You are on page 1of 16

Downloaded 09/11/14 to 201.102.102.23. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.

org/

GEOPHYSICS,

WAVE-FRONT

ALBERT

W.

CHARTS

AND

VOL.

THREE

XXVI,

NO.

6 (DECEMBER,

DIMENSIONAL

1961),

PP.

738-753,

25 FIGS.

MIGRATIONS*

MUSGRAVE?

A computer is required to calculate the complex wave-front charts which are needed in many areas. On a medium size computer wave-front charts can be constructed using up to 40 layers. Each layer can be a constant
velocity or can start with any velocity and have an increase in velocity with vertical time These wave-front charts
may be automatically plotted for use in migration in a vertical plane. At the same time that the wave-front chart
is being obtained, a list may be made which shows the depth and offset for each reflection time and stepout value.
This migration list may be used to migrate values from time maps in three dimensions. Before migrating, these time
maps should have contours of all time values even though overlapping occursas on buried foci of sharp synclines.
Thus, it is a simple matter to make a migrated depth map from any time map regardlessof the crookednessand

discontinuity of the profiles or the lack of cross-linecontrol.


INTRODUCTION

The geophysicist refers to the leading surface of


the seismic disturbance as the wave front. The
wave front is a spheroidal surface expanding from
the shotpoint with increasing time In isotropic
media, raypaths are imaginary lines constructed
perpendicular to the wave front at all times. The
wave front is distorted from a perfect sphere by
velocity changes in the earth through which it
propagates.
Wave-front charts are made by the calculation
of wave-front intersections, at a series of time
increments, with a chosen set of raypaths in a
vertical plane through the origin. In order to
perform this calculation, assumptions about the
subsurface velocity configuration must be mathematically describable. Before electronic computers were available, velocity configurations
had to be limited to those which could be represented by simple mathematical functions.
VELOCITY

ASSUMPTIONS

Velocity may vary in both the vertical and


horizontal directions. Wave-front charts are usable only in areas where velocity does not vary
horizontally or varies in particular ways as described in the authors doctorate thesis (Musgrave, 1952).

Figure 1 shows the two basic isovelocity configurations as taken from this thesis. The diagram

on the right shows an area where isovelocity surfaces are parallel to or equidistant from the reflector. Velocity has a horizontal gradient as a
result of vertical changes, so wave-front charts
and migration lists are difficult to use. Raypaths
are straight lines in this case and the velocity
changes near the origin. The special case of having
small variations with depth will allow the use of
fairly simple charts and lists. The diagram on the
left shows an area where isovelocity surfaces are
parallel to the datum plane. If the datum plane is
horizontal, velocity has no horizontal gradient
and a wave-front chart will resolve the two dimensional dip profiles. A migration list will resolve
the dip in three dimensions.
Vertical

Variations

Vertical velocity distribution in the earth is determined by velocity surveys. This determination
can be made more accurately from continuous
velocity logs. Velocity variations in the vertical
direction are much more severe than in the horizontal direction. Several continuous functions
have been used to approximate various vertical
configurations. Electronic computer programs
can use discontinuous mathematical functions
which represent the vertical distribution of velocity with any amount of detail required. Figure 2
shows a time versus velocity plot for three velocity functions. Average velocity is shown by solid
lines and instantaneous velocity is shown by

* Presented at the 29th Annual Meeting, SEG, Los Angeles, November 11, 1959. Manuscript
Editor April 18, 1961.
t Mobil Oil Company, Dallas, Texas.
738

received by the

Wave-Front

Charts

739

Downloaded 09/11/14 to 201.102.102.23. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

PARALLEL TO DATUM SURFACE

PARALLEL TO REFLECTOR

MIGRATION SOLVED BY WAVEFRONT


CHARTS AND MIGRATION LISTS

GENERALLY NOT SUITED FOR


MIGRATION BY THESE METHODS

FIG. 1. Two general classifications of isovelocity configurations.


dashed lines. The curved lines represent a linear
increase with depth function that is the basis of
the wave-front chart shown in Figure 3. The
straight lines represent a linear increase with
time function (based on the same shallow velocity
control as used for the linear with depth function)

that is the basis of the chart in Figure 4. Finally,


the lines originating at 6,800 ft/sec represent a
two-layer (discontinuousfunction that is the basis
for the chart shown in Figure 5. The two layers
have different accelerations or rates of velocity
change.

Velocity
AVERAGE

Assumptions

VELOCITY_

linear

time

= vo+

92

lineor

Depth

. $Z

(sKt-l)

INSTANTANEOUS

-c

VELOCITY__

v=vo+o
V=Vo+

-IKZ

7.2
v

= 5900

.
.
.

\\

\\

\\
\\

FIG. 2. time vs velocity plot for three velocity configurations.

--_

740

A. W. Musgrave
LINEAR

Downloaded 09/11/14 to 201.102.102.23. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

WITH

INCREASE
time

OF VELOCITY
ASSUMPTION

Although the linear-with-depth


function is
more popular because of its simplicity, the linearwith-time function has several advantages as the
basis of wave-front charts. Some of these advantages are: (1) a better empirical fit to actual
velocity data is possible, (2) multilayer charts can
be constructed, and (3) the constants are more
easily determined than for a linear-with-depth
function.
First Layer
FIG. 3. A reflection wave-front chart made on the basis
of a linear increase of velocity with depth.

FIG. 4. A reflection wave-front chart made on the basis


of a linear increase of velocity with vertical time

FIG. 5. A wave-front chart made using two layers with


a different increase of velocity in each layer.

Figure5 shows a diagram of a raypath and the


derivation of the equations used in calculating
wave-front charts. The integral equations for
time and distance are derived from Snells law
and trigonometric relations. Substitution of the
linear increase of velocity with time assumption
into these general equations gives the parametric
equations for time offset, and depth. Notice that
the time along any raypath is directly proportional to the dip (Y, that the parameter 0 equals
ICY, and that the equations for offset and depth
are those of a right cycloid whose axes are shifted
by amounts x0 and ho. A right cycloid is defined
as a curve generated by a point on a circle rolling
along a straight line as shown in Figure 7.0 is the
angle of revolution of the circle and R is its
radius. The circle generates the raypath by rolling
along a line on the bottom side of a plane situated
parallel to the datum plane and a distance above
it which is coincident with zero on the velocity
scale of the layer. The times and coordinates of
depth and offset starting from zero for each wavefront-raypath intersection must be calculated.
The values must be calculated all the way down
to the depth of occurrence of the maximum velocity which is denoted by subscripts whose first
character is M. The portion above the point of
occurrence of minimum or beginning velocity is
subtracted. The subtracted quantities are denoted by symbols having subscripts whose first
character is zero. The second character of the
subscripts refers to the layer being used. The
character s denotes length of the subsurface
coverage.

* In the oral presentation of the paper Dr. F. A. Van


Melle brought to the authors attention that this curve
is the same as the curve of the quickest descent of a
particle under the influence of gravity and is called the
brachistochrone.

Wave-Front

741

Charts

Downloaded 09/11/14 to 201.102.102.23. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

VELOCITY
--I

r-x0

DEPTH

-,

a=0
\

SURFACE--

Snells
sin a

\;

Law

sin a

dx

Or

sin

:-=

sin a

sin_
- P

REFLECTOR-

e.

sin a

Assuming

= pv

t =
dx
x

+ana=J&=

= time

V = V, + aT
measured

vertically

t=*-*
x =

R(B-sin

h =

R(l-cos01

0~

2a;

0)

-x,
-h,

Rx I
4ope

FIG. 6. The general relationshipsof time offset, and depth for any continuousvelocity function and specifically
those for the assumption that velocity increases linearly with time

Second Layer
Figure 8 shows the raypath entering the second
layer, designated by subscripts whose second
character is 2. This layer may have any velocity
at its top and any rate of velocity increase, or
acceleration, including zero. A change in these
velocity properties from those of the first layer
will change the size of the circle generating this
second portion of the raypath. The height of the
plane along which it rolls will also be changed.
The plane would be repositioned at the height of
zero on the velocity scale of .the second layer.
The raypath is made to maintain continuity
across the interface by adding its second layer
timesand coordinates to the maximumcorresponding values found on the raypath in the firstlayer.
COMPUTER

PROGRAM

A continuation of this process is used in our


IBM 650 program to carry out the tedious calcu-

lations required when velocity distributions become complex. Up to 40 layers can be handled- by
the program. Data input to the program for control is (1) a chart number, (2) the number of
layers, (3) whether a refraction (1 way time) or
reflection (2 way time) chart is to be calculated,
(4) whether either or both list and plot are required, (5) whether special raypaths need to be
calculated, and (6) information about plot symbols.
Data input to the program for calculation include (1) the initial and final raypath, (2) the
raypath increment, (3) the minimum and maximum depth, (4) the maximum offset, (5) the increment of travel time (6) the initial and final
velocities, (7) the acceleration of each layer, (8)
the depth of each layer, and (9) the scale factor
for plotting if a plot is required.
The program calculates the coordinates for
depth and offset of the intersection of each ray-

Downloaded 09/11/14 to 201.102.102.23. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

742

A. W.

Musgrave

v, = vo, + a,t,
x, = R, (cl, - stn 8.,)-X,,
h, = R, ( I-cos e.,)
2 Ji.4
TI = 0,
R,=y

-h,,

-TOI

se
0,

AT

Where
x0, = R, (%

-St eo, 1

h,,=R,c-me,,)

2e,,

To,= ~, p
P=x

AT

FIG. 7. Diagram and mathematical relations used to calculatethe

cycloidal raypaths through the first layer.


path with each increment of travel time within
the limits of the maximum values. The output
can be both a plot and a list card for each intersection calculated. Special plot symbols are assigned to intersections falling on designated raypaths and wave fronts, usually those having

values that are multiples of five. In addition to


the incremented raypaths, special raypaths may
be calculated which are incident on each interface
at very nearly the critical angle for the final
velocity of each layer. These are needed to control the shape of the wave front near the inter-

V.ot

+ c-f*

x, * R, (%l

-me.*)-X,*

h,=R,lI-cor&)-h,,

-b-h,+

FIG. 8. Diagram and mathematical relations used to calculate the


cycloidal raypath on through the second layer.

Downloaded 09/11/14 to 201.102.102.23. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

Wave-Front

face. The raypaths used here are those defined


by the AT across a chosen unit of length of subsurface coverage.
EXAMPLE

OF COMPLEX

WAVE-FRONT

CHART

Figure 9 shows a plot of time versus velocity in


a fairly complex configuration. This configuration
was used to calculate a wave-front chart, a portion of which is shown in Figure 10. This chart
was plotted by listing the IBM 650 output plot
cards on an off-line IBM 407 tabulating machine.
The platen drive of the 407 is modified for 10
lines to the inch, so plotting accuracy is plus or
minus 1/20th of an inch. Accuracy can be improved by plotting to a large scale and photographically reducing the plots to the desired
scale. The final wave-front chart is shown in Figure 11. The final plotting accuracy is plus or
minus 1/40th of an inch or less than one millimeter.
Values of depth and offset from the list cards
can be read directly into a Benson-Lehner Model
S electroplotter. A plot prepared in this way is
shown in Figure 12.

VELOCITY

FIG.

743

Charts

Both plots of this chart were calculated using


the same velocity function but different raypath
increments or lengths of subsurface coverage.
This illustrates the utility of the program in
adapting the charts to changes in length of the
geophone cable used by the field crew.
Even though raypaths and wave fronts are not
constructed with continuous lines, these charts
are quite usable for plotting dips from reflection
seismic records. The only manual drafting needed
is the addition of identification and labels for the
raypaths and the wave fronts.
MIGRATION

BY

WAVE-PRONT

CHARTS

Reflection wave-front charts are used to plot


depth cross-sections from the time and step out,
or AT, taken from seismic records along a line of
profile. The depth section can be made in two
ways from split continuous records.
First Method
When using the long trace-to-long trace
method, the cross-section paper is placed over the
chart so that the shotpoint coincides with the

(KILOFEET/SECOND)

9. Velocity configurations for a complex wave-front chart.

A. W. Musgrave

Downloaded 09/11/14 to 201.102.102.23. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

744

. .

I.0

. . .

. .

0
.O

:
YY

o-.

Y
.

0
. 0
_ ** ._

0
y
0
-. --/
-_ **
Do.

. *
0
0
*
*
.
. .
. . 0 v . .
. .
0
*
*
Y
. 0
.
0 .
*

*
0
. * Y 0
. . - *o
:
* ..o
.
. . . .O
0 *
.
.

. *
.
0

.
.
. . **
. 0
. . . .o
0
.
0
*
*
0
.
Y

.
.
.
. 0 *

. ,.
.o*
.
Y
o*
.
. .
. Cl
. *
Y O
.
0.
vy
.
*
.
0
V
Y
0
.
.
Y
. . 0 *
. .
0
.
Cl
*
0
.
*
*
0.
.
0 . - . *
.

. 0
0.
V
0
0 *
. . . . 0 . * *

.
0
0
.
. . . *
0
.
. . .
.cl
.

0
. 0
. . *
0
Y

.
.
.
0
.
. . . 0 .
.

0
.
0
. .
Y
.
. 0
Y V
0 v
.
.
. 0 *
.
0
.
.

.
.
. 0
.
0 *
.
. .
- 0 -
.
0
.
0.

*
.
.
*
.
.
.
*
. .
. .
0
.

. *
. 0 .
0
.
.
.
.
Y o
.o.
.
0
*
. .
.
*

0
.
.
0
.
. .
.
.
Y

.
0 .
.
:
O.

o
.
.
O
Y
.
Y
f
.
0
0
.
0 Y
V
.

0
.
. .
.
.
0
.
0 .

0
.
.
.
.
0
.
.
Y
.
.
*
0

0 *
.
0 .
.
- x
.
0
.
.
.
x *
*
0
.

.
.
*
.
0

.
0
.
0
.
.
.
.
.
Y
.
V
. .

0
. . 0.

FIG. 10. A detailed portion of the wave-front chart shownin Figure 11 as plotted on the IBM 407.

origin of the cha:t. The travel time is taken as the


corrected reflection time of the center traces. The
AT is taken as the difference in corrected reflection times of the end traces. These two values
locate the mid-point of the reflecting segment on
the chart. The reflection segment is drawn tangent
to the wave front with a length equal to one-half
the distance between end traces. If irregular
spread lengths are encountered, the end traces
must be corrected for normal moveout and the
AT normalized to the chart. This normalization
is done by multiplying the moveout corrected
AT by the ratio of the length of the spread to the
unit of length chosen for the chart.

Second Met hod


When using the center-to-center method, the
cross-section paper is placed over the chart so that
its origin is midway between two shotpoints.
Travel time is taken as the average of the corrected center trace times on the records taken at
these shotpoints. The difference between the
center trace times is the AT. These two values
locate the reflecting segment on the chart. The
segment is drawn as before with a length equal to
the shotpoint spacing. If this spacing is not equal
to that for which the chart was calculated, the
AT must be normalized as described before, but
no correction for normal moveout is required.

Downloaded 09/11/14 to 201.102.102.23. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

Wave-Front

745

Charts

FIG. 11. A completewave-front chart plotted by the IBM 407 for the velocity configurationshown
in Figure 9 and for a subsurfacecoverageequal to 1,320ft.
The advantage of the long trace-to-long trace
method is that the AT is taken from a single
record and is not affected by hole lag, variations
in time break and uphole time filter corrections,
and ghost effect.
The advantage of the center-to-center method
is that no normal moveout correction is necessary
and that the datum correction can be calculated
more easily.
EXAMPLE

OF WAVE-FRONT

CHART

MIGRATION

Figure 13 shows a section where reflection times


are plotted vertically below the shotpoints where
they are observed. A record section would serve
equally well. Any length of reflection segment
may be migrated by normalizing the AT. These
time data are migrated by a wave-front chart into
the depth section shown in Figure 14. With some
knowledge of the geology of the area and dips of
shallower reflections, the depth section might be
interpreted as shown in Figure 15. All of the time
data turns out to be reflected from one horizon,
the data having three buried foci and a point of

diffraction. A buried focus of a reflection results


from a syncline whose radius of curvature is
smaller than the radius of curvature of the wave
front.
MIGRATION

LISTS

Three dimensional migration may be accomplished by using a printed list from the list cards
output by the IBM 650 program. The information punched in the list cards is chart number,
raypath, interface, travel time
depth, and
offset. These cards can be sorted and listed as
shown in Figure 16. The sort for this list was in
order of raypath for each wave front or travel
time An alternate list is made by sorting in order
of travel time for each raypath. Pages of this type
of listing are put together to form the composite
list shown in Figure 17. This list is most useful
in resolving dip in three dimensions.
It is necessary to have an identifiable reflection. Before the list can be used, a time map must
be made of this reflection. All possible reflection
times must be picked and put on the map even

w~~rncwl
will NoI,(,
......
1.11* . I
M **Gee
:,,
z4
2 _
7n.a.m
7.M ::,y

Downloaded 09/11/14 to 201.102.102.23. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

CALCULATED
IBM 650
PLOTTED
BENSON-LEHNER
MODEL S

t
r3
0..
$0
/

,:.., ..
.
,. ..

,g + if
I_
I..m 8,.
0 _,:z
u _.a_
7.-.elo_ ,,....I
.~IrnU.~..~_
-n ~I.m.a

? 0. ,,,, ..~~._,_::-:~~.~~~~-.--.
.I..:..... .,p
:......y
..,, ,_ -.:.,..
;, .... ,,:....
,..
....
...
..
...
.
.
.
.
.
.
.
.
+ ., ...
....
I
**
.:.
....
.
.
.
.
.
,...
.
, ., ,.. .:... ... ..
.: ..:...,
.; ,y
..
.. .... .. ,,... ..
.. J.. $
.:
.. .;; 11,
1.j
] .:I;.,;.., ,.:: _?o,
.+

+.

i
*.

:: :.:
.. %
... .....+ : . : :: ::. . .. :+

.
.,,: .,. ....
%
:. .,.
:.: .;: ,,
,:
%

...

fl
c,
a .p

a0
FIG. 12. A nave-front chart plotted by the Benson-Lehner Model S electroplotter using the velocity configuration
shown in Figure 9 and a subsurface coverage equal to 1,000 ft.

FIG. 13. A time cross-section made from reflection records in a complex area. A record section
may be used just as well or better for this purpose.

Downloaded 09/11/14 to 201.102.102.23. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

25

20

15

IO

-5ooo-

FIG.

14. A migrated depth section from the times shown in Figure 13.

INTERPRETED
25

20

DEPTH
15

+
+
+
+
+
+

SECTION
IO

+
+
+
+
--5ooo-

FIG. 15. An interpretation of the migrated section shown in Figure 14.

Downloaded 09/11/14 to 201.102.102.23. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

though the reflection overlaps in synclines and


appears several times on the same record. These
overlaps may not be meaningful until they are
migrated.
EXAMPLES

OF THREE-DIMENSIONAL

MIGRATION

Figure 18 shows a time map of an anticline


made with values from records taken at all of the
shotpoints indicated on the grid. The migrated
depth map is made by reading travel times at the
shotpoints and interpolating the stepout across
the distance for which the list is calculated,
measured normal to the time contours. One such
time and offset is indicated on the map. These

values serve as entry to the migration list and


determine the values of depth and offset from
Figure 17. Although straight traverse lines have
been used for this synthetic example, the migration system is of even more benefit if cultural
and/or topographic features require shooting
crooked lines.
Figure 19 shows the completed depth map after
all points have been migrated. Notice how much
the depth points converge on the anticline in the
areas of steep dip.
It is interesting to see what happens with a
syncline of this same shape. Figure 20 shows a
cross-section across the east-west axis of a shallow
and a deep syncline. The synclines are narrower
in time than in depth. The buried focus effect of
the deep syncline causes some time overlap.
Figure 21 shows a cross-section across the northsouth axis of these same synclines. The steeper
dips in this direction caused considerable time
overlap which shows the buried focus effect more
prominently. Note that the overlap increases
with the sharpness of the syncline and with
greater depth.
Figure 22 shows the time map of the shallow
syncline. The steeper north-south dip causes more
squeezing of the contours than the more gentle
east-west dip. The small amount of time overlap,
because of the buried focus, is shown in the
center of the map.
Figure 23 shows a depth map of the syncline

ITIC. 17. A part of a migration list sorted in order of time for each stepout.

Downloaded 09/11/14 to 201.102.102.23. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

/
FIG. 18. A time map of an anticline showing the shotpoint grid with 5,000 ft between lines and 1,250 ft between
shotpoints. An example time (T) and stepout (AT) are shown for which depth and offset information is obtained
from the list shown in Figure 17.

FIG. 19. A depth map made by three-dimensionally migrating the time map shown in Figure 18. The
migrated location for each shotpoint is shown by the converging grid.

Downloaded 09/11/14 to 201.102.102.23. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

A. W. Murgrave

FIG. 20. An east-westcross-section


acrosstwo synclines.They are referred

to as a shallowand a deep syncline.


showing the migrated position of all the shotpoints. The displacement of the shotpoints is
greatest in the area of steepest dip.

Figure 24 shows a time map of the deep syncline. The time overlap caused by the buried focus
in the center gives a very complex picture. When
all of these times are migrated and converted to
depth, they give a very smooth syncline as shown
in Figure 2.5.
FIELD

FIG. 21. A north-southcross-sectionof a shallowand


a deep syncline. (Note the increasing effect of the
buried focuscrossingof the time limbs. The amount of
cross-overon the time plots increaseseither with the
narrownessof the syncline or the depth of the syncline.)

METHOD

OF

THREE-DIMENSIONAL

MIGRATION

The migration list made by the electronic computer can be used in any field office to migrate
time maps simply and accurately. The procedure
to follow is to make a time map as complete as
possible. From the time map read values of
travel time and AT. Lay off the offset on an overlay at right angles to the time contours. Post
depths at the migrated points and contour the
depth map. This process is simplified by making
the time contours have the same values as calculated wave fronts on the list, then migrate only
points on contours controlled by data. This elimination of interpolation is the secret of speed.
SUMMARY

The electronic computation of wave-front


charts is an improvement over other methods

Downloaded 09/11/14 to 201.102.102.23. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

c--o-o-D_

FIG. 22. A time map of the shallow syncline. Note the crossed over
or hidden contour in the center (dashed).

FIG. 23. 4 depth map of the shallow syncline showing the divergence of the shotpoint grid.

Downloaded 09/11/14 to 201.102.102.23. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

FIG. 24. A time map of a deep syncline. The crossed over or hidden contours are dashed.

FIG. 25. A depth map of-the deep syncline showing the divergence of the migrated shotpoint positions.

Downloaded 09/11/14 to 201.102.102.23. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

Wave-Front

because: (1) Mathematical assumptions may be


used that reduce the time required to determine
the constants needed for calculation. (2) The
speed of the computer reduces the time required
to obtain a chart with increased accuracy. (3) The
detailed calculations that can be handled by the
machine program allow the use of complex, discontinuous velocity functions. Many layers may
be included having almost exact correspondence
to observed velocity data. (4) Trained personnel
are not require-d to makes thencdhzuhdtions after
the program is completed and a check list of the
input information is made. Personnel may need
considerable review for manual calculations if
charts are infrequently required, but the computer program retains its speed. (5) The migration lists allow easy and efficient migration in

753

Charts

three dimensions. The lists may be made so detailed that no interpolation is necessary.
ACKNOWLEDGMENTS

The author wishes to express his appreciation


to the Mobil Oil Company for permission to
present this paper and to his associates for their
help in preparing it. Particular thanks go to G.
W. Ehlert and W. C. Woolley for their aid and
cirticisms! also to Manus Foster, Charles Hickman, and Helen Gray for their part in writing the
computer program.
REFERENCE

Musgrave, Albert W., 1952, Wave-front Charts and


Raypath Plotters: Quarterly of the Colorado School
of Mines, v. 47, n. 4.

You might also like