r,f",tui,,g sands and gravels,
but both rvere able to locate the bedrock surl,ce at
all occuPictl sit..s.'Ultimately, the geophysical survey dclineated.a buried rivcr vallcy
llllcd rvith tt," to,....,l ...qucnce of sands and
gravcls to an extenl sulficient to constitutr. a virrblc
u.utcr. s()ur.(.c ,i,; ,ir. communiry.
This
source now is developed and, with sufllcicrrt caution.
is likclv to rcrrrain uncontar,ina(cd
due to Lhe impermeable clay layer.
typical travcl-time curvc liotn thc Whatcly urcil

rs prcscntcd
Although liris time-distance ploL is frorn a sitc wherc

bccrrtx k
i.
,.r.,..,.t,
in Figure
2
3
4
5
6
1
8
9
r0
il
t2
t iitrrrrs
I
!'igun:3-38 Field
scisntogmnr Irrrrrr Whatcly, N4lsslchLrseus (icoPhonc

traccs atc
labelccl I 12. The flrst gcoltlrorrc is lrx:rrctl 3 rn
ll..trI rlrc
intervrLl is r0 m. F-irsr bicrks trr crch rr:rcc rro
irrrlic,Lrctr ,,; l:lil,,l:illlr::
Tirrring Iincs are at
l0 rls
intcrvals. Tlrc rccortl cncornltrsscs 100

nrs
3-3g.
shallower r.han the
,;;l:ff'li:l:
124
Method
Chap'
Applications Usir
the curve exhibits the classic three-layer
60-70-m depths associated with the buned valley,

layer is unsaturated sands and alluvial deposits'
case typical throughout the area. The first
gravels, silts, and the lake clays' The third layer

The second layer consists of saturated sand,
typically is bedrock.
000-
950
Southeastern New HamPshire

c
^^^
c 9(Il
AstudyconductedinNewHampshirelnlr9.14(Birch,l976)illustrateshowseismicrefracThis
.E
as well as bedrock elevations'

tion studies can be used to map water-table elevations
in a short period of time at low cost and prostudy is noteworthy because it was conducted
trough which could be conelated with a
duced useful results. Field work outlined a bedrock
erosional resistance' Analysis of water-table ele-
tr
850-
meta-sedimentary unit with relatively low
correlate with bedrock elevations and not

vations led to the conclusion that these elevations
contains implications for water flowsurface elevations in the study area. This finding
800-
paths,whichinturncontaininlplicationsforplanningtoavoidcontaminationoflocal
l'igure 3-39
using the Gr
aquifers.
detail on the
which are in
Waste-DisPosal Site
High-resolut
Typical information necessary for waste-di

geometry of the bedrock surface, and thickn
den. This information olten is best gathered
inlormation is necessary, the GRM should be
ed , Geotech
depth to bedrock'
ol'Explorati<
ics of the overburhere quite detailed
Lankston (1990) de
waste-disposal site at an
inlbrmation obtaincd usin
north-south direction, and
lithology log shows

icantly deeper (at al
soft to hard materia
lithology log is mi
loose, solt material
depthsatlineintersectionsvarybylessthan5pcrcent.Especiallystrikingisthedegreeof
velocity varias well as the position of lateral
detail obtained conceming the basalt surface
(interpieted by Lankston as layering in the basalt)'
The wetl
bott
MaricoPa Area, Arizona
much deeper at this

after striking a larg
straining geophysic
firmation of Lhe l0t
was one of several locations in the

The area near the Maricopa Mountains in Arizona
ing and provides a

not infallible, sour
ations in
Ll.re
basalt
UrritedStatespropos"aa,usitcfortheSuperconductingSulel'Colliderand,assuch,
Gravity' seismic'
infornration be collected'
required that detailcd geologic and geophysical
deposits, are only
the well.
andelectricalmethodsinconjunctionwithsur{acegeologyandwelllogswereusedto
As we conclu
determinethesubsurfacegeologyatspecificsiteswithin|heMaricopaarea.
ion study fiom one location at the Maricopa
ciation for the stren

logic conditions thi
travel-tinre cur-ve (Fig. 3-40(a)) is relatively

ed liom this seisrnic intbrr.nation is illustrated
cutvcll also is shown. The results of logging
setsThe
3-a0(c))'
are diagrammed in Figulc
The
s intefl:ace is 35 m a[ the well location'
able cost. Many
ol
the method is used

of'ten can discem
served
if you
tl
keep
approaches wcll enr
Applications Using Seismic Refraction
125
Distance (ft)
Earth's surface
c
c
Ill
l-igure 3-39 An exatnple of a geologic section derived from a derailed rclraction study
using thc Generalized Rcciprocal Method (GRM). Note that this methul provides
much
detail on the surfacc of the basalt refractor and is able to determinc lateral velmity
changes
which lrc intcrpretcd as layering in the basalt (Modificd from Lankston, Robcrt W,
lgg0,
High resolution rcl'raction seislnic data acquisition anrJ inlerprctation ,r Wrrd,
Stanley H_,
cd', Gcotechnical and environtncntal geophysics; Vcrlunrc l: Revicw and Lutorial:
Socictv
ol- Iixploration Geophysicists lnvestigations in Geophysics, no.
p_ 45 73)
-5,
lithology Iog shows this velocity change (sand versus compacted l'anglomerate) to be signilicantly deeper (at about 42 m). However, the drill rate indicates a pronounced change
from
soft to hard material at 37 m. This likely is the true location of the interface, whereas
the
Iithology log is misleading due to inaccuracies inherent in working with cuttings frorn
loose, soft material traveling to thc surface.
The well bottomed in very hard material encountered at 62 m. Seismic bedrock is
much deeper at this location (approximately 82 m), and it is likely drilling was terminated
after striking a largc boulder. In general, well information is essential for aiding and constraining geophysical interpretations. In this case the drill-rate log provides excellent confirmation of thc 1000 nr/s-2130 m/s interface depth. However, the lithology log is mislcading and provides a cautionary note. One always should use well logs as an additional, but
not infallible, source of infomration. Well logs, especially those fiom unconsolidate<j
deposits, are only as good as the expertise of the drilling company and the person logging
the well.
As we conclude rhis refraction seismology chapter, we hope you now have an apprc_
ciation for thc strcngths of the method as well as sonre of its weaknesses. In favorable gcologic conditions this approach to exploring the subsuriace returns valuable data at reasopabl(r cosL. Many ol thc unccrtainties inherent in applying seisrnic refraction disappear whcn
the mcthod is used in conjunction with other geophysical approaches. Seisrnic rellccti6n
often can discem the presence of low-velocity layers, for example. Your work will bcst bc
scrved if you kccp the inherenl problerrs of the methocl in mind. learn the standard l-icld
apprc.raches well enough to dcsign flexible surveys, and always atteurpt to enrploy adclition
Method
126
Chap. 3
Problems
PROBLEMS
3-1. Refer to Figure
site.
3-2. Figure 3-41 is a
3-3. Using the infon

geology at the si
ing sources of po
'a:
Distance (m)
2
3
0-
4
5
l0-
6
7
20-
8
9
30-
l0
lt
40-
t2
50-
E
oO
!
2130 nVs
60-
|0rns
Figure 3-41
located 3 rn
fr
ms intervals
,2
3
rlll
100
300
200
400
Distance (nr)
Sttpcr
proposcd site lbr the Supcrconducting
Figure 3-40 Seismic investigation o1'a
(c)
data
5gi5nric
6l'
(b)
[ntcrprc(ation
Collider (a) Travel-tinrr cutves at sitc MRI
LiLhologylogand-drilllatclogatrhcwellltrationidcntil'icdin(b).(Mtililicdlrorrl
Ccophysical invcstiga-
Sternberg, Ben K-, Poulrtln'
tions in suppon
Maiy M ' tnd Thot,as'
of the Arizolt 'SSC ["ol"tt
ScoLt
J''
1990'
irr Ward' Stanlcy
H'' cd ' Ceoti:chnicrl
and
envilonlnclltlr]g*pl-,y..,.,,Volunle3:Ccrltcchnicirl:Socictytr[.lixplorationCcolllll,sicists
p 2l l-22E)
lnvcstigations in Ceophl'sics' no' 5'
t0
l1
12
I
algeophysicalrnel.hods.Ir.rthencxtChapl-erweColltilluctoutilizeseisrrricl.ecl.rniquesvta
.fhis approach lras its orvn strcrrgths iuld wcaknesses, bul. lnany of
lnelhod.
|0rns
tlre rcf.lection
Figure 3-42
thesetendtocot-trpletrrenttherefractionapproach,sLlchthat[hctwo,uscdinconiunction,
cr)nstitul(' lr forrttidllrlc flir"
locatcd 3 rn
ms inlcrvals
Problems
127
PROBLEMS
3-1. Refer to Figure 3-38. Assuming horizontal interfaces, determine the subsurface geology at this
srte.
3-2. Figure 3-41 is a field seismograph from a region with horizontal interfaces. Interpret the data.
3-3. Using the information in Figure 3-42, determine as much as possible about the subsurface
geologl' at the site where the information was recorded. Be sure to include comments conceming sources of possible error.
2
3
4
5
6
7
8
t0
il
t2
--rr-l s 11ms
l
80 nrs
I| o,,,,
Figurr 3-41 Field seisrnogram. ceophone traces are labeled
locatcd 3 nr from
l0 m. Timing
r.hc
energy source. The geophone intervar is
2. The flrst geophone is

rines are at 5
,2
3
4
5
6
1
8
t0
l1
t2
<r
-----r'n,,
I
10,,,*
Figurc
3-42
locatcd 3
tr
ms inter vals.
Jau..
Field scisnrogram Gcophone lrirces are labeled l-12. The llrst geophonc is
from the energv sourcc The geophone intcrvat is l0 rn Tinring lines irrc at -5
128
3-4. Calculate the critical distances for
the
Method
Chap. 3
Problems
first and second interfaces in the accompanying diagram.
12m
800 m/s
15rn '-------------------------------
l800m/s------------------------'----------------1'
.F:
6000 m/s
3-5. Interpret the following refraction
data:
Reverse Traverse (ms)
Distance from Shot (m)
Forward Traverse (ms)
1.1
10
r4.3
15
2t.4
18.6
20
286
22.0
25
3-5 7
25.4
t0
30
47.9
288
20
35
48.5
32.2
30
40
510
3s.6
40
45
53.5
39.0
50
50
56.0
42_3
60
55
584
451
70
60
60.9
49.1
80
65
63.4
52.5
90
'71)
6-5.9
5-5.9
100
75
684
593
rr0
120
'7
.1
14.3
3-7. In the atea whr

5000 m/s. Wha
Distance from St
80
708
62'7
85
73.3
66. I
90
75.8
695
95
11.5
12.8
100
78.3
16.2
105
79
7'7.9
110
80.0
19.0
1-5
808
80.2
l0
120
8r.6
8r
20
3-8. Assume that a

overlain by a li
velocities of the
tion of the verti(
Distance from Sh
30
3-6. Provide a qualitative interpretation for the following travel-time curve, but be as specific as
possible
40
50
Problems
129
'++^+!tooeltl
"******+ol
*,
a-
*..4
4..
a*
a
E
tr
Distance (m)
3-7. In the area where the following time-distance data were collected, bedrock velocities average
5000 m/s. What is the minimum depth to bedrock?
Foruard Traverse (ms)
l0
t6l
t6l
20
JJ J
33.3
50.0
l0
50.0
40
66.1
66.7
50
17.6
11 6
60
83.9
83.9
10
90.
80
964
90.
964
90
102 6
102.6
r00
t08 9
r08.9
il0
l5.l
It5.t
120
I2r.0
121.0
3-8. Assume that a vertical contact separates two layers of moderate velocities. These layers
are
overlain by a layer with low velocity. Using the following time-distance data, detennine the
velocities of the materials, the thickness of the top layer, and, as closely as possible, the
loca
tion of the vertical discontinuity
Distance frorn Shot (m)
Foruard Traverse (ms)
t0
33.3
-1-r_
20
66.1
66.1
30
r00 0
r00 0
40
ul4
I 19.8
_50
1t4 3
125.1
I
130
Distance from Shor (m) (cont.)
Forward Travcrse (ms) (cottr./
Method
Chap.
Reverse Traverse (ms) (conr)
Problems
Distance fron
60
lt7
t34 I
t0
120 2
t31.0
25
80
123.1
140 0
30
90
133.6
r42.9
35
100
138 9
l4-5 8
40
lt0
t44.r
148 8
45
120
r49.4
r51.7
50
20
with this text illustrates
55
the east at a rate of 0.5

a classic three layer pattern The second layer in this sequence thins to
60
3-9. The refraction data in fite RMI1990001 on
the computer disk included
m for every l0 m of ground distance traversed. For simplicity, assume horizontal interfaces
how far to the east
and that the thickness of the first layer remains constant. Approximately
will the second layer no longer be detected?
Explain why the
3-10. These data are taken from a traverse across a buried vertical discontinuity.
travel time curve has the appearance it does, then determine as much as possible about velocities. the thickncss ofthe overburden, and the location ofthe vertical discontinuity.
Distancc from Shot (m)
Forward Traversc (nts)
indicates that
constrarnts ca
3-13. The time-distr
bedrock are o'
Distance from
125
125
20
25.0
250
l0
30
37.5
37
-s
l5
40
50.0
s00
70
50
64.2
62.5
25
60
69?
75.0
30
1o
742
808
35
80
192
8s.8
40
90
842
908
45
50
-55
l0
(X)
89.2
961
110
91.2
99.2
120
992
3-12. Dara file RMI
l0l
60
6-5
3-lI.
indicate
The following refraction data producc a classic two-layer pattern. However, well logs
an interoccupying
layer
low-velocity
a
with
layers
three
of
cotrsists
sequence
tha[ the typical
at
mediate position The depth to the top of the low velocity layer remains essentially constant
layer
the
third
of
the
top
to
3 m. Thc vetocity of the layer is 1200 m/s Determinc the depth
(bedrock) using only the information on the travel-tinte cunr'e. Next determine the actual depth
to bedrock based on the additional infonnation from well logs. What is the percent error
involved when using only the infornlation on the travel-time curve?
'10
15
80
8-5
90
95
r00
r05
IIO
10
56
il-5
l5
83
r20
Problems
131
Distance from Shot (m) (cottt )
Forwind Traverse (ns) (cottt.)
20
llt
25
13 9
30
t6.7
35
)94
40
2?2
45
25.0
50
21 .8
55
)a
60
295
3-12. Data file RM I 1990003 is from a seismic fraverse in an area where supplementary inforrratron
indicates that a 2200 m/s laycr underlies a 1400 m/s layer. Utilizing this infbrmation, whal
constraints can you place on the depth to bedrock (average velocity = 3900 rn/s) in the area'J
3-13. The time-distance data presenLed here were recorded in an area where steep fault scarps in
bedrock are overlain by alluvial debris. Deduce as much information as possible from the data.
Distance
liont Shot (m)
For ward Traversc (nts)
Rcvcrsc Traver sc (nts)
36
3.6
IO
t.l
l5
101
l0
20
143
t4.3
7t
7
25
t't.9
179
30
2t.4
2l
35
23.0
25.0
40
240
286
45
)/
()
304
50
2-5.8
258
55
261
261
60
21 7
21 1
65
28.6
28.6
10
295
295
l5
304
304
80
31 .4
314
85
378
323
90
387
332
95
391
34r
r00
406
35
r05
4l
-5
360
il0
42.1
36.9
lt-5
13.4
31 9
)20
441
388
Chap. 3
Suggested Readir
forward and reverse traverses'

3-14. The fotlowing time-diStance values were recorded from
refractor, use the delay-time method
bedrock
a
on
information
your contract requires detaited
Because
Palmer, Derecke, 198

Oklahoma. SocietY
Sternberg, Ben K., Po

port of the Arizol
GeoPhYsics; Volur
132
time is 40-7 ms.

io obtain depths and then construct rhe bedrock surface. Reciprocal
8.3
83
20
14.2
t6'7
30
168
25.0
40
20.3
30.0
50
213
30.2
60
256
3)
10
28.2
341
80
28.4
34. t
90
31 8
36.1
100
345
369
110
355
379
39.7
40.5
l0
120
Redpath, Bruce B., 1!

Report E-734, ExP
Geophysics, no. 5,1
Thornburgh, H. R., I
Association of Petrr
SUGGESTED READING
Lankston, Robert W.,

Ward, StanleY H., t
al: SocietY ofExPk
Musgrave, A. W.,
ed.,
CeoPhYsicists, 604
Palmer, Derecke, l98t
Redpath, Bruce B.,
REFERENCES CITED
refraction method of geophysical
Adachi, R., 1954, on a proof of fundamental formula concerning

prospectingandsomeremarks:KumamotoJournalofScience,seriesA,v.2,p.|8-23,
Hampshire: Ground Water, v. l4'
Birch, Francis S., 1976, A seismic ground.water survey in New
94
p.
100.
years ago an earthquake shook

Bonini, w. E., and Bonini, R. R., 1979, Andrija Mohoroviti6: Seventy
Zagreb'. EOS, v.60, no.41, p'699-101'
Cerveny,V.,Langer,J.,andPsencik,|.,1g74'Computationofgeometricspreadingofseismicbody
media with curved interfaces: Geophysical Journal of the Royal
waves in laterafly inhomogeneous
Astronomical SocietY, v' 38, P' 9 19'
Dobrin,MiltonB.,andSavit,CarlH.,lg88,Introductiontogeophysicalprospecting'4thed.:New
York, McGraw-Hill, 867
Kearey, P., and Brooks'
l!
Report E-734, Exp
M., 1984, An introduction of geophysical exploration: London, Blackwell
Scientihc Publications, 296 P'

tool for mapping shallow targets
I-ankston. Robert W., 19g9, The seismic refiaction method: A viable
into the 1990s: Ceophysics, v' -54, p' 1535-1542'
data acquisition and interpretation in
Lankston, Robert W., 1990, High-resolution refraction seismic
volume l: Review and tutorigeophysics,
ward, Stanley H-, ed., Geotechnical and environmental
no. 5, p. 45-73.
in
Geophysics,
Invesrigations
al: Society ot:Exploration ceophysicists
Mooney,HarotdM.,lgTT,Handbookofengineeringgeophi,sics:Minneapolis,Minnesota,Bison
Instruments. Inc.
Nettleton'L.L,lg40,Ceophysicalprospectingforoil:NewYork,McGraw.Hill,442p
Robinson, Edwin
S.,
562 p.
Sheriff, R. E. and Ge
acquisition: Cambr

SeismicRefra Burger92

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

SeismicRefra Burger92

Uploaded by

Copyright:

Available Formats

EXPLORATION G EOPHYSICS

Accompanying Macintosh Computer Software

Refraction seismology has a long and distinguishcd history. Although seismographs

frorn a field seismogram. From the geophone spacing (geophone inter-

Seismic Exploration: The Refraction

This means that if we

the subsurlace is minir

face because of its lor,r

Figure 3-l Generalized diagram illustrating

You realize, of course,

will limit our analysis

tious time to review Sr

ically refracted ray foll

the interface between

can determine the velocity

A Single Subsurface lnterface

SINGLE SUBSURFACE INTERFACE

will produce reflections, refractions, and wave conversions. At the nroment we

Vclocity = 286 m/s

Seismic Exploration: The Refraction Method

Equation 3-8 is the ba

at this point, but Eq.

care ofthis task bcfort

Next, we utilize the re

Diagrant illustrating syrntrols uscd in dcrivation ol tirnc of travel lbr critically

to change Eq. 3-9 to

Derivation of Travel-llme Equation

and the identities notei

use the relationships and quantities in Figure 3-3.

With the llnal fomr ol

tA Single Subsurface lnterface

Next, we utilize the relationships

to change Eq. 3-9 to

Using the familiar relationship

identities noted previously, we can substitute to obtain

Seismic ExPloration: The Refraction

the critically refracted ray is an

ray can be utilized in

wave arrives first at

distances fiom the en

would be the case whr

However. the travel-ti

A Single Subsurface lnterface

Times for Direct and Refracted Waves

CcoPlronc intcrval trn

Seismic Exploration: The Refraction Method

sect at a point. The I

until you are satisfied that the relationship holds. (Actuatly, if

The VrV, terms

plify Eq. 3-20 to

3-5. Please note that

tables used in this text,

A Single Subsurface lnterface

equal, and, therefore,

terrns cancel, and remembering that

(Vl- Vl) = (Vz- V) (Vz+ V,), we can sim-

plily Eq. 3-20 to

x'' (" ' = 2\Vr+V,) ''\t''

Seismic ExPloration: The Refraction

This diagram illustratcs the relationships lbr calculating critical disttnce

Critical distancc is thc nlinirnum distance

Dcrivatiorr of an cquation to calculate critical distance is straightfbrward (u'hich you

You rnay have