Professional Documents
Culture Documents
5813 - Seismic Refraction Prospecting PDF
5813 - Seismic Refraction Prospecting PDF
org/
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
ALBERT
SOCIETY
Edited by
W.
SEISMIC
OF EXPLORATION
PROSPECTING
GEOPHYSICISTS
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
Copyright ( 1967
By
THE SOCIETY OF EXPLORATION GEOPHYSICISTS
PublishedJune 1967
Library of CongressCard No. 67-22554
Publishedby
THE SOCIETY OF EXPLORATION GEOPHYSICISTS
P.O. Box 1067
Tulsa, Oklahoma
Composedand Printed by
GeorgeBanta Company, Inc.
Menasha, Wisconsin
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
To Henry C. Cortes
FOREWORD
vii
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
TABLE OF CONTENTS
Foreword ........................................................... v
Refraction Volume Committee ......................................... vii
Introduction ........................................................ xi
List of SuggestedSymbolsfor SeismicRefraction Volume.................. xiii
Section1--History of Early RefractionWork ............................. 1
Early Refraction Practices.......................................... 3
J. E. McGee and R. L. Palmer
Section 2--Refraction Models ........................................... 13
Analog Model Studiesof the Simple Refraction Problem................ 15
StanleyJ. Laster,Milo M. Backusand RichardSchell
A Method of SeparatingOverlappingRefraction Arrivals............... 67
C. R. Bruce and J. E. White
Section3---Amplitudeof RefractionSignal............................... 83
The Use of Amplitudesin SeismicRefraction Survey ' 85
P. N. S. O'Brien
Method for Calculatingthe Amplitude of the Refraction Arrival ......... 119
Glenn C. Werth
Refraction Along an EmbeddedHigh-SpeedLayer ..................... 138
J. W. C. Sherwood
The Efficient Use of Large Charges................................... 152
P. N. $. O'Brien
Modified "Gardner" Delay Time and "Constant Distance Correlation"
Interpretation................................................... 171
C. Layat
Section4--Determination of Weathering or LVL Thickness by
Refraction Methods .................................................. 19S
Multilayer Near-SurfaceRefractionComputations..................... 197
W. A. Knox
A Curved Path Refraction Method ................................... 217
John C. Hollister
Practical Application of BlondeauWeathering Solution................. 231
A. W. Musgraveand R. H. Bratton
Refraction Refinement Technique.................................... 247
H. L. Mendenhall
Section5--Interpretation Techniquesfor Refraction Work ................ 257
A. G,,R^, IT,RP,TV, T,CIQJ,s .............................. 259
The Midpoint Method of Interpreting a Refraction Survey............ 260
M. R. MacPhail
A Method of In-line Refraction Profiling ............................ 267
W. C. Woolley,A. W. Musgraveand Helen Gray
Outline of a Systemof Refraction Interpretation for Monotonic Increases
of Velocity with Depth ......................................... 290
Haakon M. Evjen
ix
x Contents
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
INTRODUCTION
I. Elements
Shot or shotpointlocation s
IV. Elevation E
Shotpoints&, S% S Ex, E,
Receivers R1, R% R E, E,
V. Horizontal distance
Multiple receiversand singleshotpoint
Multiple shotpointsand singlereceiver
Multiple receiversand shotpoints
Critical distance
Interface
Intersectiondistance(distancefor coincidenttime)
Refraction along interfacesn, n-t--k
H
VI. Depth (vertical)
Interface Hi, H%' ' H.
h
VII. Apparentthickness(vertical)
Layer h, h, h,
At element R hn,Rn
VIII. Thickness z
Layer z1, z% . . . z.
At element in, Rn, * ' '
If dipping;measured
normalto whatinterface Zn, Rn,,
Intercept time To
Interface Tox,To, To,,
For distance and interface
For critical distance of interface n Tn
XlII
xiv List of Suggested Symbols
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
XIV. Raypath/XT//XX P
R4
/
H2
/
/
/
h2 /
/ V2
2
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
Seismic
Reraction
Prospecting
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
Section
1
Section 1
INTRODUCTION
When one of the writers of this paper suggestedthat the Refraction Volume
shouldcontain an article on the refractionpracticesin the early days, it was with the
idea that it would be written by one of the old-timerswho had been on the refrac-
tion scenein the mid-twenties.When the suggestionbackfiredand he was askedto
participatein writing the paper he accepted,assumingthat a sufficientnumberof
excerptscouldbe taken from the best of the early papersto createreaderinterest
in the subject,and references to the paperswouldcompletethe job. Unfortunately,
he was wrong again. After a searchof the literature, it becameapparent that no
paperswere written, at least nonewerepublished,during the heyday of the early
refractionwork and very little regardingthat work was publishedlater. Therefore,
much of the information containedin this paper is basedon memory of discussions
throughthe yearswith someof the old-timers,plusrecentinterviewswith, and the
suggestionsof, a few whowereengagedin the refractionactivitiesof thirty to forty
years ago.
The first papersfoundon the subjectwereby Barton (1929)and Helland (1929).
A part of the discussion on Heiland's paper by Rieber points out the paucity of
publishedinformationat that time. It reads,"Dr. Heiland appearsto have drawn
the materialfor his excellentpaperlargelyfrom experiencewith earthquake seismol-
ogy, including,of course,such direct adaptationsof earthquakeinstrumentsas
the mechanicalseismographs usedin geophysicalwork.
"He saysthat he has beenhandicappedin presentinga more completepaper by
the fact that practically nothing has been published,to date, with regard to the
electricalmethodsof receivingand recordingvibrations, ."
HISTORY
pletewithout
mention
ofthefirstattempt
at seismic
workin thiscountry,
which
wasby thereflectionmethod.Drs.W. P. Haseman,
BurtonMcCollum, E. A.
Eckhardt,
andJ. C. Karcherwereengaged
duringWorldWarI, at theBureauof
Standards,
in thedevelopmentofsound-ranging
equipment
to beusedin locating
theenemyartillery.
Duringthisworkseismic
energy
aswellastheairwavewas
studied.Theideaoccurred
to themthat it shouldbepossible
to mapgeologic
struc-
tureby means
ofreflected
seismic
energy.
Theybegan
experimental
workonthe
ideain 1919.Afternumerous
tests,in manyareas,a two-trace
recordwasobtained
thatpositively
indicated
a shallow
reflection.
Thusencouraged,
theyorganized
the
Geological
EngineeringCompany in1921,withthesupport
ofsomeindependent
oil
operators,
andbegan fieldoperations
in Oklahoma.Short
testprojects
were
shotin
a fewscattered localities,
thelastbeingnearPoncaCity.Theresults obtainedin
oneor twoof thesedidnotagreewiththeaccepted localgeology.
In at leastone,
thatnearPoncaCity,thereflection response
ispoorevenwithtoday's equipment
sotheirresultstherewerediscouraging.
Earlyin 1922thebackersof theoperations
withdrew theirsupport, sothreeof thefourprincipals
returnedto otheroccupa-
tions.McCollurn tookovertheequipment andcontinuedthereflection
experiments.
He alsodirectedsomeof hisresearch
efforttowardtherefractionmethod.
After he had madesufficientteststo be certainthat the refractionseismograph
waspractical,
hesubmitted
apatent
application
totheUnited
States
Patent
Office,
in which
heclaimed,
inpart,thathewasusing
sound
waves
whichtraveled
along
the
surface
ofa high-velocity
bedwithsome oftheenergy
beingdiffracted
backtothe
surface
ofthegroundwhere it waspicked
upbydetectors
andrecorded.
Therewas
noonein thepatentoffice
at thattimewhowasfamiliar
withtheseismograph
so
thequestion
arose
astowhatsection
should
handle
theapplication.
Whenit was
noticedthat soundwaveswereinvolved,the application
wasturnedoverto the
musicsection
forinvestigation.
Thatsection,afterreviewingtheclaims,
notified
McCollumthat thepatentwasdeniedbecause the methodwastechnically
un-
sound.
Hethenpresentedsome ofhisfielddatatoprovethatthemethodworked,
after whichthe applicationwasapproved.
Thefirstcommercial
useofthemethod by McCollum wasin 1924,in Mexico,
for
theCortezOil Company,a subsidiary
of TheAtlanticRefining
Company. Some
success
wasencountered
onthisundertaking,
andthefirstproducer in Mexicoto
belocated
bytheseismograph,
theNo.1,Lot199,Amatlan,intheZacamixtle
field,
wascompleted
in 1925.
Althoughthiswasa fieldextension
well,theseismograph
gets
thecredit
forit because
it was
onalease
which
was
scheduled
tobeabandoned
on the basisof the subsurfacegeology.
Therefraction
method
worked
in oneareain Mexicoin whatmightbetermeda
negative
sense.
There
were
twosystems
offaults
approximately
atrightangles
to
eachother.Wheretwoof thefaultsintersected,
thelimestone
wasfractured
suf-
ficiently
tocreate,
insome
cases,
anoilreservoir
ofcommercial
importance.
These
intersections
weretheobjectives
of therefraction
search.
Goodarrivals
wereob-
tainedwhennofaultwasinvolved
in therefraction
path.Theenergydropped
off
considerably
when onefaultwasinvolved,
butpractically
nonewasreceived
when
Early Refraction Practices 5
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
the path passedthrough the intersectionof two of them. So it turned out that the
absence of data was the criterion for which to look.
While McCollum, Eckhardt, Karcher, and Haseman were working on sound
ranging during the war, Dr. L. Mintrop was engagedin a similar project on the
German side of the conflict, and for the next few years their paths seem to run
parallel. After the war, he continued experimentingwith the in-line method of re-
fraction shootingand in 1919 appliedfor a patent on it in Germany. He organized
a companyin Germanyunderthe nameof SeismosG.m.b.H. In 1923,that company
startedrefractionwork in Mexico for the Mexican Eagle Oil Company.It is reported
that its attempt to map the extensionof the Tomasoporidge in late 1923 was "with-
out very satisfactory results." Also, in 1923, a Seismoscrew was brought to this
country by the Marland Oil Company. It began work along the fault trend near
Mexia, Texas, where its chancesof successwere about the same as thoseof the Geo-
logicalEngineeringCompany'sreflectioneffort a year or soearlier,near PoncaCity,
Oklahoma. The results were also about the same. Work done by this crew late in
1923, along the Gulf Coast, did demonstratethat the method had possibilities.The
Gulf Production Company hired a Seismoscrew in 1924 to work in the Gulf Coast
area. In a short time it discoveredthe Orchard dome, the first salt dome to be found
in this country by the refractionseismograph.The crewfound severalother domesin
a short time. This gave such impetus to the method that others soon entered the
field. Consequently,after 1925 the history of the refraction seismographbeginsto
include several names.
FAN SHOOTING
Although the first salt domes were found by Mintrop's in-line method, the so-
called "fan shooting"techniquewas adopted by all engagedin the salt dome refrac-
tion campaign soon after its development in late 1925, or early 1926. An excellent
brief account of the developmentof this method is found in Weatherby's (1940)
paper' "L. P. Garrett of Gulf was responsiblefor this company'strial of the method
and to him goesthe credit for starting the use of refraction fans. The Germanshad
experiencedgreat difficulty with the interpretation of their profiles.Becauseof the
type of sectionin the Gulf Coast,it was almost impossibleto obtain easily recog-
nizable breaks in the velocity of the different layers and to calculate the depths of
these layers. Garrett suggestedthat two equal-length shots, one acrossa known
dome and the other off the dome, should show marked time differences.The field
trial was highly successfulbecauseof the relatively high velocity of the salt, and it
was at onceapparent that a geometricalarrangementsuchas a fan would give this
time comparisonover a large area from a minimum number of shot points. This
was a tremendousstep forward, for had it been necessaryto use profiling,as orig-
inally intended, the costof coveragewould have beenvery much greater."
In fan shooting,as the name implies, the detectorswere arranged roughly in a
fan pattern with the shotpointat the center.They were usually placedalongroads
(Figure 1) or on the banks of streams,bayous,or canalswhen available to save
set-up time. Therefore, the shot-detector distance was seldom constant for all
History of Early Refraction Work
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
2 ROAD :5 4 12
13 5
I0
II
NE
6 15
I 2 $ 4
I ! I I
CALl BRAT ION OR
SCALE-MILES
NORMAL PROFILE
-5
-4
TIME-DISTANCE POINTS
FOR NORMAL PROFILE
DISTANCE- FEET
5ooo 15,000 20,000 2:5,000 30,000 55,000 40,000 45,000
Fro. 2. Time-distancerelationsfor fan shooting.The line showsa "normal" curve of time versusdis-
tancefor the normal sedimentarysectionin the area beinginvestigated.Points below the line indicate
"fast" firstarrivalsof wavesthat havepenetrateda dome.Numberscorrespond with numbereddetector
positions
of Figure! (fromNettletonop.cit.).
The first crews had no radios with which to transmit the time break from shot-
point to recorders.Sometimesthey strungwiresfor this purpose,but usuallythey
measured,or estimated, the shot-detectordistance,calculated the traveltime of the
air wave, then set the time-break back from the arrival of the air wave. After radios
were adopted to transmit the time-break, the crews dispensedwith the cost and
time required to survey the shot-detectordistance, and calculatedit from the air
wave, taking into accountthe velocity and direction of the wind and sometimesalso
the temperaturesand the barometric pressure.The inaccuraciesinherent in these
procedurescouldnot be toleratedin today'srefractionwork, but were permissible
in the earlyfan shootingwheretime leadswerecalculatedin tenthsof a second,not
thousandths.
Radioswereadoptedby mostof the crewssometimeduringthe fan campaigns.
Oneunpublished
reportindicatesthat thiswasthe principalimprovement
in equip-
ment which permitted the secondfan campaignto be successful. While the radio
wasusedto coordinatefield operations, its primary advantagewasin transmitting
the time-breakto the severalrecordingunits, therebypermittingmuchgreaterac-
curacyin determiningthe first arrival timesfor all geophones. This increasein the
accuracywas essentialto the discoveryof deep domesbecausethe sedimentary
velocitiesincreasefairly rapidly with depth in the Gulf Coast;therefore,the time
lead causedby a dome whosetop is 6,000 ft below the surfacewould be far lessthan
that of a dome of the samesizewhich risesto within 1,000 ft of the surface.
Competitionwaskeen betweenthe oil companiesengagedin the salt domecam-
paignof the middleand late twenties.Each companyhad scoutswatchingthe ac-
tivities of other companies'troopsto try to guesswhat they werefinding.This is
wellillustratedby the followingpassage from Barton's(1929)paper.In discussing
the field procedureafter the first indicationof a dome was obtained,he says:
"Theoretically, it is advisableto repeat the shotsand to throw a fan acrossthe sus-
pecteddomefrom another angle. But, unlessthe rival scoutscan be sidetrackedor
unlessthe suspected domeis wholly under land controlledby the company,any
apparent halting to reshootmost commonlywill bring one or more rival crewson
the groundwithin forty-eight hours.The attempt, therefore,ordinarily is made to
sneaktwo or threeadditionalshotsacrossthe domewithoutseemingto be doing
any reshooting.The companythen blocksthe prospectingas quietly as possible
andlater sendsa troopbackin to determinewhetherthe relaykey stuckonsending
out the time of the explosion,whethera stray limestonebed may have beenpicked
up, or whether a salt dome is present, and if so, to outline it and determine its
depth."
One of the few publishedpapers,and a goodone,whichdescribes fan shootingis
the casehistory of the Vermilion Bay salt dome by Rosaire and Lester (1932).
Other descriptionsof the method can be found in the bookson geophysicalex-
ploration.
OTHER METHODS
came the center of the fan and the shotpointswere locatedaround the periphery,
the reverseof the normal fan shootingpractice. The advantageof the method was
deeperpenetration of the refraction path for a given shot-detectordistance than
when both the shot and the detector were at, or near, the surfaceof the ground.
The methoddoesnot appearto have beenwidely used.The samepaper introduced
a methodof detailingthe flanksof a salt dome by utilizing one or more deepflank
wells, a method which is still in use, therefore,it doesnot fall within the scopeof
this paper.
A system of ring, or arc, shootingemployed before 1933 in southwestPersia
wasdescribedby Jones(1934). While this is the only publishedpaperon the method
that is known to the writers, McCollum pointed out in a private interview that he
appliedfor a patent in 1923,whichwasgrantedin 1929,coveringthe basicprinciple
of the method.He saidhe usedit extensivelyin the 1920's.The detectorsetupwas
somewhatsimilarto that of fan shooting,however,the objectiveof the methodwas
to map refractinghorizons,not to discoversalt domes.For eachring, or circle,the
shotpointwaskept fixed at the centerwhile the detectorsoccupiedvariouspositions
around its circumference.The assumptionwas made that the delay time at the
shotpoint end of all trajectorieswas constant. Therefore, the variations in delay
time amongthe variousdetectorswereassignedentirely to variationsin depth of the
refractinghorizonat the detectorendsof the trajectories.This is an approximation
which might be permissibleif the offset distancesat the shotpoint were small, i.e.,
the refractingbed wereshallow,and/or if the bed werefairly levelin the vicinity of
the shotpoint.By a proper layout of overlappingcirclesabsolutedepths could be
determined. Gardner (1939) describeda variation of this method in which the ob-
jection just mentionedis eliminated.In it the shotpointis displacedfrom the center
of the circleby the offsetdistanceso that the refractionpath entersthe refracting
horizon at the circle'scenter. Other than the use by McCollum mentionedabove,
this methodseemsto have seenonly a limited usein this country.
EARLY FIELD PROBLEMS
Many of the routine tasks of the middle and late twenties would be considered
hardshipsby a field crew today. While the Model "T" was reliable, the roads, es-
pecially in winter, were sometimesalmostimpassible.There were no drill rigs to
dig shotholes.They weredug with an augerby two men who bent over, bore down,
and turned to the right. The holeswere usually dug to a depth of 20 to 30 ft, and
sometimesas many as 20 were required for a singlerefraction shot. As mentioned
earlier, there were no radios on the first crews, so when exact overall time was
needed,as on most profileshooting,it was necessaryto string a conductorbetween
the shotpointand the recordingsetup,which couldbe a distanceof many miles.
Quite a lot of refractionprofiling was carried out in South America during the
late twentiesin which no overall time was recorded.The interpretation depended
only on interval times betweengeophonestations. Since radio or telephonecom-
municationbetweenthe operatorand the shooterwasnot absolutelyessential,none
was provided.It then becamenecessaryfor the operatorto sendsomekind of mes-
sage to the shooterto tell him when to fire the shot. This was accomplishedby
lo History of Early Refraction Work
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
having two pocket watchesso carefully calibrated that the drift between them was
known to a fraction of a secondby the operator.To insurethat care was taken at all
timeswith the watches,they werealwayscarriedin paddedsaddle-soap cans,which
not only protectedthem from physicalabusebut alsofrom outsidemagneticfields.
When all geophoneswere in place and everything was ready for the shot to be
fired, the operatorwould checkthe watches,then sendoneof them by a native down
the "pica" (cut trail) to tell the shooterto fire on the 60th secondcomingup at a
predeterminedminute, allowing ten minutes for each kilometer the native had to
travel. The operator would calculate the time of arrival of the refraction energy,
taking into considerationthe drift of the watchesand the velocity at which the
energy would travel along the refracting formation. He would allow several seconds
for possibleerrors when starting his recording camera. This method worked so
effectively that during one 18-month contract, only two shots were missed. On
one occasionthere was a large river between the detector setup and the shot-
point. In this case,it was necessaryto give the shooterhis watch and to set the
firing time the night before,becausehe had to spendthe night acrossthe river from
the recordingequipment. The shot was successfullyrecordedwhen it was fired on
the 60th secondcomingup at 9'00 o'clockthe next morning.
Those who participated in the early refraction were the pioneersof the seismic
profession.
(1) They altered trucks to make them more useful in seismicfield work which
has led to present specializedautomotive equipment.
(2) They adoptedthe radio to field operations,and today it is an essentialpiece
of equipment in all refraction work and in most reflection operationswhere the
shootingand the recordingunits are separatedby more than a few hundred feet,
especiallyin water, marsh, and swamp work.
(3) They developedat least four refractiontechniques'
(a) Fan shootingwhich, though currently inactive, would certainly play a
major role in a new salt dome province shouldone be discovered.
(b) Arc shooting,whichnow seemsto be dormant,may still have someprofit-
able applications.
(c) Radial-refractionshooting,a methodof outlininga salt domeby utilizing
a deep well bottomed in or near the salt, is still in use.It will be discussedin
a later paper of this book.
(d) The basicprinciplesof the in-line refractionmethodof mappingsubsurface
horizons which, with some refinements,is in current use. It will also be
discussedlater by others.
ACKNOWLEDGMENTS
The authors are deeply grateful to the followingmen for their suggestionsand
comments:D. ?. Carlton, E. A. Eckhardt, J. J. Jakosky, J. C. Karcher, W. W.
Early Refraction Practices 11
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
REFRACTION
Section
2
MODELS
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
Section 2
Refraction Models
14
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
ABSTRACT
A two-dimensionalanalog model with a cylindrical structure has been constructedto simulate the
simplerefractionproblem.Experimentalseismograms showingthe first refractionare presented.From
suchrecordstraveltime curvesare constructed.The structure,as predicted,appearsas an anomaly in
the traveltime.Plotsshowingrefractionamplitudeversusdistancealsoshowan anomalyin the neighbor-
hood of the structure.
A numberof othercoherenteventsappearon the experimentalseismograms, all exhibitingthe horizon-
tal velocityof the refractor.Theseare shownto be refractedreflectionsand reflectedrefractions,many
of which containcompressional-to-shear conversions.The structuralsignificance of theselater events
is shownby traveltimeanomaliesand amplitudeanomaliessimilarto thosefor the first refraction.
Finally, a brief discussion
of the shallowrefractorproblemis included.In this case,the traveltime
difference betweenthe multiplerefractions is small,sothe seismogram hasa ringing,"mode"structure.
DESCRIPTION OF MODEL
SOURCE
BRASS I 9CM
CONTOUR
CENTER
77CIV
'STAINI,I:.S
STEEL !
WAVE VELOCITIES (M/SEC.) SCALING
P WAVE S WAVE ICM = 1000 FT.
BRASS: 3850 Z 1oo 11zS = 30 MS.
STEEL: 5290 3100
R 60 cm
/ DELAY
130
/ NORM
GAIN IO
f SWEEP 2/. $eC/cm
FILTER * L.C 500 CPS
HC 2 50 KC
Generaltheory
Model seismograms wererecordedfor a continuous profilestartingat 5 cm from
the sourceand extendingto 205 cm from the sourcein one-cmincrements.Figure
3.1 and 3.2 showtypical recordsfor moderatesource-receiverdistances(35-55 cm).
(Note that theserecordsare takenovera flat portionof the interface.)The direct,
refracted,and reflectedeventsare identified. Relevant raypaths are shown in
Figure 4. For distancesup to about 40 cm the direct wave, P, is the first arrival.
Beyondabout 50 cm the refraction,P., is the first arrival. Somewherebetweenthese
two distancesthe crossoverpoint is reached.The reflection(denotedP.x) is never
the first arrival, but is of largeramplitudethan eitherthe direct or refractedpulses.
From theserecordstraveltime curveswere constructedfor the direct, refracted,
and reflectedevents.Theseare shownin Figure5.
Having presentedtypical recordsand identifiedthe refraction,it is worthwhile
to considersomeresultsof theoreticalinvestigations describedin the literature. In
particular,we quote Jeffreys(1926) and Heelan (1953). Jeffreysconsideredthe
18 Refraction Models
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
R. rJ.,ATI'VE
AJdPL, ITIJDI:
FREO.
UENCY
(Kcps)
Fro. 2.2 Frequencyspectrumfor refractedpulseP2 as shownin Figure 2.1.
Measured at 60 cm from source.
Model Studies of Simple Refraction 19
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
DELAY
I0 and I p sec Irkers
R (pSEC)
36
37
38
39
4o I00
Norm. Gain - I
FILTER: L.C. - CPS
H;C. '- 2) KC
128
itom. Gain - I
FILTERI L.C. - 00 CPS
20 XC
rl D
(3) If the direct wave has a pulse shapef(t), the refraction has a pulse shape
ff(t)dt.
(4) The amplitudeof therefraction
is proportional
to [1--(a/a2)2]-/2.Thus,the
refracted amplitude is larger for small velocity contrasts.The amplitude ap-
pearsto have a singularityfor a =a.. However,for this casethe refractedwave
and direct wave mergeinto oneevent, the direct wave for an infinite medium.
Heelan consideredthe point-sourceproblem for two solid materials in contact
with assumptionssimilar to thoseof Jeffreys.However, he did not require that the
rangebe muchlarger than the distanceof sourceor receiverfrom the interface. For
times near the arrival of the refraction, the following resultswere obtained'
(1) The amplitude decay relationshipis given more accurately by 1/(r/'La/')
where L is the distance traveled in the refractor. Note' For the two-dimensional
situation this factor is 1/L a/'(Newlands, 1952).
(2) The relative amplitudesof the reflectedevent are proportionalto the plane-wave
reflection coefficientsevaluated at the appropriate angle of incidence. The
amplitudesof the refractedeventsare evaluatedin a different mannerfrom the
sameexpression.Heelan too finds that the refracted wave pulseis the integral
of the direct wave pulse.
The two works cited are primarily aimed at discoveringthe propertiesof the
critical refraction. However, these and other authors do briefly considerthe reflec-
tion. The primary resultsfor this case,valid near the reflectiontime, are as follows'
22 Refraction Models
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
(1) For distancesfrom the sourcelarge enoughto ignore wavefront curvature, but
insidethe critical angle, the reflectionhas very nearly the samepulseshapeas
the direct arrival.
(2) Outsidethe critical angle the reflectionpulseshapeis altered slightly by what is
variously termed "phase distortion," the "pseudo" reflection, or the refraction
"tail." This effectis discussedby Dix (1961). Even here however,the reflected
and direct waves have rather similar pulse shapes,so that the refraction pulse
shapeis roughly the integral of the reflectionpulse shape.This is borne out by
the experimental recordsin Figure 10.5.
The previousresultsare examplesof the use of "near-front" or "first-motion" ap-
proximation which yield expansionsof the solution around the arrival time for a
particular event. It is also possibleto obtain exact closedform solutions(at least
for a two-dimensionalmodel) by the method of Cagniard (1939; 1962). In this
instance, the total motion as a function of time is given for each ray. (The term
"ray" has a slightly different meaning here; thus the ray which leaves the source
as a compressionalwave, touchesthe interface once,and returns to the receiver as
a compressionalwave includesboth the P. refraction and the Px reflection.) An
extensionof the Cagniardmethod due to de Hoop (1960) is particularly simpleand
elegant. Using this method, exact seismogramscorrespondingto model records
have been computed for various distances. These will be discussedin more detail
later. However, as a typical example, the vertical displacementfor the PP ray
200
time (i sec)
160
Reflection
2 Refraction
p
120 2x
Direct P
80
40
distance (cm)
0 20 40 60 80
IMPULS!.
SEISMOORAM
P2x
SEISMOGRAM FOR
IMPULSE CONVOLVED
WITH SOURCE PULSE
i !
I
EXPERIMENTAL
SEISMOGRAM
TIHE (.ICROSECOHDS)
I p sec Markers ,
DELAY
IO6
o? 232
io8
o9 g36
IiO
III 21;0
112 2
113
11,5 8
FxG.
7.1
High-gain
records
forfirst
arrival.
Sweep
rate
one
sec/cm.
Vertical
motion.
DELAY
I p sec Harkers
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
R (,sEc)
gr7
gl9
I17
118
119
12o
Igl
12'2 261
g63
IgJ
g6T
Norm. Gain -
F I LTER: L.C. - .500 CPS
H.C. - 20 KC
Fxo.7.2 High-gain
records
forfirstarrival.Verticalmotion.
Sweep
rateonesec/cm.
Model Studies of Simple Refraction
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
127
13o
131
132
133
134
13.5
FzG.7.3High-gainrecords
for firstarrival.Sweepone/sec/cm.
Verticalmotion.
28 Refraction Models
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
136
137
138
139
IO
111.2
FILTER:
L.C.
- CPS H.C. - 20 KC ./
Fro. 7.4High-gain
records
forfirstarrival.Sweep
rateonetsec/cm.
Verticalmotion.
Model Studies of Simple Refraction 29
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
/ /
DELAY
R
310
31
/
318
3:
Nora. 6aln =
FILTEr=
FiG.7.5High-gain
records
forfirstarrival.
Sweep
rateonesec/cm.
Vertical
motion.
3O Refraction Models
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
PHYSICAL LOCATION
f OFSTRUCTURE
-0
/
I
-1.0
-3.0
I I I I I I I ,
110 1Z0 130 140 150 160 170
DISTANCE (cm)
SOUR CE RECEIVER
FIG. 8.2 Minimum time raypathfor Snell'slaw computationin Figure8.1. Not to scale.
Model Studies of Simple Refraction 31
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
OFFSET (FEET)
120,000 I$0,000 140,000 150,000
I I I I
6.5 6500
7.0
7000
(.
7.5 7500
8.0 8000
r 8.5 e5oo
9.0 9OO0
ACTUAL STRUCTURE
COMPUTED STRUCTURE
I i i I
120 I$0 140 150
OFFSET (cm)
FIO. 8.3 Structuralinterpretationof experimentaltime residuals.Solidline is actual structure.
Vertical scalegreatly exaggerated.
servedwhich is identified as P4. The raypaths for theseevents are shown in Figure
12.1. Observationof higher gain records(Figures10.9-10.16) showsthat the wave-
form of theseevents are quite stable. Indeed, a long succession of stable arrivals, all
propagatingwith the refractor velocity is evident. For reference,Figure 11 shows
theoretical traveltime for the most important events.
Some confusionseemsto exist concerningthe raypaths possiblefor a multiple
refraction. Considerthe raypaths for P4 shownin Figure 12.2. The top and middle
raypaths are possible,and both occur,as will be shown later; the bottom raypath
is impossible.The first path representsthe refraction of a reflectedwavefront. The
reflected wavefront is of coursecurved, and may give rise to a refracted event of
finite amplitude. The secondraypath showsa refraction that has been reflected.
The refractionis a plane wave (in three dimensionsa conicalwave), and thus does
not give rise to anotherrefractionwhereit touchesthe lower boundary.This latter
fact thus excludesthe third raypath. Pictures of the wavefronts for the first two
raypathsare shownin Figure 12.3at two differenttimes.The smallfront C" is the
wavefront for both the raypaths considered.If an observerlocated on C" looks
forward, it appearsthat C" is a continuationof the fronts C and C', i.e. a reflected
refraction.Looking backward,C" appearsto be a simplerefractionconnectingthe
doublyreflected
wavefront
B" withthewavefront
D in thebottom.Thus,C" con-
tains both a reflectedrefraction and a refracted reflection.It is quite important to
understandthat the wavefront D' arisesfrom the refraction of B', not the refraction
32 Refraction Models
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
A=C
Oo
.Ol
L
FiG.9.1 Peakamplitudefor first arrival,P2,versusdistance,
L, traveledin refractor.Straightline is
theoretical
prediction
ofamplitudebehavior. Noteamplitudeanomalyin neighborhood ofthestructure.
Model Studies of Simple Refraction 33
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
PHYSICAL LOCATION
OF STRUCTURE
1.0
Z. 0
4.0
/ AMPLITUDE
RESIDUAL
!
6.0
-1.0
-Z.0
IDUAL
-3.0
-4.0
i i i i i i i i i
DISTANCE (crn)
FIG. 9.2 Peak amplitudeanomalyin decibelsfor first arrival, P., plotted versussource-receiver
dis-
tance,R. (At top) For reference,the wide-bandtraveltime anomalyis shownat the bottom.
are both somewhatsmaller than P. and both have oppositepolarity to P.. Also, the
PaS refraction is larger than the first arrival P.. These predictionsare borne out by
the experimentalseismograms.Second,it is seenthat while all the refractionshave
the samepulseshapefor about the first cycle, the trailing edgesof the pulsesdiffer.
For example, the /'4 pulse is measurably different from that of P.. Finally, the
importance of the ray events involving compressional-to-shear energy conversions
is shown.Refractionsof this type make up a large part of the later arrivals.
It must be admitted that some disturbing discrepanciesstill exist between the
experimentaland theoretical results. First, the sourcepulse for the model has not
Refraction Models
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
I
10 and I sec Iarl
DELAY
R (SEC)
10
11 22
12 2
13
14
15 30
FIG.10.1Typical
model
seismograms
recorded
withtoptransducer.
Vertical
motion.
Model Studies of Simple Refraction 35
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
T ,/
18
19
23
Non.. Gain- I
FtLTER: L.. - CPS
H.. -eO Kg
26
3o
31
33
7o
PS
Norm. Gain - I
FILTER: L.C. - ,OC)CPS
H.C. - 2.0 KC
4O lOO
41 1 O2
ilo
,,,,
Norm. Gain - I
FILTER: L.C. - CPS
H;C. - 2,0 KC
;.
I0 and I I c Markers
DELAY
(,sc)
99 210
97 212
I oo
IOl 216
217
218
: :. Id
P2 PS+SP P2x ,
Norm.
FILTER; L.C. - X) CPS
#c.. 2o KC
Fro.10.6Typicalmodel
seismograms
recorded
withtoptransducer.
Vertical
motion.
40 Refraction Models
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
231
107
IO8
IiO
Ill
112
238
Fxo.10.7Typicalmodelseismograms
recorded
withtoptransducer.
Verticalmotion.
Model Studies of Simple Refraction 41
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
DELAY
(,SEC)
I0and
I.sec
r,rs .
'
116
117
118-
119
2 :
121
122 27
P PS+SP 4 2x
Norm. Giln = I0
FILTER: L.C. = CP$
#.(:. - 2,10 K(:
;
VI
DELAY I0 end I I sec Markers
(,soc)
I17 gag
118
119
,,_J
12o
121
122
gJ8
gJ9
P2x 'i
PS+SP
Norm. Gain - .0
FILTER: L.C. - CPS
H.C. - 2.0 KC
128
3o
131
133
DELAY
261
(SEC)
Model Studies of Simple
Refraction
i
I0 nd I I sec Irkers
;
.I
43
44 Refraction Models
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
I0 nd I I sec Markers
DELAY
R (,SEC)
139
: : :
.
137
139
PS SP P' '
Norm. Gain - ,0
FILTER: L.C. - CPS
H.C.
Fro.10.11Typicalmodel
seismograms
recorded
withtoptransducer.
Vertical
motion.
Model Studies of Simple Refraction 45
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
3Ol
3O'2
303
3o
3o:
3o6
3o7
3o8
3o9
310
Norm. Groin - ,0
FILTER: L.. - 00 PS
#.. - E'JO K
155 320
156 321
157 322
159 321
16o
I
163 38
39
16.5 330
SP+PS
DELAY
IO and I a see s
b, SEC)
3
166
167
168 31+3
17o
172 31+7
173 38
39
, Ill
3o
p. SP+PS P
Norm. Gain = 0
FILTER: L.. = PS
#.c. =oKc
DELAY
(.SC)
36o
176 36l
177
178
179
180
.181
182
183 368
369
370
DELAY
(SEC
390
197
393 ,
199
201
397
399 -
P2S2
Norm. Gain - .O
FILTER: L.C. - )O IP$
#..
500
.o
time (izaec
440 Rayleigh
Wave
,
P4X ,...'
.'"'..?....-
'"'/'"'
/ P6
....' ....'.;:)'
.....'
,' ..... ....;,<//.
380 / ...."...'2.?//('/.('
. ....
..' /'...'.'/
...'/....r_.:y
/ / ,/ //
.....'/..,?/
...'/....>;;''/
320
..'-).X?
.././).::.r
..//.-"2'::'"'
/...'/?.. :::.'
//'
/...'./.....:.."// ./
260
>, ...:' // -
/ ./,...<";.::';,5
Pzsz/ ../-./...-"..-.':-?/
.-
/ 4'':.""..'"?/' '
/ //'....":', ,'
///......../'
zoo
/
/.//...."....:<..;;;'
/.."...' /..'...
/ ...'...'//?'
/ ...'....' .....'
/ ...'.,.//...'/
140
.?/....::.."//
...::,"/ ...'?/
8O
.. distance (cm)
0
0 30 60 90 IZO 150 180
SOURCE RECEIVER
P S
.///// //////////////////////Qz,//////////////////////!/!
S P
. !//I//I//////li////////////////////-/////////////////
r'! o
'1111IIIIIIIIIIIIIZ . / / / / / / // / / / / / ///r/////////////////////.
,/11t//////////////,/,//,,,,,,,H ////////////////////?
SOURCE RECEIVER
SOURCE RECEIVER
SOURCE RECEIVER
Fro.12.2Raypaths
for/>4
refraction.
Topandmiddle
diagrams
represent
possible
raypaths
forarrivals
ofnonvanishing
amplitude.
Bottomraypathisnotpossible
in a flatinterface
model.
SOURCE RECEIVER
B C
SOURCE RECEIVER
Fro.12.3Wavefronts
in thelayerat twodifferent
times,A isdirectwave;B isuptraveling
reflection
frombottom;B' is downtraveling
reflection
fromsurface;B" is secondreflection
frombottom;C is
primary
refraction;
C' isreflected
refraction
(fromsurface);
C isa combination
ofthereflection
ofC'
fromthebottom(a reflected
refraction)
andtherefraction
of B (a refracted
reflection);
D andD are
successive
wavefrontsin the refractor.Compressional
frontsonly are shown.
Model Studies of Simple Refraction 53
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
Theoretical
Vertical
Motion
R = 200 cm
PS+SP
s2
PS
Time in p. Sec.
Fxo. 13.1 Summationof theoreticalray eventsto yield theoreticalseismogram.R = 200 cm. Theseevents
have been convolvedwith a typical sourcepulse.Vertical motion.
P:sS
t I I I
i
II I I i
i
i
R = 60 cm
', I
I
I I t
i
I I t
I II I
I t
I i I I I
155 175 19.5 215 255
Time in j.z.Sec.
P6x
P2x
P2 P3S
PSx
R= 80cm
Time in Sec.
Fro. 13.2 Comparison of theoreticaland experimentalseismograms for representative
distances.
Verticalmotion.Theoreticalseismograms havebeenconvolvedwith a sourcepulse.Theoreticalseismo-
gramis at top in eachpair.
Model Studies of Simple Refraction 55
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
R = I00 cm
Time in p, Sec.
Pzx P3S
R = 120 crn
Time in p, Sec.
FIO. 13.3 Comparisonof theoreticaland experimentalseismograms for representativedistances.
Vertical motion.Theoreticalseismograms
have beenconvolvedwith a sourcepulse.Theoreticalseismo-
gramis at top in eachpair.
56 Refraction Models
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
PS
R = 160 crn
,, R = 180 cm
I
I
Time in Sec.
Fro. 13.4 Comparisonof theoreticaland experimentalseismograms for representativedistances.
Vertical motion.Theoreticalseismograms
have beenconvolvedwith a sourcepulse.Theoreticalseismo-
gram is at top in eachpair.
Model Studies of Simple Refraction 57
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
P .
, \,,, !i
,
sP P4
Fro. 14. Seismograms
recorded
withsidetransducer
alonga verticalprofilein a
layer. R- 100 cm. Vertical motion.
/'S refraction
andcontributes
to thesecond eventmeasured at thesurface.
Finally,
consider thesecond
eventmeasured at theinterface.
Thisrepresentsenergywhich
left thesource
asan$ wave,wascritically
refracted,
andistravelingalongtheinter-
face.Thiswavealsoradiates
intothelayerboth/' and$ energy. The/' energymay
be tracedupwardto thesurface,yieldingtheS/' refraction
andcontributing the
remainder
of the second
eventmeasured
at the surface.This analysiscouldbe
carriedmuchfurther sincethereare a numberof otheridentifiableeventsrecorded.
However,the previousdiscussion is adequateto give someidea of how the seismic
disturbance
is formed.It is apparentthat evenin a simpleone-layermodelwith a
flat interface,wavepropagationcan be quite complicated.
Theobservation of theselaterrefractions
in thefieldoffersmoredifficultythanin
the caseof the model.For onething,the later arrivalsfroma deeprefractorare
almostalwaysobscured by simultaneouslyarrivingeventswhicharecoupled to
near-surfacelayers.Thisisa serious problem,
butcanbeovercome to a largeextent
by useof velocityfiltering.Thetwotypesofinterference of greatest
importance are
shownin Figures15.1and15.2.In Figure15.1theinterfering eventsarerefractions
fromnear-surface interfaces.Theseeventsarerelativelyhighamplitude, butcanbe
eliminatedby velocityfiltering.Thesecond kindofinterfering events (Figure15.2)
cannotbeseparated fromthedeepmultiples by velocityfilteringsincetheyhavethe
samehorizontalvelocity.However,theseeventsradiateenergybackinto the
groundso their amplitudedecaysrapidly.
A second difficultyin identifying
latereventsis attenuation.
Thereis goodevi-
dencethatearthmaterials attenuateshearwavesmorestrongly thancompressional
waves. For example, most shear waves that have been identified have much lower
frequencythan associatedcompressional waves.A direct measurementof attenua-
tion in PierreShale(McDonal et al., 1958)indicatesthat the shearattenuationis
58 Refraction Models
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
SOURCE RECEIVER
DEEP REFRACTOR
SOURCE RECEIVER
DEEP REFRACTOR
able to pick the first breaksfor theselater events,it is necessaryto baseour arrival
times on the time of somearbitrary peak or trough. In using this procedureit is
tacitly assumedthat the waveformof the event doesnot measurablychangeas the
profile crossesthe structure. The primary justificationsfor this assumptionare the
resultswhich will be presented.However, in generalthis will be true so long as the
dimensionsof the structureare large comparedto the dominant wavelength.
Only the secondand third refractionshave been considered.Since fast sweep
recordswere not recordedfor this time region of the seismogram,the times were
taken from Figures 10.6-10.16. Resolutionwas probably no better than one-fourth
tsec. From these times anomalies were constructed. The two refractions considered
each consist of two events which have identical traveltimes in the flat interface re-
gion of the model. The secondrefraction containsPS and SP. Near the structure,
both of thesearrive earlier than they would in the absenceof the structure. However,
PS arrivesbeforeSP, sincein the formercasethe structureshortensthe slowS leg of
the refraction, while in the latter caseit shortensthe fast P leg. The third refraction,
P, consistsof two eventsone of which is reflectednear the source,P', and the other
reflected near the receiver, P". Again, the structure causesboth events to arrive
early, but the event reflectednear the receiverarrivesearlier than that reflectednear
the source,primarily becausethe structureshortenstwo legsin P" and only onein
Thus, four anomalies are observedrather than two. Of these four, SP and P4'
are very similar to the wide-band P. anomaly discussedearlier. The anomaliesfor
PS and P4" are much larger, but are obscuredover a portion of the traverse. The
observationof two distinct P anomalieslendsfurther support to the earlier discus-
sion of raypaths possiblefor a multiple refraction.
The experimental anomaliesare shown in Figure 16 along with the predicted
anomaliescomputed on the basisof Snell's law. Compare this to the wide-band P2
anomaly(Figure 8.1). There is muchmorescatterin the later refractionanomalies,
but this is attributed to the lower resolution of the records from which the times
were picked. The refractionoffsetat the receiverfor thoseeventsendingin S is dif-
ferent from those ending in P, so each event must be migrated the correct amount
to fall on top of the structure. However, the differencein these offset distancesis
small on the distance scale used, so the anomalies appear to be centered at the
same place. No attempt was made to constructan interpretation from these later
refraction anomalies,though no difficulty should be encounteredin doing so.
As in the caseof the first refraction,P2, attention hasbeengiven to the amplitude
behavior of the secondrefraction. A plot of maximum amplitude versus distance
traveled in the refractor is presentedin Figure 17. These later eventsalso decay as
L-a/2, though L differs for each event. This general behavior is shown in the flat
interfaceregionsof Figure 17, though the scatter is rather severein this case.The
amplitudeanomalyin the regionof the structureis again apparent. Using this data
and the L-a/2 law, an amplitude anomaly curve is plotted (Figure 18). The results
are similar to the curve for P, but the anomaly is very much larger. This may be
possiblyexplainedas follows.The secondrefractionconsistsof two events--PS and
$P. In a flat layered model with source and receiver both on the surface, these
TIME ANOMALY
FOR SP
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
A t{sec}
ol,0
-2,0
-3,0
-4,0
f PHYSICA.L
LOCATION
OF
STRUCTURE
TLME ANOMALY
FOR PS
THEORETICAL
--'---- EXPERENTAL
DISTANCE (CM)
-9,0
110 120 130 140 1SO 160 170
At(joec)
01,0
' THEORETICAL
------ EXPERIMENTAL
-Z,0'
..3,0
TIME ANOMALY
FOR P REFLECTION .
-3.0
-4.0
-5.0
-6.0
TH EOR E TIC AL
-7.0
------ EXPERIMENTAL
-.8,0
DISTANCE (CM)
-9,0
110 120 130 140 150 160 170
I.O
O.I
ee
.01 , ,,I _ , ,
60 I00 00
L
Fro. 17. Peak amplitudefor the secondrefractionversusdistance,L, traveledin the refractor.The
straight line representstheoreticallypredictedamplitude behavior for flat interface model. Note the
large amplitude anomaly in the neighborhoodof the structure.
events have the same arrival time. In the presenceof the structure, however, the PS
event arrivesearlier than the SP event. Thus, in addition to the defocussingmen-
tioned earlier, we have two other attenuating effects.First, the energythat was in
the singlepulseis spreadover a longer time, thus decreasingthe maximum ampli-
tude. Second,as the profile movesonto the structure, the arrival time splitting in-
creasesand the two events PS and SP may destructively interfere. These two ef-
fects almost eliminate the secondrefraction near the top of the structure.
+2
A/VIPLITU DE RESIDUAL
-Z
-!
-4
db _
-8
-10
-1Z
-14
-16
I I I I I I I I I I I I I I
50 100 150 200
R (cm)
and Major, 1960). This problemhas been consideredin detail by Knopoff, Gilbert,
and Pilant (1960), who explained the important features in terms of the multiple
refractionspreviously considered.They computed typical seismogramsby the first
motion approximation.We have taken the samedata and computedexact seismo-
gramsby de Hoop's technique.A typical exampleis shownin Figure 19. The top
seismogram showsthe vertical groundmotionfor a sourceconsistingof an impulsive
force acting at the ground surface. The arrival of each of the first ten refracted
events is shown.Even with this small number of events a modelike structure begins
64 Refraction Models
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
R= 80cm
P6
PS
s
'1=
I I I I I I
154 156 158 160 162 164
TIME (,LZ.
S)
R= 70cm
I I I I I I
134 136 138 140 142 144
TIME (/.LS )
Fro.19.Comparison
oftheoretical
andexperimental
PLmode
seismograms.
Toptraceistheoretical
seismogram
forimpulsive
force
source,
containing
26rays(10distinct
arrivals).
Noteemerging
mode
structure.
Middle
traceshows
theoretical
seismogram
afterconvolution
witha source
pulse.
Lower
trace
isearlypartofexperimental
Pt.modeseismogram.
Verticalmotion.
Model Studies of Simple Refraction 65
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
SUMMARY
REFERENCES
C. R. BRUCE* ^D J. E. WHITE*
ABSTRACT
INTRODUCTION
Time-delay method
The time-delay method will be discussedfirst as the principle of this systemis
easier to visualize. The detectorsin the array will be called i and 2 in order from
the shotpoint. The normal sequenceof electrical signalsfrom detectorsI and 2 for
an ideally isolatedwavelet is shownin Figure 2a. Sincedetector I is nearer to the
shot by distanceXX, the refraction wavefront will arrive at detector I a short time
fit beforeit arrives at detector 2. The time interval fit dependsupon the array
length fix and the speedof soundV in the ith refractinglayer. This velocity is also
the phasevelocity of the refraction wavefront at the receiver site. The time interval
is given by the relationship /T--/X/V,. Recall that the objective is to combine
signalsI and 2 in sucha way that the output of the systemis zerofor wavespassing
with a certain phasevelocity and is nonzerofor waves traveling at other velocities.
This can be done by first delayingsignalI in time by an amount XT, as shownin
SURFACE
HOLE
REFRACTING REFRACTION
SHOT
INTE POINT
RAY
PATH
__
TIME
TIME
TIME
Fro. 2. Signalcompositing
processin time-delaymethod.
Figure2b, and then invertingit as shownin Figure 2c. Assumingthe wave shapes
are the same,the sum of signal1, inverted and delayed,and signal2 is zero at all
times.Since/XTis a fixedparameterin the system,wavespassingat other velocities
will not be nulled. However, the output of the systemwill be the derivative of the
actual wave shapeas it would be recordedin the conventionalmanner and, hence,
will have higherfrequencycontent.A more rigorousmathematical treatment of this
system follows.
A plane wave propagatingwith velocity V in the positive direction, which is
taken to be away from the shotpoint,is describedmathematicallyby the function
1. Thisfunctiongivesnot onlytheposition
of the wavefrontin time
70 Refraction Models
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
but also describesthe shapeof the wavelet, although the latter property will not
enter directly in the discussion.
It will be assumedthat the waveletshapeis nearly
constant withrespectto [T- (X/Vi) ] overthedetector
array.If thesignalat detector
is[[r- ], thentheinverted anddelayed signalwillbe (x/a ].
The time delay is negative in this function sinceat time T-t-AT it is desiredthat the
function be just the same as the undelayedfunction at time T. The signal from
detector 2 at positionX+AX is given by
x+/xx)
v
f T-- --fT--AT =F
Vi
is a goodapproximation.
The outputof the filter systemcanbeput into thisformby
addingand subtractingf[r-(x/v3] to give
x+
T-- AT
F=AX [ AT--,
OX OT
of of
OT OT
Separating Overlapping Arrivals 71
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
or
OT
Of_/, T '
and
0 T
o/ __
of
OX OX 0 T
or
OX: -- Vif' T
Substitutionof theseexpressions
into the equationfor the filter output gives
AX
This expression showsthe filter output is zero whenthe phasevelocity V is equal to
V and the waveletshapeis the derivativeof the originalwave shape.Althoughthis
equationindicatesa larger filter, output is obtained with a greater spacingof de-
tectors in the rray, an earlier limitation that X be small was imposed.More
specifically,
therequirement
is that besmallrelative
to thedom-
inant period for all phasevelocitiesrecordedso that the output waveform will be
the differential. In practice, X should be made as large as possibleto maintain
signal strength and still be consistentwith the above statement.
Signal-gradient mthod
The secondmethod of combiningsignalsto achievefiltering is called the signal-
gradient method becausethe instantaneousdifferenceof the signalsfrom the two
detectorsis used.The magnitudeof the spacegradientdependsinverselyupon the
phasevelocity. This dependencemay best be explainedby again referring to Figure
2a which showsthe normal sequenceof the electrical signalsI and 2. Signal 2 lags
signal1 in time by an amount whichis inverselyproportionaltp phasevelocity.
As the phasevelocity increases,the lag decreases until at infinite phasevelocity the
two signalswill be coincident,and, thus, the differencewill be zero. In this method,
72 Refraction Models
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
F(T
- -) --X--X/2.'
=I/2[f(T Vi / +f(T--X+l /_!'
if XX is small. The differencesignal denotedby 8 can be put into differentialform
usingthe approximationof the definitionof the partial differentialas before.First,
thefunction
fiT-(X/VO] is addedandsubtracted
fromthedifference
X+/xX/2' f(T---)
- f(T - x-V
_x/2}l
/J'
where the differenceis signal 1 subtractedfrom signal 2. Then by definition the
differencesignal may be written as
8= AX
ox
for/xX small, or
8- f' T
V
1.5
0.5
I
0 0.5 I 1.5 .5
f f r
where K is the integration factor. The factor K/tX is the gain of the/ channelrela-
tive to the sum channeland will be called V/since it has units of velocity. Finally,
comparisonof the sumsignaland the integrateddifferencesignalis accomplished by
summing.The result is the filter output which is describedby the expression
(,_
The filter output is zero when the phasevelocity is equal to V/, and V/can be varied
by adjustingthe relative gain of the differencechannel.Increasingthe detector
spacing/xX would increasethe differencesignal level, but/xX must be kept small
relative to the dominant wavelength of the refraction wavelet to obtain a true
gradient signal.
The amplitude of the filter output referred to the unfiltered signal is shown as a
function of the reciprocalof phasevelocity in Figure 3. The phasevelocity is nor-
malized to the filter setting. At very high velocities the output of the filter ap-
proachesthe normal signal of a singledetector. However, at velocitieslessthan one-
half the filter setting, the filter systemhas a greater signal output than a single
equivalent detector.
LABORATORY DEMONSTRATION
TxR.
--V. 2.hi
tan
V. _ V1cos
hii,. '
where i,2-sin - (V/V.) is the critical angle of incidencefor refraction from the
surfacelayer to refractor 2. Likewise, the traveltime for the refraction raypath in
refractor 3 is given by
Tx. = X 2(h
Va tan
i,a
q-h.o.
tan
i.,a
-- h -- h. ),
Va V cos i,a V cos
where i.,a=sin-1 (V./V3) is the critical angle of incidencefor refraction from re-
fractor 2 to refractor 3 and i,a-sin-: (V/V) is the angie of incidenceof the ray-
path at the first interface. The traveltimes along the two paths may be equal for
rn..any
combinationsof valuesof the velocitiesand bed thicknesses. For example,the
designof the plate model specifiedthe bed thicknessh and h.to be two wavelengths
thick and the critical anglesat both boundariesto be 45 degrees.Then, fixing any
velocity also fixes all other parameters including the shot-to-detector distance at
which the arrivals are simultaneous.
The air analog model is an arrangement of soundsourcesthat producestwo wave-
fronts which have different phase velocities at the receiver spread. This model is
ideal for the purposeof demonstrationsince only the two waveletsof interest are
present during the early part of the record. The plan of this model is shown in
SHOT DETECTOR
POINT X STATION
\ \ I
V2
V$
Fro. 4. Overlappingrefractionsituation.
Separating Overlapping Arrivals 75
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
SOURCE SOURCE
B A
21,5'
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
\ /
\ /
\ /
\ /
\\
-\\4DETECTOR
/5 SPREAD
6
SPREAD LINE
TO SOURCE A
TUBE
DETECTOR 2
',,[_---
DETECTOR
I
I iNVERTER
1 OSCILLOSCOPE
RECORDING
SYSTEM
by means of tubes. The tube on detector 1 is two incheslonger than the tube on
detector 2 so the delay is 0.15 ms if the velocity of soundin the tube is 1,100 ft/sec.
Otherwise, the system is explainedby the diagrams. Recording was done with a
cathode-ray oscilloscope and an oscilloscope camera.
Recordstaken with the air analogmodel are shownin Figures 8 and 9. The time-
SPREAD LINE
TO SOURCE A
INVERTER GRADIENT
AND INTE ORATOR
ADDER -(-2)
,j
2 ADDER
SUM
(1+2)
OSCILLOSCOPE BAND-PASS
RECORDING FILTER
SYSTEM
ADDER
250-800 cps
A B
delay method was used to obtain the wavelet separationshown in the recordsin
Figure 8 while the signal-gradientmethod was used to obtain the recordsshownin
Figure 9. The resultsof the two methodson the air model are equivalentas far as
78 Refraction Models
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
resolutionof thetwowaveletsisconcerned.
Onthenormalrecord,a, thetwowaves
interfere
strongly asseenby thechangeof waveletshapefromtraceto trace.How-
ever,whenthefilteris usedto eliminate
oneof thewavelets,asin recordb, where
Separating Overlapping Arrivals 79
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
the wave of lower velocity is filtered out, the wavelet shaperemains nearly the same
from trace to trace. The phase velocity of wavefront B is easily obtained from
record b. Likewise, the phase velocity of wavefront A is taken from record c where
the faster wave has been eliminated. In each case a small amount of the filtered
wavelet can be seen by careful observation. This residual signal is a result of not
being able to adjust the laboratory system to completebalance. A much finer filter
balance is possible with geophysicalequipment.
A two-dimensional plate model was used in an attempt to show the action of
phase-velocityfiltering in a more realistic manner. The model was constructedto
reproduce the layering shown in Figure 4. There were three regions with velocities
V, V2, and Va equal to 6,250, 7,150, and 10,500 ft/sec, respectively.Thus, the
model representeda vertical plane with the top edge being the line between the
shotpoint and the detector spread. This model was cast of epoxy resin which was
modified with various filler materials to produce the different velocities.The plate
was inch thick. The width, which correspondsto depth, was 4 ft, and the length,
which correspondsto the maximum shot to detector distance, was 6 ft. The in-
dividual layer widths, h, h2,and ha,were 0.5, 0.71, and 2.79 ft, respectively.In the
case of h and h this is about two wavelengths at a frequency of 20,000 cps. It is
frequency that determines the scaling factor of the model. For instance, if field
recordshave a dominant frequencyof 10 cpsand the modelis operatedat 20,000 cps,
then the scalingfactor is 2,000, and one ft on the model represents2,000 ft in the
field.
The equipment used to study the plate model was nearly the same as used with
the air model. The dominant frequency of the wavelet was 20,000 cps instead of
500 cps as with the air model. Barium titanate transducerswere usedboth for de-
tectors and for the sonic source. The detector transducers were placed on the
edge correspondingto the surface. The source transducer was coupled to an edge
correspondingto depth. The shot depth could be changedby sliding the transducer
along the edge.The data presentedin Figure 10 was taken with the shotpoint at the
surface. The manner in which the detectors were coupled to the model was ex-
tremely critical in the demonstrationof phase-velocityfiltering. Only two detectors
were used so it was necessarythat the coupling be easily broken in order to move
the array to the various receiver sites. Good acoustical contact was obtained by
coupling with grease.In addition, when setting the array up at a receiver site, the
two detectorshad to be carefully adjusted in alignment and pressureto get equiv-
alent electrical signalsfrom each. The slightestmisalignment causedunbalancein
the system and thus a large residual signal at the filter velocity. The reasonfor the
diffi.culty with detector alignment is clear when the detector dimensionsare scaled
to field size.The array spacingon the model was three-fourth inch. This corresponds
to 125 ft in the field if the scalingfactor is 2,000. The contact area of a detector was
one-eighth inch in diameter. This is about 20 ft when scaled to the field. A slight
misalignment will radically change the contact area and possibly the spacing.The
geophonesusedin the field certainly would not present the same couplingproblem
even if usedin a cluster having a comparablecontact area.
The plate model was carefully planned to have the two refraction signalsarrive
8O Refraction Models
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
signalsand the direct wave (whichis muchstronger)did not occurwithin the length
of the model. Therefore, the effectof phase-velocityfiltering was demonstratedby
suppressingthe direct wave and the Rayleigh wave in turn. Figure 10a showsthe
signalrecordedfrom a singledetector.The nine tracesare from stationstwo through
ten from the source.Stations on the model were 4 inches apart. The two major
signalson this recordare the direct wave with velocity 6,250 ft/sec, and the Ray-
leigh wave with velocity 2,600 ft/sec. Figure 10b showsthe record with the direct
wave suppressedby the filter. The residualsignal that can be seenon someof these
tracesis due to imperfectalignmentof the two detectors.In Figure 10cthe Rayleigh
wave has beensuppressed.Again someresidualsignaloccasionallyremainsbecause
oimperfect balanceof the system.
PROPOSED FIELD PROCEDURE
to 18,000ft/sec. Then the time delay requiredwill be 13.9 ms. Further, supposethe
wavelet has a dominant frequencyequal to 5 cpsor a dominant period equal to 200
ms. One-sixthof the dominant period, 33.3 ms in this case,is consideredsmall. When
these figures are substituted into the criterion for proper operation of the filter,
Ir- x/v)l <<dominant
period,
it isseen
thatthefilterwillnotoperate
properly
on waveletswith velocitiesbelow5,300ft/sec. It is to be noted that the array length
will dependuponfactorssuchas the area of explorationand the frequencyband-pass
of the equipment.The array length is important and shouldbe consideredcarefully.
The useof phase-velocityfiltering in a refractionprogramwould be especiallycon-
venient if the conventionalspreadcould be retained and filtering could be accom-
plished by combiningsignalsfrom adjacent sites.
After the receiverspreadhas beendetermined,whether it is a normal spreadwith
the station spacingequal to an array length or a receiverarray at each site, the
signalfrom each receiveris recordedon magnetictape. There are many advantages
in this technique.First, the normal recordcan be retained.In addition,by recording
the "complete" recordthe phase-velocityfilter setting can be varied during the data
processinguntil optimum recordsare obtained. If the signalswere combinedin the
field, severalshots may be required to get the same results.There could even be
adjustmentsmade in gain if signal amplitudesare unequal becauseof unmatched
receiversor nonuniformcoupling.Another advantageis that the reverseprofilecan
82 Refraction Models
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
AMPLITUDE
OF
Section
3
REFRACTION
SIGNAL
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
Section 3
Time has been the most usedproduct of the refraction signal, and almost all in-
terpretive techniques,presently available and discussedin this volume, actually
are concernedwith the arrival time of the signal. However, recognitionshould be
given to the fact that amplitude of the refracted signalis alsoan important charac-
teristic. For this reason this section has been included.
In the first paper, O'Brien describesthe "Use of Amplitudes in SeismicRefraction
Survey." After correctionsare made for charge weight, geophoneplant, and geo-
metrical loss,meaningful amplitudes of the refraction signal can be measured.Stud-
ies are made of the attenuation and critical distance of refractors.
With the secondpaper, Werth givesa "Method for Calculatingthe Amplitude of
the Refraction Arrival." Calculationsshow that layering above the refractor can
changethe refractionwave characterand thus affect the appearanceof an anomaly
in the refractor.
The third paper, by Sherwood,is a theoreticalstudy on the amplitudeof "Refrac-
tion Along an Embedded High-SpeedLayer."
The fourth paper, O'Brien's "The Efficient Use of Large Charges," concernsthe
amplitude of signal obtained from various size charges,its dependenceupon the
elasticpropertiesof the material in which the chargeis shot, and the predominant
frequenciesof the observedrefractedpulse.
The fifth paper, Layat's "Modified 'Gardner' Delay Time and 'Constant Distance
Correlation' Interpretation," describesthe utilization of amplitude as part of a
mapping procedureused in North Africa.
84
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
P. N. S. O'BRIEN*
ABSTRACT
INTRODUCTION
the same time and by the same personsas the time-distance interpretation. Their
major useis often as an aid to correlation,and it shouldalways be checkedthat the
time-distanceinterpretationis consistentwith the measuredamplitudes.
The following sectionsoutline the factors affecting the recorded amplitudes, the
correctionswhich normally have to be applied, sometheoretical considerations,and
a number of field examples.
I
GROUND SURFACE
2-0:
1'8
- SSON'S
o
I-6
1'4
1.2
'-,,;
'AI'O
.8
-6
.2
0
O IO 20 :30 40 50 60 70 80 90
ANGLE OF INCIDENCE i
The groundaroundthegeephone
and thegeephone
"plant"
Civil engineershave for long maintained that blocksof the earth's near-surface
act like mechanicaloscillators,each block having different propertiesfrom other
blocksand beingcoupledto its neighbourswith differentdegreesof closeness. If this
is true, then a wave incidentfrom depth will excitethe differentblocksto a lesseror
greaterdegree,and the geephoneoutput will then beinfluencedby the propertiesof
the blockon whichit is placed,and only indirectly by the amplitudeof the incident
wave. Seismologists have generallynot beenvery receptiveto this idea and in any
eventit is probablyoftenovershadowed by variationsin geephone"plant."
The motion of the ground with a geephoneresting on it is not the same as the
motion it would have had without the geephone.Theoretical treatments for a
homogeneous earth have beengiven by Wolf (1944), Pasechnik(1952a) and, most
88 Amplitude of Refraction Signal
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
The observingsystem
In order to keepto a minimum the scatterin the plotted amplitudes,it is desirable
that the overall relative gainsof the individual channelsshouldbe known to within
_+5 percent.This requiresthat the geophones shouldbe closelymatchedand that
line resistance,line leakage,and amplifier gain shouldbe regularly and accurately
Use of Amplitudes in Refraction 89
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
2'8
2'0
I'0
0'$
0'0
0'0 0'1 0'4 O & 0'1 I'0 '1 '4 1.6
cI
120 ..............
I00
b=lS
bo ,
lie I I I i
0'0 0'2 0' 0'6 '0'8 I'0 1'2 1.4 1'6
-,- STATI ON 70
25,000 feet S
34O
320
I'l
300
280
260 E1
E'I
'0 240
0 220 I'1
LJ
%o0 r3
180
i-- IOO () ()
O
00 ()
()
8O
7O
()
5O
I 2 3 4 5 6 7 8 9
TRACE NUMBER
measured. In order to reduce the corrections which have to be made to the scaled
amplitudesand, also,to keep the geophonedampingconstant,it is usually desirable
to make up the line resistanceof eachchannelto someconstantamount.
If appreciabledistortion is introduced,either by the geophoneor by the amplifier,
it may be necessaryto make someallowancefor it, especiallyif measurementsare
being made on widely different waveforms.However, distortion is not seriousin most
refraction equipmentand so this problem shouldrarely arise.
VARIATION AT THE SHOTPOINT
NECESSARY THEORY
LOWER
VELOCITY
HIGHER
VELOCIYY
Fro. 4. Ray diagram for critical angle reflectionand head wave.
tion equations(e.g. Ewing et al., 1957; Jakosky, 1950) must be solvedfor the par-
ticular values involved. This is a rather tedious and error provoking processand if
many are neededit is best to get them calculatedon an electroniccomputer.Bortfeld
(1961) givessomeapproximate equationswhich appreciably simplify the computa-
tions without losing too much accuracy.
For anglesof incidencelessthan critical, no phase changesare involved and the
calculatedamplitudesare, therefore,independentof the incident waveform.
Head waves
Except for very steeply dipping interfacesthe incident rays impinge on the high
velocity refractor at the critical angle. In this casethe plane-wave partition coeffi-
cients must be replaced by the more exact ones which allow for curvature of the
wavefront. This is discussedlater in this section and here we merely give a few
results.
The geometryis shownin Figure 4. At large distancesfrom the shot the particle
displacement
of the first-arrivalheadwave(the 'refracted'wavePP2P) is givenby
= (1)
where, K is a constant depending on the elastic properties of the media, F(t) is
the displacement potentialof the incident pulse, d is the distancefrom the shot to
the receiver, and L is the distancethe wave has traveled in the refractor.
Equation (1) hasbeenobtainedby severalauthors(e.g. Heelan, 1953).It is obvi-
ousfrom the presenceof the factor L in the denominatorthat this equationdoesnot
6'0
= POISSON'S RATIO
5'0
4-0
3'0
2'0
'4
IoO ,
1'2 I-4. 1.8 1'8 2-0
VELOCITY
I::kATIO
V2p/Vlp
Fro. 5. Variation of vertical amplitudewith velocitycontrastand ?oisson'sratio.
hold for points closeto the critical angle reflection. The range of applicability of the
equationhasbeendiscussed recently(DonatoandO'Brien, 1963)with the conclusion
that it may be applied with values of L greater than about 5-6 times the predomi-
nant wavelength of the pulse.
Figure 5 is basedon calculationsby Bather (1959) and showshow K cosi, wherei
is the critical angle,varies with the elasticparametersinvolved. Insertion of a value
94 Amplitude of Refraction Signal
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
I.O
9.O
8.0
7.0
.7
6.0
r- PPP o=o)
a. 7'0
u 4-0
2'0
I-O $eS(=1810
o)
O $ss( .
o 0-6 o.a
VELOCITY
RATIO
VI/ V2 P P
$.O
2.0
270)
I'O
VELOCITY
RATIO
VIp/v2P
FIG. 7. Moduli of the amplitudecoefficients
of head waveswhen the incidentwave is of a different
type from the head wave (from Bessonovaet al., 1960, calculatedby N. V. Stairnova).
350
310
270
230
tu
o
z
190
"'
70
30
Brekovskikh (1960) has shown that, except for anglesof incidencevery close
either to critical or to grazing, for the reflection of a curved wavefront it is permis-
sible to use the plane-wavereflectioncoefficientstogetherwith a geometricaldiver-
gencefactor. Richards (1961) has calculatedreflectioncoefficientsfor a large number
of casesand has also discussedthe effect of geometricaldivergence.He also graphs
the values of the phase change which occur for reflection at anglesgreater than
critical, but, unfortunately, he interprets this changeas a lag when it is usually a
lead. He also discussedseismicamplitudessolely in terms of the modulus of the
reflectioncoefficient.Becauseseismicpulsescontain a fairly broad band of frequen-
ciesthe attendant phaselead impliesthat wide-anglereflectionsmust be distorted
to someextent. Sincethe phasechangevarieswith angleof incidence,this distortion
Use of Amplitudes in Refraction 97
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
Table 1. Parameters and results for wide-angle reflections from a layer of velocity 20,000
ft/sec and density 2.65 gm/cm3---Refer to Figure 9
Overburden Parameters
Effect of phasechangealone
Reflectedamplitude
Phase lead
Time lead on 1st trough
Incident amplitude as a fraction of period T
peak-to-peak
45 1.115 0.122
90 1.164 0.245
135 1.115 0.366
line, very marked changesare introduced. The figure alsoshowsthe time lead of any
peak in terms of the time scale of the incident pulse. The distortion illustrated in
Figure 9 is similar to the distortionproducedin head waveswhenqO0. It is obvious
from the figure that both for thesewaves (i.e. thoseother than ppp3) and for wide-
angle reflections,amplitude measurementsshouldonly be made on peak-to-trough
excursionsin the centreof the pulse.Sinceboth thesetypes of eventsare alwayslate
arrivals they often form part of an interferencepattern and this also means that
measurementsshouldonly be made on the maximum excursionsof the pulse.
Figure 10 showsthe vertical amplitudes of the head wave and the wide-angle
reflection for a velocity contrast of 2' 1. The reflection amplitudes have been taken
from Richards (1961) and do not allow for pulse distortion. The head-waveampli-
tudes have been calculated from a modified form of equation (1) which holds to
within about two wavelengthsof the critical angle reflection (Donate, private com-
munication).For a 15-cpswaveand a refractordepthof 1,700ft the amplitudescales
are identical for both events. For deeperrefractors or higher frequency events the
amplitude of the head wave decreaseswith respect to the reflected wave in direct
proportion to the frequency and depth.
Very closeto the critical anglereflectionPPP is alsodistorted.
Use of Amplitudes in Refraction 99
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
0.7
I
I , E;L;CTON
0-6
/' V- IOOOO ft/sec
=:0'5
/i V,-2OOOO
f,/,,
o 0.4 / '/' \
h-DEPTH
INFEET
- o..
,
0'1
0 I 2 4
DISTANCE
/ DEPTH
Fro. 10. Comparison of reflectedamplitudes(Richards,1961)andhead-waveamplitudes
for a velocity
contrastof 2'1 and an incidentpulsewith a predominantfrequencyof 15 cps.Note that, near to the
criticalangle,the reflectionamplitudemay be appreciablylessthan that plotted.
DISTANCE
IN cm
27
28
29
30
31
32
33
34
PPP PP
Fro. ll. mergenceo the headwavePPP rom the wde-an]e reflectionPP or water overlying
wax. Timing marksare at onetsecintervals.Sourceand receiverare 10 cm abovethe wax.
to within about twice the critical distancefrom the shotpoint, without the appear-
ance of any large amplitude reflection.It seemslikely that this is due partly to the
correction for wavefront sphericity, and partly to the finite thicknessof the layer
togetherwith pronouncedvelocity variationswithin it.
Also, it shouldbe rememberedthat the reflectionamplitude closeto the critical
angleis practicallyindependentof the velocitycontrast,whereasthe head-waveco-
efficient increasesas the contrast decreases.For small contrasts, therefore, the head
wave may well be commensuratewith the reflection.
For distances which are not great compared with critical, there is always the
danger of confusingthe head wave and the reflection.At such distancesan effort
Shouldbe madeto identifybothevents.If thisis not possible
thena longlineshould
be shot so that both events are recorded and an estimate can be made of their rela-
tive amplitude, as was done by Richards. If this alsois not possible,then the ob-
servedamplitude decay shouldbe comparedwith the decay curves for head waves
and reflected waves as calculated from the assumed model. It should be remembered
that due to phasechangesand interferenceeffects,the reflectedwave will probably
vary morerapidly in characterthan the head wave.
At larger distancesthe head wave seemsto predominateover the reflection.This
is not expectedtheoreticallyand is no doubt due to variationswithin the overburden.
In particular, relatively high-speedlayers or an appreciablevelocity gradientmay
prevent rays at large anglesto the vertical from ever reachinga deep high-speed
layer. In addition,if the overburdenis muchlesshomogeneous
than the refractor,
Use of Amplitudes in Refraction 101
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
The manner of reduction to datum will depend on the type of survey and the
interpretation model which is adopted. The following describesa method used for
first peakson a refraction line. It is illustrative only, and the precisemethod adopted
will obviously depend upon the specific operational and interpretational methods
which are being used. After applying the correctionsthe resulting values are called
residual amplitudes. Any variation with distance of the residual amplitudes is called
residual attenuation, and is usually expressedas db/station interval.
PROCEDURE
There are three essentialsteps. The first consistsof correcting the amplitudes for
that part of the attenuation which is due purely to geometricspreadingof the wave-
front. From equation (1) it follows that this may be done by application of the
factor (dL3)/2. For conveniencethe amplitude is first multiplied by ds and then by
(1-Xc/d) /, where Xc is the critical distance.This is done so that a set of universal
overlaysmay be made from which the d correctionmay be read directly. This saves
considerablerepetitive calculation. One of a set of such overlays is shownin Figure
12.
In the secondstep, the amplitudes are reduced to those which would have been
obtained with some standard weight of charge. They are multiplied by a factor of
the form (W/W)'* where W is the standard weight, W is the weight fired, and n is
usually about 1.0.
The third step consistsof the correctionfor instrumental variations.
These adjustmentsbecomesimpleadditionsor subtractionsif relative amplitude
is measured in decibels. It is assumed that the station interval is constant and that
the stations are numbered consecutively.The unit of distanceis one station interval
and the distancebetween shotpoint and station is N units.
In order to illustrate the measurement and correction at each station we will con-
sider the amplitude at station 38 from a shotpoint at station ! 1.
The following stepsare necessary:
(1) Selectthe overlay which includesthe correct shotpoint-to-stationdistance.
That is, N = 38-- ! ! = 27. Place the overlay transparencyover the record with
its abscissaalongthe zero line of the galvanometertrace. Read the positionof
the top of the peak at the abscissavalue of N-27. This is illustrated in
Figure 13 in which the peak value is 83.3 db.
(2) Using the estimatedvalue for the critical distance,X,, useTable 2 to find the
correction30 log0(!- X/N). In our example,if X, equals10 stationintervals,
then 6.0 db must be subtracted from 83.3 db giving a correctedamplitude of
77.3 db.
(3) Add to this the attenuationsettingof the amplifier(it is assumedthat all the
102 Amplitude of Refraction Signal
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
35 30 25 20
A
35 30 25 20
STATION NUHBERS SP AT STATION O
35 25 OVERLAY No...5 20
9O
88
.TN GAIN
,,2 - 24
iBS - 24.
,S4 - 24
85 - 24 80
3G -24
17 -24 ?O
38 -34
39 -24
STATION
NUMBERS
SAON 0
2C
40 -34
13. Measuring
the amplitudeof the firstpeak.
Use of Amplitudes in Refraction 103
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
4 6 8
10 12
I
Scale One
li 0.7
0.7
0.7
0.7
1.4
1.4
1.3
1.3
2.2
2.1
2.1
2.0
3.0
2.9
2.8
2.7
3.8
3.7
3.6
3.5
4.7
4.6
4.5
4.4
117
120
123
126
0.7 1.3 1.9 2.6 3.4 4.3 129
.
Xc 4 8 12 16 20
Scale Two
X,
6 Three12
Scale
18 24 30
amplifiershave equal maximum gain) say, --24 db. That is, add 24 db to
give 101.3 db.
(4) Apply a correctionfor the sizeof the charge.If n- 1, as it usuallyis, then to
correct the actual charge of 100 lb to the standard charge of 1,000 lb it is
necessaryto add 20 db. The final correctedamplitude for station 38 is there-
fore 121.3 db. Figure 14 is helpful when obtaining this correction.
(5) Plot this value on linear graph paper against station number. A scale of
10 db= 5 cm is often about right. If a delay or intercept time plot is kept,
then it is recommended that the amplitudes are plotted with the same
abscissa,as there may be frequent crossreferencesbetweenthe two plots.
Corrections for recordingsystem.--Ideally, the geophonesensitivities,the line re-
sistancesand leakages,the amplifiergains,and the galvanometersensitivitiesshould
all be sufficientlywell matchedsothat followingthe stepsgiven abovewill provide a
true measureof the relative ground motion (except for geophone'plant') at the
various stations.
In areas where the geophoneplant introduces a considerablescatter the instru-
mentsmay often be sufficientlywell matchedfor their effectto be ignored.However,
in someareas the standard deviation of the scatter due to plant may be lessthan
1.0 db and then it is usually essentialto allow for instrument variations.
By usingthe samegeophone,or geophonegroup,with the sametrace, by measur-
ing the line resistanceand leakage,and by calibrating the amplifiers--all of which
takes only a few minutes a day if the necessaryinstrumentsare available--the cor-
rectionsmay be quickly obtained.
However, if these measurementsare not possibleit is desirable to perform a
statistical analysis to determine any consistentinstrumental variation, so that it
may be allowed for. Assumingthat the effect of geophoneplant is random, then
taking the mean amplitude of each trace for a number of spreadsshot in the same
direction will reduceits effect considerably.A plot of this mean amplitude against
trace number will showthe scatter due to the instrumentssuperimposedon a slope
whosegradientis the mean of the residualattenuationfactorsfor the variousspreads.
In order to eliminate the effect of attenuation it is necessaryto assumethat the
attenuation factor is the same in both directions. Then, by taking the mean of the
two mean-amplitude curves for shots in oppositedirections,all that is left is the
scatter due to the instruments--it being assumedthat this is sensiblyconstant over
the period during which the data werecollected.
A typical plot is shownin Figure 15. In this casecorrectionswere addedto bring
all the tracesup to the arbitrary level of 75 db, and thesecorrectionswerethen ap-
plied to all the originalstation amplitudes.
I000
5000
I00
I0
50
+30
+20
+10
Use of Amplitudes
DECIBELS
in Refraction
-IO
74 ="-'
7)
/ SHOOTING
N--S
76
1-72
SHOOTING
70
76
74
NEANOF BOTHDIRECTIONS
I 2 4 5 6 7 8 9 I0 II 12
TRACE NUHBER
Fro. 15. Mean instrumental corrections.
Use of Amplitudes in Refraction 107
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
Method one' Use of routine records.--On the amplitude plot in which the corrected
amplitude in decibelsis plotted against distancein station intervals, variations in
the geometricterm give either a convexor concavecurvature superimposedon the
linear effectof the residualattenuation. By applying a number of geometriccorrec-
tions, eachfor a different assumedcritical distance,the best value for Xc will give
the straightestline. And the gradient of this line will give the attentuation factor.
For this methodto be effective,the value of Xc/d mustbe as large as possible,there-
fore the shootingdistancemust be as small as possible,though it shouldnot be so
small that equation(2) fails to apply (seesectionentitled "Head Waves").
One procedureis to fit the correctedamplitude data to a curve of the form
logan= a-- bN -- c/N,
whereA,= Ad/2(d-- Xc)a/.For eachvalueof X, therewill be corresponding valuesof
b and c. The best values of X and/ are thosefor which c=0. Figure 16 showsthe
results of such an analysis for an actual survey. It gives 1,300 m for the critical
distance and 0.055 +0.029 decibelsper station interval for the attenuation factor.
A quicker, though lesssensitive,method is to calculate the standard error of the
linear regressioncoefficient of logAc on distance. The minimum standard error
identifies the best straight line and hencethe best values of X, and
0"8
0-4
0-0
-0.4
-0.8
>0-08
Zo.o4
O
STATION
I 2 3 4 $ 6 '
I I I i i INTERVALS
m 0.00
300 6OO 9OO 12OO 15OO 18OO 21OO METRES
CRITICAL DISTANCE XC
///1
//!.1
I.t
z
o /Z . -I
, o ILL'l/ I
.oo.,
,o / ,,./'/- '" / ,o
.
II
x / 2' x
1 '00
ii 0 __ ii o
0 0 ,n 0 ,n 0 ,
length of the arrival was about one station interval, so that equation(2) may not
hold for total distances which are smaller than 41-42 station intervals.
15
4 IO 20 30 40 50 60
DISTANCE IN kms
2-8
APPROXIMATE EXPECTED VARIATION
FOR CONTINUOUS INCREASE
OF VELOCITY
2-6
2-4 OMEASURED
VALUE
2-0
I-8
'- 2-Odb PERWAVELENGTH
- o
20O
160
120
.
O 200 400 600 800 IOOO 12OO
DISTANCE IN FEET
An integratedcasehistory
This is a drasticallyshortenedaccountof a previouslypublishedpaper (O'Brien,
1960b). It is included here becauseit exemplifiesthe use of amplitude measurements
to distinguishbetween various geologicalmodels,each of which satisfiesthe time-
distance data.
The problemwas to explainand make useof a strong,late-arrival refractedevent
which had approximately the same velocity as an earlier, weaker arrival. A rather
poor interval velocity log was available which indicated that there was only one
formation with the correct velocity. It was therefore supposedthat the strong late
arrival originatedin the samerefractor as the earlier arrivalmwhich had been quite
certainly identified as comingfrom a thin Permian Limestone--but that it was de-
layed in the overburden either by multiple reflection or by P-S conversion.These
two possibilitiesare illustrated in Figure 20. With these two models,and with the
velocity information available from the well survey and the refraction records,
estimatesweremade of the expectedtime delay of the late event and of its amplitude
relative to the earlier arrival. These results are shown in Table 3, which also shows
the residual attenuations of the two events.
Becauseof uncertainties in the overburden velocities, the calculated amplitudes
could be in considerableerror. But by no stretch of the imagination could they be
changedsufficientlyto bring the calculatedand observedratios into anything like
parity. Further, both modelspredict that the residual attenuation should be the
samefor both events. This is most certainly not true; the first arrival attenuates at a
rate applicableto a thin layer and the later arrival attenuatesat a rate applicableto
a thick, porouslayer.
Therefore,in spite of the fact that the delay times agree,it is certain that neither
modelis correct.And, in spite of the fact that the well surveyshowedno other high-
velocity layer, it is certain that the late arrival must originate in a deeper, thicker
layer of about the samevelocity as the shallower,thinner layer which gave rise to
the first arrival.
The depth and structure of this deeperlayer was thereforeconstructedfrom the
time-distancedata. Its upper surfacewas found to correlateextremely well with the
top of a thick sand-shalesequence(Millstone Grit Series,Upper Carboniferous).
Subsequentcontinuousvelocity logsshowedthat the original well survey was most
misleadingand that the interval velocity of the Serieswas only 5 to 6 percent less
than the refraction velocity.
It remained to confirm that the third model is consistentwith the observedampli-
tudes.Unfortunately,it is not yet possiblequantitatively to estimatethe amplitude
of the arrival from a thin layer. However, qualitative argumentsindicate that the
arrival from the Grits shouldbe severaltimes greater than the earlier arrival from
114 Amplitude of Refraction Signal
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
P P
LIMESTOND
MODEL THREE
the thinner Permian Limestone. For, on the one hand, the Limestone overlies a
lowervelocity mediumand thereforethe reflectionfrom its bottom surfaceis always
out-of-phasewith the head wave, and will severelyreduceits measuredamplitude.
On the other hand, the Millstone Grit Seriesis underlain by the higher velocity
Carboniferous Limestone and the reflection from its bottom surface will not always
be out-of-phasewith its headwave; its amplitude,therefore,will not be soseverely
Deeper refractor
ObservedMultipleP-Sconversion
(Millstone
reflection Grits)
Time interval .100 to . 110 .105 .100 (. 100 to . 110)
CONCLUSIONS
By far the most important information on the seismicrecordis obtained from the
time-distance data. But an amplitude study is never a waste of time. At the least it
gives the seismologistanother dimensionfrom which to view the interpretational
problem.It broadenshis mind and giveshim a better graspof the physicalprocesses
which take place in seismicprospecting.
On occasion,an amplitude study providesthe critical evidencenecessaryto dis-
criminate betweenotherwiseequally valid models.And it may require the rejection
of previoushypothesesand suggestmore correct ones.It may even suggesta varia-
tion in conventionalfield procedure.For instance, as indicated in a later paper in
this section,it is well known that surfaceshotsgeneratemuch more equivoluminal
(S) wave energythan they do irrotational (P) wave energy. Sincesurfaceshotsare
often an economicnecessityit may well be worthwhile to try to detect the large
amplitude S wave by using horizontal geophones.As may be seenfrom Figure 6
there is a further amplitudeincreaseto be expectedfor either smallvelocity contrasts
(usingSSS) or large contrasts(usingSPS). The major drawbackwould appear to
be the attendant increasein noise,due to the fact that any S wave is a late arrival.
The interpretational model in refraction work always consistsof an overburden
~
the tendency towards the use of 'tighter' filtering and AVC is not so welcome. It
should always be proven that a real and necessaryimprovement in the signal-to-
noiseratio is producedbefore deliberately distorting the signal, for this distortion
may render uselessthe recordedamplitudes.
Frequencymeasurementshave not sofar providedmuch information. But in many
areas quasi-sinusoidsare recorded and their frequency must reflect well-defined
geometric properties of the subsurface.The normal mode analysis of the wave
equation shouldprovide the necessarytheory, and careful casehistoriescombined
with ultrasonic modelsshould enable an interpretational method to be worked out.
RECENT ADVANCES
Since this paper was first prepared there have been two significant advancesin
the study of the amplitudes of refracted waves.
The first concernsthe arrival from a thick refractor. Werth and Herbst (1963)
have calculated the amplitude of the head wave from the Mohorovii discontinuity
(Pn) for a number of undergroundnuclearexpansionsand O'Brien (1965) has cal-
culated the amplitude of the refracted wave for a distance of 20 km from an under-
water chemical explosion.Both these calculationsgave values very close to those
observed.This demonstratesthat it is now possibleto calculate absoluteamplitudes,
as well as relative ones,when checkingthe amplitude predictionsof a layered model.
The secondadvance concernsthe arrival refracted from a thin high-speedlayer.
Rosenbaum(1965) and Donato (1965) have treated the casefor layers thinner than
about one half of a "wavelength." The former indicates that the geometrical, power
law attenuation for such a layer should not be simply dTM,aS stated earlier in this
paper, but d/2L/a. Both authors give expressionsrelating attenuation to refractor
thicknesswhich are potentially useful in the interpretation of field surveys. Spencer
(1965) has presenteda theory for layers of a few "wavelengths"thickness.This ex-
plains the en echelonor "shingling" character of some time-distance graphs and
provides the analysis necessaryfor calculating the amplitude and attenuation of
arrivals refracted from such layers.
ACKNOWLEDGMENTS
REFERENCES
Vassil'yev,Y. I., Kovalev, O. I., and Parkhomenko,I. S., 1958a, Study of the crystallinebasementby
the refractedwave methodunderconditionsof partial masking,I: Bull. Ac. Sc. U.S.S.R., Geophysics
SeriesNo. 3. English Translation by AGU.
1958b, Investigation of the incompletely masked basement by the refracted wave method,
II: Bull. Ac. Sc. U.S.S.R., GeophysicsSeriesNo. 5. English Translation by AGU.
Washburn,M. W., and Wiley, H., 1941,The effectof placementof a seismometer on its response
char-
acteristics:Geophysics, v. 6, p. 116-131.
Werth, G. C., and Herbst, R. F., 1963, Comparisonof amplitudesof seismicwavesfrom underground
nuclearexplosions in four differentmediums:Jour. Geophys.Res.,v. 68, p. 1463-1475.
Wolf, A., 1944,The equationof motionof a geophone
on the surfaceof an elasticearth:Geophysics,
v. 9,
p. 29-35.
Zvolinsky,N. V., 1958, Reflectedand head wavesarisingat a plane interfaceof two elasticmedia-II,
Equation21: Bull. Ac. Sc.U.S.S.R.,Geophysics
seriesNo. 1, p. 1-7, EnglishTranslationby AGU.
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
REFRACTION ARRIVAL]'
GLENN C. WERTH*
ABSTRACT
NTRODUCTON
rko(r,z, t) = axt-
RoRo)
for R0 < axt
for Ro >_ at
where
0-- displacementpotential,
r, z-- cylindrical coordinates,
R0=distance to the point of observation,
t = time,
a- compressionalvelocity.
ACTUAL
,4,
iRECEIVER
SOURCE r
VIRTUAL &
SOURCE
Zo
C T./I 'T'
(2) /IO . x L -
/
/ /I
//
/
/
/
/
/
/
/
/
/
/
/
IMAGE /
SOURCE
R
0
A (a*) atr
R2
remainder for t > R/a,
where
R 2 = r 2q- (zq-z0) ,
M q-
8 M
remainder<
(-) R A(a*)
and
at(z-+-Zo)
R(R-+-An.)(z-+-Zo) R R '
/Xn = at-- R,
A (a*) = coefficientof reflectionin the auxiliary plane problem
(note that A(a*) is a function of time),
and M and M,. dependonly upon the propertiesof the coefficientof reflection.
Now
where/ is the angleof incidence.A (v*) then provesto be independentof time and
equal to the plane-wavereflectioncoefficient.Hence,
0 for t < R/a
ql = A (cos/)
sin/ for t > R/a.
R
q0P
= [ sin
R0
I 0 for
t<Ro/a
/'(/Xn) fort> Ro/a,
wheref'(/Xn) is the derivativeoff(/Xn) with respectto its argument.The expressions
for the reflectedwave (lessthan critical) become
0 for t < Rfal
qlP ---- sin 1
Zl(COS
1) f!(An) for t > R/a1.
R
q = Ao* f(a.),
(1 - ,'')/' r/'L/"
where
q= displacement
alongthe ray in the uppermedium,
2px'[X'(1+ 2m'y ') + X.'(p- 2m'y
2)]2
A10* = '
[,2(1
+ 2m,'-- p)2+ px'X.'+ x'X'(1+ 2m'2).].
f(/Xn)= reducedpotential,
r-horizontal distance source to receiver,
L- "glide" distance,
'y= al/a.,
P=Pl/P%
=
x'= (1- ,) /',
l=compressionM
velocity
upper
medium,
1
b=shear
velocity
upper
medium,
ISimilarly
for
lower
medium
p= densityupper medium, 2.
.410* = head coefficient.
(1 - ')/'
In the derivation,it ws ssumedthat r is sufficientlylre so that the reflection
beyondthe criticalanledoesnot rrive withinthepulselengthof interestfor the
hed wve. ote that the hed wve hs the time dependenceof the reducedpo-
tential itself,rather than of the derivativeasdo bodywves.
In an earlierpper, eeln (1953)derivedan equivalentexpressionfor the head
waveusingthemethodof steepest descent to evaluatetheinteml.The eometricl
spreading term and the time dependence on the sourcere the same.ead co-
efficients
for twoparticularexamplesaregivenas6.1 and17.0by eelan in hisTable
$. The headcoefficients for the sameexamplesclculatedby Zvolinskii'sformul re
6.05 and 17.$7.Zvolinskii'sformulais moreeasilyprogrammedfor machineclcul-
tion and the criteriafor validity of the appoximation
aremorestraightforward.
Downwardtransmissionoecients
In anypracticalproblemtherearelayersabovethecriticallyrefracting layer,and
the appropriate
calculations
mustbe madefor theireffects.Firstof all, in the near-
frontapproximationtheproductof thetransmission
coefficientsmustbecalculated.
The formulationof the transmissionand reflectioncoefficientsby Nafe (1957) is
convenient.
124 Amplitude of Refraction Signal
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
k =
(amplitude
at)I2-(CT)Q'T'I
amplitude at TM =
2-r(SQ)
=
[( a.
R2)(a2R
1+----
a R
1+
aR cos
Q./(R3-}- R) R3
k __.
Q2/R R -}- R
1P
OR
qo- [T.C.]qi---[T.C.]-----IP
0 = i-- 60
a0
co al 1 a
i-- [T.C.] -- -- OlP= [T.C.]- OlP
ao co ao
and therefore,
al
qo'-- [T.C.] qlP.
2 sin a
qZP
cos2 tan a q- 2 sin' tan 2'
Also, (sina)/a= (sin(r)/b wherea and b are compressional and shearvelocitiesof the
surfacelayer. A time derivative is taken for velocity-sensitiveseismometers.
Summaryof amplitudefactors
Hence, for a thick refractor, ignoringfor the moment the instrument responseand
126 Amplitude of Refraction Signal
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
Sourcefunction
The sourcefunction or reduced displacementpotential is calculated from down-
hole measurementsin the vicinity of the source.Particle velocity gagesare used and
the displacementis obtained by integration. The reduced displacementpotential
f(r), wherer=t--r/a, is calculatedfrom a finite differenceapproximationof the
expression
Or;X,r /'
whereu(r, t) is the displacement
at the distancer, anda is the compressional
velocity.
The f(r) socalculatedis usedin the near-frontapproximationas the sourcefunction.
Of coursef(r) shouldbe calculatedfrom a measurementfar enoughfrom the source
for the motion to be propagating accordingto the elastic-waveequation. To insure
this conditionit is preferredto calculatean additionalf(r) from a measurementat a
different distance. If the reduced potentials calculated from such measurements
agree, then the propagation from one to the other was elastic. The use of the equa-
tion assumesa spherically symmetric source and the measurementsmust be dis-
counted after the arrival of reflections.
Attenuation
steady-stateplane wavesis
e-aR ,
where
a--w/2aQ,
R-- distancepropagated,
w-- circular frequency,
a = wave velocity,
Q= attenuation constant.
A number ofexperimentshavebeencarriedoutwhichsupport
attenuation
propor-
tionalto frequency(seereviewby KnopoffandMacDonald, 1958).Contraryto the
usualstatements,Futterman(1962)hasshownthat a lineardependenceof a on w
overanarbitrarilylargerangeoffrequencies
isconsistent
withalinearwave-propa-
gationtheory.A phenomenologically
chosenlow-frequency
cutoffmust be intro-
duced into the attenuation coefficient.For a calculational convenienceFutterman
haschosen for theamplitudedependence theexpression
shownbelow.By invoking
causality hehasderiveda phaserelationfromthisamplitude relation.By "causal-
ity" it is meantthat thetraveltime(distance
dividedby velocity)mustelapsebe-
forethe wavecanarrive.Thisrelationbetweenamplitudeandphaseis sometimes
known as the Kramers-Kroenigdispersioncondition.Futterman's attenuation ex-
pressionis
I 2aQ
exp w(1- e-:)R
q-iq51
where
=wR/ap,
-1
{1 rQ
ap=a 1[lnq
q-lax
q-(e-:Eii"X)q-e:Ei(--
a= velocity in the limit of zerofrequency,
ap= phase velocity,
ag= group velocity,
X-- /c,
w,=cutoff frequency(very low),
R= distance,
In q/:0.5772157, Euler's constant,
Ei(--x): exponentialintegralfunction,
Ei(x) = conjugateexponentialintegralfunction.
For x_>15, the normal rangeof use,
128 Amplitude of Refraction Signal
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
=a 1 1[ln
3'-4-
Inx]}
-1
=a 1 1[lnTq-lnxq-1]}
The dispersionrequired for a causal wave to develop is very small. The phase
velocityat onecpsis lessthan 1/1000 greaterthan the velocityat 20 cps.
In practical application, it is convenient to set aavaoa, the normal com-
pressionalvelocity. Note that R/a is the traveltime and qbmay be considereda
time average for the path involved. The cutoff frequency is chosenfar below the
passbandof interest. By Fourier inversiona propagationoperatoris obtainedwhich
can then be used to convolve with the source function to include attenuation effects.
This operator physically representshow a wave that starts as a delta function loses
high frequenciesby attenuation (absorption)duringpropagation.Of course,a value
of qbmust be chosenconsistentwith the measuredamplitudes and frequencieson
the records.
Instrument response
For a completeanalysisthe impulseresponseof the instrument includingseismom-
eter should be obtained. The seismometercan be placed on a shake table, and con-
nected to amplifiers and recording system. The amplitude and phase of the shake
table are comparedto the amplitude and phase of the final recording.By Fourier
inversion the impulse responseof the system is obtained. This impulse responseis
then convolved with the sourcefunction and attenuation operator.
IOOO
- \
lOO
IO
- EXPLOSIONS IN TUFF
- ( Logon 5kt
() Blanco(x0.427)
() Antler(x 1.745)
- -----Theory for Hardtack line
, I , I I I I , I ,, I I ,
I0 I00 200 300 400 500 600 700 800
Distance (krn)
Fro. 2. Amplitude of the first half-cyclefor nuclearexplosions
in tuffBlanca
and Antler have been scaled to 5 kt.
The geologic
model
In order to study the effect of layers above a refractor on the character of the
refraction arrival, we want to find a case which tends to minimize all other effects.
We choose a thick refractor so we don't need to be concerned about additional
arrivals causedby waveguideaction in the refractor or by deeper refractions.We
choosea refractor having a thick layer above so that we need not be concernedwith
wavesrefracting in or off as shear.Waves that travel such a thick layer as shearare
delayed sufficientlyin their time of arrival so that they do not interfere with waves
traveling as compressional.
The Pennsylvaniancarbonates("Penn lime") in the EdwardsPlateau countryof
West Texas reasonablysatisfy theseconditions.The "Penn lime" starts a carbonate
sequence2,000 to 3,000 ft thick extending through the Ellenburger. We will not
calculatethe effectsof the two low-velocityshalestringers(the 30-ft Mississippian
and the 60-ft stray formations).Above the "Penn lime" lie the homogeneous Per-
mian and Pennsylvanianshales,10,000ft thick. Arrivals at the surfacerefractingoff
the "Penn lime" as compressional wavesdo not sufferinterferenceby wavesrefract-
ing off as shear.
Above the Permian lies the sand and shalesof the Trinity, the limestoneof the
Comanche, and the Edwards limestone. The geologicmodel is given in Table 1.
The thicknessesand compressionalvelocities of the Permian and deeperformations
have been taken from the continuousvelocity log of the Magnolia Morrison well in
Val Verde County, Texas, near the border of Crockett County. The upper part of
the model is taken from the Delta-Gulf Lee well in Crockett County 20 miles north
of the Magnolia Morrison well. Compressionalvelocitiesare estimated and checked
against uphole plots. The shear velocities and densities are estimates based on
compressionalvelocitiesand lithologic descriptions.This sectionis south of the area
VALLEY MODEL
TYPICAL RAYS : iP,---S
SURFACE
6Off Edwards Lime
16,000 ft/sec
Comanche
85
" 12,000
250
9,000
Converted
///ect--pGhost
Waves
of the Edwards Plateau country in which the Permian shelf (dolomitesand anhy-
drites) subcropsagainst the Trinity. The effectscalculatedhere would be enhanced
even more if the subcropwere includedin the model.
Our problem is to study the effect of layers in the Edwards limestone,Comanche,
and Trinity formations on the character of the refraction arrival from the Penn-
sylvanian limestone. These layers are shown plotted in Figure 3 for the valley
model, assumedfor the source,and in Figure 4 for the mesa model, assumedfor the
seismometer.
Relativeamplitudecalculation
The compressionalwave from the explosiontravels down through the Edwards
limestone,Comanche,and Trinity formations. At each interface part of the energy
is convertedinto shearwaveswhich travel down and at subsequentinterfacesare in
part converted back to compressionalwaves. The processis illustrated in Figure 3.
In the near-front approximationeach ray is traced separately,keeping track of the
time delay of converted waves behind the direct P wave (the wave that remained
compressionalall the way). The amplitude of each ray is computed and super-
imposedas outlined in the sectionentitled "Theory." The interest in this problem
is in the effect of the converted waves in forming the character of the Penn lime
refraction arrival. As such, only rays arriving within, say, a tenth of a secondof the
direct P wave need have their amplitudes calculated. The product of the downward
and upward transmissioncoefficientsand the resultant vertical amplitude control
the amplitude of the converted waves relative to the direct P wave and the char-
acter of the arrival.
132 Amplitude of Refraction Signal
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
MESA MODEL
TYPICAL RAYS'--P;---S
SURFACE SEISMOMETER
4Oft
--Edwards Lime
16,000 ft/sec
60
II, 000
30
16,000
I00
12,000
125 /
/ 16,000
.Comanche
85 12,000
I00 16,000
Trinity
250 / 9,000
/
/
Permian
I O, 000 12,500
Direct Converted
Waves
P Wave
Fro. 4. Mesamodelassumed
for the seismometer
location.A fewof the wavepathsthat
were computedhave been drawn on the figure.
Some6,000 rays were calculatedfor the given model, using a digital computer.
The result is mosteasilyshownby plotting the amplitudeof the convertedwavesat
their time of arrival relative to the amplitude of the direct P wave taken as unity.
We shallcall sucha plot the "impulseresponse."In the computerprogrameachray
is describedby an amplitude, time, path identification, and horizontal distance
traveled while goingdown and up. Each ray starts with unit amplitudeof displace-
ment potential. Since all waves involved travel essentiallythe same distance, no
attempt is made to computedifferencesin geometricalspreading.
The paths are determinedby requiringcompressional motion at the critical angle
in the Permian. A few paths for the sake of illustration have been drawn on the
valley model (Figure 3) and mesamodel (Figure 4). All possiblerays transmitting
down from the source or surface reflected above the source and then transmitted
downwerecomputed.All possibleup-transmittedwaveswerealsocomputed.
At the seismometer,the vertical and horizontal componentsof displacementare
computedfrom the potential and the angles.The resultant impulse responsesare
displayedin the bottom panelsof Figure 5. Rays arriving in the samemillisecond
have beensummed.Rays too small to showon the plot are omitted.
These impulse responsescontain no multiples defined as waves undergoingtwo
reflectionsand the critical refraction at the Pennsylvanianlimestone.One classof
multiples was computedand found to be too small to be plotted. This classwas
compressionalwavesreflectedfrom the surfaceas compressional, reflectedagain at
the Comanche-Trinityinterfaceas compressional, and then arriving at the receiver.
Calculating the Amplitude of Arrival 133
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
WAVE
ONLY
COMPRESSIONAL
WAVE
WITH
SURFACE
REFLECTION
COMPRESSIONAL
A ARRIVALS
T SEISMOMETER
SHEAR
,f,f, _ ARRIVALS
AT
SEISMOMETER ......../
COMPLETE
I)!./
SOLUTION
I-- O. IOO sec- I-O. IOO sec I O.I OOsec---t F--- O. IOO sec ----1
We believetheshearmultiples
arelikewise
small,but wehavenotcalculated
their
amplitudes.
Multiples
areimportant
in reflection
prospecting
onthePlateau,but
thecalculation
withcompressional
multiples
leadsusto believethatwecanneglect
multiples
inrefractionprospecting.
Theremaining
amplitude factors
couldbecalcu-
lated,but,sincein thisproblem
weareinterested
in relativeeffects,
weshallnot
calculate them.
In principle
eachofthese
raysshould
beconvolved
withthesource
function
(re-
duceddisplacement
potential),the attenuation
operator,
andthe instrument
re-
sponse.
These
functions
arenotnormally
known
ingeophysical
prospecting.
Asis
donein synthetic
seismograms
forreflection
prospecting,
wewillsimplyassume
an
equivalent
excitation--equivalent
to thecombined
effects
of source,
attenuation,
and recordinginstrument.
As with theoretical
reflection
seismograms,
weshouldtry severalsuchfunctions
134 Amplitude of Refraction Signal
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
EXCITATION FUNCTION
wn
O0
TW
WO
W
_jo
_..O_
Calculating the Amplitude
n.-Z
of Arrival
0.-
o
cD c
o.
135
136 Amplitude of Refraction Signal
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
The angleof incidenceat the baseof the Trinity now changeswith slopeof the
structure. Since the convertedwaves depend on this angle, the characterof the re-
fraction arrival is a function of the slope of the refractor. This characterchangeis
displayedin Figure7.
The rays that arrive at the seismometerhave refractedoff over an interval on the
refractor as shownin Figure 4. This interval changeswith the slopeof the structure
and has been marked with bracketsin Figure 7. The slopemust be essentiallycon-
stant over this interval for our waveformsto be valid. As can be seenin the drawing,
this conditionis easily satisfiedfor the uphill refractions.The interval becomesvery
large for the downhillrefractionsand althoughthe slopeis essentiallyconstantover
this large interval the waveformsbecomevery sensitiveto angle.If the angleis less
than 38.6 degreesthe wave in principleis totally reflectedby the Trinity formation
and no arrival is shown.Becauseof the large refraction interval and the sensitivity
to angle,the downhillwaveformscan only be consideredto representa trend.
The relationship between the refraction character and the structure is clearly
evident. On the uphill side, energyis shifted to earlier times, giving a very strong
secondleg. On the downhillsidethe energyis shifted to later times,increasingthird
and fourth legsuntil they are dominant. If the slopebecomessteepenough,the en-
tire refraction arrival disappears.Of course,the record would contain later arrivals
such as the reflectionfrom the Pennsylvanianlimestone.If they arrived too early
they might mask the effect describedhere. In principle, reflectionsbeyond the
critical angle,refractionsfrom deeperbeds,and internal reflected-refractions could
be includedin the analysisby developingmore elaborate calculationalprocedures
than thosegiven here. Model studiesby Laster (elsewherein this volume) have
shownthat structurescan focusand defocusthe wavesmaking significantchangesin
the amplitude of the arrival. No attempt hasbeenmade to calculatechangesof that
type.
CONCLUSION
Knopoff, L., and MacDonald, G., 1958, Attenuationof small amplitudestresswavesin solids:Rev.
Modern.Physics, v. 30, p. 1178-1192.
Muskat, M., 1933, Theory of refractionshooting:Physics,v. 4, p. 14-28.
Nafe, J. E., 1957, Reflectionand transmissioncoefficientsat a solid-solidinterfaceof high velocity
contrast:Bull. Seis.Soc.Am., v. 47, p. 205-220.
O'Brien, P. N. S., 1960,The useof amplitudesin refractionshooting--acasehistory: Geophys.Prosp.,
v. 8, p. 417428.
Werth, G. C., Herbst, R. F., and Springer,D., 1962, Amplitudesof seismicarrivals from the M dis-
continuity:Jour. Geophys.Res.,v. 67, p. 1587-1610.
Werth, G. C., and Herbst, R. F., 1963, Comparisonof amplitudesof seismicwaves from nuclearexplo-
sionsin fourmediums:Jour. Geophys.Res.,v. 68, p. 1463-1475.
Zvolinskii,N. V., 1957, Reflectedand head wavesemergingat a plane interfaceof two elasticmedia:
1 Izv. Akad. Nauk S.S.S.R., Ser. Geofiz., v. 10, p. 1201-1218.
-- 1958, Reflectedwavesand head wavesarisingat a plane interfacebetweentwo elasticmedia:
2 Izv. Akad. Nauk S.S.S.R., Ser. Geofiz.,v. 1, p. 3-16.
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
J. w. C. SHERWOOD*
ABSTRACT
INTRODUCTION
The modelunderconsideration
is illustratedin Figure 1. A localizedseismicsource
is locatedin medium1 at a heightH abovethe centerplaneof a high-speed
layerof
medium 2 with a thickness 2d. The dilatation and shear-wave velocities and the
densitiesfor the media1 and 2 are (a,/, m) and (a.,., p.),respectively.
The prob-
lem is to determinecharacteristics of the disturbancethat will be generatedat a
point situated a horizontal distancer from the sourcecenter and a vertical distance
z from the centerplane of the layer.
SOURCE DETECTOR
d (o,o)
-- C2,
2, jO
2
d r
The source
/x= A
{Jo(kr)exp[--(k2-- o2/a2)
TM[ z q- H I](k - w2/a)-/2kt
dk. (2)
B{}=(iA/2)
[Jo(hr
sin
O)h
sin
0{}]dO,
and C is a contourequivalentto the path from 0 to (,r/2) and from (,r/2) to [(,r/2)
Symmetricvibrations
The sourcesof strength(A/2) at the positions(O, --//) and (O,//) yield distur-
banceswhichare symmetricalabout the centerplane (z-0) of the high-speedlayer.
The total symmetric radiation in media 1 and 2 in the negative z region can be
expressedin terms of dilatation and rotation potentials (/x, W), (/x2,W2), respec-
tively. The displacementsw and q parallel and perpendicularto the z axis are given
by Ewing, Jardetzky,and Press(1957,p. 9) as
q = (oa/Or) + (oW/Oz)
and
w = (oa/oz) - (6)
whereit is to be understoodthat equations(6) are valid when either the subscript1
or 2 is given to eachof the variables.
In order to satisfy the boundary conditionsof continuity of displacementand
stressat the interface (z---d) it can be shownafter sometediousoperationsthat
the potentialsmust take the followingform:
/x= B{exp[--ihcos0I z q-//I ] q- C exp[ihcosOz]},
W = [OB/Or]
{ D exp[ikcoskz]},
A.= B{E cosh[ih.cosO.z]},
W.= [OB/Or]
{F sinh[ik.cosk.z]}. (7)
In these equations
cO= hlOtl = ha. = kl/l = k5,
and
(M) = , (o)
cos2b - 2 sin2bcos p2cos2b2
cosh
X 2-2 sin2
b2
cos
b.
cosh
Y
P P
2g
-- cos 0 cos 2 . cos0 sinh X cos2 sinh Y
142 Amplitude of Refraction Signal
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
and
A ntisymmetricvibrations
Theseare initiated by the sourcesof strength(A/2) and (-A/2) at the positions
(0, --H) and (0, H), respectively.By analogywith equation(12) the antisymmetric
disturbanceswhich are refracted along the layer and then radiated as dilatation are
contained in
ZX.4
= B{([Q]/[P])exp[-ih cos0(H- z - 2d)]}. (13)
The determinant[P] is identicalwith [M] [see(10)] exceptthat (cosh)is everywhere
replacedby (--sinh) and vice-versa.Also[Q]is formedfrom [P] by merelychanging
the sign of the first element in each of the first and third rows. Finally the equation
[P] =0 is the normal modeequationfor antisymmetric vibrations.
G = (2fi/a)[ 1 q- (p,t2//p2t2
2)[(ot2//
2)-- 1]},
H-=- [1 -- (/22/0t22)][1
-- (p2/22//pl/12)]{2-
[1 -- (pl/12/p2/22)][1
--
and
x = exp (i02).
Where the choiceof sign occursin equation (15) the upper and lower signspertain
to the symmetricand antisymmetricmodes,respectively.Equation (15) has been
investigated for the specificcasewhere
(a,//,) = (a/)= 3'/, (pv.//pl)= 1.2, (ot2/Otl)-=-(/2//1) = 1.5,
which leads to G= .91358 and H-- 1.79094. At low values of g the only solutions
which infer unattenuated propagationin the horizontal direction are'
Symmetric: x .4870i/2 or 0.= (,r/2) -- iln (.4870/2)
(16)
Antisymmetric: x - 2.6277i/g or 0.= (,r/2) -- iln (2.6277/).
DIRECTION OF PROPAGATION
zr
WAVE-/ '
FRONT
a) UNATTENUATED
MODE b) ATTENUATED
MODE
Fo. 2. Amplitude variation of modes.
in Figures
3 and4. It istobenoted thateach
mode turnsaroundat some
upper
frequency
limit.At these points,
attenuatedmodesareexpectedtooriginate
(Sher-
wood,1958b),which will,ingeneral,
radiate
energy
fromthelayer.
Attenuated modes
PHASE
VELOCITY
I _ C
FIRST
I
ATTENUATED
MODE
.8
........ FREE
mm
PLATE
mmm mm mmm mm
.4
'-UNATTENUATED
MODE
,0,:'(wd/a2):
.5 1.5 2 2.5 .0
0
-I (sinO
2 )=a0[2
.8
ATTENUATION
CONSTANT
Fro. 3. Symmetricalmodes.
146 Amplitude of Refraction Signal
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
PHASEVELOCITY
1.2
1.0 _ R!sin2)02
.8
.6 FIRST
ATTENUATED
ODE
.4 ___:.
....... .....
.2 - NUATED
IODE
.4
.6i.
..
.8i,, i1
1.0 ).
...'";::TTENUATION
OF
SECOND
MODE
-I (sin2)=o 2 ..'
1.2
ATTENUATION
CONSTANT
are, respectively, the phase velocity and attenuation constant normalized to the
velocity
The symmetric mode in Figure 3 exhibits two interesting plateau levels at phase
velocitiesof approximately.94 and 1.01 times P-wave velocity, a2,in the high-speed
layer. Also shownis the first symmetric mode of the high-speedlayer situated in a
vacuum. It is to be especiallynoted that the first plateau level of the symmetric
attenuated mode possesses essentiallythe samephasevelocity as the thin isolated
plate. It alsoseemssignificantthat the attenuationfactor (aw) [seeequations(17)
and (18)] remainsfairly constantover this plateauregion.In other wordsone might
expect approximately nondispersivepropagation over the normalized frequency
range for which the first plateau persists.
The dominant attenuated antisymmetric mode of Figure 4 does not display
particularly distinctive characteristics.Perhapsthe most notable feature is its close
resemblancewith the phasevelocity of the first antisymmetricmodefor an isolated
layer.
The frequencyvariation of the attenuation constantis noteworthy. For each of
the dominant attenuated modesit risesrapidly from zero to a maximum, and then
returns gradually towards zero. Due to the rather high-maximum attenuation
Embedded High-Speed Layer' t47
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
(a/U)
= 0(w sin 0)/0w
= sin 03-3-c0(0sin 0/0w).
(20)
The normalizedgroupvelocitiesfor the unattenuatedmodesare drawn in Figure
5. The variationsare somewhatalarming,particularlythe excursions into velocities
greaterthan the P-wave velocity,a, in the high-speed
layer. Evidently, groupveloc-
ity cannotbe equivalentto energyvelocityin this instanceand this is apparentlyin
conflictwith a theoremdue to Blot (1957). Possiblythis discrepancyis due to the
fact that the amplitudesof the modesincreaseexponentiallywith distancefrom the
layer, a situation which is not completely realizable.
The problem of a meaningful definition for group velocity in the case of an
attenuated mode also presentsa problem. Someauthors (Oliver and Major, 1960;
Gilbert and Laster, 1962) appear to have obtained not unreasonableresults by
essentiallyignoringthe presence of the attenuationfactor.On the otherhand,others
who have employed this same apparently plausible approach have come to con-
siderablegrief. Suchan operation,in fact, onceproducedan apparentconflictwith
Einstein'stheory of relativity and arousedconsiderablecontroversyearly in this
century.A very interestingaccountof this subjecthas recentlybeenproducedby
Brillouin (1960). It only remainsto say that neglectionof the modal attenuation
factorsin the problemunder investigationhere doeslead to groupvelocitiesgreater
than a. Application of methodsdevelopedby Phinney (1961) and Rosenbaum
(1960) wouldcorrectthis situation.The procedureconsistsof extendingthe numeri-
cal analysisto solvingthe normal mode equationsfor complexrather than merely
real values of frequency.The solutionseffectively definesurfaces,the saddlepoints
148 Amplitude of Refraction Signal
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
!
u NORMALIZED
CROUP
VELOCITY
ANTISYMMETRIC
MODE
NORIALIZED
FREQUENCY,
-4
MODE
-8 I
I
I
I
-12
Signal velocity
Lavergne's(1961)data is replottedin modifiedformin Figure6 togetherwith the
theoretically determined phase velocity for the first symmetric attenuated mode.
Strictly speakingwe shouldnot anticipate goodagreement.In the first place the
theoretical computationswere performed for a model with elastic parameters
differingconsiderablyfrom thoseemployedby Lavergne.Secondly,the theoretical
phasevelocity is plotted rather than a signal or energy velocity. However, this
shouldnot be a major drawbackover the frequencyregionswherethere is relatively
low velocity dispersion.Inspectionof Figure 6 shows,in fact, that not unreasonable
agreementoccursover the low-dispersionregionfrom (2d/X) =0.2 to 1.0. The main
differenceis a shift in frequencyof the transitionregionfrom thin-plate velocity to
infinite medium velocity.
VELOCITY
(12
Cp
i
!
i
0 .2 .4 .6 .8 1.0
is thinplatevelocity. (2d/X)
,
are Lavergnes data points.
Fro. 6. Comparison
of the theoreticaland observed
velocitydispersion
of the dominant
symmetricmode.
150 Amplitude of Refraction Signal
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
Intercept time
Lavergne (1961) extrapolatedhis refractionpicks to zero offsetfrom the source
and measuredthe intercept time. This measurementwas performed for a range of
valuesof the layer thicknessto wavelengthratio, (2d/X). Theseexperimentalvalues,
however,differednoticeably from intercept times, T, calculatedin the conventional
manner using the formula
T = (2///a) cos0,
where (23)
sin 0 = a/c.
Here is it assumedthat both the sourceand detector are a distance H from the upper
surface of the refractor. The propagation angle of the dilatation head wave, 0, is
supposedto be real, while c representsthe velocity of the refracted wave. It is
important to note, however,that equations(23) are actually only approximations.
The actual angle of propagation is, in our nomenclature, 0, and this is in general
complex.It is related by equations(8) to the modal parametersplotted in Figures
3 and 4. The real part of 0 specifiesthe direction of propagationof the dilatation in
the upper medium. The imaginary part of 0, however,implies that the wave velocity
is actually lessthan a (see Sherwood,1958a, p. 208). If the complexnature of this
angle0 is taken into accountit is a simplematter to showthat the intercept time is
actually given more accurately by
r = (24)
It wasinitially hopedthat the useof formula (24) in placeof formula (23) would
Embedded High-Speed Layer 151
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
It has been shown how attenuated normal mode theory accountsfor the main
featuresobservedin experimentalstudiesof refraction along a high-speedlayer. The
theory showssatisfactoryagreementwith experiment for the variation of the phase
velocity and the attenuation with the thickness-to-wavelengthratio. On the other
hand it appearsthat satisfactoryvaluesfor the energyor signalvelocitiesat very low
frequenciescannot be derived usingthe analysisemployedhere. From this point of
view it would seem advantageousto reanalyze the problem employing either the
techniquesof Phinney (1961) and Rosenbaum(1961), or someother comprehensive
method.
REFERENCES
P. N. S. O'BRIEN*
ABSTRACT
INTRODUCTION
UNDERGROUND EXPLOSIONS
if it is very soft. For most rocks,no doubt, both shattering and flow will take place.
Outside this surface, which is usually called the equivalent radiating surface, the
stressesare small enoughto apply infinitesimal elastic wave theory. Sharpe went on
to solvethe problemof the wave motion due to an arbitrary radial stressapplied to
the surfaceof suchan equivalent radiator, which he assumedto be spherical.Later
(Sharpe,1942b), he made somepioneermeasurementscloseto the shot which con-
firmed the major conclusionsof his theoretical analysis. Sharpe'sanalysiswas for a
rock with a Poisson'sratio of 0.25. Blake (1952) extendedthe theory to cover all
valuesof Poisson'sratio. This wasa very necessarystep as many near-surfacesedi-
mentary rocks have a Poisson'sratio of about 0.4, and even higher values are not
uncommon.
ATION PRESSURE,P
I.SxlO5
0.4XlO5
I
I
j ,
.577
at the sametime tend to eject the tampingmaterial. On the surfaceof the equivalent
radiator, the initial pressureis probably greater than the later pressuredue to the
expandinggasesand will still be largely impulsivewith a duration very much less
than the width of the observedseismicpulse.The pressurein the gaseswill decay
due to movement of the tamping material, but, unlessthe chargeis so closeto the
surface that the tamping is rapidly ejected, it will not change much over times
comparable with the duration of the seismicpulse. The decay time constant will
dependprimarily upon the weight of explosiveand the nature of the rock and will,
perhaps,be of the orderof one-halfsecfor a chargeof 100-11; weight.
.oxoS(o t.)
TIHEHARK50
Fro. 2. Pressure
closeto a water-tampedcharge(fromIto and Sassa,1962,usedby permission).
Efficient Use of Large Charges 155
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
The valuesof T, V, and a are known, and so a value for a may be computed. For the
example in Figure 3a, it is found to be 3.8 ft. This is the radius of the equivalent
radiator. From the observedshape of the dilatational pulse at 10 ft the shape of
the radial stressmay be calculated at the surfaceof the equivalent radiator. Figure
3b showsthis calculated radial stress'it is not a reliable picture, becausethe neces-
sary integrationsdepend rather critically on the preciseform of the observedpulse,
but it servesto show consistencybetween predicted and observedstresses.
In order to predict how the seismicsignal varies with the weight of chargefired,
we need to know how the limiting stressand the size of the equivalent radiator vary
with weight of charge. If we know these, then Blake's theory may be applied to
compute exactly the variation in the seismicpulse. The amplitude of the limiting
stresswill remain constant,for this is determinedby the tensilestrength of the rock.
The variation in the size of the equivalent radiator and in the pulse width may be
found by using scalinglaws. There seemsto be no discussionof the derivation of
theselaws in the open literature, but it can easily be shownby dimensionalanalysis
that, for a given type of explosivein a homogeneousrock, (1) the volume of the
radiator is proportional to the weight of chargefired and (2) the time scaleof the
15
o IO
z
DECAYING SLOWLY
FROM HERE ON
' o
TIME IN MILLISECONDS
16
14
stress
pulseisproportional to thecuberootof theweightof charge,providedthat the
factor (distance/W/3) is kept constant.
It isnotpossible
to measure
directlythesizeof theequivalent
radiator,butmeas-
urements havebeenmadeof thesizeof thecavityproduced by an explosion,
and
Figure 4 showsone such set of results.
The cavity radius,R, is givenby the equation
R = BW TM
12
m 8
-- 7
,/
uJ 4
The stresspulse at the equivalent radiator is roughly of the form given in Figure
3b. This may be approximated to by an exponentialpulse with a relatively long
Amplitude of Refraction Signal
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
time constantplus an initial impulse which lasts only one or two ms. In the sub-
section on energy release,we give someevidencethat the initial impulseprovides
only a small amount of the low-frequencyenergy, and we, therefore, represent the
limiting stressby P(t) where,
P(t) = 0 fort < O,
P(t) = Poexp(-t/a), t70.
P0 is a constantfor all chargeweightsbecauseit is closelyrelated to the strength
of the rock. On the other hand, a scalesaccordingto the secondof the laws given
aboveI and is, therefore,given by a= (TW) 1/ and is possiblyof the order of one-
tenth secfor a one lb charge.
The frequencyspectrumof such an exponentialspike is given by
= +
We may thereforesubstitutefor P and a in Blake's equation and obtain the de-
pendenceof the particle velocity on the weight of chargefired. Putting aa=DW,
where D is a constant of proportionality, we have
.40,= w(PK/pV2)(DW/r) {ws-+-(TW)-2/a}-/s{
(w/V)s-+-1/rs /
(2)
{ 1 + (1 - 2K)(coaDW/Va)
s/a+ KS(coaDW/Va)4/a}-/s.
Deductionsand applications
It is instructiveto examinethe approximationsto this equationfor low and high
frequencies. In refractionshootingws>(TW)-s/a,sothat, if wehave (waDW/Va)<<I,
A0,is directly proportionalto W for all o.Referenceto Blake'spaper showsthat in
this rangeof frequenciesthere is no phasedistortion,and so,if the weightof charge
is doubled,the seismicpulseamplitude will also be doubled.For high frequencies,
on the other hand, phasedistortiondoestake place, and also A0,becomespropor-
tional to the cuberoot of the weight of charge.
Providing the first of the above conditionsis fulfilled, the variation of A0, with
chargeweightdependson the parameterx=(w/V)(DW) us,in the mannergivenby
the function
lO
IOO
80
60 -
4o
_1 CALCULATED
o OBSERVED
2o d POISSON'S
RATIO
IO
8
I STANDARD
ERRO,
6
I
27-5 55 137-5 275 550 IIOO 2750
CHARGE WEIGHT IN POUNDS
Fro.7. Seismic
amplitude
asa function
ofweightofcharge
firedin a single
hole.
160 Amplitude of Refraction Signal
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
weight had beendivided into five equal parts and loadedinto five holes,eachsepa-
rated from its neighborby at least the diameter of the equivalentradiator, then the
amplitude of the refracted arrival shouldhave increasedby a factor of about 2.5.
From which it follows,either that the amount of dynamite could have been reduced
in this proportion, or that readablerecordscould have beenobtained under noisier
backgroundconditions.It may be noted that the pulse width for the larger charge
was significantly greater than that from the smaller charges,an observationwhich
supportsthe frequencydependenceof the scalinglaw.
Figure 6 showsthat, when is lessthan about 0.4:,the linear law fails to hold for
valuesof x greater than about one. Utilizing this conditionthe radius of the equiv-
alent radiator for the shotsin Figure 7 wascalculatedto equal4:.5W 1/3ft, whereW is
in poundsweight. This is quite a typical value for rockswith velocitiesaround 5,000
to 6,000 ft/sec and is rather more than the value of 3.8 W 1/ ft quoted earlier for a
soft sandstone.
By plotting a graph similar to the oneshownin Figure 7 the maximumchargeto
be fired in a singlehole may be found by observation.2 However, a good enough
value may quickly be calculatedfrom the equation
X- (co/V)(DW) 1/ -- E,
whereE liesin the range0.5 (softrocks)to 1.0 (hardrocks).V is the P-wve velocity
at the shotpoint,D / is usuallyabout 3.5-4.5, and cois the highestangularfrequency
of interest in the observedarrival. Assumingthis to be twice the predominant fre-
quency,it may be taken as 2rT, whereT is the durationof the first half-period.
Equation (2) showsthat, for a givenweightof charge,the seismiclow-frequency
mplitudeisproportionalto thefactor(PKD/pV). For K= 1 (a= ) thismaybere-
written as 4 PD/V, where $zis the rigidity of the rock. Sinceit is well known that
softerrocksgive larger amplitudesit is apparent that the variation from rock to rock
is greater for the ratio D/IaV than it is for elastic strength. For a given rock the
distant amplitudeis directly proportionalto D, that is, directly proportionalto the
volumeothe equivalentradiator. It may be supposedthat the crackingand failure
of the rock is due mainly to the initial impulseimparted by the detonationwave in
the explosive.Sincethis is greater for explosiveswith greater detonation velocities,
such explosiveswill give greater seismicamplitudes than those with lower detona-
tion velocities.This has been amply verified by experiment (Atchisonand Roth,
1961). Other characteristicsof the explosiveare important, for exampleloading
density and available energy,but the most important singleparameter is probably
detonationpressure,and this is very closelyrelated to the easily measureddetona-
tion velocity.
Seismic energyrelease
Measurementscloseto properly tamped shotsshowthat the energyin the radiated
P wave is essentiallysphericallysymmetric.For a soft sandstone,measurementson
pulses,suchas that indicatedin Figure 3a, indicate that the total radiated energy
(potential plus kinetic) is about 4 percent of the available chemicalenergy in the
explosive.Due to absorption, the seismicpulse at 10 ft containsonly 2.5 percent oi
the available energy and, at 100 ft, only 0.50 percent. The energy in the low-fre-
quencyband relevant to refractedarrivals is probablyno more than 1/50th of this
last amount.
Similar measurementsfor shotsin clay showthat the total radiated energy close
to the sourceis nearer to 9 percent.
Thesefiguresare much the sameas thosewhich may be deducedfrom the observa-
tionsof McDonal et al. (1958) and Ricker (Van Melle, 1954) for explosionsin shale.
An extensiveset of measurementsby Fogelsonet al. (1959) on strain wavesdue to
explosionsin a granite-gneissindicate a seismicefficiencyin the range 10-18 percent.
This increasein efficiencyis probably due to the increasedstrength of the rock.
Most of the energycloseto the shot is of high frequency;from equation (2), there-
fore, and the fact that energy is proportional to pVA ', it follows that the total en-
ergy is roughlyproportionalto (P'DvaW/a)/pV.Comparisonof their pulseshapes
162 Amplitude of Refraction Signal
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
r 25-0
z
n ; ,
020-0 O - O
O O
15'0
O O
I0'0 O O
O O O
D O
-,.,- 5-0
/ o
0
u 2'5
.,e.. 2. o
u
I-5
z
>' ?
/
U
0
0'5
"- 0
IO IOO IOOO
Fza.8. Spectral
amplitude
fora pulseobserved
100ft froma one-lbcharge
in softsandstone.
Nocorrection
has been made for absorption.
Efficient Use of Large Charges 163
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
E- 8'r2pVW'
i0
Al'df,
Table 2. Percent energy release from explosions in water and in soft sandstone.
Upper frequencylimit in cps
Charge
weight 10 50 100
in pounds
Water Rock Water Rock Water Rock
figures are also given for underwater explosions,these being discussedin a later
section.The low-frequencyspectraon which these estimatesare basedare not very
reliable, and so the figuresin the table are not very reliable. However, the general
trend is not in doubt, and more accurate results would not be of much extra use
becauseof the large variation between one rock and another. For instance, by
assumingthe fl.0 law it may be estimated (very roughly) that shots in granite
releaseonly about one tenth the low-frequencyenergy of shotsin sandstone,and
shotsin sandstoneare only aboutone-quarterasefficientasthosein clay.
Departurefrom sphericalsymmetry
The above theory has been developed for a spherically symmetrical source. A
chargeof 200 lb weight loadedinto a singlehole will commonlybe about 20 ft long
and 4 or 5 inchesin diameter.The equivalentradiator for sucha chargewill resemble
an ellipsoidof revolution with a major axis some three or four times its minor axis.
Usami and Hirono (1956) have shown that such a radiator will still behave essen-
tially as a sphericalsourceproviding the observedfrequenciescorrespondto wave-
lengthsgreater than two or three times the major axis. In long-distancerefraction
surveysthe predominant frequenciesare usually between 10 and 20 cps and the
maximum frequencyof importanceseldomexceeds30 to 40 cps. For a shothole
164 Amplitude of Refraction Signal
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
where W is the weight in poundsand d is the depth in feet. In order to prevent emis-
sion of secondarypulses,experienceshowsthat the chargemust be detonated at a
depth no greater than one half of that calculatedby putting R-d in the above
equation(Knudsen,1958; Griffith et al., 1962).
Providingthe chargeis detonatedat the top, as safetydemandsthat it shouldbe....
Efficient Use of Large Charges 165
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
If the chargeis deep enoughfor the gasesto expandand contract again then the
first bubble pulse will be radiated at a time T after the shockwave where
T -- 4.4 W/a(d q- 33)-5/0 sec.
T is, thus, nearly always greater than the period of the first refracted arrival, so
that it is only whenlater arrivalsare beingused(as they alwaysshouldbe) that it is
necessaryto explodethe chargeat a shallowdepth. The above values for R and T
are for TNT, they will not be much different for other explosives.
It is the initial shockwave which generatesthe observedrefracted wave. Even for
a 1,000-lb charge, which is much greater than is normally necessary,the time con-
stant of the shockwave is only about two ms, soits seismiceffectis purely impulsive.
It followsfrom the scalinglaws for the shockwave that the seismicamplitude, A, is
proportional to W (O'Brien, 1960) where n varies slightly from one explosiveto
another, but is usually about two thirds. A recent paper by Muller et al. (1962)
presentssomevery goodexperimentalresultsfrom explosionsin lakes whichprovide
a value for n of 0.65___ .013. Ideally, therefore, the maximum seismicamplitude will
only be obtained by splitting the chargeinto a number of smallerunits, in the same
manner as for undergroundexplosions.However, this is not often worth doing, for
the financial saving is small and the operational difficulty is usually great.
Seismicenergyrelease
O'Brien (1960) has shown that the percentageof the shock-waveenergy in the
seismicfrequency range is given approximatelyby the factor 400 k2 F/a, where F
is the highestobservedfrequency,1/a is about 5.0X10 -5 W TMof a second,and k is
about 1.75 (not 1.2 as statedin the paper). This factor will not hold for very low fre-
quencies,becausethe analysisassumesno rarefaction to be generatedby the explo-
sion, and it is fairly certain that this is not true.
Arons and Yennie (1948) state that only about 25 percent of the shock wave
energy is actually radiated, the rest being lost in turbulence and other dissipative
phenomena,and so the useful seismicenergy from an underwater shock is about
1.5X10 -2 W/F. Table 2 shows the energy release as a function of W and F and
comparesthe values with those for underground explosions.
The figuresin the table show that underwater explosionsare far more efficient at
producing low-frequency energy than are underground explosions.But this is not
the whole story. The ghost reflectionis much greater for underwater explosionsand
this will considerablyreducethe refraction arrival. Also, the water bottom often has
a very high reflectioncoefficient,with the consequencethat only a small proportion
of the incident energy reachesthe underlying refractor. The secondfactor is espe-
cially critical and can easily reducethe efficiencyof an underwater explosionto less
than that of one underground.
Field experience on this point is limited and contradictory. Certainly, much
smaller chargesare needed when doing large-distancerefraction shooting at sea
than for similar distanceson land, but this may be due to the relative simplicity of
submarine
geology.Steinhartand Meyer(1961)reportthat muchlargeramplitudes
areobtained
fromunderwater
shotsthanfromquarrybasts,but pointout that
166 Amplitude of Refraction Signal
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
If the water is shallow enoughto avoid producing a bubble pulse it is often con-
venient to fire on the water bottom. Sometimesthe rigidity of the bottom is very
small, and the acousticimpedanceis not much different from that of water. In this
caseit obviouslymakeslittle differencewhether the chargeis on or off bottom. How-
ever, for hard bottoms, the essentialasymmetry at the charge location causesthe
production of a large amount of shear wave energy. In such a case the primary
S-wave amplitude may be greater than that of the primary/' wave. Even if vertical
geophonesare beingused,the wavesconvertedfrom the initial S will certainly com-
plicate the seismogram.Becausefiring on the bottom doesnot increasethe amplitude
of the primary P wave it is goodpractice to fire only suspendedcharges--unless,of
course,the interpretation is made more sure by the presenceof identifiable shear-
wave events.
roughfigureswhich apply particularly to TNT (Brode 1959), but they serveto give
an idea of the order of magnitude of the forcesinvolved.
Becauseof the high acousticcontrast between air and rock, it is obviousthat most
of the blast wave producedby an air shot will be reflected at the ground surface.
This will be especiallytrue on outcropsof hard rock.
Becauseof the rapid attenuation of the peak pressureof the blast wave and be-
cause rocks and soils are relatively strong in compression,chargesexploded a few
feet above ground produce practically no damage. Even for chargeslaid on the
ground, only a very small amount of cratering is produced. For instance, a 400-lb
charge fired on a baked clay surface produced only minor cracking and a surface
depressionof a few inches.
There is little information on the relative efficiencyof underground and surface
shots. In reflection shootingTsvetaev (Hermont, 1948) found that underground
shots were 15 to 20 times as efficient as air shots. However, Poulter (1950) claims
that if proper attention is paid to the spacing,a pattern of air shotswill require at
most twice as much explosiveas a singleundergroundshot.
Buffet and Layat (1960) have reported on an extensiveset of experimentson
surface chargesin the sand dunes of the Sahara. They found, in common with
Tsvetaev and Poulter, that more seismicenergy was generated by a given weight
of charge if it was suspendeda few feet above the ground surface. This may be
becausea slight raising of the charge from the ground surfacerapidly reducesthe
peak pressurein the blast wave to below the elasticlimit of the rock and also spreads
the incident wave over a larger area. One would expect that the necessaryeleva-
tion of the chargewould be quite small and that raising it to greater heightswould
produceno increasein efficiencyand would eventually reduceit. Further, one would
expect that the height for maximum efficiencywould be roughly proportional to the
cube root of the weight of charge. This leveling-off effect was observedby Buffet
and Layat, who alsofound that it occurredat a greater height for a greater charge.
For a 25-kg charge, heights greater than about three m gave no improvement.
The exact value for the limiting height will obviouslydependupon rock type.
Miller and Pursey (1954) have shownthat for a surfacesourcesmall compared
with a wavelength the seismicamplitude is proportional to the applied pressure
and to the area of the source. For explosions,the sourcearea will be proportional
to W TMand the low-frequency pressurewill be proportional to W /. This latter
proportionality follows from the fact that the initial blast wave is essentiallyim-
pulsive, with a duration which is much shorter than those observedon a seismic
record and which dependsupon W /. We should,therefore,expect the seismicam-
plitude from a singlechargeto dependdirectly upon the weight fired. In the field, we
would expect the dependenceto be somewhatlower than a first power becausethe
shape of larger chargeswill usually be different from those of smaller chargesand
the area scalinglaw will not be followed exactly. Buffet and Layat found the ampli-
tude to be proportional to W /4. However, they compared the different charge
weights for a constant height. Since charge efficiency increased with height, the
exponentfor a more properly scaledexperiment would be even closer to unity. For
168 Amplitude of Refraction Signal
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
CONCLUSIONS
Underwater shots are usually more efficient than underground shots, and the
latter are more efficientthan surfaceshots.The ratios dependmarkedly on the pro-
cedure adopted at the shotpoint and on the near-shot conditions. For instance, if
there is a high acousticcontrast at the water bottom, only a small proportion of
the initial energy will reach the refractor, and an underground shot may then
become more efficient than an underwater shot.
For underwater shots seismic amplitude is roughly proportional to W 2/a, and
little advantage is to be obtained by splitting large chargesinto a number of smaller
ones.For large undergroundexplosionsa greater seismicamplitude is obtained if a
number of widely spacedholesare drilled and a maximum of 200 to 300 lb of explo-
sive is placed in each. Whether it is better to put a greater amount of explosiveinto
a singlehole, or to put a smaller amount into a number of holesdepends,of course,
on their relative effectson the cost per line-mile.
Surfaceshotsshouldalso be divided into a number of small individual pieceseach
placed at a sufficientdistancefrom its neighbour.
The most efficient shotpoint procedure depends so much upon local conditions
that it is impossible to lay down any hard and fast rules which will be generally
applicable. It is hoped, however, that the discussiongiven in this paper will help
exploration seismologistswhen they seek to discover the most efficient procedure
for their particular area.
ACKNOWLEDGMENTS
The author wishesto thank Mr. L. H. Tarrant, for his critical reading of the first
draft of this paper, and the Chairman and Directors of The British Petroleum
Company Limited for permissionto publish it.
REFERENCES
Ito, I., and Sassa,K., 1962, On the detonationpressureproducedat the inner surfaceof a chargehole:
International symposiumon mining research,v. 1, G. B. Clark (ed.), New York, PergamonPress.
Knudsen,W. C., 1958,Elimination of secondarypressurepulsesin offshoreexploration(a modelstudy):
Geophysics, v. 23, p. 440-458.
McDonal, F. J., Angona, F. A., Mills, R. L., Sengbush,R. L., Van Nostrand, R. G., and White, J. E.,
1958, Attenuation of shear and compressional wavesin Pierre shale: Geophysics,v. 23, p. 421-439.
Miller, G. F., and Pursey, H., 1954, The field and radiation impedanceof mechanicalradiators in the
free surfaceof a semi-infiniteisotropicsolid:Proc. Roy. Soc.,A, v. 223, p. 521-541.
Muller, St., Stein, A., and Vees, R., 1962, Seismicscalinglaws for explosionson a lake bottom: Zeit. f.
Geophysik,v. 28, p. 258-280.
Musgrave, A. W., Ehlert, G. W., and Nash, D. M., 1958, Directivity effect of elongatedcharges:Geo-
physics,v. 23, p. 81-96.
O'Brien, P. N. S., 1957,The relationshipbetweenseismicamplitudeand weight of charge:Letter to the
Editor, Geophys.Prosp.,v. 5, p. 349-352.
1960, Seismicenergyfrom explosions:Geoph.Jour., v. 3, p. 29-44.
Passechnik,I. P., 1961, Seismicmethod of detectingand identifying nuclear explosions:Bull. Ac. Sc.
USSR, Geophysics Series(Issuedin Englishtranslationby A.G.U. September,1961).
Poulter, T. C., 1950,The Poulter seismicmethodof geophysicalexploration:Geophysics, v. 15, p. 181-
207.
Pursey,H., 1956,The powerradiatedby an electromechanical wave source:Phys. Soc.[London]Proc.
B, v. 69, p. 139-144.
Sharpe,J. A., 1942a,The productionof elasticwavesby explosionpressures.I. Theory and empirical
field observations:Geophysics,v. 7, p. 144-154.
1942b,The productionof elasticwaves by explosionpressures. II. Resultsof observations
near
an explodingcharge:Geophysics, v. 7, p. 311-321.
Steinhart,J. S., and Meyer, R. ?., 1961,Explosionstudiesof continentalstructure: Washington,Car-
negie Institute of Washington.
Usami, T., and Hirono, T., 1956, Elastic wavesgeneratedfrom a spheroidalcavity whosewall is sub-
jectedto normalstressof harmonictype: PapersMet. Geophys.Tokyo, v. 7, p. 288-321.
Van Melle, F. A., 1954,Note on "The Primary SeismicDisturbancein Shale" by N. Ricker and W. A.
Sorgein the July, 1951, Issue of the Bulletin of the Seismological
Societyof America: Bull. S.S. A.,
v. 44, p. 123-125.
Weston,D. E., 1960,Underwaterexplosions as acousticsources:Proc.Phys. Soc.,v. 76, p. 233-249.
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
AND
INTERPRETATION
C. LAYAT*
ABSTRACT
INTRODUCTION
* CompagnieGnrale de Gdophysique.
172 Amplitude of Refraction Signal
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
sina__-a__-tana, cosal
AI B'
I II
I i
I i
I I
I
schematically
represented
and exaggerated.
Delay Time and Correlation Interpretation 173
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
$ X
M
Fro. 2. Illustration of the basicormulrelative o rcmctcpath:
where X is the horizontal distanceSR, and Tas and Ta are the delay times relative
to path SA and BR; the points A and B are associatedwith the layer having
velocity V. Ta and Ta are designatedas the delay times at the shotpointS and
any geophoneR.
Intercept.The interceptT0 relative to a given path SABR with a traveltime
T and an assumedvelocity V of path A B is
X
T0 = T- (3)
V
SR
To = T- (6)
V,
The origin chosen for the relative curve determines a relative intercept Tot.
Writing:
T0 = To + C; To = T0 -- C (7)
C is a constant for all the spreadsshot from S; it is called the "difference constant
relative to S."
The delay time at the point of the marker is either the delay time Tdsat the
shotpoint for the shotsfrom S, or the delay time Td at the geophonefor the shots
correspondingto the spread including
Now:
Fro.3. Illustration
of shotpoint
S delaytimecalculation
fornonreversed
shot.
and
(9)
E Tot
- TaR= Tor R.
By algebraicaddition,since
T%= Tas
(lO)
T%d-T%= 2T%= Tore
-- C.
Moresimply substituting
T%=Tos; (theinterceptreadontherelative
curve
at
a particular
pointdetermined
fromS in "fieldposition"
bydisplacing
bytwooffset
distancesin the direction of the shooting).Finally,
=
s 2
If onlyanapproximatevalueVaof themarkervelocity
isusedforthecalcula-
tionsofintercepts,
it iseasily
demonstrated
thattheformula givingthedelaytime
is transformed into
Ts= Tos-C
2SI;(
. 1 1) 2 V V
(12)
Therelativeanddelaytimecurves
arenolonger
parallel;theyhavea certaincon-
vergence
whichcanbealgebraically
measured
in thedirection
of theshothaving
a gradient,definedby the expression
AT0- ATa
where zXT0ris the differencebetween the relative intercepts read on the relative
curve (offsetposition)betweentwo points of a zXXdistance,and/XTdis the difference
between delay times at the correspondingpoints.
If in a particular area the offset "a" can be assumedas independentof the direc-
tion of the shot
a--2)
SE;
and nearly constant; if, in addition, the velocity V of the marker doesnot vary,
the corrective term
rd
In other words, if the values of delay times for several shotpointscan be deter-
mined with this formula, the gradient of convergencebetween each pair of such
points can be established,and hence, a determination of true velocity V is pos-
sible and the correctiveterm to be applied to the delay times is obtained. The value
of delay times can be calculatedfor all shotpointsS providing a point corresponding
to which is displacedby two offsetsin the direction of shootingis within the pro-
file itself (that is the seismographspreadin the field).
In general the offset "a" is not accurately known, and an approximate offset
"a'" must be used.It is possibleto showthat this error in the estimate of the offset
has the following consequences' the delay time curve drawn with offset a' is de-
termined from the actual curve (Figure 4) by substituting for point ' of the
marker, point ' with the same abscissaas ' and an ordinate which is the mean of
and '.
If the dip variations are small, when the error of offset is not very large, the two
curvesare practically similar.
The final delay time curve will be drawn by interpolating the relative curve
between the various delay time points calculated and by extrapolating it at the
extremitiesof the profile,taking into considerationthe mean convergence measured.
ShootingiN bothdirections(doublecontrol).mlf the shotshave been recordedsys-
tematically in two directions(direct and reverseshooting),each directionmay of
coursebe interpreted separately and the results compared. However, it is much
more interestingto exerciseinspectionat the time the shotsare taken and to control
the resultsin the two directionsby meansof a suitableinspectionprocess.
Mean relativecurve.--Consideringthe two relative curvesin offsetpositionob-
tainedfor eachdirectionof shootingthe algebraicconvergences
in relationto the
Delay Time and Correlation Interpretation 177
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
/
/ /
/ /
/ /
/ /
/ /
/ /
/ /
/ /
//
/
If the data are plotted on the same diagram and a singledirection of shot, the
convergenceswill be equal but of opposite sense.The mean relative curve ob-
tainedbytakingtheaverage
oftheordinates
T%andT%ofthedirectandin-
verserelative curvesat eachpoint will have a zero convergence;
it will be parallel
to thedelay curve.
The measurementof the entire convergencebetween the two relative curves
alsopermits immediatedeterminationof the real velocity without requiring any
calculation of delay times'
AT0, -- AT % 1 1
.... (14)
2AX V V
Convergence
curve.--It is interestingto plot the curve obtainedfor eachpoint
by calculating
thedifference
T%--T%.Theslope
of thiscurvecalledthe"con-
vergencecurve" (which is a straight line when V is constant) determinesthe
real velocity of the marker for each point, as well as the computation of the
correctionterm applied to the delay time values'
Sd S R Si
/!
1
1
1
/
/
1
1
/
/
1
To= T% -- Ca. i
[from
equation
(7)]
Assumingnow that a shot is fired from Si, the intercept correspondingto a geo-
phoneplacedat S is by definitionequalto the inverserelativeinterceptT%.
From the principle of reciprocity'
To= T0 T08
'
Hence,
T0rh-Ca= T0,8.
Likewise for the inverse shots with similar definitions'
T0,ah-Ci= T0,
8d .
And,according
to thissamereciprocity
principle
T%i-- (suppose
S at
Finally the differenceconstantsmust verify the relations
Delay Time and Correlation Interpretation t79
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
Operating
Improvementsin field techniqueto which this is applicablehave been discussed
in a previouspaper (Layat, Clement, Pommier, Buffet, 1961).
Asmarkerhorizonsmay be difficultto distinguishbecauseof their similarvelocity
characteristics,it is essentialthat doublesubsurfacecoveragein each direction of
shooting should be recorded (i.e. at least two shots at different distancesin both
directionsfor every geophonelayout). This is obtainedpracticallyin the field by
usingdoubleinput geophonecableswith 12 traceson eachside of the recording
truck and by advancingthe spread12 tracesinstead of 24. In someinstances,if
severalmarkersare to be investigated,or if the usefullength recordedfor a marker
is quite short, an increasednumberof shotsshouldbe considered.To represent
them, it is convenientto plot a diagramsuchas Figure 6; the central line indicates
the profilein the field with geophonebases;from eachshotpointa slantingline is
drawn on which are designatedthe segmentscorresponding to the basesactually
shot from this shotpoint.
Such a diagram may have varioususes;programmingof the work in the field,
control of reverse returns, interpretation with intersectiondistance curves as on
the selectedexample,statisticalinformation(chargesize,amplifiergain, quality),
and others.
The final resultsof the interpretationof a profileby this method are given as a
delay time curvefor eachmarker investigated.Figure 7 showsan exampleof such
curves;it refersto the sameprofileas Figure 6.
CONSTANT DISTANCE CORRELATION
Generalities
Mi
5 & 3 2 I
M3
500 -
Morker I
60O
,oo
800
i
,,I !& i 13 i 12 i II I I0 I 9 I 8 I 7 I, 6 I 5 I 4 4 3 I 2
.
,I I
Amplitudes
S R Rc
.
over quite a long distancebeyond the critical point. Then the two arrivals separate,
the refracted one arriving first, while both quickly loseenergy.
The tests made in the Sahara have been essentially based on the study of re-
fracted arrivals in the neighborhoodof the critical point. The identification of the
wide angie reflection in this zone is often difficult, becauseeither its lossof energy
is too rapid, or the curvature of the time-distancescurve is too small. A confusion
arisesbetween this reflected arrival and a refracted one coming from a slower me-
dium. It should be noted that the apparent asymptotic velocity of the reflected
Amplitudes
I
I
I
I
I
I
s IR
c R
c N
$ R
Sample records
On Figures11 and 12 are samplerecordstaken in the Saharashowingeventsin
the neighborhoodof the critical zonewhich correspondto the basement.The traces
closeto the critical distancehave been shownin this figure at a distance of around
12 km. The arrival of energyis abrupt; it is sometimesprecededby a damped trace
showingnoticeablyhigher frequencies,the origin of which is not clear (bottom
recordof Figure 11). The large elongationspersiston sometraces,then the ampli-
tudes decreaserapidly.
The seismograms of Figure 13 havebeenrecordedin areasvery distant from those
of Figure 11 and 12, but have a similar aspectcharacterizedby suddenarrival of
high amplitudesdue to the basement.The lossof energyhereis not very noticeable;
moredistantshotpointsshouldbe usedin orderto seeit conspicuously. The critical
33 34 35 36 37 38 33 40 41 42 t.3 44 45 46 47 48 49 50
S.R 197
$.P. 196
&P. 196
197
36 37 36 39 40 41 42 43 44 45 46 4.7 48 49 50 51 52 53
To,, =
25 25 27' 29 43 45
S RctX
Tc
c i
Rn-tT"
XR
n
n
c H
M vM
= Tc.-M Xc=2Tdc
TOc
Fro. 14. Determinationof delay time at the criticalpoint.
186 Amplitude of Refraction Signal
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
Xc
T0c = Tc = 2Tdc. (17)
V
The delay time at C is thus determined at once from Tc and Xc provided that
the precisevelocity of the marker VM is known.
But the accuratedeterminationof Xc is not alwayseasy. On the other hand, the
offset in spite of the assumptionof small dips varies with the marker depth. If
velocity VM is not accurately known, the error made in the term X,/V:u in the
delay time estimation is not constant and the curve drawn on the basis of these
values is distorted.
This difficulty may be resolvedby no longerconsideringthe critical point R, but
a geophoneRK near Rc and locatedat a fixeddistanceXK from the shotpoint(Fig-
ure 15). Applying to this geophonethe same computationsas that applied to the
critical point, i.e. determiningthe delay Tds definedby
2Tas = T
V
The curve determinedfrom the delay time values Tds is thus related to the
curve of true delay times as the midpoint of the chord Ta, TaK.
If the variations in dip are gradual and the distanceR, R remains small, the
erroris negligible.The errormadein the constantterm X:/Va by taking an approxi-
ITc
RCXc RK
tX
TK
K'
I TdK
/ Td$,K
=TK- XK-_Td
2TdsK VM c + TdK
FzG. 15. Principleof constantdistancecorrelation.
Delay Time and Correlation Interpretation 187
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
D= 14.6 14.6 14.25 14.$ 14,6 14,25 14.95 14,6 14,6 14,6 14,6
mate value Va instead of V is also constant and the curve obtained from the
valuesTaskhasexactlythe shapeof the delaytime curve.It is the useof sucha
fixed distance XK which has led to the name of "Constant Distance Correlation"
for this method.
Figure16 representsa sectionconstructedby takingfor eachshotpointthe trace
at the nearestdistanceof RK; times are corrected(subtracted)by Xz/Va taking
into account small but unavoidable variations of X distance due either to topog-
raphyor to the existenceof traceswhichare disturbedby noiseor not usabledue
to equipment.In the presentcasethe distancechosenX is 14.6km. plusor minus
one trace (with the exceptionof two).
The correlationbetweenshotpointsgivesno difficulty. There is no doubt that a
sectionsuchasthe presentonedoesnot giveall informationrequiredfor a thorough
studyof the marker.The interpolation
betweenadjustmentpointsandin particular
the locationof faults resultsfrom a careful investigationof the recordsincluding a
number of additional shots fired further off where the marker is identifiable as first
arrivals.
It is to be noted alsothat the time scaleof interceptsrepresentsdoubletime for
the delay times and is exaggerated.
The adjustmentof the delay time curveinvolvesthe determinationof the true
velocityV of the markerbed.It canbe doneby usingdifferentvaluesfor Xr. As
a matter of fact, the two curves(C1) and (C2) obtainedfrom two valuesof
188 Amplitude of Refraction Signal
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
)
FG. 17. Determinationof the markervelocity.
ATa = (19)
2 Va
delotimecurve
v--6000% .+ v=625o + v=6ooo-/.
800
,a..--.,-.,-!--.--,------.,,-,J
--- 1 ' ' --900
, I .' I; -*-l-r:__::_--=-----:::-----::--:
deloy
times
corrected
forconvergence
_1,oo
o
delay
times forVa = 6000m/s.
calculated end0'XK= 4600m {A
12200
m
!I 18800m
_J____l+ AY J
vM- va - AX
Convergence
curve
AX
Delay
time
curve
Assumed
velocity
forcalculations
:1
=5100we/&
k ,.
I I I
2 I I
:5 i I
4 I I
5 I I
CrossingM2 ' M 3
X2
X3
CrossingM2-M3
Fro. 20. Exampleof shotdiagram.M,, Ms, and M4 are surveyedby correlationat constantdistance
X, Xs, andX4; M3 is studiedwith both Gardnerand constantdistance(X3)correlationmethods.The
dotted line indicatesthe vanishingof marker M3 by lossof energy.Length of eachspread'4.2 km.
5oo....c
. o _- * .._ _ _ M!
" o .... M2
_
800
15 16 i3 12 II I0 9 8 7 6 5 / 3 2 I
I I I I I I I I I I I I I I I I
Crossin
9 M2' M3
X3
X2
3 2 !
Xi
X2
X3
CrossingM2-M3
FIG. 22. Example of shot diagram.M, and M2 are surveyedby correlationat constantdistancesX
and X2; the deepmarkerMa is studiedwith both Gardnerand constantdistance(Xa) correlationmeth-
ods.Length of each spread'4.2 km.
Theuseofmagnetic
recording
isveryadvisablel
if notto saynecessary.
In addi-
tion to playingback the individualtraceswith appropriategain, the useof a mag-
netic unit makes possiblethe systematicrealization, at the Central Playback
Office,of sectionssimilar to that shownin Figure 16.
CONCLUSIONS
600---c
700
n
800 M3
12 I! I0 9 8 7 6 5 / 3 2 1
I- I I I I I I I I I I I I
REFERENCES
DETERMINATION
BY REFRACTION
Section
4
OF WEATHERING
OR
METHODS
LVL THICKNESS
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
Section 4
196
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
W. A. KNOX*
ABSTRACT
INTRODUCTION
Geologicalconditions
Understandingof the surficialgeologyis a material aid in interpretingseismic
refractiondata. A brief review of typical complexnear-surfaceconditionsand their
associatedvelocity distributions follows.
In the simplestlayering, the weatheredzone constitutesa singlelow-velocity
layer near the surface.More complicatedare the glacialdrift depositsof northern
and easternUnited Statesand Canada, where the velocity of the unweathereddrift
may be intermediatebetweenthe very low velocityof the weatheredlayer and the
relatively high-velocitybedrockbeneath.
Fossildrainagechannelsor depressions filled by detrital materialmay presenta
* WesternGeophysical
Companyof America,LosAngeles,
California.
197
198 Determination of Weathering Thickness
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
problem when the enclosingrocks are near the surface.The infilling material may
have a velocity appreciablylower than that of the surroundingrock, yet its velocity
may be much higher than that of the usual surfaceweathered material. Pakiser and
Black (1957) report a situation in southern Utah in which low-velocity Triassic
Shinarump Conglomerate fills erosional channels in the Moenkopi Formation.
Thralls and Mossman(1952) describea similar examplein southernKansasin which
Quaternary fill occupiesthe many irregularities of an older erosional surface on
flat-lying sandstoneand limestone.Hawkins (1961) discusses a three-layermantle-
rock problem in the vicinity of Canberra, Australia.
Buried low-velocity zonesare not uncommonand, when encountered,can prove
troublesome
in refractionsurveys.Pakiserand Black (1957) illustratean exampleof
high-velocity mudstone lensesfound imbedded in the lower velocity Shinarump
Conglomerate which, in turn, is contained in high-velocity Moenkopi channels.
Burg (1952) citesa velocity reversalin Pleistocenedepositsof the Williston Basin,
North Dakota. Velocity-reversalproblemsassociatedwith buried fossilweathered
layers have been encounteredin southern Saskatchewan,Manitoba, and northern
Alberta. In the Far North are found permafrost lenses,having a velocity of 13,000
to 15,000ft/sec, lying on unfrozen rocks with a velocity but one half to two thirds
as great.
Differing from a velocity reversal only in dimensionsis the so-calledstringer
problem in which refracted arrivals from an upper, thin, high-velocity bed mask
slower arrivals from underlying strata. One such example is given by Thralls and
Mossman (1952).
Climatic cycles,both long and short term, influencethe velocity of near-surface
layers. In mid-northern areas, surficial material freezesto several feet in winter.
Particularly in swampy areas,frozen material constitutesa stringer with velocity in
excessof 10,000ft/sec overlyinglow-velocityunfrozenearth. The ice stringer may
disappear in summer.
Bedrock near the surface may comprise a sequenceof alternating shales and
poroussand members,the latter being water-saturated. When a deep river channel
is cut into the strata, ground waters are drained from porous members of the
sequence,lowering the interval velocity within those members. A buried low-
velocity zone is thus produced extendingfor some distance away from one or both
banks of the channel. An uncorrectedseismicmap then showsa false synclinal halo
along the river course.Such an effect, attributed in part to post-glacialerosion,has
been observedalong the Peace River in northern Alberta and the AssiniboineRiver
in Manitoba.
COMPUTATION FORMULAS
Introduction
Vi
Z-- Tj
2 cos io
where
Shot depthcorrection
Shot depths may be large comparedto thicknessesof the layers of interest. The
time-intercept equation above must be modified to take into accountthe depth of
shot. In Figure 1, AE is the surface,AB the depthof shot(D), and BCDE the path
of a refracted ray. It is evident that the unsymmetricalpath BCDE can be replaced
by the isoscelestrapezoidrepresentedby the dashedlines.If the intercept time, due
to a shot at B, were introducedin the equation above, the computedthickness,Y,
would be in error by just half the shot depth. Correct thicknessis Z- Y-}-D/2.
Multilayer Near-SurfaceComputation 201
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
Voo
A E oo
' Di/2
-I
Y
/
o Vo
c D
Vl
Equations
A listing
oftypeequations follows,
adaptedtofit variousconditionswherein
the
shotmaylie above, within,or beneaththerefractor of interest.
Eachgroupof
equationsrefers
toa figure
illustrating
assumedtravelpaths andthecorresponding
refraction
graph.Becausetheso-called"weathered
layer"issimply thefirstmember
ofasequence ofrefractors,
nospecialconsideration
willbegiven toit thatwould
not
alsobe givento any otherrefractor.
where
Voo
GEOPHONE
SHOT
Vo
Vl
V$
tr
TO
Fro.2. Raypathsandrefraction
graphfor shotin layerVo.
Zo - T Vo/cosio, (4)
x = (r- ro)/(2 cosi,) + (D- go)/2, (s)
Z = (r- ro- r)/(2 cosi2), (6)
z = (r- ro- r.- r)/(2 cosi): (7)
where
Voo
GEOPHONE
Vo
Vl
V2
I/V4
I/V
T4
I/V2
T2
tr I/Vl
T
To = (Zo/Vo) cosi0.
Deeperlayersarepeeledoffusingequations(6) and (7).
It is apparent from Figure 4 that direct rays from shot to geophonespreadfollow
diagonalpaths in layer Zt, therefore, the first part of the refraction graph is non-
linear with intercept equal to the uphole time tr; the true intercept is the intercept of
the asymptoteto the curve.Equation(8) appliesprovidedthe true interceptis used.
Figure 5 illustratestypical refractiongraphsresultingfrom variationsin shot depth.
2O4 Determination of Weathering Thickness
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
GEOPHONE
Vo
SHOT,
V$
I/V3
I/V2
T3
T2
tr
Fro.4. Diagonalandrefracted
pathsfor shotnearbaseof layerV.
.,oo
:o
.o,o
i ..-
.oo-/
Ti T
r = ,0250
tr
=.0575
T.040
T = .0564
.oo D
t r = .0404 T
.060
tr = .0454 T .040
TI=.0564
.020
T,..o5o7
.o6o ,. .......
x
Fro.5. Change
inappearance
ofrefraction
graphs
forshots
at various
depths.
GEOPHONE
Vi
V2
V$
V4
Thus, tr> T provided the shot is at or below the V0--V boundary; if tr__< T1, the
shot is above the interface. This is a relationship of considerableimportance in
interpretation of refraction data since it enables one to detect a hidden layer.
Sometimesan interpreter will attempt to substitute tr for To, mainly as a short cut
in the computations. The table below showsthe error in secondscommitted using
this approximation. Relative errorsrange from 13.4 percentfor V0- V down to less
than two percent for the highest velocity contrasts.
Table 1
VELOCITY REVERSALS
Introduction
Zo = T1Vo/cosi01. (13)
TIME IN SECONDS
.020 .040 .060 .080 .100 .120 .140 .160
I I I I I I I
5O
I00
150
200
250
_z 300
50
400
450
500
550
600
Fro.7. Time-depth
graphillustrating
velocity
reversal.
Multilayer Near-Surface Computation 209
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
To computethicknessof layer Z,
(1) Given' velocitiesV, V+,
210 Determination of Weathering Thickness
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
GEOPHONE
Vo
Vi
0T
V2
V3
I/V4
tr I/Vl
intercept time: T,
required: Z.
(2) ConnectV with V+ Figure 11. Read/V, + on the appropriatescale.
(3) Connect V with T Figure 12. From/V, +, draw a secondline through the
intersection of the first line with the a axis. The extension of the second line to the
Z scale gives Z.
CONCLUSION
Accurate results are possibleby use of shallow refraction methods provided the
data are properly interpreted and computed.Establishmentof a systematiccompu-
Multilayer Near-Surface Computation 211
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
GEOPHON
Vo
SHOT Vl
JE PATH ,UMED PATH
(Va = Avg.of Vi
T$
tr
/// I/Va
FG. 10. Raypathsand refractiongraphfor shotabovevelocityreversal.
tation routine will permit the calculationsto be made with ease by the average
explorationfield crew.A numericalexample(Figure13) is nowpresentedby way of
illustration.
The refracted traveltime graph is shownin the body of Figure 13, uphole time-
depth graphsfor shotpoints1 and 2 just below, while a profile of the overburden
layersis drawn at the top. The distancebetweenshotholesis about 2,000ft but the
geophonespreadis 1,600ft longwith the nearestgeophones beingplaced200 ft from
each shothole.
At shotpoint1, three velocitiesappear: 3,500; 6,320; and 8,400 ft/sec (true
velocities:6,000, 8,000 ft/sec). At shotpoint2, however,only two velocities,5,720
and 7,700 ft/sec, are evident. Despite the absenceof a low-velocity intercept, there
is a hiddenlayer of 3,500 ft/sec material beneaththe shot as can be seenfrom the
upholetime-depthgraphand the fact that tr is lessthan T2, wheretr is the uphole
time at shot depth of the refractionrecordusedfor the refractiongraph.
Depth to the base of the V0 layer can be obtainedby inspectionof the uphole
time-depthgraphs,yieldingZ0-30 ft and 25 ft for shotpoints1 and 2 respectively.
The depth could also have been computedfrom T at shotpoint 1 although this
couldnot be done at shotpoint 2 becauseT is absent.
Thicknesses for layershavingvelocitiesV and V2canbe computedwith the aid of
equations(5) and (6), respectively.Becausethereis no V interceptfor shotpoint2,
212 Determination of Weathering Thickness
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
EXAMPLE
CHART I CHART 2
4000
-- io,ooo Vn 5OO0
--9000 Nn'n+i
= V"';i
-- 8000
6000
9000
- ooo .50--
- . I0,000
.40 m
--4000 -
.30--
_
14,000
--3000 .20m
Vn+i
,10--
2000
Nn,n+i
-- iooo
Vn
FG. 11. Nomogramfor computingsin i..--
Multilayer Near-Surface Computation 213
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
150--
Vn 600 --
I00 --
400--
.090--
350 --
.080--
.070-- 300 --
.060--
250 --
.050--
200--
.040--
150--
.030
--20,000
18,000
16,000
14,000
.020
-- 12,000
-- io,ooo
_
-- 9000
_
-- 8000
_
--7000
.010 --
40-- 6000
.009 --
_
.008 -- -- .80
--5000
.007-- 30--
--4000
.006--
--.90
.005--
20-- --3000
.004-- -- .95
.003 --
L--2000
I0--
.002 --
Tn N Zn Vn
m. 12. Nomommfo compudng]sythickness,Z.
214 Determination of Weathering Thickness
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
SURFACE
Vo = I000
50-- - 50
V 3500
-- IO0
I00--
-- 150
150--
V2 = 6000 -- 200
200--
--250
250--
V3 = 8000
I I
500 1000 1500
.35o i I
,300
.3oo --
25O
.25o --
.200 --
z
o
_z
150
_
.15o -
T3
=.
106 I00
T3 = .083
Ta=.078
tr: 039'050;/ Ta:
t r:
.043
.032
,,:1o29
SHOT POINT I
o 500
I GEOPHONE
SPREAD I000 1500 2000
SHOT POINT 2
060 -
,(3 OU o, .060
t = ,032 -I.oo
0
I
50 I00
I 00 0
SHOT DEPTH IN FEET SHOT DEPTH IN FEET
3,500-Va=2
Vi=3,500 /(,6,320 1) = 2/(.158
q-5,720 q-.175)
= 6,000
Va=6,320 -Va=2/(,.8,400
5,720 + 7,700'
= 2/(.130
q-.119)
= 8,020
Va = 8,400 7,700 N01 -- .29 N12 = .58
Ds = 60' 50' N0a = .17 -/Via = .44
tr- .039 .032 N0a=.13 N:a = .75
Ti= .029 .025
Ta= .078 .043
Ta = .106 .083
Zo = 30' 25'
T0 = .030 .025
T0a= .030 .025
The symbol qdenotesthe operation of aligning Nn. n+. V,, and T as instructed
in the directionson the nomograms,Figures 11 and 12.
I would like to thank Mr. Carl H. Savit for his patient help, suggestions,and
encouragementduring preparation of this paper.
216 Determination of Weathering Thickness
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
REFERENCES
JOHN C. HOLLISTER*
ABSTRACT
INTRODUCTION
z = (Zm + A)sinna- A.
Differentiation gives
dz -- n(Zm q- A) sinn- a cos (s)
I/Vm
Tm
I/V o
Xm
Zm V= C(z+A)I/n
dz Vdf
dx
Fro. 1. A single-layer
systemwhosevelocityis definedby Banta'sfunction,V=C(zq-A) /',in which
C, A, and n are constantsand is depth.
Xm = 2n(Zm + A) sinna da
o
The integral, I,, may be evaluated by Simpson'srule for various n's and various
A/Zm's or variousVo/Vm's.A family of curvesfor Is versusn can be plotted, each
member being identified by a particular A/Zm ratio. These plots will be used in
computation as we will see later on.
To find Tmfor a particular ray, we refer to the basiccurvedray expressionindi-
cated in Figure 1,
foz dz V cos a
t =
n(Z
+.4)
Vm o
sin n-2 a da.
which, with a plot of It/1,, versusVo/Vmfor various n's (Plate I), will determine n.
A methodfor determiningvaluesfor Voand Vmwill be discussed in AppendixII.
Having determined n, we can evaluate the other constants,A and C. From (7)
we find
(Vo/Vm)n
A/Z m -- , (10)
1 -- (Vo/Vm)n
and from (6)
Xm
=
(1 q- A//Zm)I,
Knowing Z, and A/Zm, we have A. The value of Is is obtained from the plot I
versus n for various A/Zm's (Plate II), and the value of X'mis that used to find
I,/L,.
The physical significanceof A is indicated in Figure 1. It is the distanceabove the
groundsurfaceat which V would be zero were the regionentirely filled with material
governedby V=C(z+A) TM.
The value of C comesfrom (3) or
C = VolAnt". (12)
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
220
!.8
2.3
1.5
2.2'
1.100
1.7 }.....
0.2
O.3
Determination
PLATE I
of Weathering
6
Thickness
IT:Y FUNCTION
HART
$.9
125
150
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
2
0
!
2
3
Curved
4
Path
PLATE II
Refraction
6
Method
7
V.-ALUE.':
8
9
221
222 Determination of Weathering Thickness
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
The constantsC, A, and n are independentof the path whereasIx, It, and A/Z,,
are parametersof the particular path.
SINGLE-LAYER SYSTEM
Now that the constantsof the material are known, reductionto datum beginsby
the selectionof the proper value of the depth, Zm, from the surface to the datum.
Not only must we choosethis depth, Zm, but we must also compute the corre-
spondingvertical time, r, to be subtracted from the reflection times. This vertical
time is
rm
=fozm
dz v
1 - 1In c
--
The correspondingX and Tm for the imaginary ray (V absent) can be found
from (6) and (8) and the plots Ix versusn (Plate II) and It versusn (Plate III) since
C, A, A/Zoo,and V and n are known.
The time differencebetween the imaginary ray penetrating to Z and emerging
at X and the actual refracted ray emergingat X can be written as
Z' + A' X,,,'
T,--T- I'
V, V=
Tm I/V
/v2
Xm X
A .r--,,,.
cco
' !1 I=C(z+A)
i/n
,, .
Z'm // V2 = Consfont
>C(ZI+A)
I/n
. :X
0m
FIG. 2. A two-layersystemwhoseupperlayervelocityis definedby V=C(zq-A)/'and whosesec-
ondlayer velocityisconstantandgreaterthan V=C (Zq-A) lln . C, A, and nare constants,zisdepth,
and Z is the thicknessof the upperlayer.
or
Equation (15) leadsviaa plot of It'-Ix' versusA'/Z' for variousn's (Plate IV) to a
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
224
.
.
2
3
.........
i........
i........
i.......
Determination
4
........
i.......
i......
PLATE III
of Weathering
6
7
Thickness
8
9
10
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
3.0
2.0
1.o
Curved
4
Path
FOR
PLATE
5
Refraction
6
FOR.!.VAL.
Method
7
LES..CF..
REFR,ACT.iON:
.AIZ
8
I'L'ATION
225
226 Determination of Weathering Thickness
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
T1 -- (17)
i - 1In C
I/
II Xrn/2
I
;Zrn= ds V=C(+A)I/n
Fro. 3. The single-layersystemof Figure 1 shot from a deep hole. The inverseslopeat the inflec-
tion point of the time-distancecurveequalsthe maximumvelocity, V,, of a ray whosegreatestpene-
tration, Zm, is equal to the hole depth.
Curved Path Refraction Method 227
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
Throughoutthe foregoingdiscussion,surface-to-surface
timesand distanceshave
beenpresumedand no allowancefor the depthof shothasbeenmade.If the shot-
holesare shallow, then uphole times may be added to refraction times without
appreciableerror and Vomay be computedby dividingthe shotdepth,ds,by the
uphole time, tuh.
For deepshotholes, a moreexactproceduremaybefollowed.Reference ismadeto
Figure3, whichrevealsthe rather interestingrelationthat the inverseslopeat the
point of inflectionof the time-distance
curveequalsa Vmcorresponding to a Zm
equalto thedepthof theshot,ds.The coordinates of thisinflectionpoint,X,/2 and
-t I
IXl ]Xm-XI
L. Xrn
4x,lx,F--
V=C(z+A)
I/n
i' /
/ V2=
cnstant
/ >C(+A]
/n
Fro. 4. The two-layersystemof Figure2 shotfroma deephole.To restorethe refractiontime, T,, to
its surface-to-surface
value,a time,h, isadded.The surface-to-surface
distanceismeasured froma polar
x to the left of the shothole.
228 Determination of Weathering Thickness
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
ACKNOWLEDGMENTS
The original study which resulted in this paper was made in 1955-56 during a
review of low-velocity-layer correctionproceduresin a student seminar on seismic
exploration at the Colorado Schoolof Mines. To members of that seminar, Messrs.
C. W. Kerns, R. C. Adamson, and M. N. Qureshy, go my sincerethanks for their
stimulating discussions and for the many hoursof hand calculationsspentin prepar-
ing the forerunnersof Plates I through IV. Also my thanks go to ProfessorsL. J.
Prince and C. R. Baer of the Mathematics Department at Mines for writing the
program used in computing the data for the plates, and to Mr. F. S. Lobato for his
drafting of theseplates.
I am particularly grateful to the anonymousreviewer of the original manuscript
for his thoughtful and significantsuggestionswhich have been incorporatedin this
version.
REFERENCE
APPENDIX I
Single-layer system
1. Plot the Tin,X data on ordinary rectilinear paper, and using French curvesor
a spline,draw the time-distancecurve.
2. Using a slopemeter(seeAppendixII), find Vo (the inverseslopeof the travel-
time curve at the origin).
3. Measure V with a slopemeterat some convenientpoint (X, T) on the
curve away from the originand using(9) and the It/Ix versusVo/V plot (Plate I)
find n.
4. Now using the n determinedin step 3 and expression(10), computeA/Z,.
5. From (11) and the Ix versusn for variousA/Z,'s (Plate II), find Z which with
the value of A/Z,, givesA.
6. Find C from (12) which completesthe determinationof the constants.
7. With the requireddepth to datum (either fixed or floating) as a new Z, find
the corresponding verticaltime,r, from (13).
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
I..iJ
Curved
Path
Refraction
Method
229
230 Determination of Weathering Thickness
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
Two-layer system
1. Find the values of Vo,n, A, and C as outlined in steps1 through 6 above. Then
determine V2.
2. Determine Zm from (14) and the correspondingXm and T using (6) and (8)
plus the plotsIx versusn (Plate II), and It versusn, (Plate III).
3. From the traveltime curve, measureTr correspondingto Xr.
4. Find It'--Ix' from (15).
5. From the It'--Ix' versusn for variousA/Z,'s (Plate IV) determineA/Z, .
6. Finally, calculateZ and r from (16) and (17).
APPENDIX II
WEATHERING SOLUTION
ABSTRACT
This paper describesa method for determining the vertical traveltime to a pre-
determineddepth below the surfacein an area wherevelocity increaseswith depth.
The method is normally used for correctingreflectionrecords for weathering and
near-surfacevelocity variations. The mathematical solution is credited to the late
Mr. E. E. Blondeau with GeophysicalResearchCorporation and, therefore, bears
his name. Publicationscoveringparts of this subjectincludeBanta (1941), Handley
(1954), and Duska (1963). The mathematical theory for this method is included in
the Appendix of this paper.
The first caseof the Blondeau solutionis demonstratedby referring to Figure 1.
The assumptionis made that velocity is proportionalto depth to the 1In power (V
=CZ /'). By mathematical manipulation, it can be shown that the slope of the
time-distanceplot on log-logpaperis equalto !- 1In. This slopeis referredto as B.
Also, mathematically it can be shown that there is a function F such that the rela-
tionship betweenB and F is as shownby the tabulated B-F Table, Figure 2, or the
plotted curve, Figure 2A. Since l_<n< o the quantity B, the linear slopeof a line
canvary from 0 to 1. As a practicalmatter we usuallyuseslopesbetween0.2 and 0.98.
The function F representsthe ratio between the horizontal distance (X) and the
vertical depth (Z). It is also equal to the ratio between the refraction time (T) and
the vertical time (tv), that is, the raypath time (T) at distance(X) and the vertical
time (tv)from the surfaceto a depth Z. The depth to which it is desiredto determine
a vertical time may be selected;thus a peel-offof a certain thicknessof material, for
example to a depth of 1,000 ft, may be utilized for the problem area. At the top of
Figure ! the quantities X for distance,T for raypath time, Z for depth, and t for
vertical time are illustrated. Below this is showna plot on log-logpaper of refraction
time versusdistance.The linear slopedy/dx of the line is equal to B. The value of F
may be determinedfrom the B-F Table, Figure 2. Now, if a Z has been chosen,then
the productFZ will yield X. From the plot, shownin Figure 1, the value of T can be
read at a distanceX. T divided by F will give tvwhich is the vertical time. However,
it is seldomthat a geologicconditionis encounteredwherethe velocity gradient is so
* Mobil Oil Corporation,GeophysicalServicesCenter, Dallas, Texas.
231
232 Determination of Weathering Thickness
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
I
I
I
T
IogT
log:X:
dy
=B --.,. F from B F table
Z F = X -- T from plot
T
- = tv
Fro.1. Blondeau
raypathmodelanda time-distance plotona logarithmicscaledepicting
relations
usedto determine
verticaltime (tv)to desired
depth(Z).
uniformthat time-distance
pointswillplotasa singlestraightlineonlog-log
paper.
Therefore,provisionhasbeenmadefor severalotherpossible solutions.
The secondcase,shownin Figure 3, is referredto as the normal knee case.The
slopeof the second
portionof thelinebreaksup asshownin Figure3 andis dueto a
decrease in velocitygradientbelowthe depthZ. Now two slopes mustbe deter-
mined.SlopeB will yield a valueof F and slopeBswill yielda valueFsfromthe
B-F Table. The depthto the interfaceZ canbe determinedby the distanceto the
normalkneeX dividedby thevalueF of thefirstlinesegment. The depthZ to the
interfacemay be subtractedfrom the desireddepthZ to determinea/xZ. This
may be multiplied by F to give a/xX. zXXis now addedto X to determinea final X
to yielda corresponding
finaltime T from the log-logplot. Returningnowto the
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
Blondeau Weathering
233
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
234
I
m.I'ON
T
'ON 3ABn3
I
,
3An3
,
I
,
Determination of Weathering Thickness
I
zz
oo
Blondeau Weathering 235
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
log 'Z'
time at the knee, T may be divided by F to give a vertical time, tv.Then the total
time T minus the time T yields a AT which may be divided by F2 to give a Xtv.The
total vertical time to the depth Z may be determined by adding these two portions
of vertical time tq- At, = t.
The third case,shownin Figure 4, is somewhatmore complicatedand may be re-
ferred to as the reversedknee case.Here the slope of the secondline breaks down
relative to the slopeof the first line. Again, from the two slopes,it is possibleto de-
termine the values of F and F2 for the two line segments.Depth to the interface will
be Z again. This time the deeper material has a higher velocity than the shallower
zone; therefore, it is impossibleto observea first arrival which stays wholly in the
shallowermaterial and reachesthe depth Z. In order to determinethe depth Z, the
distance to the knee Xk is divided by the value of F., the F function for the second
236 Determination of Weathering Thickness
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
I
TI
tv I ,z
I
Atv
I
I' j
I i
T I .
A,T iI ',.-'
! ',,,/,
log T I I
A
i
' II AX
I
Xz X
log X
___
dy,.Bi ' FI
dx dy,
dx:,= BE , F2
Xk
F, = Z, Z - Z = AZ (AZ)F,= AX
(Z) F = X, '--T XK+ AX = X '--" T
T, T-T AT
-- =
F, =tv, F F =Atv tv, + Atv = tv
Fro.4. Blondeau
raypathmodelanda time-distance
plotona logarithmic
scaledepicting
relations
used
to determineverticaltime (tv)to desireddepth(Z) for the reversedkneecase.
AT and divide this by F to get a Ate.Again, in order to get our total vertical time t
the times t and At must be added together.
Figure 5 is enclosedto give a generalsolutionof n number of breaksin the line. It
shouldbe explainedhere, however, that drawing many detail breaks in the lines for
relatively shallowdepthsusually causesa lot of work for nothing, becausea general
line can often be drawn that will yield the same vertical time within a few milli-
seconds.In order to show the value of this method, two examplesare includedfrom
the Delaware Basin of West Texas.
Figure 6 showsa sampleplot and calculationsheetfor point A where the refrac-
tion timesversusdistanceare plotted in the upper portion of the diagram. The calcu-
lations are made on the right-hand portion of the diagram, and the plot of the final
logT TK
I i
X XKX2 X$ X
log X
Fro. 5. GeneralBlondeausolutionfor multiple knees.
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
238
z
o
o
<
,..
o
o
Determination
(SON003S)
o_
31NIl
of Weathering
Thickness
--o
o
'
'--c
Blondeau Weathering 239
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
information is made in the lower portion of the diagram for comparisonto the up-
hole survey. The refractiondata have beenapproximatedby three straight lines, and
the B values have been determined. They are determined by use of the horizontal
distance acrosstwo cyclesof the logarithmic paper which measure 1.0 on the linear
scale. Then using this linear scale on the vertical axis, the intercepts are noted,
and the difference between the intercepts gives a value of B for the linear slope.
Thus, the left intercept of the line through the first portion of the data is equal to
.438, and the right intercept of this line with the secondcycle of the paper which
occursat a horizontal distanceof 10,000 ft is projected back onto the same scaleand
read as .816. A differencein thesegives a value of .378, and from the B-F Table a
value of F is then read to be 2.74. The values for F2 and F3 are determined in the
same manner.
The values Xs, T8 for the shortestrefraction time are used to calculate the Z, and
tv, which are the coordinatesof the first point on the vertical survey from the refrac-
tion calculation. The values at the first knee Xk and Tk are equal to the value X
and T sinceit is a normal knee. These values are used to calculate Z, tv,the second
point on the calculatedvertical time survey. A parallel line to the first line on the
refraction curve should connect these two points, thus allowing a check that all
calculations are made correctly.
Since the values X. and T. are the values of a reverse knee, it is necessaryto
establishthe valuesX. and T..This is doneby first establishingthe/xZ, the vertical
distancerepresentedbetweenthe two knees.From this Z2, the depth of the second
knee is determined. Also/xX is calculated and used to determine X. By knowing
X., T.may be read then from the extensionof line 2. From this/xt the vertical time
in the secondlayer and t2the total time to the baseof this layer can be calculated.
Next the/xZ.,the thicknessof the third layer, and ZF, the depth of the final refraction
point, may be calculated.The /xtv2,the time in the third layer, and tF, the total
vertical time of the final point on the calculatedvertical survey, are alsodetermined.
As on the first curve, this givestwo valueson eachportion of the curve,and the slope
of eachline is parallel to that of the correspondingrefractionline, thus checkingall
calculations and establishinga time-depth curve representative of all measured
refraction information. In this casethe depth rangesfrom 186 ft to a total of 1,338
ft. It is noted that the vertical times, plotted with circles,from an uphole survey
along this profile check reasonablywell with the exception that the times are
slightly shorterthan thosemeasured.Although the hole was shot at a depth of ap-
proximately200 ft rather than at the surfaceas prescribedby the BlondeauTheory,
the surface-to-surfacetraveltimes consistingof the uphole time plus the refraction
time(s) are plotted versusdistancesfor the refractionsurvey.The empiricalmethod
usedfor adjustingthis deviationfrom theory is to shift the calculatedtime-depth
curveby the amountof time requiredto move the curveto the upholetime at the
200ft depth.Whenthisadjustmentis made,it canbe seenthat the calculatedvalues
agreefavorablywith thosedeterminedfrom the upholesurvey.This methodof cor-
rection allows a simple and effectivecorrectionfor hole depth and shalloweffects
that are not accountedfor by the Blondeaumethod.
Figure7 showsa similarsetof informationto that shownin Figure6 and a corre-
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
240
Determination
of Weathering
Thickness
.,-
Blondeau Weathering 241
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
SUMMARY
REFERENCES
APPENDIX
or that in general
sin ik,m __
1
Vk Vm
rearranging
Vk
sin i,m -- ' (1)
Vm
io
ro Vo
rI Vl
r V2
V3
rm
Vm
X/2 J
I /
/
I /
I /
/ dz
/
dz /
/
dx
dx
tan i =
dz
Fro. A-2.
Equation (2) statesthat the depth to any point on a raypath is equal to the maxi-
mum depth of penetrationfor this particular raypath multiplied by the sinni,where
i is the anglebetweenthe raypath and the vertical at the point Z on the raypath.
Differentiating equation (2) with respectto i yields
dZ = /Zm sin-1 i cos i di. (3)
fO
x/2 fOa'/2
dx = l,Zm sin n i di
fO
r/2
X -- 2nZm sinn i di,
244 Determination of Weathering Thickness
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
and defining
F = 2n
fo'12sin ' i di
yields
X -- ZmF.
The depth-distancerelation is
Zm = X/. (S)
The expressionfor surface-to-surfacetraveltime can be derived as follows, from
Figure A-2'
cosi = dZ/Vdt,
and solvingfor dt,
dt = dZ/ V cosi.
Substitutingequation (3) for dZ and rearranging,the followingexpressionis ob-
tained'
foa'l'fo/2Zm
dt=
sin'*-1
idi V
T -
2nZm
fo Vm
sin"- i di,
and defining
G = 2n
f0'/2sin *- i di
yields
gm
Go
Substituting for
Vm: CZmTM
Xm G
y S :--- Zml-l/n,
CZmI , c
T = -- . (7)
C
Solving for X,
1/B
(8)
The X componentof instantaneousvelocity acrossthe surfacecan be obtained by
differentiatingequation(8) with respectto time,
= T 1/B-1. (9)
dT B
G TF
dT
: __.
B ( T (1/)--1
'
dX FX
-- T(1/B)--I
dT FBT 1/s
dX X
o
FiG. A-3.
and
d! sin i 1
(11)
d V Vm
Substituting(11) into equation(10) yields
dx X
-- V, (12)
dt BT
t=f dZ V
lv----
fozm
1 ----
CZI/n
dZ = 1/C
foZm Z-11ndZ,
and integrating
lv = Zm1-1In.
C(1- i/n)
Rememberingthat B- !-- i/n,
1 1 Zm
tv -- ZmB -' (14)
CB B Vm
Substituting
Zm=X/F from equation(5) and Vm=X/BT from equation(12) into
equation (14) yields
T
showing
therelationship
betweentheverticaltimeandtotal traveltime.
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
H. L. MEND ENHALL*
ABSTRACT
Conditions
existin thenear-surface
rocksin manyof theareasexplored
by thereflection
seismograph
that profoundly
influencethe traveltimes
of the deeperreflections.
Theseconditionsexistat depths
considerablybelowthe depthsof present-day economic shotholedrilling,and quite often a measureof
theseeffectsis a prerequisite
to the correctreflectioncorrelations.
The salt problemof the Western
Anadarko Basinisanexample ofthisproblemandseveral otherareasarepostulated asexamples. Refine-
mentsin the reversedin-linerefractionprofilingtechniqueare described whichcan be employedto
solvetheseproblems. The refinementsareapplicable to all reversed
in-linerefraction
profiling.
CORE HOLE,
COREFK)L
C TOP BLAINE
TOP BLAINE
lo 15 2o25
15 2o5'
VELOCITY
+ O00
BASE STONE
BASE CORRAL MARKER
STONE CORRAL
MARKER
1.1.1
.288
Fro. 2. Time dip of the nearly flat StoneCorral marker in the coreholesin Figure 1.
QUA'ERNARY
S S S S S S S S S S S S S S S S
S S S S S S S S S
S S S
Fro. 3. DelawareBasinproblem.
250 Determination of Weathering Thickness
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
Sl
TIM
E /,T
Z:KTo=IT-x.
TO / | vn/
DISTANCE
S G
Fro. 5. Center-pointcomputation.
Refraction Refinement Technique 251
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
'rIME
DISTANCE
Vt
V2
V3
Fzo. 7. Refraction interference zone.
252 Determination of Weathering Thickness
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
I I I / I /
\\ \ III \\ .JLII // x. I , // II / /
, IA' _.. /
()---(To/2).
A4-(To/2)B
- T-n---RIGHT
SIDE
ISCENTER POINT DETERMINATION
(SEE FIG.- )
()--- (To/2)A-- (To/2)B= DIP FROM
Afi DETERMINATION
FROM A TO B, ETC. FOR BC AND
CD
Fro.9. Combination
of/xtiandcenter-point
computations.
Refraction Refinement Technique 253
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
$1 $2 R R2 R3
'"' ................................r......,.....
t t I I
I I I
\ I / /
// _*///s.P.
(It will be shownlater that it is practicalto assumeV and any errorin suchassump-
tion will be eliminated.) In order to obtain the data representedby this diagram,
a spreadof geophoneslaid betweenR and R. recordedthe refracted wavesinduced
by the explosionin S. Also, a spreadof geophoneswas employedfrom R. to Ra to
recordthe refractionsset up at S. The slopesof Lines CD and D'E then represent
time dips as describedby Barthelmes (1946). However, the differencebetween D
and D' is a measureof changesproducedby usingdifferentshotpointlocationsince
R is the commonreceivingpoint for both shots.This differencebetweenD' and D
is the time dip in the refractor at the shotpointend (proper horizontaldisplacement
employed)and is shownby A B. The dip from B to A is then equal to D--D'. This
presupposes that D'E is determinedfrom the samephaseof the samerefraction as
that producingCD. In order to seethe purposeof plotting A B let us assumethat an
error (which could be causedby the computer, wind noise,refraction interference,
etc.) was made in determiningD' and the value D" had been used instead. We
would have an erroneousdip D"E but at the shooter's end we would have the
erroneous"shooter'smove-dip" B'A which contains the same error made at
with the sign reversed. The error D'D" is then exactly compensated.The name
Self-Adjusting Procedure is used since by this procedure any error which might
be made at any point will be compensated.
We have presupposedthat we know the correctcorrelationof eventsto compute
the data shown in Figure 10. (Confirmation of correlationscomesfrom similar
computationsof recordsmade at S and S, etc., and shot at R, R, Ra, etc.) It is
obvious,for example, that if we shot at R into a spreadof geophonesat S to
we shouldmeasurethe dip A B. Any error in the V assumptionwill produceiden-
tical errors in both determinations.Minor discrepanciesin the two measurements
are causedby the computer, wind noise,or refraction interference.Discrepancies
in the order of .040 secwould indicate "leg jumping."
It was stated above that the data shownin Figure 10 were computedusingan
assumedrefractor velocity. Compensationof errors in this assumptionmust come
from averaging of data computed from waves which have traveled in opposite
directionsin the refractor. Barthelmes(1946) quite adequatelycoveredthis point.
Therefore, data recordedat S and S, etc., and shot at R, R, Ra, etc., when com-
puted and properly averagedwith the reverseddirection of shooting,becomesthe
compensatorfor errorsin the assumptionof V.
Figure 11 showshow all data is computed,plotted, and averagedto obtain the
254 Determination of Weathering Thickness
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
Xd (DI3PLACEMENT DI3TANCE)
I- '-I ---- EAST
'r ......... 1....... i ......................... '...............
/ I /
\ /
/ \ /
\\ . //
procedure
andextraequipment
areobviously
necessary
butthese
problems
arenot
insurmountable and can be handled economically.
ACKNOWLEDGMENTS
Theauthorexpresses hisgratitude
to PhillipsPetroleum
Company for allowing
thismaterialto bepresented
andparticularly
to W. H. CourtierandC. D. Hierfor
very helpfulsuggestions
in organizing
thispaper.
REFERENCES
Barthelmes,
A. J., 1946,Application
ofcontinuous
profiling
to refraction
shooting'
Geophysics,
v. 11,
p. 24.
Gardner,
L. W., 1949,Seismograph
determination
ofsalt-dome
boundary
using
welldetection
deep
on
domeflank: Geophysics,
v. 14, p. 29.
Widess,
M. B., 1952,Saltsolution,
a seismic
velocity
problem
in theWestern
Anadarko
Basin,
Kansas,
Oklahomaand Texas'Geophysics,
v. 17,p. 481.
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
INTERPRETATION
TECHNIQUES
Section
5
FOR REFRACTION
WORK
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
Section 5
This sectionis the longestand most important part of this volume, and, therefore,
has been divided into several subdivisions.These are separatedfor the following
reasons:
A. A few of the papersthat did not fall in any of the other classificationsand are
consideredas general interpretive techniquesare included in the first group
under that title.
B. The secondgroup includesgraphs and nomographsused for refraction inter-
pretation. Theseaids for making rapid interpretationsare describedquite well
by three papers.
C. The third sectionis devoted to a delay-time method which has beencoveredby
two papers,one of which includesmore information on its application.
D. The fourth section covers the wavefront techniquesthat utilize an instanta-
neoustrace of the wavefront surfacein preferenceto the calculation of a point
alongthe raypath as is usedin most of the other methods.
E. The fifth sectionis concernedwith errorsin refraction interpretation.
258
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
Section 5A
259
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
M. R. MAcPHAIL*
ABSTRACT
INTRODUCTION
PROCEDURE
Wesuppose
thattheproblem
is two-dimensionai,
thatthetime-distance
curves
have been reversed, and that they consist of straight-line segmentsso that the as-
sumption of plane interfaces is justified. The analysis or interpretation may be
carried out in the following steps'
* Humble Oil & RefiningCompany,Houston,Texas.
260
The Midpoint Method of Interpretation 261
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
2x/(V./V) - - 1. h = V.T -- X.
8. The process
may be extendedin thisfashion,the depthof the top of the nth
bed below the surfaceat a distanceof X/2 being
whereh,h.,h_.havealreadybeendetermined
andwhereh_l is givenby the
equation
18,000
17,000
16,000
15,000
10.0
14,000
9.0
13,000
12,000 8.0
11,000
7.0
10,000
6.0
90OO
25,000
80OO
5.0
vi 20,000
7000
15,000
4.0
6000
10,000
3.0
5O00
2.0
4000
5000
4OOO
1.0
3OOO
0.5
300O
0
The Midpoint Method of Interpretation 263
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
We wish now to estimate the error of the method. This would undoubtedly be a
formidableproblemin the generalcaseof n dipping beds,but for a singleinclined
interface, which is all we shall attempt, it is not too difficult. The method is ob-
viouslyexact for a horizontalinterface.Furthermore,by symmetry the error must
be an evenfunctionof the dip; therefore,whenexpandedin ascendingpowersof the
dip it shouldstart off with a second-degree term. Our problemherewill be to calcu-
late this term.
It will be convenientto take the origin O at the point midway between shot-
point S and point of observationR (Figure 2); thus, if X is the distancefrom S to
R, SO= OR= X/2. The refractoris indicatedin Figure 2 by the line REF, whichwe
have drawn dipping downwardto the right at an angle a. Let the depth of REF
below O (measuredalong the vertical) be h. The perpendiculardistancesfrom O,
R, and S to REF are then
X
h cos a and hcosa+msina
2
respectively.
Supposethe true seismicvelocitiesof the upper and lower bedsare V and U2
respectively.(We wish to reserveV2 for the averageof the apparentvelocitiesof
the secondlayer and thereforedenotethe true velocityby U.) Then by Snell'sLaw
the angleof incidencefor refractionalongREF is qb=sin - V/U. The distancea
wave travels in the upper bed is
(hcosa
x )
2
sina secqbfrom S to REF
and
Xcosa--
(
hcosa
2
sina
) tanqb-
( hcosa-[---
2 sin
a) tan 4
= Xcosa- 2hcosatanqb;
\x\\\
s x/2 0 x/2
Surace
Velocity
Vi
X cos :forREF
Velocity
U2
Fro. 2. The two-layercasein whichthe originis taken at the
point midway betweenshot S and detectorR.
Asin thepreceding
sectionwelet V be the averageof thesevelocities;
that is
V = (Va + Vu)/2. (3)
The procedure thenis to determinea depthhm(m for midpoint)from (1) in which
a is put equalto zeroand U to V; thus,
V2= V[1/sin
(4
+ a)+ 1/sin
(4- a)]
= V sin cos/(sin - sin )
= U cosa/(1 - sin a cosec
)
= s('-./2+...)(+.cosec +...)
= u( + . + -.. ) (S)
where =cosec -=(2U-V)/2V . To the same degree of approximation,
that is to termsin the squareof the dip, (5) gives
V = r( - . + ... ). (6)
Hence,
(2r - r)/2r. (7)
In (1) let us write for convenience
F(V) = 2V -- V/VV, (8)
and let us substitutein (8) the value of U2 from (6). By Taylor's theorem
(U) ()
( -- ,
,()/
+ ..-) + ....
(9)
To the samedegreeof accuracy
cosa 1 - a/2. (10)
Then substituting(6), (9), and (10) in (1) we get
r x( - ./2)/r( - .) + h( - ./2) [(r) - r.(r)/r]
= x[+ (- ).]/r+ h[- (r'/ + ).]. ()
where for short we have written F and F' for F(V) and dF(V)/dV2 respectively.
In the samenotation equation (4) reads
r = x/ + h. (2)
Subtracting(12) from (11) we have
F(V) V = 2V - V/V
266 General Interpretive Techniques
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
and
= v2/v2(v22- v2).
Hence,
E= IX x/V'2-
V
V'- h3V?-
V.'- 2V'1o/2'
V '_1
(14)
This is our expressionfor the error in depth. As regards units, the dip a is to be
measuredin radiansbecauseof the approximationin (5) and (10); V and V. may
be measuredin any appropriate units, such as ft/sec, provided of coursethey are
both in the sameunits; E will then be in the sameunits as X and h.
A numerical example may be instructive. Supposea is one degreeor about 1/60
radian, V is 6,000 ft/sec, V. is 10,000 ft/sec, X is 6,000 ft and h is 1,000 ft; (14)
givesE-0.6 ft. For a dip of two degrees,E would be four times as large, or 2.4 ft,
and so on. Thus, for dips of only a few degrees,the errorsdue to the mathematical
approximations of the midpoint method are undoubtedly small, and possibly
negligible,comparedwith the errorsto be expectedfrom other sources.
SUMMARY
The time-distance relations for n refractors of zero dip have been adapted to
provide an approximatesolutionfor the caseof n plane-but-dippingrefractors.The
error of the method has been estimated by comparingthe exact and approximate
solutionsfor the two-layer caseand was found to vary as the squareof the angle of
dip. The resultsare containedin equation(14). A numericalexampleis given.
REFERENCE
W. C. WOOLLEY*, A. W. MUSGRAVE*,
^D HELEN GRAY*
ABSTRACT
I, INTRODUCTION
267
268 General Interpretive Techniques
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
SM R/ XR-
RB SN Va = Average Velocity to Refractor
VR = Velocity of Refractor
P
Va Refract._
V
r Q
RT - Reciprocal
-I Va
i.' CriticalAngle
or sin V
()sinL= Vo
VR
(2)AB-x(AO+OB)
sinL
RB rface
COS G:
AO + OB
(3) co= R-
2 cos L
(5)XR= VaTRsinL
cos a:
or (6) XR= Va
TR
sin
cos
L
(7)R = 2Vocos
TRL
(8)AE= X.._R
2
(9)CE= XR
2
tan L
XR
(0)Slope
=R=va sinL
RetfOCtf
Fro. 1. Definitionof termsand principalrelationships
statedby Hales.
In-line Profiling 269
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
R.T.
SM RADistance RB SN
dist--
SN
1 'd%
clt.
VR
.t
---'-dist [: Subsurface
Coverage
X '
SM RA E RB
ME MA+ AE-X+ XR
2
2.0
1.0
5000
0
7 9 II 13 15
Shot or Receiver Location
FIG. 3. Dashedlinesshowtime-distance
plot for in-linerefractionprofile;odd numberis shotpoint
numberand evennumberis recordinglocation.All reversedtimesmusttie. Solidlinesshowtime-loop
with times shifted to make a continuous line. Primed numbers indicate location of translated zero time
coordinateof (east or north) shots.Double primed numbersindicate locationof translatedzero time
coordinate,time plotted downward,of (west or south) shots.
posite direction. When this processhas been completed,a time loop is then estab-
lished, as shownin Figure 4, and a slopemay be set off so that TR and XR may be
read at eachgeophoneposition.In order to usean accurateplotting scale,a diagonal
plot is used,as shownin Figure 4 betweenthe dashedlines, (i.e. 1 cm-0.040 secfor
the time scale and 1 cm=400 ft for the distance scale. Note that time must be
multiplied by 10,000when calculatinganglesof slope).The diagonalplot is inclined
approximately equal to the refraction velocity slope.
The adaptation and applicationof Hales' Refraction Method concernsthe follow-
ing'
1. From existingvelocity information, a solutionis made on paper to show the
refractors to be expected,
2. In a part of the prospectivearea having a normal geologicsection,shoota long,
(40,000 ft), reversedline to checkthe solutionfrom (1) above,
3. Lay out field procedure,
4. Play back field tapes,
5. Interpret refractor time data,
6. Calculate and plot depths and positionsof refractors.
The above points will now be developedin more detail:
1. Using available velocity information in the area, which may be a conventional
velocity survey or an acousticvelocity log with check points, these data are pro-
cessedthrough a Mobil developedwavefront program on an electroniccomputer.
This program can handle up to 40 layers. The computer output yields data for the
preparation of time-distanceand depth-distanceplots similar to Figure 5. The effect
of attenuation is neglectedand the figure showsthe refractors to be expected and
the particular timesand distancesfor whichthey will be eitherprimary or secondary
(later) arrivals.
2. Next, it is advisableto check the above solutionby shootinga long reversed
"normal line" in the prospectivearea. This line couldbe 40,000 ft or longerdepend-
ing upon the subsurfacevelocity configuration and depths to the refractors of
interest. For this line and also for subsequentshooting,production is materially
increasedthrough the use of two or more shooters.
3. From the results of (2) above, lay out field procedureto obtain the desired
refractor(s), and, if necessary,the weatheringdata. It must be rememberedthat the
refraction energy doesnot travel a vertical path to a correctionplane but is at some
angle to the vertical. This angleis the arc sine of the (subweatheringvelocity)
/(refractor velocity), seeAppendixII.
4. A field monitor record is obtained at the time of shootingtogether with a re-
cordingon magnetictape. Thesetapescan be correctedand processedon playback
with a tilt of an approximaterefractor velocity. Also severalgain settingscan be
played togetherwith band-passfiltering and phasecorrectionto give definitedata
enhancementand to increasethe reliability of picking the same refraction event.
With longin-line spreads,it is possibleto have a widevariation in frequencycontent
of the refractors.For this reason,it is imperative that the recordingsystemhas no
phasedistortion.However,undermany conditions,the quality of the refractionin-
formationcan be improvedby filtering. For this, phasecorrectedfilters are needed.
SeeFigures 6a and 6b.
In-line Profiling 273
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
2.0
i.0
...........
1
I :5 5
R'AXR
E R'B For very low dip areas
Va h=VaTRcosi.
2
XR
or 2teni.
(6)
E,L
TI Va TR sin i. XR (7)
AE= 2 or 2
FIG.4. Diagonal
plotof timeloop,fromFigure3, shown
between
longdashed
lines.Methodshown
fordetermining VR,Va,andslope
are shown in the lower box.
(tan/).Equations
fordepthanddistance
forrelatively
flatareas
274 General Interpretive Techniques
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
distance
FIG. 5. A combinedtime-distance
plot and raypathchart showing'criticaldistanceto eachrefractor;
which refractor is a first break and the time relation of the refractors at all distances. Refractors are
assumed to be flat and without attenuation.
Shat 2.0
Paint Receiver
- 15ooo__
B,S 3.0 3.S 4.0 4-__
5ooo
300OO
o(S. P. )
4 0000
eu($.p. I )
a:
$5ooo
3o0oo
5000
20000
2.0
FIG. 6a. Record sectionof a reversedrefraction profile showing
relative strengthof refractionsand their velocities.
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
276
General Interpretive
Techniques
43W 39W 35W (:3:3) :31W 27W (25) 2:3W (21) 19W
41W 37W32W 29W 23W 2lAW I?W
SURFACE LOCATION OF RECORDING SPREADS
EAS SHO rS
3,.4
:3.6 :3.4
- ..... .. .
2ND REFRACTOR WEST
SHOTS
EAS' SHOFS
:3.8
........
4.0 - - .-- -
. , , RDREFRACT(,R ESTSHOT
FIG. 7. Time-distance
plot of secondand third refractorsfrom recordsprocessed
as the last recordin
Figure6b. Dashedlinesindicateinterpretedresultsthat wouldbe obtainedif shotpointsin parentheses
were shot.
ready for conversionto Hales' "time loop." Figures8 and 9 are "shift-plot" sheets
usedfor listing data from the time loop. AlsoseeFigures3 and 4. On the first line are
listed the shot and receiver locations. The second line lists the raw times from the
analysis, recorded for each cable. These times include an additive correction that
wasinitially subtractedfrom the tracesin taking the "tilt velocity" from eachtrace.
On the third line, a backoff of 0.020 sec,determinedfrom the records,is subtracted
from line two to give in line four the refraction break times that should occur on a
field recordif we had sharpbreaksand accountingfor filtering and phasedistortion.
Beginning with a plot time of zero on the south (or west) shots, at the extreme
south (or west) shotpoint position, our "continuouscables" are established.Since
reciprocityis required,shift timesfor southshotsbecomeplot timesfor north shots
and vice versa (compareFigures8 and 9).
6. Calculationsfor depthsand positionsof refractorsmay now be handledby the
two-layermethoddescribedearlier,seeequationson Figure 4.
Digital processing'The electroniccomputerprogram calculatesdepth and hori-
zontal position from corrected and shifted refraction times based on the mathe-
matical technique describedby Hales. Adaptation of this technique to machine
computationwas made with the programreplacingthe graphicalportion of Hales'
procedureby solvingsimultaneouslinear equations.Other modificationswill be de-
scribed below.
ISHOT
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
25w Sa 54. 33
X /5000 Zoo
-- Tc -.020
I - .OZO
Shift
Plot
.9/0
X
-0.521
.7
2o
.7ZOZ/
/.43
/.064
/.0
A/I.I25 Z 5000
, ! I , I ,
Tc
-.ozo
Shift
X 1260o 0000
..................... ..__ ,
SHTI 2I
Eli E sp
Tr
2G
,2.2. 2.27/ 2.293 2 1 2.337 Z 5 2.82 ZO .27 .
- I !
Tr 2.21 ,g7 2,2ql Zgl 2350 2.37 2.404 g.4g 2.445 ,62/ ,
Tc 020
Tr-Tc 2.23/ 2.z2
Plot ,805 .8 ,
X 000 0
i,
Shift /.800
Plot .05
Data for the programcan be punchedinto cardsdirectly from the shift-plot sheets
codedby the field crew. The card format and identificationof the data are controlled
by the program. The groupingand sequenceof the data cards is controlled by the
geophysicistor operator,within the variationsallowedby the program.
The programcalculatesin cycles,the sizeof whichis controlledby input data. At
the end of each cycle, someof the data is droppedand an equivalent amount of new
data is added. This means that the total amount of data may exceed the memory
capacity of the computer.
A calculationcyclecontainsthree major phases.The first of theseis the determina-
tion of refractor velocity, VR, and the equation of a least squaresline which is
equivalentto the centerline betweenthe two limbs of the time loop.
The secondphase of the calculationcycle computesdepth and position values
from the forward and reverseprofilesseparately.This is doneby first calculatingthe
slope,Va sini. This slopeis projectedfrom a forwardor reversetime to intersectthe
centerline. The time and distanceat the point of intersectionof this slopewith the
centerline, previouslydetermined,are usedto determineXR and T. Thesequanti-
ties re then usedto calculatedepth and positionusingthe value of V obtainedin
the first phaseof the calculationcycle.The value of Va is determinedby input and
may either be constant or variable.
The third phaseof the cycleis the one which shiftsthe data in memoryto incre-
ment the calculation cycle.
The averagevelocity, Va, in the first layer may be determinedin the following
ways: (a) from the first slopeof the time-distancecurve, (b) estimatedvelocityfrom
lithology as projectedfrom surroundingwells, (c) vertical velocity surveysin the
area, (d) syntheticvelocitiesfrom wells where depthsare known and where a refrac-
tion line has been shot acrossthe well, (e) calculatedvelocity by measuringtime in
onewell with the shotpointat the critical distancebeyondanothernearby well, (f)
anomalousevents on the two limbs of the time loop.
Each of the above methodsfor determining Va will be discussedin more detail:
a) Refraction: Hales stated that the early part of the time-distancecurve, as
shownin Figure 2, could be used to determine Va. This is true only if there are no
thin high velocity memberswithin the upper layer. However, if the velocity is
lower under a relatively thin high velocity layer, as is the casein many areas, the
averagevelocity must be determinedby oneor more of the other methods,
b) Lithology' Reasonablevelocitiesmay be estimated for a sectionif it is known
that it is all clastic,all calcareous,or high in anhydrite content, whether it is geo-
logically young or old, etc.,
c) Vertical Survey' If vertical surveysare used(it is helpful to have acousticlogsin
orderto seea mixedzonecontaininghigh and low velocitylayers). It shouldbe noted
that the averagevelocity alongthe critical wavepathmay be higherthan that along
the vertical. This may be due to a pseudo-anisotropic effect causedby the raypath
seekingthe shortesttime path in a mixedlayer--high and low velocity--zone. Or it
may be due to true anisotropiceffectin which the velocity is higher horizontally
than vertically through the same material. This anisotropiceffect seemsto be a
function of density (Kaarsberg, 1959). We do not seeits effect in the Gulf Coast
clastic sectionbut we do encounter it in the Cretaceouslimes,
280 General Interpretive Techniques
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
Va- / VR2
Va
All of the terms are the sameas have been usedexceptX which is the distancebe-
tween the shotpointand the new hole, and T which is the time from datum plane to
the geophone.It may be seenthat severalvelocity determinationsmay be made by
drilling one coreholeat a proper locationbetweenold holes.
(f) AnomalyCorrelation'Correlationof anomalieson the two limbsof the time loop
determinesthe averagevelocity, as shownin Figure 4. If the anomalieson the re-
fraction lines are definite, they are probably the best velocity information, sinceby
using them velocity changesmay be determined ahead of the drill.
The way in which the digital program varies Va can now be explained.Each time
the program reads data, the values of Va to be used to calculate the first and last
points in the limits of the secondphaseare also read in. The program linearly in-
terpolatesa value for all points betweenthe first and last. The interpolation is done
on the basisof point number. The interpolatedvalue of Va is usedto calculatedepth
and positionfrom both the forward and reversetimes at a given input distance.This
meansthat the program doesnot correctlyvary Va unlessthe variation is causedby
changinglithology near the surfaceof the ground. The way in which Va actually
varies alonga line of data shouldinfluencethe choiceof step size.
The program also contains an optional routine that will apply a correctionto the
input times.The initial value of this correctionmust be input with the time-distance
values. An increment to be added or subtracted from this initial correction must also
be input. This increment can be changedthree times in the first and last calculation
cycle and can be different for eachintermediate cycle.
The accuracy of results from this method is most seriouslylimited by its require-
ment that velocity be constant down to the refracting horizon. Compensationfor
bending of the raypath, which is more normal, can be partially made by applying
I n-line Profiling 281
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
correctionsto either or both the input times or Va. Sucha correctionprocessis itself
proneto errorin eitherdepthor positiondependingon how the correctionis applied
versushow the actual raypath was bent. This is discussedfurther in the sectionon
model studies.
A meansfor making a better compensationfor this limitation was made by writ-
ing a programto calculateand subtractfrom deeprefractor times, the portion of
traveltimes aboveshallowerhorizonswhosedepth wasknown. This program accepts
depth, position,and velocityinformationconcerningthe shallowerhorizondirectly
from the output cardsof the previousprogram. The program will output the cor-
rected times and distancesin addition to either the times adjusted to distancesthat
are multiplesof somechosendistanceor the timesand distancesfurther correctedto
a datum plane below the depth of the shallowhorizon.If the datum plane option is
elected,the datum-correctedtimes and distancescan be adjustedto even multiples
of a chosendistance. The distance adjustment can be elected to be either to the
nearestmultiple or to the next multiple. In the latter case,the first adjustment is
made to the nearestmultiple. After this stripping process,the output cardscan then
be input to the Hales' refractionprogram.
Multilayer refractionproblemscan be accuratelysolvedby thesetwo programsif
the data are not taken in an area where velocity logsshowlarge decreasesin interval
velocity with increasingdepth. In areaswherethis conditiondoesexist,the program
can still be used to make accurate shapemaps.
Data machineplotted' The output of the computerprogram is a scaledvertical
crosssectionfor which the forward and reverse times are calculated and depth and
position points plotted independently.These points are machine averagedand
plotted as a third symboljoinedby linesto give the final profile.Depths can then be
read at convenientlocationsalongthe profile and postedon the basemap.
Model Studies: Computational models, with controlledinput, were used to test
the modification of Hales' method and these have brought out some of the limita-
tions of the method and pointed out someof the pitfalls.
In general,careshouldbe takennot to set Va toolow. It seemsto be wiseto evalu-
ate as accuratelyaspossiblethe averagevelocity of the raypath, rather than usingan
averagevertical velocity. An even highervelocity than the actual velocity may be
desirableas it appearsto be better to have the offset as nearly correct as possible
eventhoughthere will be a resultingerror in depth.
Presentedare three examplesof the model studies. Example 1, Figure 10, is a
simpleanticlineon the 1st Refractor.The depthvariesfrom 4,000ft at 0 distanceto
3,000ft at 18,000ft and back down to 4,500 ft at 35,000ft. The sidesof the anticline
arestraightlines.The Va to the 1stRefractoris 7,400ft/sec and the refractorveloc-
ity is 13,000ft/sec.
The slopesof the anticlineare straight in orderto avoidexcessivelycomplexcalcu-
lations.The anglesof the slopeswerecalculatedand then the slopesof the raypaths
in each direction on both sidesof the anticline were calculated. The raypaths were
then drawn and the distanceswere measuredfor each travel path through the upper
layer on both ends,and alongthe refractor.Thesedistanceswere then converted
into times.
282 General Interpretive Techniques
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
EXAMPLE I
ASYMMETRICAL ANTICLINE- I ST REFRACTOR
, T' o,DATUM
PLANE
o Va=7400'/s o o
o
= " I STREFRACT
OR =
TRUE SCALE
Fro. 10.Modelstudies:
Example1, showing
physicaldimensions
andvelocities
of program
input.
---:X)O' - 3000'-
o LUT:ON
---$50(f TIMEs -3500'.
20000' 15000'
I I
DISTANCE
EXAMPLE 2
FLAT AND FAULTED- I ST REFRACTOR
FAULTWITH 150'THROW,
FROM-:5850'TO-4000'
AT :50000' DISTANCE FOR HIGH SIDE
o
1:3
o
'o
o
0 'o
0
o
c o o m
DATUM
' I I I I
PLANE
Va= 7400
o = TRUE SCALE
I ST REFRA:TOR
COMPUTER DATA:
Vo= 7400s
VR= (12992-1:5015S)
NFON= I0
NN
Since only one velocity existsin the model above the refractor and the refractor is
flat on both sidesof the fault, it is possibleto obtain a very accuratereproductionof
the model with the exceptionof this "draping effect." The fault itself occursat the
extreme edgeof the "drape effect," on the upthrown side. It may not be possibleto
produceresultscloseto the fault on the downthrown side, and the depth closeto the
fault will have to be deducedfrom the generaldip farther away from the fault.
Example 3, Figure 14, is a multirefractor problem with the sameconditionsfor the
1st Refractor asin example2. The same45 degreefault cuts a flat 2nd Refractor with
400 ft of throw. The east and west sidesof the fault are at -7,600 ft and -8,000 ft
depths. The fault cuts the 3rd Refractor with 800 ft of throw and the depths on the
east and west sidesof the fault are -15,200 ft and -16,000 ft. The velocities be-
tween the 1st, 2nd, and 3rd Refractorsare 7,400ft/sec, 11,000ft/sec and 14,250ft/sec,
respectively.The averagevelocity vertically from datum plane to the 3rd Refractor
COMPUTERSOLUTION
. J:- - ,,
REVERSE
TIMES
; ? -
-_....... : ..... --'--- - .
5,1000, I
30,000"
DISTANCE
I
25,000'
284
I
General Interpretive Techniques
--
I
--
HJ. d3a
Hd3Q
I
'
I
T'
I n-line Profiling 285
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
EXAMPLE 3B: PROGRAMMED WITH ONLY THE 1ST REFRACTOR STRIPPED OFF
REFERENCES
AB= AOCOS
sinLC+ OBsinL = (AO+OB)
COS :sinL
- = XR
Tb cos=: = CO
CD =2r
R
or :;)r =
COS
AE
sin 2L= r
AE
r= sin2L
2AE A B R
so 2r= sin2L = 2sinLcosL = COS C
COS
T c . CAE= COB = . L AB cosa: (AO+OB)sinL
R= 2sin LcosL= cosa: a si cos L
since they are inscribed
by the same arc CB = (AO + OB)/2 cos L
Thisangularity
correction
isusedtocorrecttheTc'sthatareassumed tobevertical
travelpaths;
when
actually
thetravelpathisat someangieto thevertical.
Thisangle isthearcsineofthesubweathering
velocity/refractor
velocityor5,700ft/secdividedby 13,000ft/sec,(26degrees).
Receiver Shot
Weathering WSurface
B. Weathering
500' /
Sea
Level
/
/
//
Refractor Velocity= 1:50007Sec.
S FrstRefractor
, 13,000'
Fro.17.Appendix
II' Weathering
correction
forrefraction
shooting.
Angularity
correction
= (Pq-r _ (qq-sh
\5,700/ \13,000!
Er- W Er- W
P .... (0.198)(10-4)(E- W)
(cos26) (5,700) 5,700
Ea Ea
r = - (0.198)(10-4)(Ea)
(cos26) (5,700) 5,700
then
p q- r = (0.198)(10-4)(E q- Ea -- W)
(E -- W) (tan 26)
q= = (0.3752)(10-4)(E- W)
13,000
Ea tan 26
s= = (0.3752)(10-4)(Ea)
13,000
I --x - -
q (I)T' """
* hcosl.
V.
I -1 '2)
T' R
+hcsi'
V"
.I.hcOf.
I. VRsini ' V.
(3) T VR - X + h cot L.
Oeophone
VR
(5) co TVRh - X
V.
sini- V_g
(6)
VR2
-V"2"
TV
a-
X
V; V. h
. vR
2. Va
a
Va unknown (7) ?. h
assume that
' h
X, h, VR, and T
are known
///,
Fro. 18. AppendixIII: Developmentof the equationfor Va obtainedby
recordingin a well and shootingX/2 beyondanotherwell.
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
HAAKON M. EVJEN*
INTRODUCTION
In the early 1930's,Dr. Evjen wrote a treatiseon "The Geometryof Refractionand ReflectionShoot-
ing" whichto thisday remainsa classicin its field.Most of the seismologists
of the ShellCompanies over
the entireworldhavelearnedpart of their tradefrom this treatise.The bookwasreleasedto Dr. Evjen
after he left the serviceof Shell,but circumstances
havepreventedhim from publishingit, whollyor in
parts. It was, therefore,an honor as well as a genuinepleasurefor this writer to condense,with Dr.
Evjen's permission,one of the chaptersof the book.It dealswith monotonicincreasesof velocitywith
depthin a way whichis entirelysatisfactory fromboth the practicaland the theoreticalpointsof view.
NOTATION
Vz
=VI(L
+qZ)ll/q L
The rate of increaseof velocity with depth dVz/dZ is a function of Z and the
constantsVx, L, and q. Differentiationof equation(1) gives
* Retired from ShellResearchand DevelopmentCompany.Now deceased.
, Introductionby F. A. Van Melle, Shell DevelopmentCo., Explorationand ProductionResearch
Division, Houston.
290
Monotonic Increases of Velocity with Depth 29t
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
dZ [1+()Z1(1-q)/q
dVz_() (Vlq-1
=(-)\z/ (2)
and
d2Vz
Vz-q
dZ --(-) / (-) V1q--- I(1
(I-q)(V11_q - q)lVz2q
, Z Vl2q
-1
(3)
Equation (3) showsthat for q< 1, the secondderivative is positive,and the velocity
increasesfaster than linearly with depth; for q-1, it is zero, and the velocity in-
creaseslinearly with depth; and for q> 1, d2V/dZ is negative,and the velocity in-
creasesmore slowly than linearly with depth. In the last case,there is a wide range
of possibilities:For large q, equation(2) and Figure 1 showthat the rate of increase
with depth decreasesrapidly with increasingdepth.
An important property of this family of velocity increaseswith depth concerns
the radius of curvature R of the path. We have
R =
ds dZ dZ dZ dZ(Vz)dZ
- = = = U--- --, and
di (cosidi)d(sini)
() dVsin/
dV d
dZ
Rsini= Vz . (4)
dV
From (1),
= 1+ z =
dZ (L q- qZ)
therefore,
R sin i = L q- qZ. (5)
This is the property that allows constructionof the trajectories, not rigorously
but with any degreeof approximation desired.Although the family of velocity dis-
tributions treated here has been described in the literature--for instance, in H.
Kaufman's(1953) comprehensive tabulationof the characteristics of a largenumber
of velocity distributions--this particular property has, to our knowledge,not been
brought out.
A simple, accurate, but not very elegant, procedure for the construction of an
x F. Kalisvaart, who was so kind as to read the manuscript,pointed out that stating formula (1) at
the beginning,as we did for the sake of concisedevelopment,makes the reader wonderwhat led to this
particular function. Dr. Evjen's book showsthat it was obtainedfrom the propertiesof the radius of cur-
vature of the raypath, as given in formula (4). R= U(dZ/dVz) and R sin i= VzdZ/dVz, the vertical
distancebetweenthe centerof curvatureand the corresponding point of the raypath. For an exponential
increasein velocitywith depth Vz-- V exp (Z/L), we find R sin i-L.
For a linear increasein velocity Vz= Vq-kZ, we have R sin i=(V1/k)q-Z=Lq-Z. This leadsto con-
siderationof the more generalformulationin which R sin i=Lq-qZ= VdZ/dV. From this follow
dZ(L+qZ)=dV/V. Integrationyieldsformula(1).
292 General Interpretive Techniques
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
vz x I0'a,ft/sec
oo
IO
z/;:,,ooo ft/sec
V INEAR
PARABOLICCII
q' :) %.LVI /k IZ,000'
L' '4,400' _Vz(6000+O.SZ)ft/sec
.,6ooo(,+ 2z '/2.
\ 14,400/
zo ;6000+2500T) ft/sec
FIG. 1. Velocity-depthfunctions.
approximate trajectory is to strike, for each of a number of zones,an arc with the
radiusR = U(dZ/dVz) arounda centerlocatedon a line, L+qZ abovethe depthZ,
which is the middle of the zone for which the radius is assumed to be constant. This
shows the merit of the relation R sin i=L+qZ. The constant radius means that
the processapproximates the actual velocity distribution with a number of zones
of linear increaseof velocity with depth, for each of which the centersof the trajec-
tories are easily located. Figure 2 shows a more elegant, geometrically oriented
construction which also yields a good approximation with a moderate number of
zonesof linear speedincrease.The center,Co,and the radius,R0, may be thought to
belongto a very thin surfacezone(not shown).This construction yieldstrajectories
with a continuousfirst derivative. It must be noted that, in this age of electronic
computers,the paths can be obtainedby machinemethodsthat yield, for any point
along a path starting with a given ix, X, and Z as parametric functions of i or of V.
di (6)
sin ix
and
, di. (7)
t= -- sinix
Monotonic Increases of Velocity with Depth 293
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
$ o
ZONE
I
ZOIE ---
ar, (8)
dx= tan
idZ= (U'_ V').
V= ,sini dV = ( V) cosidi = (U ' -- V')/2di.
Vx sin ix sin i
(9)
\V/ sin
i)qdi,
,
sin i
(0)
dt = - = - = di,
Vz (Vz cosi) (U ' - V')TM
etc., yieldsintegral (7).
Generalforms with V as independentvariable follow from (8) and (10), namely,
x = dV (11)
(U'- V)-
and
f (
t= vs \-/ ,(v.- .)-/.V (2)
294 General Interpretive Techniques
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
x=
q
cot ix-- coti
sinj)
sine q
and the time of travel
t= q
(q -- 1) U IX q---coti
L
q sinq
sinq i L
ix q
cotix
Hence, for the total time, T, in terms of the shootingdistance,X,
T = (q-- q1)U IX 2L
q
cotix . (13)
Therefore,if the T--X graph is given, the parameters,q and L, can be determined
from this equation,provided the equationis satisfiedby the graph to a sufficient
degreeof approximation.Over limited regions,this requirementcan always be
fulfilled.Therefore,a generalmethodof analysisof normalT--X graphsmay be
basedon equation (13). This analysismust begin with the first part of the time-
distancecurve, for which valuesof q and L are determined.At the point at which
these values cease to hold, the time-distance curve must be reduced to a lower
horizon,whichcan be donewith the aid of formulas(6) and (7) or (11) and (12),
as can be seenfrom Figure 3.
A B C
II
REFERENCE
Kaufman,H., 1953,Velocityfunctions
in seismic
prospecting:
Geophysics,
v. 18,p. 289-297.
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
Section 5B
E. J. NORTHWOOD*
ABSTRACT
INTRODUCTION
sin i = Vz/V,,,
* StandardOil Companyof Texas,Houston,Texas.
298
Nomogram for Curved Ray 297
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
where
6OOO
7OOO
I I
8000
9O0O
o
I0,000
I I000
12,000 -.
13,000
80
14,000
'0
15,000
30 40 50 60 70
LEGEND
AI vocm)tms wAcxo
2.125- -2.125
2.0- -2.0
-i.O
A B
5000 2500
DISTANCE
6000
I I I \ \
/ I I i ' \ L'
.... t--- -.-.--\ -
7000
I i
I0,000
,ooo
13,000
80*
4,ooo
}0
15,000 ;50 40 50 60 70
LEGEND
F1o.3. Nomogram
showing
themethod
ofdetermining
thetruevelocity
ofa refractor
whose
position
is assumed
andwhose
apparent
velocity
is known.
G whichliesonthetrajectory
of 15,000
ft/sec.Thetruevelocity
of therefractor
is
obtainedbymoving alongtheisovelocity
layeroftheoverburden
fromG toH, i.e.,
horizontally
fromGtoa curveofequalanglethatis10degrees
lessthanthecurveat
G.ThispointH iswhere
a horizontal
refractor
would
havethesame
angle
ofinci-
denceandthe sameoverburden velocityasthat of a refractordipping10 degrees
at
G. It musttherefore
havethesamevelocity.NowthepointH isalmostonthetrajec-
torycorresponding
to anapparent
velocityof 20,000
ft/secorprobably
aboutthe
equivalent
ofa trajectory
of19,500
ft/sec.Thisvelocity
valueequals
thatofa hori-
zontalrefractor
at H andalsoequals
thatof a refractor
dipping10degrees
at G.
3OO Graphs and Nomographs for Interpretation
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
t0 t0 to
I 5l + t t
5000 2500
=oo o7o F
2000
4000
Fro.5. Refractor
calculated
fromstraight-lineapproximations
shownin Figure4. Portion
of nomogramof Figure1 is shown
positioned
at A.
4. Wecannowcompute
thetruetotaltimefromA toB using
thenomogram
of
Figure1 andtherefractorof Figure5.
5. Figure
5 also
shows
theoverlay
ofa portion
ofthenomogram
positioned
at A.
6. Nowif therefractorwereflat thentheenergyarrivingat A wouldtravelalong
thetrajectory
of13,340
ft/secthroughCtoA.However,therefractor
dipsat13de-
greessowemovehorizontally
fromCtoD (i.e.,adistance
of13degrees
measured by
usingthecurves
ofequalangle).
D liesonatrajectory
ofabout11,500
ft/sec.
Point
E is theemergent pointfor therefractorasshown.
7. The time fromA to E is 0.775sec,asgivenby curvesof equaltime.
8. Onecanusethesame
procedure
at B to obtaintheemergent
pointF andthe
time B to F, which is 0.172 sec.
9. Using
thevelocity
oftherefractor
of 13,340
ft/secandthemeasured
distance
E to F, thetimeE to F canbecalculated
(0.882sec).
10. Totaltraveltimeis thusthesumof A to E, E to F, andF to B, whichis 1.829
sec.The observedtime is 2.125sec(Figure 2).
Wetherefore
mustmakeadjustments
tooneormoreofthethreeparameters,
the
depth,thedip,orthevelocity
oftherefractor.
Thegreatest
error
inourapproxima-
tionswasprobablytheoverburden
velocity.
Astheoverburden
isthickest
at A it
wouldcausethe greatest
errorthere.Solet usproceed
asfollows:
11.Measure
theapparent
velocity
at A resulting
froma shotat B.Thisis10,750
ft/sec.
12.If westeepen
thedipif therefractor
tosay16degrees,
wemustmakeadjust-
mentsto its velocityto honortheobserved apparent
velocity
at A. Theapparent
velocityat B isnotassensitivedueto thelargevelocitycontrasts.
13. In orderto determinethevelocityof therefractor
asdrawnin Figure6 that
wouldsatisfy
boththedipandtheapparent velocity
of 10,750ft/secobservedat
A, weagainusethenomogram ofFigure1.Theemergent pointontherefractor
is
G.Thedipoftherefractor
is16degrees.
Therefore,
move 16degreeshorizontally
to
H andreadthevalueofthetrajectory
onwhich H falls.Thevalueis13,000ft/sec.
14.Using13,000
ft/sec,
wecanagaincomputethetotaltraveltimeasbefore.The
302 (raphs and Nomographsfor Interpretation
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
2500 500
70*
/ 60*
J 40o 50'
Fro.6.Second
approximation
ofrefractor
showing
calculations
ofitsvelocity
fromitsposition
andapparent
velocity
at A using
thenomogram
ofFigure1.
50 e 60 e
Fro.7.Thirdapproximation
ofrefractor
showing
calculations
ofitsvelocity
fromits
position
andapparent
velocity
atA using
thenomogram ofFigure
1.
Nomogram for Curved Ray 303
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
2000
4000
40' 50*'
Fro. 8. Usingfirst approximationof refractorit is possibleto obtain a better measureof its velocity
usingits apparentvelocityat A, whichis 10,750ft/sec, and the nomogramof Figure 1.
CONCLUSIONS
REFERENCES
Dix, C. Hewitt, 1952, Seismicprospectingfor oil: New York, Harper & Brothers.
Musgrave,Albert W., 1952,Wave-front chartsand raypath plotters:Quarterly of the ColoradoSchoolof
Mines, v. 47, no. 4.
Slotnick, M. M., 1950, Graphicalmethod for the interpretation of refractionprofile data: Geophysics,
v. 15, p. 163-180.
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
E. J. STULKEN*
ABSTRACT
/
/
/
(V)
Vo,2,S
/
/
/
\\ Vo,2,R
11,2
SCV _ VR
x
..
'.V2
V2
0
-i'..'"(SS'
"'-
+'"''1
+ +'-'-VT)./2
FiG. 2. Determinationof a seconddippingrefractioninterface (2-2) by measurement
and geometricconstruction.
layer.1 At O, the intersectionof the rays through the V2 layer, a circleis drawn with
radius (SS'V2/V-kRR'V/V-kS'O-kOR'-VT)/2. The rest of the construction
proceedsas before.
This construction can be extended straightforwardly to n layers provided all
boundariesare plane and dip directly in the direction of the refraction profile. (See
Figure 3.) From time-distancedata one can constructray segmentsthroughthe first
layer, the positionof the baseof this layer having beenpreviouslydetermined.At the
first interface, Snell's law applies so one has at hand the information necessaryto
constructray segmentsat each end of the profile through the secondvelocity layer,
the baseof this layer having likewisebeen previouslyestablished.Again the Sne]l's
law constructionsapply in establishing raypaths through the third layer. This
procedurecan be continuedthrough all n--1 previouslyestablishedlayers. Further-
more, the rays through the nth velocity layers can be drawn and projected to the
point of intersection,O in Figure 3, as in the two-layer case.With O as center and
the quantity
V.(s/Vt -+-s/F, -+- -+-S._l/V._t + rt/Vt -+-r/V2 -+- r,_t/V,_l
q- S'O/V,, q- OR'IF,, -- r)/2
as radius,an arc is swung.Where this arc cuts the projected nth layer rays at M and
N, tangentsare erectedand extendedto their point of interesection,P. As before,a
line throughP perpendidularto OP constitutesthe desiredline (n -- n)--the base
of the nth velocity layer.
Concentriccircleswith radii V and V, are centeredat the foot of the first ray segment.A per-
pendicularto the boundarywherethis ray segmentterminatesis droppedfrom the intersectionof the
ray with the V.ocircle.The centerof the circleand the point wherethis perpendicularintersectsthe V
circle determine the secondray segment.
Constructions, raphs and Nornographs 307
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
io
o 4 x 12(
KILO
FEE'IF O 24
(,a)(X,H)GRIO
('C)(X,H
) GRIDVI (FT/ms)
OVERLAY
ON(Vi,ViT)
SHEET
28
0 4 18 20
ViS(FT
/msI vl -
(b)T LINES
AND
V2 CURVESON( VI ' ViT)SHEET (d)EXAMPLE OF NOMOGRAM IN USE
310
n_
_
>
.
Graphs and Nomographs for Interpretation
NOllg3la
3AVM
Constructions,Graphsand Nomographs 311
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
.. $
i
v! =
v;. io,5oo'/s
iii)
(i)
a. Protractor
appliedto refractions
through
'a singlelayer
FIO.6. Applications
oftherefraction
rayplotter.
312 Graphs and Nomographs for Interpretation
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
THE G FACTOR
= + 1/a(v,,
After evaluatingthe G factorsand time intercepts,one can computeHI,Sand Hx,.
Theseare the vertically measureddepthsof the refractinginterfacedirectly beneath
S and R, respectively.
It followsthat a similar applicationholdsfor refractionsthroughn layerswith
horizontal interfacesand a dipping base.
Computationsare facilitated if time intercept (To) linesare addedto the G graph
asshownin Figure8. Theseare placedsoas to serveasmultipliers.By followinga G
factor acrossto the appropriateTo line one can identify the depth of the refractor.
A novel way to read the Gg,yfactor is to placeblank sheetsof translucentrecti-
linear graphpaperin positionas shownin Figure9. On onesheetthe bottomscale
representsV. The vertical scalerepresentsV;/2 cosi,the G factor. On the other
sheethorizontallines are given Vi values. The (Vi, Gi,j) sheetis laid atop the Vi
sheetso that its origin lies on the V=0 line and a selectedvalue of the Vy scale
falls on a selectedVi line on the underlyingsheet.The value on the G scalewhich
Constructions, Graphs and Nomographs 313
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
7OOO
gijFACT.OR
FOR
IEFRACTION
COMPUTATIONS
EXAMPLE OF USE
gijISVlVj/2
V
SCA cuu AT
I .7'/s 1,,v3-
1 rresents velocityifi lb4 ith layer
j repmntsvelocity
in lb4Jfitm--
in whh the rfroctlan I)Orallels the
4 On v. IH .1 '
2 4 6 8 I0 Gij(FT.
PER
12 SEC.)
,nthousands.
14 16 18 20 22
lies on this V line is the multiplying factor for converting delay-time elementsto
depth elements.
Figure 9 illustratesa settingfor V=6,000 ft/sec and Vi= 10,000 ft/sec, which
indicates G,i = 3,750 ft/sec.
COINCIDENT REFRACTION CHART
ffd,n,n+l
: 2 Zi tan(ii,n+[
q-ii,n
).
Figure 10a showsa ray diagram, construction,and identification of symbolsfor n
layers. The symbolXa .... + denotesthe distanceat which the time-distanceline with
slopel/V,, intersectsthe line with slope1/V,`+. It is often termed the critical refrac-
tion distancebut is more properly the distanceat which the arrival times for refrac-
Constructions, Graphs and Nomographs 315
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
tions from the nth and (n-+-1)th velocity media coincide.The angle ij,+ pertains to
the ray of a refraction which penetratesinto the medium having a velocity V+ and
travels parallel to the boundary therein. It denotesthe angle which this ray makesin
the jth layer with respectto a normal to the boundaries.The angle ii,, pertains to
the ray of a refraction penetrating into the nth layer and traveling parallel to the
boundariestherein. It denotesthe angle with respectto the normal, which this ray
makes in the jth layer. Zj denotesthe thicknessof the jth layer. (See Appendix for
derivation of relationship.)
Define
1
Fj,n,n+l = 2 tan (/j,n+l -- ii,n).
Then
Xd,n,n.l
= Zj.Fi,n,nq_lo
j=l
316
Graphs and Nomographsfor Interpretation
X
i
,,
,4-
._
.--
i
i
o
,._
r-
Constructions, Graphs and Nomographs 317
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
I0.0
8.0
6.0
4.0
2.0
1.0
o
I,I
.f.
"' .6
.4
.2
,I
0 .2 .4 .6 .8 1.0
Vj / Vn+I
FIC. lob. Chart for computingcoincidentrefractiondistances.For refractionsalongthe nth and
(nq-1)th layersthe V and V+ velocitysegments
of the time-distance
chart intersectat
Xa,n,n+l
= Zj'Fi,,,,,,+.
318 Graphs and Nomographs for Interpretation
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
, ............
'- ....... . , ,, SINi or-, COS
i, end
TAN
i ' .... ,, . ' '' ' ' 09 o
o.r-,. ', o." '.. .:'.:.'....
.'i'.:
i:,,o.'-,..-i
.......0.4
,,.
,- ....,,o.5
' ,, ,,-.',,06
....... ,,o.?" ,,,,.,
o.s,.','....,,.-',v
. ........
'=',
.r.., ' .... _"_i..i'
0
.... ' -'-0i- ,":
' ';o.::,'."'
..
'...... ' .... - - "' B(,)'. '.................. -.:
:: "- "':"'":' 0- ' - "-..'
.5 .-'0,. ...........
-'' ' ' ...... '.4 ' . '..,....-..:
'50 ,, ,, ..: , . ,,-.:, . ........ ....,,
,, ..,:.:':
.;"...':.
i':'........
=.. .'::.
ii.;'
,:'":.'
... i:;.:i'.
.... :'.
::;;'
. .';':.'
'.,'.'
...... ':.:,....
........
. : .i , ',"' ,'............
, .."'". ._, ,,....
,, -. '_..............:""-
,= '-:'';'.' .-'
....
:'"" '........
....... ,.'.......
-.-. "'"'"'
."'.......
'-' ,...... '"i' '
:' "'
''i '. :.:'i.:...... :.-. - '--'.....' ' '.'.' ,',' or, FEET PER SECOND "
......-- ". .............":" -A ........go& ........ oo ....... o ._'. .,,oo
...
e',-' '. "' -" -,' .": " ' -' ; '- -'.....
';.':;'.,.".
:.,-.:i'.':
.:":.:.'.':. ,;..:::::''-:,;
'-'.,:
':,;'
' ' "..,' ',' ' '':'.
............. '"'" '..... "...."":' :'"' '.','
','.'.:,
' .,:,:i ' "so-
. .....
:,.:;::.
:- :':'';,.:...... ' ':.'.'-.::'.':,' :'.", i'.','. ".,,' '.:':,'- ' ," "'. .....' ..........
....... .... . ":' :;":'
'-'.:"
:.'".;::.
' ' 5.
' -:..':i _ . '.- - ' -: "" ' ' "-" ' ' - ....... -: . ; -. ....
:..,
........' .......... . ...., ..................
. '..............
..... . ,.: . .. . ': ' .....'.........
,...., .,,.,,..... ,. , , .
_ :..A,........... .-............ ..:.:.....,.._. . . _.....
. . _-...
.......... '."- ' "' .Oo
'
7o. ...... :':. -'." .'...-';':- .' ........... :-. ..... ".' : ....... - ................. ., , .;',.:? :,:
.......... '--"............... ' .......... '.,:: . . - ' ........ i ';',.,'::. ' ................
' . ....... :':. :. ::i'-:. , ." '; , ,':'. ' ' .. ,:..... '-', ':' ,'.-, ..'..... :,' ,'
........ ,- . . _ . ; : . ..... . .,. ,. . ., .... .
...... F '": ..... ' " '. ' - .. _' .. :" ..- " - _ .'. : ' .... ' , '. ' .... - ....... ':_'.... ' ' 65-
65-
2.0"_.,TAN,
...... -.- : .'- ' - . ::.--.' .' '. . . .- -.' ' -- . ' '-' ' -: ' '- - -' . 2.o
' ,":.:i:i-::i:;;i._-':'.!,
, :!':i"1":.:-
_:'.::.'::-"':.:'-:',.
- ."-:.::.;'.':.":i':.,'"-
': "."..,__'.".;
......
-:.-"-.":'__.:":'...'::!.'
.'.:,'.';""/':.:o
-' '" ' -:' ' - -.?-.:
;.' ' - :F:;.'"..' .... 55-
so. :;.'i
......o'
"';'"'' "'-: ...."' ' ..... '":
"':'',:' ' . .':
..... ,i."
:'i;.... ' ' ", ,., . ,' ,, ,. . ',..:..:-
':'..,:-,,;,.:.,:.,:,;:'-'
". : ' ,7:r..;
........... -. -..o45-
,:
.
-.
...
.....
. . . .
,.
..
'1 I'"1
-I;Ia''I ':'' le, PLANE
SINGLE-LAYER
PROBLEMS "'' '1 - '1 1 5''' ''....... ' ....... io'
tion from the base of the nth layer coincideswith the minimum refraction distance
at which surface-to-surface
refractionsoccur.It is given by
X,
n = 2 hi tan i/,n+l,
j----1
wherehi is the thicknessof the jth layer and ij,n+=sin-1 Vj/V,+. Thesetangents
can be read directly from the general-purpose chart in Figure 11 or from a trigono-
metric table--upon computingVi/Vn+ and finding thesevaluesin the sinecolumns.
Alternatively, one can construct a single-linegraph of tan i;.+ with respect to
Vj/V,+ and read valuesdirectly.
CHARTSYIELDING H//X FOR CURVED REFRACTIONS
Supposerefractionsfollow circularor cycloidalpaths, as in the caseswhereveloc-
ity isrespectively
a linear[Vo+Kz]ora parabolic
[Vo(1-kKz)
/2]function
ofdepth.
If one knows Vo, the velocity at the referencesurface,then the chart in Figure 12
permits immediateevaluationof H/X. This is the ratio of refractionpenetration
(H) to distanceof seismometerfrom shot (X) and is given in terms of sin/o--the
ratio of Vo to the apparentrefractionvelocity observedat X. The solidcurveper-
tains to the velocity Vo+Kz and is given by the parametricequations
H/X = (sec io -- tan io)
.5
.4
forVz:Vo+Kz
.3
for Vz =Vo'ql+Kz
.2
o .2 .4 .6 .8 1.0
sin io or vo AT/AX
Fro. 12. The ratiosof maximumrefractionpenetrationH to spread-length
X where
velocityis a linear or parabolicfunctionof depth.
Constructions, Graphs and Nomographs 321
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
1.6
1.5
1.4
HL 1.3
Hp
1.2
I.I
1.0
0 20 40 60 80
io (DEGREES)
Fro. 13. The ratio of depth HL computedfor refractionswhere V,= Vo-{-Kzto depth Hr computed
where Vz=Vo(1-{-Kz)J/2.Here, HL/H, is expressedwith respectto io=sin-VoZXT/zXX.
and
a tt/X obtainablefrom Nettleton's(1940,p. 258 and 355) expressions for p and X uponnotingthat
i= r/2 and that K= (VH--Vo)/tt wheresubstitutionis appropriate.
4 This relationshipand later onesinvolving Vo(1-{-Kz)/ follow readily from raypath relationsde-
velopedby Houston(1939) uponletting iH=r/2 or sin- VulVae,as the caserequires.Note that angles
are in radians.
322 Graphs and Nomographs for Interpretation
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
1.0
.8
H
.6
1.0
.4 .8
.6 K
.2
2 4 6 $ I0
T
is K2/4 and its intercept on the VH2 axisis Vo. If the scatter is systematic,perhaps
the assumptionthat Vz= Vo+Kz is not justified.
DETERMININGro AND K IN [z = [o(1 q-Kz) '/2 FROMREFRACTIONS
Constantsof the parabolic expressionof velocity can be evaluated with the aid of
the chart in Figure 15. If surface-to-surfacerefractions are received through a
medium in which V:= Vo(l+Kz) /, then at the shot distance X, where T is ob-
served,
X/T = Vo(' -- 2io q- sin 2io)/(2- -- 4io) sin io.
As in earlier sectionsio is the angle the incident and emergingrefraction rays make
with the vertical. For Volsin io we can write AX/AT, as before, and obtain
(X/T).(AT/AX) = (- -- 2io + sin 2io)/(2' -- 4io).
This quantity is plotted with respectto sin io or VoAT/AX in Figure 15. The ordi-
nate (X/T). (AT/AX) can be determinedfrom observations.Using sucha determi-
nation, say p in the figure,onereadsVoAT/AX off the chart. This readingmultipied
by AX/AT yields a value of Vo.If refractionrays actually conformto the parabolic
velocity formula, repeated determinations, using X, T, and AT/AX at different
spreaddistancesshouldyield a goodrepresentativevalue of Vo.
To obtain K we use a well-known expressionfor KX:
KX = (KVoT/2 + 2 cosio)/sin io.
For sinioonecanwrite VoAT/AX'andfor cosio, [1--(VoAT/AX)]/. Solvingthis
for K yields
K = 2[(ZXX/Vo.
ZXT)
--- 1]/'/[X- zXX.r/2zxr].
[.0
.8
x fit
T fix
.7
0 2 4 6 8 I0
/T
sin io=Vo-
15. A graphof XzXT/TX with respectto VozXT/zXX
(or X/V. T vs. Vo/V) applicable
the determinationof Vo where V--Vo(i+Kz)X/L
324 Graphs and Nomographs for Interpretation
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
SLOPE
=(Sech
KT/2)/V
o
sP >R
//r
iv
=vo+
K,
(VR;VR>V. )
I I I I I I
2.00
1.50
KTo
'X,(i )
1.00
.50
KH
Fro.17.UsingNettleton's
x chartto determine H to a horizontalhigh-speed
refractor
abovewhich
V,= Vo+Kz.a represents
sinio;andc, xn. Uponfindingsinin d onecancomputeH.
depthH. Thetwopairsofbrackets
contain
identical
functions
ofi, withi takingon
thevaluesioandin, respectively.
TheyareNettleton'sChi (x) function,anda graph
of thisquantityappears
in Figure17.It canbe seenthat
KH/Vi = sin i,-- sin io.
This, together with
Zero = X(io) -
formsa pairof equations
that enables
oneto findrefraction
depthH fromthe x
graph.KTois thedifference
betweentwoordinates,andKH/Vi is thedifference
326 Graphs and Nomographs for Interpretation
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
This paperdescribes
a simplegeometricconstruction for determiningrefraction
pathsand refractordepthsfor one,two, and n layers(seeFigures1, 2, and 3). It
alsoillustrates
a nomogramfor quicklysolvingone-layerrefractions
in thefield(see
Figure4). With thisdeviceonecanreaddepthof penetration directlyforparticular
valuesof V, V.,T, andX. The refractionray plotterdescribed
in Figures5 and6
constitutesa useful drafting aid.
Severalchartsareincludedwhichcan expediterefractioncomputations. The G-
factorchartin Figure7 yieldsthevelocityfactorby whichrefraction
intercept
times
Constructions, Graphs and Nornographs 327
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
f (i H)
--f(io)
sin2 i
sin2
i
0I
o, o. o. o.
02
o.
0.$
o.
0.4 KHsin2i
015
o.
016
o'.
0.7 0.8
o'.
o.9
o
I
.o
sinZi
sini
and
T+ = 2 Zi cosii,,+/Vi+ X/V,+.
Here, ij,, pertains to the refraction ray penetrating into but not through the
nth layer and denotesthe angle which the ray segmentin the jth layer makes with
the vertical; ii,,+ is the correspondingangle for the refraction into the (n-t-1)th
layer.
Figure A illustrates the raypaths to which theseequationsapply for refractions
through one layer and through two layers (for n-2, in other words). We note in
particular the coincidentpoint Xd.,.,+where refractionsthrough (n-l) layers
arrive at the same instant as--or coincidewith--those through n layers--Xd,2.ain
Figure A.
At the coincidentpoint X .... + the two times T and Tn+ are equal, so we multi-
ply the resulting equation by V, rearrangethe terms and write the nth term of the
secondsummation separately. Thus:
n--1
Xd,2,3
il,3
(v)
-(il,3+
il,2)
2'. Zitan
i---1
illustrated for n= 2.
+ 2Zn COS
in,n+/(1 -- sin in,n+),
which is readily reducible to'
-- 2 Z tan (/j,n+l
J(a,n,n+l- x ' --i,n).
j=l
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
NEAR-HORIZONTAL BEDS
CARL H. SAVIT*
ABSTRACT
INTRODUCTION
REFRACTION EQUATIONS
A X B
Hu Hd
2Zu cos i
Tou- , (1)
V
2Zd cos i
Toa = , (2)
V
-
sini = Vsec , (4)
Now,if wesubstitute
the velocityratiofor the trigonometric
function(3) andre-
arrange,(5) becomes
4Z2/F = 4Z/V q- ro,. (6)
We notethat 2Z,/F is the reflectiontime of the refractorat shotpointA and can
thusbereplacedby TRz,,sothat finally
TRz2= (4/V)Z q- To, (7)
and similarly
Teza = (4/V2)Za -4-Toa, (8)
cos i
Z-- // (12)
cos (i + )
If we now substitutethis value into (1), we obtain
2H, cos2 i
To,= _ (13)
V cos (i + a)
which may be transformedinto
ro =
2cos(i-a)
_
[ cos2i
V cos(i - ) cos(i + )
. (4)
cos i sin a
= + (s)
cos(i- a)cos (i + a) cos a - sin i
The approximateequationcorresponding
to (14) is
2 cos(i - a)
To, H,, (16)
V
or squared,
Profiling Using Average Velocity 333
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
1--sin 2(i--a)
Tou' . 4 H . (17)
V
Equatingtwo expressions
for the ray constant,we obtain [sin(i-- a) ]/F= 1/Vu,
whencesin(i-a)=F/V,. We thustransform(17) into
To 4H a (18)
2
ic is seeno be a rearrangementof ). similar processmay be usedto obtain
10), te don-dip equation.
The relative error of te approximation for bot up- and dondip is exactly the
expression1) as can be sccny subsfimfin -- for in 14) and cncc 1).
Clearly, the error is zero for = and increasesith incrcasin . Table 1 displays
e error for dips to 1 derees and for velocity ratios from . to .8.
0 5 10 15
To permit speedysolutionof equations(9) and (10) or, for that matter, of (7)
and (8), oneneedonly preparerectangularcoordinategraphsheetswhosescalesare
squared.Figure 2 illustrates one of a set of forms which has been in regular usesince
1958. Times and depths are marked at distancesfrom the origin which are propor-
tional to the squaresof their values. Velocities are labeled with values of twice the
ratio of scaleddepth increment to scaledtime increment. The factor of two is re-
quired as the slopein (9) and (10) is (2/V) . Preplotted on theseforms is a selection
of standard reflectiontime versusdepth functions.If a different function appliesfor
an area, it may be plotted on one of theseforms and that form duplicatedfor compu-
tation purposes.Separateformshave beenfound convenientfor depth ranges1,000-
4,000 ft; 3,000-10,000 ft; 6,000-20,000 ft; 300-1,200 m; 900-3,000 m; 2,000-6,000
m; and 3.0-15.0 km.
334 Graphs and Nomographs for Inerpreaiort
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
7000
9000
1000
300o
0.1
1000 2000 2500 3000 3500 4000
VERTICAL DEPTH (FEET)
... .................
:_.=,..=.::::.:: ::: :::: ::::: ::::..: , :,-:, ..... ::::.::::::::::::::
................................ ..... ::.: . . .. _ ...::::.... :..::.:....::..........:.............:z::::....:...:::..:::::_
SPREAD
( plottecl
fromlefttorcjht)
............
/.O ........................
/?'_4'0 SPREAD ( plotted
fromrght
toleft)
DIRECTION.......
_,..,4.,.,@...
....................
,..,J.._.40'/V
DIRECTION
Plane:
atShot
Hole
........
_ ........ atEnd
Group
.....
.,...,...,._,_
........... Plane'
atEndGroupO_O_
........
atShot
Hole
...........
DIPDIRECTION
.....
O'a/'....(ua-dlpod
....die)ii Velocity
FU,ct
........
...?.....[ DIP
.-_.,.-,.,,,;,;:;;;::
...........................
.:::,:
...... ..........
::::::::::::::::::::::::::::::::::::::::::
::::: : ,::::,,...:.,:,;,,
:::.::....=
.........
,:::._::
..............
:.............
_.,
va Vm..... (a,gl. 5TO........
io+g)
.........
:.$,,0 o< ooo
Fee
co o
O.....................................................
Dd_IdVo COt(O-g)
co$= a
Fro. 2. Refraction computationform for rapid determinationof vertical depth to the refractor from
observedphasevelocitiesand intercept times. Similar formshave beendesignedfor other depth ranges.
Profiling Using Average Velocity 335
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
a=--i
On a cross section, the two depth points are plotted on rays drawn at angles
(i-i-a) and (i--a) from the downdip and updip shotpoints.The downdip shot-
point is the one from which the updip shot is made. Connectingthe plotted points
producesa trace of the refracting horizon. Dip of this horizonmay be measured(or
calculated by finding the arctangent of the difference of depths divided by the
distance between immergent and emergent points) and this determination is in-
dependentof that made from the velocities alone. The degreeof correspondenceof
these dips is a measureof the validity of pairing of velocities and of the validity of
the basic geometric assumptions.
Empirically, criteria have been adopted that permit acceptanceof an interpreta-
tion if the agreementis within one-halfdegreeor 20 percent, whicheveris greater. If
the agreementis not sufficientlycloseto permit acceptance,other pairings are at-
tempted. If no reasonablepairingsare possible,it is probable that a fault or a change
of dip occursover the profile. Fault interpretations or two or more dip segmentsof
the samehorizonmay be computedif V is known or can be assumed,so that the dip
angle can be determined from V and the apparent velocity, and each refraction
segmentmay be plotted separately. If separate computations are made when there
is a simplechangeof dip, an additional error is committed. In most cases,the error is
negligible, especially as it is in the same direction for both profiles of a reversed
pair.
When small departuresfrom the averageline of the time-distanceplot suggestthat
the refractor has an irregular or undulant surface,that surfacemay be approximated
by first calculating and plotting the average straight-line for the surface and then
usingtheperturbation
methodof Dix [1952,p. 271].
Unreversed profiles shot in areas for which no seismic information other than
average velocity data is available may neverthelessyield some depth information.
Average velocity in the upper sectionmay, for example, be very crudely determin-
336 Graphs and Nomographs for Interpretation
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
able from a knowledgeof the generalarea or possiblyonly from the age or lithology
of outcrops.A rough idea of the thicknessof sectionwill then be producedby going
through the computationprocedureusingthe observedrefractionvelocity and in-
tercept time. Surprisingly,often under such circumstancesit is possibleto make a
statement such as: The presenceof a refractorat a depthnot lessthan about10,000fi
suggestsa su2]cientlydeepsedimentarysectionto warrantfurther exploration.
CONCLUSIONS
All of the analysis and design of the computation procedure was performed in
collaboration with John T. Brustad who was then a member of the Western Geo-
physicalCompanyof Americaresearchstaff. Without the resultsof his inspiredand
persistentefforts, the mathematicsof this paper would have been much lengthier.
My sincerethanks are expressedto Curtis Johnsonand Donald Blue for their helpful
suggestionsand careful checkingof the mathematics and the text.
REFERENCES
Dix, C. H., 1952, Seismicprospectingfor oil' New York, Harper & Brothers.
Nettleton, L. L., 1940, Geophysicalprospectingfor oil' New York, McGraw-Hill.
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
Section 5C
337
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
L. W. GARDNER*
ABSTRACT
INTRODUCTION
,.u: :..,., .
DISTANCE
BETWEE m?m]:'z.:
REFRACTION
RECORD
ARRIVAL TIMES
CORRECTEDTO DATUM 1':,
I
I
I
, t+t'G
+: ,
JM
Fro.1.Graph
oftraveltime
versus
distance
illustrating
dipanddelay-time
relationsin a typicalgeologicsection.
Either first arrivals or later arrivals may be used and interpreted similarly but
independently. However, continuity of good dip control is a requisite for good re-
sults, and this meansthat eventsgenerallymust be identifiable and cotrelatable.
Arrival time readings,T,are made on leadingpeaksor troughschosenaccording
to typical character.In somecases,there may be questionas to which cycleto read
on a given set of arrivals. In these cases,chooseone and note question of cycle.
The dip indication will be correctregardlessof choice,and the questionof absolute
value can be deferred until the late stagesof interpretation when the relationship
to other data will be apparent, and choicecan be made in line therewith. The com-
mon time interval from beginningsof events to troughs or peaks read can be
accountedfor in the depth chart. Intercepts, To, are calculatedby'
To, = T,-
Ti = Observedarrival times;
X= Shot-to-detector distances;
Refraction Seismograph Profile Interpretation 341
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
V,,=Assignedmarker-horizonvelocity;
ERi--Receiver elevation above referenceplane;
Es=Shot elevation above referenceplane;
Vc= V/cos sin-(V/Vm.) ;
Vr= Near-surfacevelocity; and
Wc=Weatheringcorrections
as determinedfrom specialrefractionweathering
shotsor uphole time measurements.
This refersall data to the chosenreferenceplane. If it is chosennear the ground
surfacelevel, complicationis avoidedin computingoffsetdistancesand all correc-
tionsare kept relativelysmall.A featureof the techniqueis that an approximate
value for V,,, sufficessincecorrections
for error in this are easilyappliedsubse-
quently.Therefore,
muchdifficultyin handlingthe primarydataof interpretation
is avoided.
;i = 1. (6)
i is interpretedas the proportionof total thickness
overlyingthe markerhorizon
whichis assigned velocity,V, regardlessof whetherit occurs
in a relativelythick
uniformlayeror is distributedasthin layersthroughoutthe section.Similarrela-
tionswerereportedby Ambronn(1928)andJones(1932).
Unlessdetailed information is available from well velocity data, we believe best
resultsareobtainedin assuming
that all bedsmaintaintheirvelocities
laterallyand
thin or thicken in accordancewith relations (5) and (6). We therefore compute
depthsandoffsets
in accordance
with the followingrelations:
342 The Delay-Time Method
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
Xn
= (t'--a)'y't5,
tansin
-(V,/V.,)
(8)
y,/5i
cossin
-
Vi
(Vi/V.,)
If more detailed information is available or may be reasonablyinferred, it is
possibleto express(7) and (8) in terms of variationsfrom conditionsat a reference
location Thus, H, hi, t', and Xcnmay be replacedby quantitiesrepresenting varia-
tionsfrom their referencevalues.The/5i may then be assignedto allow for different
rates of thickeningor thinningof variouslayers.If the referenceplaneis not chosen
near the surfaceof the ground,adjustmentfor this must be madein equation(8),
sinceH strictly shouldrepresentthe depth below the ground surfacerather than
below the referenceplane.
EA ST
Sl
/ Sk5
J---SHOT PO,NT
POSITIONS Sl4 31,..
'-,................{,..... INTERCEPTS
AT
'-"................-"-'2'................. ....... ._............
' ...........
_:............
RECEIVER POSITIONS
v m = 22,000 FT./SEC.
.75
15S.I'"-O
Fi:'$
ET
IPuINT SHOT--"
SI4
POSITIONS) j13 512
3\Tmo(3) am
.,_TO(13). To(2) _To(i)
(CORRECTED FOR
TRUE
MARKERVELOCITY).70
"-'-- --' --' I'-'"'.
- .__
(6's-6'R)
__ .X__!'
INTERCEPTS IN
OFFSET POSITION
.--....
.7,5
'ts
/--1--'"-----"-----'
AVERAGE
DELAYECU'__p_lR _
TiM -' RELATIVE
DELAY
TIME CURVES
.3O
.4O
Fro. 2. Refractionprofileinterpretationprocedures
by the delay-timemethod.
Refraction Seismograph Profile Interpretation 343
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
profile, and approximatedelay-time values can be obtained from this estimated dip
and the intercept values. The offset distancesto be used are determinate from these
values and formula (8), subject to improved approximations.Intercepts are now
replotted in offsetpositions,shotpointlocationsalso beingoffset.In this replotting,
the data are smoothedfor irregularities repeated for shotsfrom oppositedirections,
which thus are indicated to be near-surface influence.
The next step is to constructrelative delay-time curves which have no absolute
fixed reference time but show a continuous dip relationship. One such curve is
drawn parallel to offset intercept values for shotsin one direction. A similar curve
is drawn for shots in the oppositedirection. The average dip of these two curves
representsthe true dip in terms of delay time regardlessof whether or not a true
value of the marker horizonvelocity, Vm,was chosen.The final absolutedelay-time
curve must be parallel to this averagerelative delay-time curve within the accuracy
of the data. The problem of establishingits level remains.
If the relative delay-time curvesfor oppositelydirected wave travel are parallel,
the originally chosenmarker horizon speedis correct. If they diverge it is in error,
and adjustments in intercept-time values and in the value of the marker horizon
speed are necessary.The interpreter must be alert to the possibility of a sharp
divergencein velocity representinga changefrom one marker to another. The ad-
justment to be applied to an intercept value is the divergenceof either one-way
relative delay-time curve from the average delay-time curve for the distancefrom
offsetshotpointposition to offsetreceiverposition. If, in order to achieveparallelism
of the two oppositely directed relative delay-time curves, it is necessaryto rela-
tively decreasethe time at the arrowhead end, the true velocity of the marker is
lessthan that assignedand the intercept values must be decreased.The method is
illustratedin Figure 2 for the S spread.The magnitudeof the correctionis/xs-/xR,
and it reducesthe value of the intercept. Note that the correctionis calculated only
over the length of the baseof the trajectory betweenoffsetshot and receiverposi-
tions. This is that part of the trajectory that is, in fact, affected.
The adjustedinterceptvalues,togetherwith the averagedip in delay time shown
by the average relative delay-time curve taken between offset shotpoint and de-
tector positions,sufficeto determine absolute delay-time values of these two posi-
tions for eachspread.Only one delay-time value per spreadneedbe computedsince
the othersare related through the dips shownby the averagedelay-time curve. The
method of computingdelay times is illustrated in Figure 2 for the $ spread.The
adjustedintercept, To,is split into delaysunder the shot and receiverpositionsin
accordancewith the dip, , indicated by the averagerelative delay-time curve.
The final delay-time curve is drawn through the set of absolutedelay times thus
determinedessentiallyparallel to the average delay-time curve. If incompatibility
in thesedata is found, adjustmentsconsistentwith tolerancesin data may be used.
Questionsof original readingsof eventsto a cycle may be reviewedand taken into
account.Possiblefaulting or changesof marker may showup. If there are crossing
profiles,delay-time values at tie points should match. Finally, these delay times
are convertedto depthsthroughformula (7).
The true marker velocity need not be known accurately to carry through this
344 The Delay-Time Method
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
When refraction data are obtained on two markers, shallow and deep, it is evi-
dent that the lateral velocity variations representedby variable delay times to the
shallow marker must affect determinations on the deep marker in a manner analo-
gous to the influenceof weathering upon either reflection or refraction data. The
refraction data on a shallow marker yield a measure of delay times down to it,
without discrimination whether the variations are due to variations in vertical
velocity of the overlying sectionor to variations in its depth, with consequentlat-
eral velocity variation concentrated at the level of the interface. These two types
of variation are called Case I and Case II, respectively,in the following treatment
of the problem of applying corrections.
(8)
(9)
From (9) we obtain'
- !
C = t'Xa- l XC
=h
Ics
sin- cos
sin
-1(Vl/Vm,2).l
(V,/V,2) (10) V
= hi
[cOssin-l(Vls/Vm'l)
cssin-l(V/Vm')'
1Vls 1) --
V1
. (1
c = - ....
V 2cossin- (V/Vm,2)
and
At'= - h
V cos
sin zXV . .
- iV/Vm,O
(13)
+, \\ h .,., . I
IXa Xy Vm,
I
2
Fro. 4. Variations in deepmarker delay times for undulating shallowmarker.
c =
cos
.
sin
-1
(g/gm,)
= F. zXt'.
cos
sin
-x
(g/Vm,)] V.
cossin- (V/V,,.)
A#
(7
In both CasesI and II the correctionto be appliedis of the sameform, C-F. zXt',
with the only differencesin procedurebeing the use of appropriateF values and
relative offsetdistances.For CaseI, F = 1.0 can generallybe used.For very shallow
velocity anomaliesthe shallowhorizon delay-time curveswill require little or no
offset,and there will be little or no differentialoffsetbetweenthe pointsat which the
correctionsderivedfrom the shallowhorizoncurvesare appliedto the deepdelay-
time curves.In CaseII, F-0.5 can generallybe used.The shallowmarker delay-
time curvesare offseta distanceXcl. The deep marker delay-time curvesare offset
a distanceX,; the correctionis applied, and they are then offsetto a distanceXc2.
The method is illustrated by the followingprocedures'
1. Plot a profile of shallowmarker interceptsunoffseton crosssectionpaper.
2. Draw one-way shallowrelative delay-time curvesoffset appropriateamount
as discussed above.
3. Draw averageshallowrelative delay-timecurveaveragingcurves(2).
The Refraction Seismograph Profile Interpretation 347
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
4. From divergencesof (2) determine and record true shallow marker velocity
and correct intercepts for marker velocity deviation from the assignedvalue.
5. Compute and plot shallowdelay times at each end of each trajectory.
6. Construct a final delay-time curve by making slight adjustments to dips and
delay times, if necessary,to bring about overall agreementalong the profile.
7. Post delay times on a map, adjust at crossingprofiles to effect ties, and con-
vert to depths.
8. Plot deep marker intercepts unoffset.
9. Draw one-way deep relative delay-time curves offset by an amount less than
or equal to Xa as discussedabove.
10. Subtract F times the final shallow delay-time curve from the deep relative
curves.
ACKNOWLEDGMENT
REFERENCES
Ambronn,R. (translatedby Cobb, Margaret C.), 1928, Elementsof geophysics: New York, McGraw-
Hill Book Co., Inc.
Gardner,L. W., 1939,Seismograph prospecting:U.S. Patent 2,153,920(application1936).
1939, An areal plan of mappingsubsurfacestructureby refractionshooting:Geophysics, v. 4,
p. 247-259.
Jones,J. H.,1932,The diffractionof elasticwavesat the boundariesof a solidlayer: Proc.Royal Soc.
A, v. 137,p. 325-343.
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
PROFILE INTERPRETATION
K. M. BARRY*
ABSTRACT
INTRODUCTION
DELAY TIME
I- slope
= sin(ic-e)
= I/Vr
I
I
I !
I I
I x Ix DISTANCEX
i I I
i I i
I I I
I I i
i I I
O[ , ,, ,, ,, I ,, ^ ^ * zX . . . DATUM
',\', /
' ' Vr
Fro.1. Raypaths
andT-D curveassociated
witha homoclinal
refracting
horizon.
T = to q- tc q- tcx. (2)
For a refractorof gentledip (i.e. 0<10), ADEX, so that in termsof the delay-
times (la) and (lb), the total traveltimemay be expressed as:
T = X/Vr d- lio d- licx. (3)
At zero distance,the interceptof the time axis will be'
r [=0= = o + ; (4)
thus
= r- X/V,. (5)
It shouldbe observedthat I is not the sameinterceptas Ia, the apparentintercept,
whichis produced
by the intersection
of the time axiswith a line throughthe
refractedarrivals. This is becauseI is basedon a slopeof 1/V, whereasthe slope
whichyieldsIa is a functionof both V andthe dip of the refractor.Thislatter
35O The Delay-Time Method
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
slope=sin(i,_+0)/Vrsinic.Fora horizontal
refractor
these
intercepts
willbeidenti-
cal,aswillboth/oB andcx.Theintercept I canbereadilycalculated
from(5),
but to determinethe depthto the refractor,it is necessary
to partitionthisinter-
ceptintoa delaytimeat thesource andanotherat thedetector.
PARTITION OF INTERCEPTS
IX
I I
DISTANCE
I I
I II
I
I I
Fro.2. Raypaths,
T-D curve,andintercept-distance
curveassociated
witha refractor
which
is partlyhorizontalandpartly homoclinal.
Delay-Time Profile Interpretation 351
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
OFFSET DISTANCE
Constantvelocity(referto Figure 3)
$= - v?)TM, (0a)
= 5Z - V?)l/VVV. (0b)
DATUM
Vl
I sj
Vn
tural changesof the refractor being attributed to variationsin the thicknessof Zn.
It is alsopossibleto assumethat the thicknessesof severalor all of the layers vary
by someproportionalamount of the total depth.
/tos
=--{cosh
-(Vr/Vo)
--cosh
-x[Vr/(Vo
+ kZ)]
- (v.- Vo.)/Vv
+ [v?- (Vo+ kz).]/Vv}. (lib)
2. Linear increaseof velocity with time'
V = Vo+kt = (Vo'+2kZ)
1
s = -2k { Vo(V
.- Vo")'/- v, sin-x (Vo/VO
- [V2Vo"+ 2kZ) - (Vo.q- 2kZ)2/'
q- Vr2sin- [(Vo..nt-
2kZ)/./Vr]}, (12a)
Vr
/tos
= 2k{sin-
[(V2
nt-
2kZ)/'/V]
--sin-
(Vo/Vr)
- (Vo/V)(v.- Vo.)/.
q- [VrVo'q- 2kZ)- (Vo2q-2kZ)']l/'/Vr'}. (12b)
The equationsfor any continuousvelocity functionare derivedby an integration
of the generalrelationshipsof offsetdistance,
S=
f0VdZ/(V?- V TM, (13)
Delay-Time Profile Interpretation 353
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
DATUM
Vo
Vr
Fro. 4. Geometryof the critical raypath througha sectionhaving a
continuouslyincreasingvelocity.
and traveltime
t=
fozVrdXf[V(Vr2-- V2)1/2]. (14)
V is the velocity distribution applicableto the area, and (-t-S/Vt. The resulting
expressionsfor offset and delay-time are cumbersome,even for the simple linear
velocity functions.Their application in the generationof curvesis simplifiedby the
use of a computer or by means of auxiliary functions such as used by Nettleton
(1940, p. 264), for equations(11a) and (lib).
SHOOTING PROCEDURE
INTERPRETATION
Delay-TimeProfileInterpretation
-,
i
O
0
0
0
(3)
0
i
0
'
0
0')
0
I
0
0
0
0
i
0
*-:
0
355
356 The Delay-Time Method
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
8000
9000
DEPTH
(FEET)
.._..._..1.072
=I/2 I
I0000 --
10420'
k'
I000 m
The sign associatedwith cid is negative if the relative curvesrise with respectto
the average curve when consideredin the direction of wave travel, otherwise, the
sign is positive. The delay times are sensitiveto a changein Vr so it is necessaryto
recomputethe intercept values if the refractor velocity requires correction.If the
correction does not produce a significant change in the offset distances,then the
averagerelative curve will still show the proper attitude of the refractor.
It is possiblethat the refractor may not maintain a constant velocity but will
exhibit lateral variations in Vt. In this case the interpretation is accomplishedin
discretesegmentsor graded accordinglyas V changes.
E. The intercepts are then partitioned into their delay times. The horizontal
positionsof the point of entry and point of emergencefor a particular raypath are
located on the average relative time curve. The time difference,/Xr, between the two
pointsis then read and the delay timesare computedby meansof (9a) and (9b). (In
the example, the /Xr, on the average relative time curve, between the offset posi-
tions of shotpoint lB and the first detector of spread lB is 0.059 sec. The delay
times are (2.145+0.059)/2= 1.102secand (2.145-0.059)/2= 1.043 sec).The larger
of the two delay times is plotted at a position which correspondswith the greater
time on the average relative curve, the other at the position of lesstime.
This procedureshould be carried out for at least two intercepts of each spread
and for all the spreadsin both directionsof shooting.(For clarity, only a few critical
delay-time points are indicated in the example.) The resulting delay-timepoints
shouldparallel the averagerelative curve, and the intervening values are obtained
from a fit with the average curve.
Where two adjacent spreadsoverlap there will be several common detector posi-
tions. The waves arriving in the same direction from the two associatedshotpoints
will yield different intercept times at a commondetector.The amount by which these
intercepts differ will be the time differencebetween the offset locations of the shot-
points on the average relative curve. It is thus possibleto determine the relative
time of those shotpointswhoseoffset position falls beyond the ends of the average
relative curve, that is, beyond the limits of two-way control. This is valid, how-
ever, only if the refractor velocity remains constant between the points of entry of
the rays.
The completionof the final step is accomplishedby assigninga depth scaleto the
delay-time scale,utilizing the previouslyderived delay time versusdepth function.
(From the curvesof Figure 7, the delay time of 1.102 secgivesa deptlxof 10,420ft,
and that of 1.043 sec,a depth of 9,600 ft.)
REFERENCES
APPENDIX
Constantvelocity
Referringto Figure 3, it is seenthat the geologicsectionis presumedto consistof
n layers,eachof which hasa constantvelocity V and a thicknessZi. j-- 1, 2, ., n.
It is apparent that for the jth layer,
tan 0.
From Snell's law
sin 0 = V.//Vr.
Thus,
tan 0 = V.//(Vr 2-- Vj2)TM,
and
By definition, the delay time associatedwith the segmentA C of the jth layer is
.4c = A C/V./ - S.//Vr.
Then,
cos0 - Z./tan O/Vr = Zi(1 -- V./sin O/Vr)/V./cos O,
but
V./ = Vr sin O,
so
= - (10b)
Delay-TimeProfileInterpretation 359
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
Varying velocity
In orderto derivethedelaytimepertaining
to a geologic
section
thatisapproxi-
mated
byanycontinuously
varying
velocity,
V, it isnecessary
toobtain
anexpres-
sionfor thetraveltime,
t. Thegeneral
equations
for traveltime
andoffsetdistance
follow from a considerationof Figure 4. Thus,
AS = AZtan0, and VAt = AZ/cos0,
or, by usingexpressions
involvingSnell'slaw,
AS= VaZ/(Vr
2-- V2)/, and at = VrAZ/[V(V
'- V)/].
Makinga summation
overthetotalpathresults
in thegeneral
equations
S----
0ZVdZ/(Vr2__V2)l/2 (13)
and
By usingthe transformation
u = v/v = (Vo + kZ)/V,
such that
du = kdZ/V,,
s = (wJk)
f0Wu/[V(- v/v)'l,
or
S
v,foUaU/(
k _u)
'
Integrationresultsin
Vr Vr 0
k{-(- u),}
0 =- 1[- (Vo
+kZ)/V]'},
Vr
S
{(- Vo/V,)/
- [- (Vo
+kz)/v,]'},
S {(v,- Vo)
/- [v,- (Vo+ kz)q/}. (11a)
360 The Delay-Time Method
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
Using
becomes'
thesame
transformation
asbefore,
thenthetraveltime
equation,
(14)
t = VrfoZ
a/[v(- vVv?),q
or
Application
offormula
(129)fromA Short
Table
ofIntegrals
(Peirce,
1929),
yields-
1 0
t =--{sech
k
-u}z
or
1 1
t = --k {cosh
- (1/U)}z
o= --k {cosh-
[Vr/(Vo
q-kZ)]},
1
t = --k {cosh
- (Vr/Vo)
- cosh
- [V/(Vo
q-kZ)]}'
Fromthe definitionof delaytime'
o = t - SIVa,
so
/oB
=--{cosh
- (V/Vo)
- cosh
- [V/(Vo
q-kZ)]}
1
kV,{(v?-Vo),
- Iv?-(Vo
+
or
1
oB= --k {cosh
- (VrfVo)
- cosh
-x[Vf(Vo
q-kZ)]
-- (Vr2- Vo2)l/2ffVr
+ [Vr9'- (Vo+ kZ)2]I/2/Vr}. (11b)
2. Linearincrease
of velocitywith time'
V = Voq- kt = (V? q- 2kZ)/.
From the given relation it followsthat'
VdV- kdZ,
sotheoffsetdistance
of equation(13) becomes'
s=T v'.av/(v'.
- v)
Delay-Time Profile Interpretation 361
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
S { (V/2)(Vr
2 V2)
/2+ (V?/2)sin
- (V/VO}
Then,
1
S= 2k
{[Vo + 2kZ]/2[Vr
2- (Vo + 2kZ)]TM
o
t=
V
kfoZ
dV/(V - V) TM.
Vr
t = {sin
-[(Vo
q-2kZ)I/2/V,
.]--sin
-1(Vo/Vr)}.
Since
tio = t - S/V,
it follows that
oB
=- {Vsin
-[(Vo
q-2kZ)I/2/V,
.]- V,.
sin-1(Vo//Vr)
-- (Vo/2V,.)(V,.
- Vo.)
/'q- (V/2) sin- (Vo/V,.)
+ + 2z) - (Vo + TM
-- (V/2)sin
- [(Vo
2q-2kZ)/2/V,.]},
or
/ioa
= {sin
-1[(Vo
'q-2kZ)ll:/Vr]
- sin-X
(Vo/V,.)
--(Vo/V,.2)(V,
.--Vo)
TM
+ [V?(Vo
q-2kZ)-- (Vo
+ 2kZ)2]t2/V'}. (12b)
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
Section 5D
Wavefront Methods
"Wavefront Methods" utilize the intersection of wavefronts drawn from the shot-
point and wavefrontsdrawn from the receivingpoint or points to form a loci of solu-
tions, which can be interpreted from multi-raypaths and multi-wavefronts.
The first paper, "A General Wavefront Method," by Rockwell, discussesthe
broad picture utilizing wavefront methods and includes some novel and special
applications.
The secondpaper, "Refraction Solutionsand Wavefront Targeting" by Schenck,
demonstratesa special wavefront technique which is unusual in its approach to
solving a refraction problem independent of refractor velocity for the reversed
profile.
The third paper, "Outlining of Salt and Shale Massesby Refraction Methods," by
Musgrave et al., describesthe mapping technique for nonconformableor diapiric
type intrusive masses.
362
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
DONALD W. ROCKWELL*
ABSTRACT
INTRODUCTION
Over the years a number of very competent papers have been publishedon the
subject of the interpretation of refraction data. The approachesto the problem
have ranged from highly analytical mathematical methods through combinations
involving a certain amount of straightforward arithmetic, combined with simple
graphicalrepresentation,to methodswhich are nearly completelygraphical. Among
these latter are found the so-called wavefront methods. Whatever the merits of the
other approaches,it has been the author's experiencethat the wavefront methods
are peculiarlysuitedto the interpretation of actual field data, especiallywhen several
nonplane layers are involved.
The objective of this article is to present a general wavefront-interpretation
method. Many of the various techniqueswhich together make up the method are
quite frankly borrowedand adapted as neededfrom all pre-existingmethodsknown
to the author, and the attempt is made to showhow thesemay be weldedinto a prac-
tical, highly fieldworthy set of toolsfor the practicingrefraction interpreter.
Certain writers must be cited at the outset for their original contributions to
wavefront technologyas applied to refraction problems.Among the early writers,
Thornburgh (1930), Ansel(1930), and Gardner (1949), may be mentioned,and more
recently,Baumgarte(1955), and Hagedoorn(1954 and 1959). There is a considerable
2.25
/
/
1.75
xd2
.25 --
LEGEND
.75
FIRST ARRIVALS
LATER ARRIVALS
PROJECTIONS
.5O
; ; ; ; ; REFLECTIONS
.25
DISTANCE X ----
/
/
/
/
/
Fro. 1. Basicreflection-refraction
travel paths and t-x relationships.
cometo appreciate
the visualadvantages
of wavefrontmethods
andthe flexibility
which they offer.
Radiating wa,efronts
Radiatingwavefrontsystemsare nothingmorethan setsof wavefrontsemanat-
ing from a point source.The bestexampleis the roughlysphericaldisturbance
spreading downwards andoutwards froma typicalseismic surface
shot(seeFigure
2). Seenin two dimensions,
theseare the familiarplottingchartsusedfor reflection
--TOTAL
, TIME I S
? . .2 .\ 4 . .2 ., ,
...
....,.-. ._,..,:--,.... ..... 5, 5"-X'=" ' -%"/2 ''
FIGURE
2a(AROUND
CIRCULAR
WAVE FNTS,REAL
ST) IMAGINARy-,
(ARND CEIVER),IN CANT
1 L E FIGURE
2b RADIATING
W-FM IN E
TOTAL REFLECTION TI. VERITY LAYE
6o. R.P.
FIGURE
2, HI-VECITY
DIFFCTI
FIGURE 2c EXPANDING ENDI CIU
WAVE F IN A
Fro. 2. Commonlyencountered
radiatingwavefrontsystems.
366 Wavefront Methods
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
Directedwavefronts
Directed-wavefrontsystemsare characterizedby an actual or apparent line
sourceof energyand are typically plane or near-planeparallel systemsin contrast
to the nearly concentricaspectof radiating systems.
The typical "bow wave" or "refractedwave" (Mintrop wave) is the commonest
exampleof sucha system(seeFigure 3a). Theseare almostexclusivelyassociated
with the recordingend of a refraction survey, where they really exist, while the
real existenceof the radiating system is at the shot end.
The ultimate sourceof energy in a bow-waveor directedwavefront system is
usually a radiating wavefront traveling in an overlyinglower velocity medium.
When the latter encountersan underlying high-velocitymedium at a certain in-
cident angle (the critical angle), it suffersthe maximum possiblerefraction and
propagatespractically along the interface separatingthe two media. The wave-
front moving in the high-velocity"refractor" acts as a line source.Through the ac-
tion of Huygens'principle,whichholdsthat eachpoint on the wavefrontof a wave-
propagateddisturbanceacts as a sourceof minute sphericalwavelets,the directed
wavefront or bow wave is awakenedin the upper medium. This is accomplished
through the coalescence of successivelyinitiated sphericalwavelets originatingat
the interface as the wave in the high-velocitymediumprogresses alongthe bound-
ary. This phenomenonis depictedschematicallyin Figure 3a.
An interesting discussionof the "volume" involved in a minimum traveltime
path in simplerefractionis givenby Hagedoorn(1959).
General Wavefront Method 367
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
The angle which the coalescedwavefront (the refracted or bow wave) makes
locally with the interface is a function of the velocitiesin the two media. It is known
as the critical angle, ic, and is given by'
where V is the lower velocity and V. the highervelocity. This is clearfrom a study
of Figure 3a. For example, the particular wavelet emerging from a center at time
1.0, at 0, has progressedas far as the 1.2 secwavefront position at P in 0.2 secat
velocity V, while the disturbancein the refractorhas progressed
to P at velocity
V. over the same time period.
The directedwavefrontstravel obliquelyupward from the refractor,undergoing
in all but the simplesttwo-layer casessomechangesin directionor inclinationof the
wavefronts,but arriving at the surfaceas a systemof essentiallyparallel and nearly
plane wavefronts which sweepacrossthe detector array set to record their arrival.
If the velocity in the near-surfacemedium is known, it is clear that the attitude
of this directedwavefront systemcan be reconstructedquite approximatelyfrom
the arrival times associatedwith the wave's passagedown the line of detectors.Two
useful piecesof information are provided:
(1) The total traveltime from the shot sourceto any given detector(and by inter-
polationto any surfacepoint).
(2) The approximateconfigurationof the wavefront in the plane of the sectionat
any given time after the explosion,i.e., the particular "directed" wavefront
systemcreatedby the geologicalconditions.
CONSTRUCTION OF WAVEFRONTS
Radiating wavefronts
Radiating wavefrontsfor any given observedor inferred vertical velocity dis-
tribution can be constructedaccordingto already well-known methodsdescribedin
the literature. For constant velocity media, they are, of course,mere setsof circles
(Figure2a). For a linearincreaseof instantaneous
velocitywith depth,described
by
the relation
V, = Vd + aZ,
one can describethe wavefront for any given traveltime by utilizing the parameters
(seeFigure 2c)
h = Va/a(Cosh at - 1),
r = Vd/a(Sinh at),
Directedwavefronts
The individual directed-wavefront system is constructed from observed time-
distance data applying a knowledge of the overlying velocity field. In contrast to
the radiating systems which generally remain or are considered constant for the
prospect, the directed-wavefrontsystem is unique and individual and must be
at least partially reconstructedfor each new portion of the recorded data. The
attitude of the wavefrontsis directly related to structure, i.e., to the dip of the re-
fractor and to the dispositionof the overlying layers.
The first interpretational task is always to construct this directed system im-
mediately below the surface (in the first layer).2 Figure 3a illustrates how this is
done. Assumethat along a given recording profile a given set of time-distancedata
are recorded and that these are known or believed to be refractions from a single
subsurfacehorizon. Also, the average velocity in the first layer is known to be
V= 5,000 ft/sec.
Supposethat for the particular problem we are trying to resolveit is sufficientto
show a particular wavefront system at intervals of .100 sec. Let us begin with the
wavefront correspondingto the time 1.400 sec. This may be an observedvalue or
one interpolatedbetweennearby observedtimes.It is convenientin this caseto keep
the wavefronts in even multiples of .100 sec or some other regularly increasing
time interval and to let the surfacepositionsat which these times are observedfall
where they may.
Now, at the surfaceposition correspondingto t= 1.500, obviously the 1.4 wave-
front positionis displaceddownwardand to the left towardsthe shot by a distance
' It is not necessaryto reconstructthe completesystem,as Mr. Frederick Schenckshowselsewherein
this volume.(SeeRefractionSolutionsby WavefrontTargeting,by FrederickL. Schenck,p. 416.)
370 Wavefront Methods
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
correspondingto .100 sec times V or 500 ft. Similarly, at the surfaceposition for
observedtime 1.600, the 1.4 wavefront is .200 times 5,000 ft/sec or 1,000 ft away.
We need only draw small circlets at each of these surface positions with radii
proportional to the correspondingincreasein time or At, expressedas a distancein
order to locate quite closelythe positionof the wavefront 1.400 which we set out to
construct. For more or less detail, we use greater or smaller time increments. Of
course,when the t-x curve is a straight line, as in this example, and the velocity is
constant,the resultingwavefront will be plane and the set or systemof wavefronts
will be parallel, so very little drawing will sufficefor the constructions.If the t-x
curve is not straight, the wavefront must not be plane, so enough points are se-
lected to show its essentialconfiguration.
After drawing the 1.4 wavefront, the others may be drawn in, each intersecting
the surface at the observedtime correspondingto the wavefront time value. This
is all there is to drawing an emerging wavefront system. A little practice and the
systemfor an entire long t-x curve can be drawn quite rapidly with as much or as
little detail as is desired.
How much of the system should be drawn in will depend on the problem to be
solved, the approximate expecteddepth of the refracting interface, the validity of
the time-picksand corrections(i.e., whetherminor time variations are significant),
whether one or more velocity layers are involved, etc., but this will becomeapparent
from later examplesand from practice.
tend the t-x curve from S back towards S without projection. These new data will
permit us to completethe directed-wavefrontsystemon the basisof observed(At's)
rather than on the risky basis of extrapolation of the last observedslope of the
t-x curve correspondingto "first-arrival" data from S.
Having done either of these, we now have two overlapping wavefront systems,
one radiating and one directed. Since the wave initiated in one system and con-
tinued in the other and the transition took place at the refractor, the possibleposi-
tions of the refractor must lie along a locuswhere traveltime values in the two sys-
tems are equal, i.e., where equal time wavefronts intersect.This curve is drawn in
Figure 3b and appearsto be a parabola. If we considerthe shotpoint as focus and
the zero-time line of the refracted-wave system as directrix, it is obvious that the
curve is a parabola, i.e., the locusof points lying equidistant from a straight line
and a point.
The fact was first pointedout by Thornburgh(1930), whoreferredto the parabola
as a "coincident time curve" along which any point is at equal times from the shot-
point along two different travel paths. He pointed out that these coincidenttime
curvesweretangent to the refractor and intersectedthe surfaceat the "breakover"
or critical distance, xa (a changein velocities signalling a changein first-arriving
wavefront systems).
For plane refractors, the Thornburgh system was ultimately simplified into the
"parabolicmethod" in which only a singleray was constructedand the positionof
the zero-time wavefront determined along this ray to serve as a directrix for the
direct determination of the parabola, whose focus was at the shot. See the ray
r- on Figure 3b. For the simple plane two-layer case,it is evident that with V
passingthrough the origin of the t-x plot, the distanceto the breakoverpoint or knee,
being V.txa, is also the distancefrom this point along an emergingray to the zero-
time line or directrix. This will only be so for a constant overburdenvelocity, with
no shinglingor lossof legsin the overburdent-x curve.
Another ray of interest is the ray r- VX T0, drawn back from the shot position;
this ray, together with any other ray (the xa ray, for example),convenientlylocates
the directrix.
Thornburgh made his depth interpretations by determining the loci of these
parabolasand near parabolas(in deeperlayers) at each shotpoint.Presumably,he
had techniquesfor adjusting nonplane interfaces to fit deviations from straight-
line t-x data between these control points but these are not elaboratedin his classic
AAPG paper.
Severalother commentsabout the wave systems(Figure 3b) are in order:
(a) A flat refractor will, of course,be tangent to the parabolaat its lowestpoint,
where the ray from the shot impinges on the refractor, and refraction along the
interface begins.The observedvelocity V2 will be true refractor velocity in this case.
(b) Exactly the samerecordedtime-distancepointscouldbe causedby refractors
of a wide range of velocitiesdipping up or down but thesewould still lie tangent to
the parabola. If dip is known, true velocity can be calculated or vice versa from
relationshipr = it___
a wherer is emergingangle,a is dip, and ic is the critical angle.In
terms of velocity,
372 Wavefront Methods
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
r I r 1
+ a = sin -x -- sin -
-- V (apparent) V (true)'
and the signof a indicatesdip awayfrom the shotpoint(-k) or towardshotpoint(--).
(c) If dip changesinside of the knee, then the t-x curve lying immediately ad-
jacent to and just outside of the point where the first critically refracted ray, the
critical ray, reachesthe surface,must be used to construct the parabola or intersect-
ing wavefront systems.This must be extendedby phantomingfrom first-arrival data
using auxiliary or second-arrivaldata. See Figure 3b, line marked 13,000 ft/sec
velocity. If not, the parabola will not give the true locus of possiblerefracting sur-
facesbut only what it would have been had there been no dip changein the "blind
zone." The differenceis shown by the additional directrix and parabola dashedin
lightly on Figure 3b.
Applyingthe dip formulas,assumingV true to be known at 10,000ft/sec,
5000 5000
4-a- sin- - sin -
13,000 10,000
- sin - .384 - sin - .5
= 22.6 -- 30.0
Thus, instead of a flat refractor at --2,000 ft, the true refractor position is shown
by dotted line tangent to dotted parabola. The importance of constructingfrom
data pertaining to actual subsurfacecoverage(within "blind zone") and not
projecting is thus clearly shown.
The principal utility of the parabolic method is in representingall possibleposi-
tions of the refractor leaving the way openfor later selectionof dips and/or true
velocities.
We shall review later the extensionof the parabolic method to more than two
layers and analogousmethodsfor other velocity distributions.
30
RT
'
20
I I
i
CRITIC DELAY
A(S) RI RAY C(R) .TIME R
e o
/
b
to times TARand TBRon the t-x graph. The t-x data beyondR1 and R.permit us to
draw these two directed wavefront systems to which the pair of complementary
wavefronts with time TAR and TB 2 belong, utilizing the simple construction
method given on the right-hand side of Figure 3a. These wavefronts intersect at
depth and fix the position of O on the refractor. As a matter of fact, each O is fixed
only by the real portions of the wavefronts immediately above a particular O and
image portionsimmediately below. The remainderof the wavefrontsare superfluous
and need not be constructed, except as they may assist in keeping track of the
wavefront time values.
Classicor reversedrefractionprofiles
The basic understanding of wavefront interpretation of the typical in-line
reversedrefraction profile can best be obtained by studying the solutionof a simple
example of a three-layer casewith dipping but essentiallyplane refractors having
discrete velocities.
Figure 5a showsthe type of data which is usually recorded in the field. After
correctionfor weathering and, if necessary,to a referenceinterface,a the plotted
t-x curves are studied, refractor changesmarked, and general velocities drawn in.
The velocity lines need not be straight and are primarily for preliminary refractor
identification and assessmentof velocity magnitudes. If reversed traveltimes,
R.T., do not agree, some adjustment of data plots may be required, taking into
a The term "referenceinterface"is perhapsbetter than "referenceplane." It may be usefulto clarify
a few pointswith respectto the useof weatheringand elevationcorrectionsand correctedtime-distance
plots when the interpretationis to be accomplishedby wavefrontmethods.
The essentialquantity usedin the wavefront techniqueis the arrival time (i.e., observedwavefront
value) at a specificlocationin space.This may be at the surfaceor on someother arbitrary plane or
velocity interface.
Becauseof the scaleinvolvedin mostfield work, it will be foundmostconvenientto subtractweather-
ing delay timesfrom observeddata prior to commencing the graphicalinterpretation.Weatheredlayers
are frequentlytoo thin to be convenientlyhandledgraphicallyon the samesmall-scalesectionrequired
for most refraction depth sections.The sameis sometimestrue of other thin near-surfacelow-velocity
layers. Therefore it is usually better to remove arithmetically those rather large near-surfacetime
delays.This may sometimesbe donegraphicallyby makingspeciallarge-scalenear-surfacesectionsbut
in the caseof simpleweathering,sinceraypaths are nearly vertical, the subtractionof delay times is
usually sufficientand introducesno significantdistortionof the horizontalpositionof the wavefront
arrival.
It is not necessary to havea horizontalor evena planesurfaceor referenceplanewhenmakinga wave-
front interpretation.Thus, the base of true weathering(or of the near-surfacelow-velocitycomplex)
may usually be used,howeverirregular it may be. "Elevation corrections"are thereforenot only non-
essentialbut shouldgenerallynot be made. If eachsignificantdiscretevelocity interfaceis viewed as a
reference"plane," the interpreterknowsat oncewhich onesmay be besthandledby removal of delay
times abovean interfaceand which onesleft for interpretationusingwavefronts.The criteria to apply
are whether correctionshifts the data significantlywith respectto the subsurfaceand whether the
remaininglayersare thick enough,at the scaleused,to allow accurateconstructionof wavefronts.
The writer feelsthat it is generallypreferableto plot arrival-timedata, correctedas required, on
conventionaltime-distanceplots,eventhoughthe usualusesof suchplots,suchas calculationof appar-
ent velocities,intercepttimes, delay times,critical distances,etc., do not enter directlyinto wavefront
methods.A time-distanceplot is still a usefulway to visualizethe data as a whole, identify refractors,
checkparallelismand penetration,checkand ascertainreciprocaltimes,and, mostimportant,selectand
combinesegments of t-x data into thecomposite
arrival-timeplotswhichare thebasisof a wavefrontinterpre-
tation.
For thisreason,t-x plotshavebeenincludedin the illustrationsin thispaper.The readershouldrecog-
nize that apparentvelocitiesmarkedthereon,exceptpossiblyoverburdencurves,are of no particular
significance
in the interpretationby wavefronts.
376 Wavefront Methods
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
RP. S2
\ At ml V.103(2</ -- -- TWO-WAY
COVERAG
\, -n- -- NE-WAYCOV....
FIGURE
S(O) FIRST
ARRIVALS
(S
CTA
(s3sAwnj x
arrivals are pickable and may be used to guide extrapolation. Figure 5a showsthe
location of typical auxiliary data obtained at some added shotpointsbetween S
and S..
A shotin the middle (S3)practicallycompletescoverageon V.", but thereis a gap
in the first-arrival coverageof V. (downdip), becauseof its greater depth to the
right. This may be filled if desired, by an additional shot from S4, etc. Unless the
approximate depths and attitudes of the refractors are known in advance and
"shootingon paper" has allowed the seismologistto preplan his coverage,it will
often be necessaryto add additional shotsto avoid or reduce the use of projections.
Thus, the data pertaining to the same refractor are all consolidatedinto a pair
of reversedadjusted time-distancegraphs for each refractor with at least an R.T.
at one end as preciselydeterminedas time and the objectivesof the survey permit.
It is helpful to identify the various portions of this combined t-x curve as being
based on actual reliable first breaks, secondaryevents, values paralleled or phan-
tomed from first or later events,or outright projections,supportedor not by recipro-
cal time or common intercept time data, since these criteria will, in turn, determine
the relative reliability of the various portions of the depth interpretation and of the
true velocity determinations.
One-way data
Certain portionsof the reversedwavefront systemswill be found not to overlap.
However, usinga radius proportionedto the At multiplied by true refractor velocity
(At. V. or in this case .250X10,530 ft/sec plus or minus) and starting from the
extremeintersectionpoints of reverseddata (C' and C"") the refractor coverage
can be extendedby intersectingthe one-way wavefronts to the left and right of the
portion of the refractor already establishedfrom reversedor two-way data.
a Paradoxically,it seems,an arbitrary or erroneousincreasein R. T. raisesthe refractor positionin
the middleof the spreadand vice versa.There can only be oneR. T. exactlycompatiblewith a complete
set of data.
General Wavefront Method 379
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
The secondrefractor
Figure 5b illustrates the wavefront constructionused to map the third layer or
second refractor. Horizon // is transferred to a new section below the same t-x
curve to avoid congestionand confusion.As before, the compositet-x curves have
been adjusted and extended with auxiliary data from auxiliary shotpoints offset
beyond the endsof the profile.
The R.T.a is observedto be 2.225 sec. A systemof wavefront arrivals at integral
times of 1.7, 1.8, 1.9, etc. secis marked on the t-x curve from S and the comple-
mentary systemof R.T. minus n./Xt or n..100 secfrom S., using the extendedpor-
tion of the V3' curve basedon parallel phantomingfrom S data inside the point
xc.3.The complementarywavefrontswill have values1.625, 1.725sec,etc.
Emerging wavefronts are drawn as before, this time with circlets proportional to
n..100 sec 5,000 ft/sec with centersat the ground positionsof the selectedarrival
times. Note that the emergingwaveis travelingin the mediumof velocityV, so5,000
fi/sec is still usedto drawtheemergingwavefronts.
These wavefronts are extended only deep enough to obtain a good intersection
(without projection) with horizon//. These intersectionsnow mark the positions
on // where the correspondingarrival times would have been observedwith the
detectors(but not the shot) locatedon//. (// is thereforea referenceinterface as
discussedabove.)
Now usingradii proportionalto n./xt. V. (assumedto be not greaterthan 10,530
ft/sec, if no more valid vertical velocity is known), "arclets" are drawn around
centersat the intersectionsof the wavefronts with //. Tangents to the "arclets"
are connectedup and joined to their respectivewavefrontsat the interface.If this
systemis extendeddeepenough,somepair of wavefrontsshouldbe found to inter-
sect with a total time of RTa= 2.225. The intersectionwill be a point on the second
refractor//. (seepoint 0', Figure5b). If severalsuchpointsare located,the refractor
attitude is fixed and the true velocity determinedas beforeby dividing the length of
diagonal0'0'"' by the time differencesfor 3./xt. This turns out to be 5,160ft/.300
sec- 17,200 ft/sec- Va true.
Note that if the auxiliary data from S5 had not been available, the wavefront
systemsbasedon first-arrival data from S andS.wouldjust barely haveintersected
at a total time of 2.225 (near 0') and, although the approximate depth to the
refractor would be known, neither its true velocity nor attitude would be deter-
380 Wavefront Methods
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
minable from observeddata. Either the velocity or dip of the V3 layer would have
to be assumedto completethe interpretation.
The usual procedurefor nonoverlappingreversedsegments,if no other data are
available, is to project the wavefronts the least amount possibleto effect intersec-
tion (or project the t-x curve and construct the wavefronts, which is the same
thing) and then assumethe refractor is plane in the vicinity of the intersections.
Somevalue of horizontal V can be found by trial and error (or basedon the angle
2icbetweenwavefronts)which will yield a plane refractoror geologicallyreasonable
dips on the two one-way segmentsof data.
This emphasizesthe importance of securingadequate reversedcoverageon the
principal refractorsif a reliable interpretation is to be obtained.
In the present case,a reliable velocity and attitude of V3 can be obtained using
the auxiliary data and the one-way coverageutilized to extend the //2 horizon to
completion,in the samemanner as was done for//.
Advantages
of thewavefrontmethod
for reversed
refractionprofiles
The advantagesof this techniqueare many but the following stand out'
1. The t-x data are useddirectly and need not be smoothed,convertedto straight-
line velocity segments,or projected over long distances,especially acrossthe blind
zone, to obtain intercept times.
2. The R.T., an observed quantity, is used to fix the depth rather than the in-
tercept time, which is usually a projected value (although it need not be and can
be establishedby phantomingif data are available).
3. The true refractor velocityis determinedby the depth interpretationrather than
assumedin order to make the depth interpretation. (On the other hand, as in all
refractionmethods,the accuracyof the solutiondependsdirectly on the accuracy
of the V or other overburden velocities used.)
4. One-way and incomplete data are handled in a routine manner related to the
interpretation as a whole.
5. The physical reality of the wavefronts helps the interpreter to visualize the
geophysical-geological aspectsof the problem and remain aware of the location,
quality, and sourceof his data, hence,of its relative reliability.
6. The systemdovetailsneatly with other interpretation systemswhich may be
used from time to time, such as the aplanatic and parabolic methods, and even
with delay-time and intercept-time methods. Especially in the solution of com-
plex field examples involving faulting, diffraction, unusual velocity distributions,
incomplete data, etc., some of which will be describedbelow, the method will be
found to be both flexible and compatiblewith other specialtechniques.
to depictthe requiredwavefronts.
A table of h and r valuesfor successive
.100-
secintervalsis usuallydetailedenough.A simplecompasscan be usedto draw
the wavefront circle for successivelydeeper-centeredwavefrontletsat depths h
belowdatum (seeFigure6b). If muchworkneedsto be done,a simpleplasticdevice
suchas is shownin Figure6a canbe madeof two-mmthick vinyl-plasticsheeting.
The calibration of the h and r arms is outward from the pivot and is marked in time,
proportionalto distance
(or depth)givenby the standardh andr formulas. Ordi-
narily,an extendible
compass will do aswellasa draftingarmfor drawingcirclesof
radiusr, whenonlya smallamountof datais to behandled.
REFERENCE PLANE
t= .247
PIVOT
WAVEFRONT . 247
PENCIL HOLE
SPRI
SLIDER'
382
I
Wavefront
I
Methods
General Wavefront Method 383
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
not neatly fit any particular analytic expressionfor increasing velocity can be
handled by wavefronts if the required precisionand available detailed velocity in-
formation justify the effort involved. For example, an unconformablelayer of con-
stant velocity underlain by an acceleratingvelocity could be handled with ease. An
example would be a thick water layer in marine work, deep glacial drift, or simply
an unconformity, such as low-velocity Tertiary section over older rocks. The re-
verse situation is occasionallyencounteredof an acceleratingoverburden (a Gulf
Coasttype section)aboveseveralmassiveolderlayersof near constanthighvelocity
and is alsowithin the practical applicationof the method. The wavefront method, in
other words, readily permits the stripping of the first layer by lowering the wave-
front intersectionsto any desired referenceinterface.
SEVERAL SPECIAL WAVEFRONT TECHNIQUES
SI .5 1.0 ! 14
/,
? ,000
Fro. 7. Aplanatic-curve
methodappliedto salt-domedelineation.
A number of such curves are drawn for traveltimes to various positions in the
well and the simplest geologicallyacceptablecommon tangential surface to these
aplanaticcurvesis a solutionfor the interface.In Figure 7, only the salt wavefronts
used to draw the curve for a time value observed at R4 are shown to avoid confusion.
Curves for two other values, at R and R6, are shown without full construction lines.
If the shot location remainsfixed, only the salt-velocity wavefront need be moved.
For a different shot, the overburden wavefront is recentered on the new shot. As
many shotsas neededto give the desireddetail may be taken. To demonstrate over-
hang, the shotpoint shouldbe relatively far from the well and the well seismometer
positionedat numerousplacesnear the top of the salt plug. Severe overhang may
require placing the shotpoint far enoughaway to employ minimum time raypaths
which curve down into the overburdenand actuallyrise up undertheoverhangbefore
entering the salt.
Excellent velocity control, both in the overburden and in the salt, is necessaryfor
386 Wavefront Methods
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
z
o
'h' SCALE
General
I
o.
o
o
b
IIII I I I-----
Wavefront
Z9' +O9=Z
Method
o
o
-o
o
o
b
387
388 Wavefront Methods
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
the t-x curve within the "blind zone" must be fixed either by pickable secondary
arrivals or by parallel phantomingfrom a more distant shot which will permit the
t-x curve to be reconstructed. Lacking this, the first arrivals beyond the knee,
xd23,
mustbe projected.
Parabolasmay be convenientlyusedfor one-waydata shot in tandem, for "semi-
reversed"data, as describedby Dix (1952), or for one-waydata shotboth waysfrom
the same shotpoint. In this way, the common tangent to two adjacent parabolas
drawn for the samerefractor may be consideredto be the interface and the dip and
true velocityapproximatelyestablishedwithout the necessityof reversedshooting.
This method would work best where parabolaswere not too distantly separatedand
structureis simple.
If the data are suchas to permit the first interface to be established,then an ap-
proach similar to the parabolic method can be used in establishingthe second
interface. No exampleof this is given. The proceduremay be summarizedas follows.
A radiating wavefront system for two discrete layers, (analogousto top two
layers of Figure 2b) is drawn in; of course,only half the wavefront need be drawn.
After drawing circlesin the first medium, Huygen's wavelets may by used in the
second, or a number of refracted rays are drawn in for incident angles less than
ic(sin-V/V2) for the first layer and equal traveltime segmentsmarked off on these,
which are then joined by a smoothcurve. The directedsystemfor Va is drawn in as
before for the two layers as was the casein Figure 5b. The intersectionsof equal
valued wavefronts will yield a parabolalike curve to which the secondrefractor is
tangent.
Interpretationsusingcombined
reflectionand refractiondata
Many refraction surveysare carried out in an area where reflection data already
exist. Sometimesthe existing reflection data are inadequate or inconclusivebecause
they are not goodor continuousenough,correlationsacrossfaults are not clear,posi-
tions of unconformitiesare in doubt, multiple reflectionsare suspected,or "base-
ment" is not clearly established.In such cases,certain lines may be reshot by the
refraction method.
Also, it is commonpractice to follow up a reconnaissance refraction survey with
detailed reflection shooting, often making use of the same recording lines.
Reflection data are often recordedduring the refraction survey itself on the near
spreads.In particular, wheneverrecordingproceduresare employedwhich are de-
signedto obtaingoodsecondaryinformation(and they alwaysshouldbe sodesigned
if there is any chance of obtaining such data) much useful reflection and late-
arriving refraction data may be obtained. The advent of the newer techniques
of magneticrecording,variable densityrepresentations,etc., coupledwith the useof
AGC and reflection filters, practically insuresthat such data may be obtained in
most areas.
UGHS
IFLECTION DATA
OTAINED WHILE
$HOOTING REFRACTION
REFLECTION
HI
175 2.0
! i i
1.75
VI 5,000'S /
-
.20' 5,000'/$
can also be carried out usingwavefrontsbut rays are frequently easier.The analogy
to multilayer refraction constructionis obvious.)
The Va refractor unreversed bow-wave system is then drawn in, using multi-
layer techniquesas shown in Figure 5b, and pertinent text. As before, the intersec-
tion of the reflection dip segmentwith the wavefront system permits a reasonably
reliable true velocity to be calculatedand the refractioninterfacett2(Rx) is extended
as far as one-way data will allow and adjusted up or down for correcttotal refraction
traveltime.
Shootingbetweentwo wells
Figure 10 showstwo wells, a mile apart, one of which has a velocity log. The log
and previousexperiencein the area indicate that two goodrefractors exist with con-
trasting and easily distinguishablevelocitiesand consistentcharacter. The problem
is to determinewhetherfaulting existsbetweenthe wellsor near either of them which
would affect the choiceof a third well site. Two shotpoints,estimated to be properly
placed to securefirst arrivals from the two refractors, V and V3, are selectedand
one-way data only are obtained from spreadsplaced to yield the desiredsubsurface
coverage.Using the apparentvelocity (V2 data), the wavefront system(solid lines of
Figure 10) is drawn in employingan averagevertical velocity given by the well logs
(or from refraction overburdendata if no log exists) of 5000 ft/sec. This systemis
extended to below the first horizon H as identified in the two wells "A" and "B".
The wavefronts bracket the well calls HA and HB. The interpolated times T and
TB of the wavefrontsat thesedepthsare calculatedgraphically. Vm true velocity is
then given by the distanceHxA-- Hz, divided by the time difference, Tx-Tx,. This
velocity is applied as before through intersection of the bow-wave system at ap-
propriate time (and distance)intervalsfrom Hz or/-/x to detail the interfaceH.
This done,the sameprocedureis appliedto V3' data, refractingthesewavefronts
as they pass through H. The horizontal refractor velocity for H turned out to be
12,200 ft/sec. The velocity log indicatesthat a truer vertical averagefor the entire
lower sectionwould be 10,000ft/sec, so this was usedin spacingV3' wavefrontsin
the lower section.
Accordingto the data, the Va refractor velocity is 16,700 ft/sec. A shift of the
General Wavefront Method 393
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
I.S
Usingonewell to establish
reversed
timesfor longoffsetshots
Many times,a singlewildcatwell existsaroundwhichmoreshootingis planned.
When great depthsand unfavorablylow-velocitycontrastsare involved,very long
offsetsmay beneededto registerfirstarrivalson the refractoror refractorsof interest.
This is the casein southwestTexas, for example,where the attenuation of arrivals
394 Wavefront Methods
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
o ,o
i i i [ i i i' i
, ,.... (' 1
/ '
(900 OOOOl (IlgOI
" iii/ _
0 S4TO
B14.
7.I SEC.
. S3I .
SEC. Vt'
FIG.12.Half-refraction
surveys
to, from,or througha well,showing
equivalence
of aplanatic
and intersection
techniques
from take-offpoints(T.O.P.'s)
Thus, the sameT.O.P. A in the well may be used, even if the data were not actually
recordedat or shotfrom this point. The wavefrontsare drawn back from the surface
as was the casein Figures 10 and 11, to establish the time of a wavefront passing
through the well call on the refractor at T.O.P.A.
An alternative method, if no well exists,is to establishthe refractor depth at some
point nearT.O.P. A or in lieu of T.O.P.A. This couldbe doneby shootinga reversed
refraction line in a plane normal to the half-refraction profile. If reflections are
recordable,an expandingspreador velocity profile may be recordedat someplace
along the half-refraction profile where subsurfacerefraction coverage is planned.
The increasein interval velocity obtainedfrom a plot of interval velocitiescalculated
from the velocity profile often will indicate which reflection correspondsto the re-
fracting interface, and will permit a fair estimate of depth to the refractor. In addi-
tion, the V.P. will furnish overburdenvelocity data for usein interpreting the refrac-
tion data; the reflection dips will permit confirmation of true refractor velocity or
confirm the identity of the refractor-reflector.This method was successfullyusedby
the author in Mexico (1958) in connectionwith half-refraction surveysacrossthe
Golden Lane Reef. Of course,there must be at least one good reflection below the
reflector-refractorin order to calculate the interval velocity by means of a velocity
profilesotrue basementcouldnot be identifiedin this way.
InterpretationsA lternativemethods
By whatever means the t-x curves of Figure 12 are finally obtained, and the
T.O.P. A fixed, the interpretation may proceedby either of two good methods' the
aplanatic-curvesor the intersectiontechnique. They are both very practical and
equally reliable.
Where the t-x data are obtained at short detector intervals, for example by
shooting 24 traces at a time from a long-offsetshotpoint, the emerging directed
wavefront system is constructed as heretofore. A radiating wavefront system is
centeredaroundthe T.O.P. (or a compassmay be usedto draw smallarcsof circles)
and the appropriate intersectionswith the bow-wave system made. Thus, for a
particular wavefront 1.6 sec, Figure 12, since the wavefront for t-1.4 sec passes
throughthe T.O.P., an arc of radius .2 sec(1.6--1.4 sec)times 16,000ftfsec is drawn
around T.O.P. A to intersect the directed wavefront 1.6. Other arcs cut successive
wavefronts1.3, 1.5, 1.7, 1.8, 1.9, etc. It soonbecomesapparent that the refractor is
concaveupward which violates the straight-line travel-path assumptionin medium
V2. In this case, each successiveintersection is made from the next previously
establishedboundary wavefront position in short stepsapproximating travel at V2
along the curved interface.
Further drawing of the V horizon showsthat it becomesconcavedownward to-
wards the left as could have been anticipated from the t-x curve. Now the shortest
time path must cut throughor "penetrate" the structure. If this effect is marked, the
intersections should be redrawn using radii from a more distant T.O.P. such as
T.O.P. A or B so that the path in the refractor approximatesthe minimum time
"penetrating" path.
The second method which may be used is the aplanatic-curve method. Three
aplanatic curvesare shownon Figure 12, onefor the top of the structureand onefor
each flank. A radiating overburden wavefront is drawn or centered around the
surfacepositionof the recorded(or interpolated) time; segmentsof arcs are drawn
in, or a radiating refractionwavefront chart of the refractor velocity is usedas be-
fore to intersect several of these overburden wavefronts to give the locus of all
possiblesolutions totaling the observed time, t. In this case, it is convenient to
label the overburden curves backward from the origin instead of outward from it
and the refractor arcsoutward from the T.O.P. (seeconstructionfor curve t equals
3.2 of Figure 12). At the intersectionsforming the aplanatic curve, the two sets
have the same time value. The same limitations on the choice of T.O.P. locations
apply with respectto a stronglyupcurvedrefractor and/or penetratingrays as do
for the "intersectionmethod" (i.e., the minimum traveltime path shouldfollow the
interface for concaveupward refractor horizonsand "penetrate" when this is indi-
cated; therefore, T.O.P.'s should be selectedaccordingly). Thus, curve t-2.5 sec
shouldbe constructedfrom T.O.P. B as also curve t-3.2 sec, but t-2.7 sec should
employ T.O.P.C.
Notice that, in this application,only the recumbentlower limb of the aplanatic
curve is drawn, sincethe refractor lies beneathit, whereasin the salt domeof Figure
7, the apexor upper branchis the significantpart.
Notice also the overburden wavefronts are tangent to the full directed bow
waves usedin the other constructionand that the point of common tangency will
coincidealmost exactly with the point given by the intersectionmethod. Thus, the
two systemsare equivalent. For isolated data and widely separatedshotpoints,the
aplanatic method is better as data would be insufficient to establisha reliable bow-
wave system. In this case, the actual values of the times at the shotpoints are
used instead of interpolated times from a curve. For more closelycontrolled or de-
tailed t-x data, interpolatedpoints will be accurateenough.
Isolated aplanatic curvesunder offsetshotpoints
Many times data are recordedfrom isolatedoffsetshotpointswell beyond the area
of immediate interest or where the continuoussubsurfacecoverageis obtained. For
example,the V.'data of Figure 10 do not afford coverageback to the shotpoint
In Figure 5b, without the auxiliary coverageaffordedby data from S3and S4,the V.
refractor would not have beenmappedat eachend of the line
A very good single-depthdetermination, however,can be made under these out-
lying points by drawing an aplanatic curve from the last controlled point on the
refractor or from some other position where a timed wavefront is established.For
example,in Figure 10, the wavefront .820 secpassesthrough// in the well. Using
this time and the true V velocity an aplanatic curve can be constructedunder
showingthat the interface dips down to the left and should not be projected far
from the last controlpoints of the refractiondata.
General Wavefront Method 399
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
In Figure 5a, an aplanatic curve could be drawn in using a wavefront time of 1.0
secfrom C or of .750 secfrom C to provide a depth under &. In Figure 11, depths
under the long-offsetshotsS3and S4couldbe estimatedquite closely.This otherwise
wasted data can be utilized by the aplanatic method and might be the clue to another
structure or a better interpretation. The depths at long-offset shotpoints, for ex-
ample,might or might not agreewith previouslymappeddata, nearby well data, etc.
Is there an unsuspectedhigh in the area or is somethingwrong with the interpreta-
tion? Isolated depth points may answerthesequestionsor at least suggestthem for
investigation.
wave beyond the right-hand fault, and certain minor diffraction patterns, in order
not to complicate excessivelyan already complex diagram.
The shotfrom S2generatessimilar waves(solidwavefrontsof Figure 14a). In this
case,sincethe bow wave strikes the sheernear-vertical fault plane, DE, outside the
critical angle, total reflection takes place along this plane giving a reflectedrefrac-
tion in contrastto the 90-degree(critical-angle)refractionobtainedwhen shooting
from S into fault BC. Alsoshownis the directly penetratingwave leavingthe far side
of the horst beyond BC, which, however, does not reach the surface before it is
overtaken by the bow wavesfrom AB and diffractions from the left-hand upper
corner of the fault block, C. The bow waves from A B are, strictly speaking,not
bow waves (i.e., at critical angleic) becausea minor amount of penetrationtakes
place,but they gradually approachcritical-anglerefractionwith increasingdistance.
The overburdenarrivals V would precludethe picking of many of thesearriving
wavessincethey are not first arrivals. However, by usingoffset shotpointssuch as
SPs, much of the t-x curve could be observed as first arrivals. The reflected-
refraction can often be picked out of the secondariesin actual practice, since it
moves out in the opposite direction. The diffractions start strongly but weaken
with horizontal distancefrom the point sources.The amplitude of the penetrating
wave from shotpoint S may be expectedto be greater than that from shotpointS.
where total reflection from fault DE means that the remaining energy sourceis
largely reducedto diffracted wavesrounding the cornerat/.
We shall assumethat through the use of offset shots and secondaries,the curves
from & and S.can be reconstructed.We may note that near-side diffractions above
C and D indicate the location of thesepoints, which fall directly below the minimum
arrival time. A different combination of dips, depths, and velocities, might show
more or lessdiffraction data but never the complete curve, which in all probability
could not be picked with confidenceexcept where it was a first arrival.
ShotpointsSa and S4,locatedabove the horst,would yield traveltime curves as
shown with very little V. data, too little for reliable interpretation, and without
overlapping reversed coverage.If these shotpoints are moved apart, they "drop
off the block," and penetrationoccurs.The maximum separationof Saand &, with-
out penetration,would be in shootingfrom Sa' and &' whichwouldonly increase
first arrival coverageslightly. Though probably not useablein practice, unlessex-
traordinarily good secondarieswere obtained, these Sa and S4 t-x curvesillustrate
the true updip and downdip apparent velocities acrossthe upthrown block. These
velocitiesof about 10,800ft/sec and 13,800ft/sec are projectedup to the "penetrat-
ing-wave" curves to illustrate that both of these show higher apparent velocities.
If a fault block is too narrow for its depth to allow a goodpenetration-freesegment
of high velocity to be observed,it is often possibleto shoot another profile aligned
alongthe long directionof the block, obtainingin this way t-x curvesfree of pene-
tration effectsfrom which true depth and true velocity may be calculated.
Figure 14b showsthe resultsof a standard wavefront interpretation of the prob-
ably pickableeventsfrom Figure 14a. Note that the coveragefrom downblocksis
limited by a shadowzone of diffractions.If those are weak or attenuate rapidly,
somedownfaultedblock refractionsmay be pickable after the diffractionarrivals,
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
I
I
I
I
I
/
I
I
I
I
I
I
I
I
I
I
I
I
/
I
I
//
/
/
/
I
I
I
Io
I
I
I
I
General Wavefront Method
I
I
i
o
o
'o
40
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
404
/
/
*
Wavefront
Methods
J
.-
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
General
Wavefront
Method
405
406 Wavefront Methods
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
The emergingwavefrontsalready drawn from the full t-x data (Figure 14b) now
intersect the adjusted horizonCD at points I, 12,Ia, etc. of Figure 14d.
Since these are high-velocity diffracted arrivals with a sourceat the toe of the
fault block points B and E, a similar treatment to that of the low-velocity diffrac-
tionsmay l:e used,only in this case,radiating W.F.'s basedon V2-12,000 ft/sec are
used.One suchhyperbolais drawn for W.F.'s from S2. It is a branchof an hyperbola
with focus at I and 12. Others could be drawn.
On the other hand, the time from C0 to 12, for example, is quite well known since
the W.F. value and position on the refractor is establishedat these points. This
time, alongminimum time path CoEI2,includesE. Therefore,the locusof all possible
positionsof E must be an ellipsewhere the sum of the distancesCoEand EI2 is equal
to t(Co to 12). V2. This ellipseis easily drawn by usingradiating W.F.'s around 12
(sameoneas usedfor hyperbola)and C0and joining the pointsof intersectionwhere
W.F. times total this traveltime. The figuresin parenthesismark time back from I.
so equal-valuedwavefrontsintersecton the ellipse.Note that the ellipseand hyper-
bola both intersect at E. Other ellipsessuchasfor the time from C0to I could be con-
structed, which should give a corroborationof the location of E. The entire lower
portion of ellipseis drawn with constructionlines to show its relationship to C0 and
12 asfoci. In actual practice, only a short segment,such as the solid portion, is re-
quired and a few intersectionsof circular arcsestablishthis.
On the left side of Figure 14d, the ellipse for time Coto Ia is shown without con-
struction lines. B must lie along this curve.
If reflectedrefractionsor 90-degreerefractions(critical-anglerefractions)originat-
ing from the fault plane itself are observed,they will appearas straight-linesegments
with high or negativevelocitiesdependingon the hade of the fault (see t-x curve,
Figure 14a).
The reflected refraction wavefronts are drawn as any other with values labelled in
decreasingtime and the intersectionwith the direct-arriving refracted wave system
previously establishedis noted. Both these are directed W.F. systems of nearly
plane wavesso they intersectin a line. The intersectiontakes place at the fault plane
where both systemsof coursehave the same time value. The line of intersections,
DE givesthe trace of the fault. If a later phaseof the reflectedevent (a secondary)
is picked,the dip of fault will be correct(D'E ) but it will be displacedtowardsthe
shotpoint(S.).Usingthis dip and point D, E can be established.
Possibly secondarydirect refractions in the shadow zone can be used to extend
CoC8or C9Cn towards E and B. Since diffractions weaken with distance, identifica-
tion of the secondarydirect refraction is quite often feasibleespeciallywhen faults
are large. As a last resort CoCs,etc., may be projected to locate E, etc.
When the fault hade is large the refraction is not reflectedbut is produced,as a
bow wave, from the fault plane itself. If the event is identifiable thesecorresponding
wavefrontsmay be reconstructedat the approximate depth of B and intersectionat
common times with the direct bow wave from & established. If the equivalent
first arrival time is used, the line of intersectionswill pass through B and the dip
of the fault plane may be establishedat the angleic to the W.F. system(seeleft side
of Figure 14d). Only a small portion of a reflectedor fault-plane refraction event,
therefore,may be quite significantif recognizedand used.
408 Wavefront Methods
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
6,000s CALCULATION
OFDEPTHPOINT
..........
/f/ / , /?/ 7500's TIME =1.600 //
0
COS
SIN-I
/-I.l.t
x
tVV 7500=9820
////X
7 ' ' TIME- 640
-0'-I./'
I
./
I0,000DEPTH
ERROR
=180'=
ERROR
2'/.
INOFFSET
//
= 920'=14%
// / / ./
,ooo /7' I -
o.ooo // / ,.s,.o=)='V
//
..'
/ /
,'!7 .....
.......//Z ....... . ----.
F....
o
o
5000'
'' ," / / I
_SJ /' /' ./ / I
,o _, ,",C, Z-
o
,000 o
data, which can subsequentlybe smoothed out for field use, is the best procedure.
(For delay-time methods, the equivalent procedureis to calculate layer by layer
delay-time versusdepth curves to use instead of time-depth curves or carry out a
microintegration of CVL's for variously slanted raypaths in order to construct true
overburdenwavefront charts from the integrated offsetand depth increments.)
Refractionarrivals
Refraction arrivals, since their path is largely horizontal, have the seriousdis-
advantage for overburden-velocitydeterminations of selective travel along the
higher speedbeds. Also, velocity inversionsdo not show on the first-arrival curves
but do contribute delay time which is misinterpreted as excessdepth to the next
lower refractor. Where velocities increase,the penetration of the overburden ray-
paths is not sufficientto samplethe entire overburdenvelocity field before the bow
waves from the higher velocity refractor or refractors take over as first arrivals.
For this reason, overburden functions should be computed on off-structure lines
where a deeperpenetration is possible.Even then, they will be incorrectwhen used
on structure. Figure 8 showsthe maximum depth of penetration for the last pickable
overburdenarrival (at xa) for a linear-with-depthvelocity field. The analogouscase
General Wavefront Method 411
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
Di.17ractions
Diffractions serve more to check overburden velocities than to furnish them. If the
overburden function in use is correct, diffractions of good quality will often con-
verge on a point sourcewhen interpreted by the wavefront method. If the source
can be localized,throughother means(well data, for example),or if shot-enddelay
time and the proper horizontal traveltime can be removed from total observeddif-
fraction times, then the residualtimes constitutethe same type of data as a velocity
profile and an x2-t2plot shouldyield averagevelocity and depth. This may require a
successive approximationapproach.
Wide anglereflections
Wide angle reflections,if observed,can often furnish a grossestimate of average
overburdenvelocity. It is sometimespossible,usingdata from severalshotpoints,to
determine quite closely the traveltime of the critical ray, to, where the wide-angle
reflectionbegins(seeFigure 1 and 16) and its distancexc. At this point, the reflec-
tion and refraction t-x curves are tangent. These data, togetherwith the intercept
time, to,permit an averagevelocityto the refractorto be computedwhichis probably
more accurate than can be obtained from refraction data alone. The formula is
simpleand similar in form to the velocity profile formula:
Velocity profiles
Velocity profiles or expandingreflection spreadsare excellent sourcesof over-
burden data for refraction use when reflections can be obtained in the area. Besides
providingoverburdenvelocity data, they often assistin identifying refractorsand
checkingtrue horizontal refractor velocity values by providing independent dip
information, as was discussedabove under "Combined Reflection-Refraction Inter-
412 Wavefront Methods
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
- =ll-Xc?'-
V c;_tc.
to
PARALLEL
PHANTOM
to REFLECTION
Xc
SHOT POINT SHOT POINT
B A
The systemsmay be directedor radiating and may involve the same or different
velocity fields,but there are nearly always two and never more than two systems,
one centeredaround an actual (or fictitioussource)and the other arounda receiver,
or receivers(or a fictitious line of receiversin the form of wavefront arrival times at
positionsalonga knownsurfaceor interface).
While our aim hasbeento instruct in the useof thesewavefront techniques,more
than to attempt a formal classification of them, it appearsthat, viewedin this light,
they do fall naturally into a generaland logical system.This system,which we may
call the General Wavefront Method, is summarizedin Plate I and may assistthe
reader in recalling the essentialpoints and areas of application of each separate
technique.
It may be of more than academicinterest to note that the various end results
achievedby the methodsare curvesof familiar types--points,lines,circles,ellipses,
parabolas,hyperbolas;theseare summarizeddiagrammaticallyon Plate I.
Any problemwhich would requirethe useof more than two wavefront systems,or
of two velocitymedia,is first reducedto the simplercasebeforea solutionis possible.
These multilayer problemsare easily handledby drawing the wavefront systemsin
the first layer to intersection with the first interface, (which must be known or
previouslyestablished),
thento intersection
with the next established
interface,and
so on, until the conditionthat all remainingtraveltime can be divided betweenonly
two wavefront systemsis fulfilled. The analogybetween"layer stripping" or reduc-
tion of wavefronts to a lower interface and establishingof a take-off-point,T.O.P.,
lies in the fact that both are to reducethe problem to its simplestelementsinvolv-
ing traveltimes expressableby only two wavefront systems.
The flexibility offered by the ability to combinefreely any or all of these tech-
niquesin handling a singleinterpretation problemis truly remarkable and will be
appreciatedby thosewho have tried to handlenonstandard,nontextbookfield data
by many of the other refractionmethods.The essenceof this flexibility is the fact
that all the pointson the depth sectionare locatedin bothtimeand spacemakingit
possibleto changesmoothlyfrom one techniqueto another as the conditionsor
type of data require.
REFERENCES
o
o
o
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
FREDERICK L. SCHENCK*
ABSTRACT
INTRODUCTION
4.500
00'
4.000
Wavefront Targeting
>"
../.o o
// #ii
417
418 Wavefront Methods
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
and added to the other. Thus, the centersof the circlesat the bottom of Figure 1,
representingthe fourteenth and eighteenthtarget circles,determinepoints defined
by Thornburgh's principle. The centersof a whole seriesof such circles gives the
complementarysolution.
The target is maneuveredsothat eachsuccessive circleis properlytangentto both
wavefronts, and the location of the center of the target is marked on the cross
sectioneachtime. The resultis a seriesof pointsspacedat equal traveltime intervals
alongthe refractinginterface.The processis both rapid and accurate.The comple-
mentary targeting solutiongivesan approximatepicture of the local relief on the
refractor. The complementarysolutionis independentof the refractor velocity: No
refractor velocity evaluation is required for achieving the solution.
REFRACTOR VELOCITIES
The top of Figure 1 shows that three different velocity zones make up the re-
fractor. These zonesare recognizedin the complementarytargeting solution at the
bottom of Figure 1 by the spacingof the targeted points at equal time intervals.
Simple algebra,applied to time and distancemeasurementsalong the shortestend-
to-end travel path (the minimum trajectory), determinesthe boundariesof the
velocity zonesand the velocitieswhich obtain.
THE COMPLEMENTARY PRINCIPLE
Now can be stated the first principlementionedin the introductionof this paper'
The refractorvelocityvalues,determinedfrom time and distancemeasurements alongthe
minimum trajectoryof the complementary solution,are always correctregardlessof
erroneous valuesusedfor the overburdenvelocity.This important statement is readily
verified from complementary solutions achieved by deliberately applying incor-
rect overburden velocities.
The principle can be deducedby simplelogic. For example,if we selectan over-
burden velocity that is too slow, the two complementary wavefronts will be shifted
away from their respectiveshotpoints.This resultsin the complementarysolution
being stretchedout proportionately to the velocity error. The AT of the target is
alsoincreasedin like proportion, and the number of AT intervals along the minimum
trajectory is unchanged.Similarly, if the overburdenvelocity is too fast, the wave-
fronts will be shifted toward their shotpoints,and the minimum trajectory of the
complementarysolution will be compressedin proportion to the velocity error.
Since the AT of the target also is proportionately decreased,the velocity measure-
ment along the minimum trajectory is constant.
In casesinvolving severe relief, erroneousrefractor-velocity determinations can
be made through incorrect interpretation of the route of the minimum trajectory
on the complementarysolution. Fortunately, as will presently be seen,such errors
are readily detectedand corrected.
The complementarysolutionis erroneouswhen achievedby using an incorrect
overburdenvelocity. The criteria and a method for establishingthe correct over-
burden velocity will now be developed.
420 Wavefront Methods
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
ONE-WAY TARGETING
WavefrontTargeting
421
422 Wavefront Methods
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
xxx
.oo
/
/
FIO.5.Criteria
fordiagnosing
velocity
errors
frommismatch
ofthetwoone-way
solutions.
eachof thetwodirections
of energytravel'If correct
velocity
values
areused,the
twoone-way
solutions
coincide,
andthisproves
thecorrectness
ofthesolution.
Con-
versely,
if anincorrect
velocity
isused,
thetwoone-way
solutions
willnotcoincide,
and error is established.
Figure
5sketches
thecriteria
fordiagnosing
velocity
error
asestablished
byfailure
ofthetwoone-waysolutions
to coincide.
Refractor
velocity
errorcauses
thetwo
minimumtrajectories
torotateoutofcoincidence.
Overburden
velocity
errorcauses
thetwosolutions
to shiftlaterallyoutof coincidence.
Thedirection
of rotationor
shiftindicatesthe natureof the error,as shownin Figure5.
CORRECTING VELOCITY ERRORS
Whenanerroneous
refractor
velocity
is used,theminimum
trajectories
of the
twoone-way
solutions
arerotated
awayfromeach
otherassketched
ontheleftof
Figure
6.Letebethevertical
angle
included
between
thetwominimumtrajectories,
andletthesense
ofebenegative
when
therefractor
velocity
V.'used
insolving
the
-- GEOPHONE
ARRAY
Fro.6.Left:Diagram
illustrating
formula
fordetermining
thecorrect
refractor
velocity
fromtwo
rotated
one-waysolutions.
Inpractice,
notravel
path
construction
isrequired,
andonly
angle
eneedbe
measured.
Byconvention,
eisillustrated
here
asanegative
number.
Right:
Construction
fordetermining
thecorrect
overburden
velocityfromtwoshiftedone-way
solutions.
424 Wavefront Methods
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
Thornburgh,
H. R., 1930,Wavefrontdiagrams
in seismic
interpretation:
Bull. A.A.P.G.,v. 14,p. 185-
200.
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
BY REFRACTION METHODS
ABSTRACT
Flank wells are used for further refraction shootingwhich yield more accuratevelocity information
and more detailedprofilesof the domalmass.A map from this integratedinformationpermitsexploita-
tion at a minimum risk, even thoughevery locationis essentiallya wildcat.
INTRODUCTION
Earlier papers by McCollurn and LaRue (1931) and Gardner (1949) have de-
scribed various methods for delineating salt domes. The method described here
utilizes many of the ideas developedin those papers as well as new techniquesto
give an integratedmappingprogram.The salt and shalemassesusedare hypothetical
examples encompassingmany of the special features that the authors have en-
counteredwhile interpreting numerousanomalousfeatures. The examplesinclude
specialproblemsinherent to water-coveredareas.
This paper developsmethodsfor mapping massesconsistingof five steps,three in
phaseone, and two in phasetwo. The massesmay be definedas high or low velocity
material existingat an angle with the normal depositionalsection.
In phaseone, called the Exploration Phase,the first step is to locate domal masses
by reconnaissance reflection shooting. The secondstep is to shoot two refraction
lines to determinethe depth to, and the location of, the top of the massif it is salt. In
the third step, longer and more detailed surface-to-surfacerefraction lines are shot,
and locationsfor at least three flank wells are made on the map drawn from these
profiles.
In phasetwo, called the Development Phase,the first step is to shootsalt proxim-
ity profiles to determine the distance between the flank wells and the salt-sediment
interface if salt is present.For the secondstep, profilesof the sideoppositethe flank
well are obtained which give more detail and yield greater depth control where the
anomalousmassis salt. Where salt-shalemassesare present,flank wellsmay be used
to obtain someinformation about the amount of shale on the oppositeside of the
dome.If a flank well encountersshale,it may be deepenedto allow a locally detailed
outline of the shale.In the caseof the shalemassalone,deepperiferal wellsmay be of
EXPLORATION
Reflection
Salt or shalemassesare originally located by reconnaissance
geophysicalmethods.
The reflectionseismicmethod is best for this purpose.Figure 1 showsthe massde-
noted by an absenceor cutout of reflections.Even though a detailed reflectionpro-
gram couldbe used,this negativeindicationwouldstill not be asaccurateasthe map
obtained from the refraction seismic method.
Refraction
In addition to good instrumentation and accurate time measurement,two other
quantitiesare necessaryfor domal massrefraction work. They are: (1) the distance
betweenthe shotand phoneand (2) the velocitiesof the massand the sedimentsout-
side the mass.
Distance
.
o
o
o
o
Outlining of Salt and Shale Masses 429
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
WATER VELOCITY
DETERMINATION
OF WATER VELOCITY
=: TEMPERATURE 67 F
DIFFERENCE
Velocity
The velocity of the salt is usuallyquite uniform and about 15,000ft/sec. However,
it has beenfound to vary as much as four percent from this value.
Shalemassvelocity varies with depth, age,and compactionbut is usuallybetween
6,500 and 8,500ft/sec and may be asshownin Figure 4.
Accurate control of sedimentaryvelocity is necessarysincevelocity varies with
depth and sincethere is usually a large ratio of sedimentary travel to domal mass
travel. A vertical wavefront chart may be calculated from (1), the best vertical
velocity information available in the area (well velocity survey or expandingreflec-
tion spread (Musgrave 1962), (2) a horizontal velocity survey as describedlater,
which may be fitted by a mathematicalrelationshipand extendedin depth,and (3)
a surface-to-surfacerefraction line, sometimescalled a normal line. Since the third
method can be shot at the time of the reconnaissancerefraction lines, it will be de-
scribedin somedetail (Figure 5).
A surface-to-surfacerefraction line, showneast of the dome, is shot in a regionof
normal sedimentarysectionin the area near the dome. A plot, Figure 6, is made of
first-break times versusdistance.For long lines, diagonalpaper (suchas shownbe-
tween the dashedlines in Figure 6) may be made in order to maintain plotting ac-
curacywithin a reasonablespace.As first breaksbecometoo weak to pick, first peaks
or troughsare used.This is justified on the basisthat the phasevelocity in the first
wavelet remainsnearly constant.Lines are drawn through this data with the slopes
changingwhere necessaryto obtain a best fit. A continuousfirst-breakline (solid
line) is derivedusingtheseslopes(Figure 7). This line shouldfit the first breaksas
well as possibleand shouldparallel the slopesof later events,thus being representa-
tive of the first breakshad they been obtained.
From this "break line," a set of constants(Figure 8) is derived describingthe
vertical velocity by use of the cycloidal raypath method (Musgrave, 1952). The
constantsinvolve an initial velocity, an acceleration,and a final velocity or depth
for eachlayer. On the velocity versusvertical-time plot, the instantaneousvelocity
is representedby straight-linesegments.The velocity in eachlayer may either vary
linearly with time or be constant.However, the average-velocitylines after the first
layer are slightly curved. With theseconstantsa wavefront chart is computedand
plotted by an electroniccomputer.Up to 40 layers can be usedin the programwhich
has beendevelopedfor the computer.The wavefront chart (Figure 9) will fit the ob-
Outlining of Salt and Shale Masses 431
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
, ( I = NORMAL
INTERBEDDED SHALE)
I
+-SHALE MASS
i i
i
4 i i
i
"
ml
m !
mm ,m
m!
I m m
n
IIIm m
miI
6
Z
,
mmmmmmmi
m'mm m !
, )mmmm
':' mm!
III iI!
I
8 II1II I I
m
m+'l
'11
I0 14'1III
o
o
o
o
o
o
o
o
hj
o
0 Zo
0 /0
/ 0 _JO
oo oooooo <o
o / .. ;;o
-o-' ,o
o c) O
o Z o
. o o
o
o
o
o
o
,5ooo o
o
- .
-6.0 --"'
ISTPEAK
/
-2.0 /
IST
TRO
-I' '
/
16 20 24
I I i
DISTANCE
8
i
- K ILOFEET
3 I 3 4 44 4e
i I 2 56 C
i
VELOCITY FT./SEC.
VELOCITY
5000 7000 9000 I 1000 13000 DEPTH CONSTANTS
0 0 -- ------- "-"
I I I I I I V-- 5250
A=5400
.5
z V=8272 %
o _
V=8272 A
_
-5000
i _ A = 3480
_
V=9838
. i.o _
V-8000
A=O
V- 8000
V=10150'/
- -I0000
_
:15 _
z
_
A=4200 ,
o - -15000
2.0 - -20000
TVo
sinc[.o
= AX
sin 00AX
Vo AX AT:
Vo
s,nCo__S,n
o:,_sn
P: Vo v, - v.
AT
P= "X
&T= I000 p
V, 1 I
5 IO 15 20 25 :50 35
20
Fro. 11. Wavefront chart showingboth wavefrontsand raypaths--note the lossof the raypath whose
AT=0.100 by total reflectionat the velocity interface.
436 Wavefront Methods
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
wavefront chart should represent all the information available on the sedimentary
velocity.
Reconnaissance
refraction
The next step in developing a picture of the anomalousmassis to shoot two sur-
face-to-surfacerefraction lines (Figure 12), crossingeachother at right angles.Their
intersectionshouldbe asnear the centerof the domeaspossiblefrom the information
available. The purposeof theselinesis to determine whether the domeis salt or shale
and to locate the center of the dome more accurately and to determine the approxi-
mate depth of the top of the dome. For this shooting,two boats are used.The re-
cordingboat tows a 2,000-ft long or longermultiphonefloating cable,while the shoot-
ing boat occupiesshotpoints(spacedfor the length of the cable) along the line of
survey.
The derived first-break lines are plotted (Figure 13) for the two refraction lines in
the same manner as the normal line describedearlier. For future crosssection plot-
ting, times are recordedfrom theselinesfor eachshotpointdistance.Also, times and
distancesare recordedfor changesof slopewhich occurbetweenshotpoints.If a sub-
stantial lead is establishedover the normal time (for example,if the time is short by
0.050 sec) then the dome is a salt dome.
Procedure
for mappinga saltdomebysurface-to-surface
refraction
The approximate depth to salt may be determined by the two-layer refraction
method as indicated in Figure 13. The dip may be determinedby the deviation of
Outlining of Salt and Shale Masses 437
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
./ To roY,
0.339x 5ooox_,5ooo
2 Vi5000_50002
9020
,/
/ /
Fro. 13.Diagonalplot of two shortrefractionlinesacross
the top of a saltdome.
theapparent
velocityfromthatof saltvelocity(approximately
15,000ft/sec),and
corrections
canbe madefor depthandpositioncorresponding to the amountof dip.
A preferable
method(Figure14) of determining thedepthto saltis by useof the
wavefrontchart, and the time and distanceto the salt break. One-halfthe time is
markedon a transparenttime-distance scalecalibratedat salt velocity.With the
striphorizontal,
thistimeis movedverticallydowna lineequalto one-halfthedis-
tanceto the salt break until a time match is obtainedbetweenthe strip and the wave-
frontchartat theproper/XT.The.depthto thesaltin thiscaseis 1,065ft.
With the depthto saltdetermined, andthe time versusdistancedata previously
recorded,verticalsections areplottedfor thefirsttworefractionlines.
A pieceof transparent paperthathasbeenblocked in 1,000-ftsquares at thescale
of the chart is set over the chart, and the startingpoint is set as describedabove.
Assuming that theprofilehasbeenplottedout to 6W (Figure15),thetime1.537sec
is observedfrom 6W to 6E. The average/XTat 6W is 0.149whichestablishes a pivot
pointonthewestshoulder
of thesaltwitha sedimentary
timeof 0.535.A saltstrip
issetwitha timeof 1.002at thispivotandthelocus
6W-6Eisplotted
fromtime
coincidencebetweenthe saltstripandthe wavefrontchartlocatedat 6E. The time
from6E to 8W is1.820sec(Figure16).The T at 6E is 0.096with a sedimentary
time
of0.530;therefore,
a timeof 1.290issetonthesaltstripat thepivotontheeastside.
The locus 6E-8W is then drawn as before with a wavefront chart at 8W. It can be
assumed at all timesthat an exit or entry point on the salt belowthe samesurface
locationremainsconstantaslongastheshotpointor receiverontheotherendis not
movedmorethan a shortdistance.If the cablelengthis toolongthencorrections can
458 Wavefront Methods
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
I000
I
/2DISTANCE
TO
SALT
BREAK
DISTANCE-FEET
2000
I
:5000
I
3000'
.067 AT
4000'
,,,I
/
/
I
/-
I'
/
/
/
/
Outlining
0
u,,I
d
I
I-
bJ
I
i.
0
,000
of Salt and Shale Masses
439
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
44O
bJ
//
//
//
//:
//
//
//
//
Wavefront
0
I--
LI.!
1:3
I.iJ
Methods
0
OCli
Outlining of Salt and Shale Masses 441
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
e' o o
o o o o
o o o
o o oo oo o o oo o o oo o oo oo oo o o ooo oo o
o o
o
o o
o o o o o
o
o o o o
o
o o o o o
o o o o o o
o o
o o
o o o o
fflo o
o v
0u 0 o
n "
o o
o o 0 -
o o o
o
o/ o o oo o, ..-,
o ooo o o o o oo
o o o O%o o o
t ooo c o o o
<zo o oooooo,oOO
o o o! ooooo o
'o
O\%o2 o
o
o -
oOOfo
oo o n
o u o oo o
o o o o
o 0 o o
o
o o o o o o o
o o o 0 o o o
o o o o
o o o o o,
o o o o o o
o o
o o o o o o
o o : o o o
o o
o o c o o o
0o o , o o_1
o
o o
o o
o o
o o
o o o o
n o o
o
_l
o
o
o
442 Wavefront Methods
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
Procedure
for mappinga shalemassbysurface-to-surface
refraction
The refraction seismicmethod of mapping shalemassesis similar to the technique
used for mapping salt domes. However, the mapping of shale massesdiffers very
greatly in onerespect'shalevelocityis lower than that of the surroundingsediments
while salt velocity is higher. An analogywhich might illustrate the point here is to
think of the salt as a conductivebody in an electrostaticexperiment where all the
current lines collectand go through the conductingbody, and to think of the shale
as a nonconductivebody in which the current lines spreadand go around it. It may
be apparentthat it is moredifficult to map a shalebody with refractionmethodsthan
to map a salt body.
In mappinga shalebody from the surface,only the upper limits may be mapped
unlessthe sidesare nearly vertical.
In an area where a large tabular body of shale is encountered,whosevertical
-- 15000'
VELOCITY LINE
INTERVALVELOCITY-THOUSANDS
OF FEET/SECOND
5 6 7 8 9 I0
.2.
'i I I i
.4
.6
.8
.
. .__ .....
2.0 '
Velocity
As the A-! well is drilled, a continuousvelocity log is run to total depth before
eachcasingstringis set. If no shalemassis penetrated,a velocityline (Figure 18) is
444 Wavefront Methods
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
DISTANCE-THOUSANDS OF FEET
I0 20 30 40 50 60
8.0
I
w6.0
4.0
2.0
4 ,ooo,
o
RAYPATHS/1000'
/
IN SEC.
IN SEC.
shot away from the dome at distancesof 1,000, 3,000, 5,000 ft and at 5,000 ft, inter-
vals thereafter out to 40,000 ft, at approximateright anglesto the strike of the salt.
Shots are recordedwith a phone at 1,000 ft intervals to total depth in the well from
each of the above distances. This processmay be greatly simplified with multi-
level phones.
A plot of timesversushorizontaldistance(Figure 21), determinedby water breaks,
is made for every level. The distancesto even 0.1 sec are read on each curve, and
thesedata are usedto constructan observedwavefrontchart (Figure 22).
If a shalemassis penetrated of sufficientmagnitude to yield a velocity appreciably
lower than expectedfrom the surroundingsediments,the velocity line may be used
to map the contact betweenthe low velocity shaleand the normal velocity section.
A wavefront chart such as shown in Figures 11 or 22 may be used for the sand-
shale section outside and a shale strip inside the shale mass in the same manner as
they are usedfor the mapping of salt domes.All of the loci of least time paths which
solve the time-distance relationship now occur inside of the shale mass.When all of
the loci have been drawn along the profile, a tangent to theseloci will approximate
the shapeof the shalemass.
Salt-proximity profiles
At the sametime that the velocity line is shot, two salt-proximitypoints (Figure
18) are shot into phonesevery 500 ft down the hole to total depth from locations
Outlining of Salt and Shale Masses 445
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
DISTANCE - KILOFEET
0 4 8 12 16 20 24 28
-4000
:8000
--
-,ooo
FiG.22. Observed
wavefrontchartconstructed
by joiningpointsof equaltimeasreadfromFigure21.
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
446
bJ
b
0
o
b
0
o
_..
Wavefront
b
0
o
Methods
.L3- H.La3Q
bj
i.-
I--0
,
C)
b
0
o
b
0
'
o
o
o
b
0
o
--
Outlining of Salt and Shale Masses 44?
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
near the top of the dome. One salt proximity or "S.P." point is at the center of the
surface-to-surface lines,and the other near the far shoulder.A tangent (Figure 23) is
drawn to the loci plotted from these data. This profile showsthe proximity of the
salt to the flank well in greater detail than any of the other shooting.Unlessthe well
was drilled into salt, the upper sedimentarytime and salt outline is usedfrom the
surface-to-surfacerefraction shooting.The loci (Figure 23) are obtained by pivoting
the salt strip about a pivot at the top of the salt and finding coincidenttimes on
specialwavefront charts which are made for eachlevel at which phonesare located.
The velocity layering is determined from the acousticlog and the shootingon the
15OOO' A-2'--
B
A-2-A
A-2-C'"""
VELOCITY LINE
velocity line. Small wavefront charts are then calculated above and below each
phonelevel with an electroniccomputer.From the salt proximity profile an offset
bottom hole location, A-2, is recommendedwhich will explore the sedimentscloser
to the salt, in this caseunder the overhang.Salt will be enteredat a vertical depth
of about 11,000ft (Figure 26). It will penetrate about 1,500 ft of salt for usein shoot-
ing a fan of lines(Figure 24) acrossthe dome.
If a thin shale layer or shale sheath existsbetween the well and the salt it is diffi-
cult to detect. It can best be detectedif SP-! and SP-2 are widely separatedand the
shale can be traversed in two reasonablyseparateddirections. If the shale is not
detectedthe effectis to map a reducedamountof salt. This may give a falseamount
of salt overhang.
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
448
Wavefront Methods
Outlining of Salt and Shale Masses 449
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
DISTANCE- KILOFEET
4 8 12 16 20 24 28
,. , , ' '
55 ,oo .
' ER
PE
I -I$00'
8.841
E-88
18800
'
Fro. 26. Methodfor obtaininglocion the far flankof the dome.(Final horizontalvelocitycontrolwould
be obtainedby blendingthe observedhorizontalvelocitysurveysinto the B-! and C-! wells,Figure29.)
ure 26) is plottedfor eachline, andan outlineof the saltis drawntangentto all loci.
A plot of the loci for the lower phone on line A-2-E is shown. The number on each
locusrefersto the approximate
shotpoint
distance
in thousands
of feet.Eachpar-
ticularlocus(Figure26) is obtainedby settingthe total time on the salt strip at a
pivot point whichis at the true vertical depth of the phoneand at a horizontalloca-
tion projectedinto the sectionfrom the directionallog. A wavefrontchart, repre-
sentativeof the sedimentaryvelocity in the directionof the line, is set at the exact
distanceof the shotpointandthe locusis drawnasbefore.Whentwo or morephones
are used,a verticalzXTmay be usedto determinethe vertical angleof approachof
the wavefrontin the salt. The loci and the verticalzXT'stogetherdeterminewhether
the velocitiesusedare reasonable.If a mathematicalwavefrontchart is used,hori-
zontal zXT'scan alsobe usedto checkthe salt tangentpoint on eachlocus.
450 Wavefront Methods
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
DISTANCE- K ILOFEET
0 4 8 12 16 20 24 28
, /
Fro. 27. Method for obtaining a horizontal sedimentarytime strip at --6,000 ft from a
wavefront chart for use in plotting horizontal loci.
pivot height is set which is the vertical differencebetween the phone level and the
contourin question.It makes no differencewhether it is up or down. The time is set
on the salt strip on the elevated pivot over the projected position of the phone. The
sedimentary strip is pivoted at the shotpoint location and the locusis drawn as be-
fore. The matching times on the strips arcing around their pivots solvea rather com-
plex, three-dimensionalmathematical relationshipfor all possibleleast time paths in
that plane. The line drawn tangent to a group of loci for each contour level deter-
mines the correct solution for each contour and is the outline of the salt.
A map (Figure 29) is made from a set of loci maps from three flank wells.The map
should now be of such detail and depth that development of the petroleum reserves
can be completed at a minimum cost. Certainly this is a worthwhile goal around a
piercement-typesalt domewhere every locationis a wildcat.
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
O
u.>.
Outlining
Ld
I
of Salt and Shale Masses
451
452 Wavefront Methods
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
Outlining
o
33d
NI
H/d30
o
of Salt and Shale Masses
o
o
453
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
454
bJ
0
0
0
o
Wavefront
133.4
NI
Methods
Hld30
0
-
o
0
o
Outlining of Salt and Shale Masses 455
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
Shotpoints can be spaced wider than for salt dome mapping. They should be
spacedabout every 5,000 or 10,000 feet along radial lines about the drilled location.
Longer lines are neededfor mapping shale than for mapping salt. Normally lines
should extend about 40,000 feet outside of the estimated shale outline. The number
of radial lines and their directionsdependson the problem. Since the shale is often
lessdomal shapedthan salt, normally it is sufficientto have lines every 20 degrees
coveringan arc about oneline wider (20 degrees),in both directionsthan the desired
coverage.Geophonesmay be spacedfarther apart in the shalethan in the salt. It is
desirable to have more than one geophonelevel recorded at all shotpoints. The
spacingshouldbe closerif the shaleinterface is closeto the well. Necessaryspacings
may be determinedat the well site by the vertical zXT;that is, the differencein time
per 1,000 ft between vertically spacedgeophones.This quantity is actually a mea-
sure of the inverse of the apparent vertical velocity. If the apparent velocity ap-
proachesshale velocity, the signal is approaching the geophonesnearly straight
down the hole. When this result is obtained, the geophoneinterval and the shot-
point spacingsshouldbe extended.If the vertical/XT changesappreciablybetween
shots,the spacingshouldbe shortened.In caseslike thoseshownin Figures 30 and
Table 1
outline
indicates
theshale
interface.
If thins
isagradational
velocity
change,
theout-
line will representthe approximate center of the gradation. Table 2 lists the times in
secondsfor a different profile shot into well A over a different hypothetical shale
massalongsidea salt domeshownin Figure 31. This profileis much flatter and more
difficult to map. The loci and raypaths are shown as before. Note that the shale is
only definedabout two milesfrom the well sinceit is so flat, but the minimum shale
outline is known to be a mile beyond this point by inferencefrom the last locus.
456 Wavefront Methods
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
Table 2
Depth (ft)
12,000 1.620
15,000 2.060 2.220
18,000 2.540 3. 250 4.130 5.020
SUMMARY
ACKNOWLEDGMENTS
.
The authorswish to thank Mobil Oil Company for allowing them to present this
material and to thank all thosepeoplewith Mobil Oil who helped in developingthe
many parts of thesetechniques.
REFERENCES
Section 5E
458
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
NOTES ON
E. J. NORTHWOOD*
ABSTRACT
:Depth
error
Fro. 2. The error Te in pickingfirst arrivalscausesa computeddeptherror of BB'
using the critical distancebisectorconstruction.
DIP=
To
N
Fro. 4. The condition of maximum error for three intermediate layers of known velocity which
fail to registeras first arrivals.
based on vertical well surveys, computations would tend to give depths which are
too shallow becauseof the fact that in a refraction survey the trajectory through the
overburdenis neither vertical nor horizontal. The amount of anisotropycan be found
from oblique shootinginto a well survey.
Another sourceof error is the incorrectpairing of velocitiesusingreversedprofiles.
Normally the only cluesof the pairing of velocitiesis the reciprocaltotal traveltimes
and the geometry of the problem. Changesof dip along the sectioncan create prob-
lemsin the geometry,and, even though the refractor is controlledin both directions,it
may not be controlledat the samelocation at depth. However, if a correctknowledge
of the true velocity of the refractor is known, then there is not sucha problem. The
only way to reducepossibleerror due to this causeis by moredetailed shooting,i.e.,
move overlapping of reversedprofilesso that the objective refractor is continuously
controlledfrom both directions.It may also be possibleto correctly pair the arrivals
from their character or frequency, but this is a somewhat risky procedureat the
present stage of our knowledgeof factors affecting character and frequency.
CONCLUSION
(1) Obtainasmuchcontrol
onoverburden
velocities
aspossible,
particularly
with
respectto velocitydecreasing
with depth.
(2) Shootlargecharges
to insurethecorrect
pickingof firstarrivals.Reproducible
recordingtechniquesassistin this.
(3) Obtaincontinuous
controlof theobjective
refractor
frombothdirections.
REFERENCES
Cholet,
Jacques,
andRichard,
Henri,1954,
A testonelastic
anisotropy
measurement
at Berriane
(North
Sahara):Geophys.Prosp.v. 2, p. 232-246.
Dix,C.Hewitt,1952,Seismic
prospecting
foroil:NewYork,Harper& Brothers.
Hagedoorn,
J. G., 1954,A practical
example of an anisotropic
velocity-layer:
Geophys.
?rosp.,
v. 2,
p. 52-60.
Hawkins,
L. V., andMaggs,D., 1961,Nomograms
fordetermining
maximum
errors
andlimitingcon-
ditionsin seismic
refractionsurveywitha blind-zone
problem:
Geophys.
Prosp.,v. 9, p. 526-532.
Khalevin,N. P., 1956,Contribution
to evaluationof precision
in interpretation
of dataof refracted
waves method: IzvestiyaAkademii
NaukU.S.S.R.(BulletinoftheAcademy ofSciencesoftheU.S.S.R.),
SeriyaGeofizicheskaya,Geo.
physical
Series,
no.8, p. 912-919.
Kleyn,A. H., 1956,Onseismm wavepropagationin anisotropic
mediawithapplications in theBetun
area,SouthSumatra:Geophys.
Prosp.,v. 4, p. 56-69.
Postma,G. W., 1955,Wavepropagation
in a stratified
medium:
Geophysics,
v. 20,p. 780-806.
Soske,
Joshua L., 1959,The blindzoneproblem in engineering
geophysics:
Geophysics,
v. 24, p.
359-365.
Steinhart,
J. S., andMeyer,R. P., 1961,Minimumstatistical
uncertainty
of theseismic
refraction
profile:Geophysics,
v. 26,p. 574-587.
Uhrig,L. F., andVanMelle,F. A., 1955,Velocity
anisotropy
in stratified
media:Geophysics,
v. 20,p.
774-779.
Zirbel,N. N., 1954,Comparison
of break-point
andtime-intercept
methods
in refraction
calculations:
Geophysics,
v. 19,p. 716-721.
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
FIELD TECHNIQUES
Section
6
FOR REFRACTION
SHOOTING
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
Section 6
468
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
SCREEN I NG LAYERSt-
M. E. TROSTLE*
ABSTRACT
Presentedat the 30th Annual Meeting of the SEG, November 19, 1960, Galveston,Texas.
* GeophysicalService,Inc., Dallas, Texas.
469
470 Field Techniques for Refraction Shooting
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
J 4
> ooo'-e?,ooo'
PLEXIGLAS - I000'
iALUMINUM - 500'
Vp FT./SEC. IZ,OOO'
VS FI/SEC
Vp=21,200FT/SEC.
TIME
-DISTANCE
PLOT I
FOR REFRACTIONS IN THE PRESENCE OF A SCREENING LAYER
20 40
I i025-5
5)CiS
PASS
60
BAN
i I 80 I00
54,00O
55,000
56,000
57,000
58.000'
59,000 , .............
, - .....
, _.
61,000
62000
63,000
65.00O
66,0O0
67,OOO
69,00
OF A SCREENING LAYER
7t,
(.O25-5..5)CPS PASS BAND
75,
A secondmodel study was made at a later date which includedan anomaly on the
deeprefractor(seeFigure3). In thiscase,the modelwasmadeof steeland brasswith
a steelscreeninglayer with an equivalentthicknessof 750 ft.
The energy sourcewas placed below the screeninglayer with detector positions
along the upper surfaceof the screeninglayer and on the surfaceof the screening
layer with the detectorpositionsthe same.In both instances,a time anomaly, which
could only be associatedwith the anomaly on the deep refractor, was observedat
the samedetector positions.
The positionof the anomaly and the arrival times proved that the primary energy
traveled a refracted path rather than a reflectedpath. Figures4 and 5 illustrate the
time anomaly observed.
472 Field Techniques for Refraction Shooting
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
L
1
159
134 j1 STEE
BRASS
;OURCE
STEEL
in the two models mentioned here the screeninglayer was less than one-half
wavelength thick.
The model work proved that energywhich traveled a refractionpath in the deep
refractor could be recordedin the presenceof a thin screeninglayer. Many questions
o o o
o
_
o
oo
o o
L
ANOMALY ON DEEP REFRACTOR
i
I I I
20 140 160 180 2OO
DISTANCE (cm)
SOURCE ON SURFACE
o o o
oo
-I-
LOCATION OF
ANOMALY ON DEEP REFRACTOR
I I I
I00 120 140 160 180
DISTANCE (cm)
One of the first field tests was carried out in the Delaware Basin of West Texas
where the layering can be summarizedas follows(seealsoFigure 6):
0--1,000 ft--Sand, gravel, and clay (6,000 ftfsec).
1,000-3,500 ft--Anhydrite (19,500 ft/sec).
3,500-12,500ft--Sand, shaleand thin lime (13,800ft/sec).
12,500-15,500 ft--Dolomite and limestone (21,000 ft/sec).
Available information indicates that high velocity and density characterize the
sectionfrom Devonian (12,500 ft) to basement(15,500 ft). This zone, and possibly
the basement rocks, can be expected to act as one refracting layer when long wave-
lengths are considered.
In Figure 7 a display of five refraction recordscovering 57,185 ft is shown. Be-
fore analyzing these field records,a descriptionof the recordingprocedureseemsin
order.
Each spread consistedof 24 seismometergroups spaced approximately 500 ft
apart. Each group containedten 4-cycle seismometerson a commonbase. These
seismometers fed energyinto preamplifiersof a specialdesignat eachgrouplocation.
Texas Instruments' 7000 KB seriesamplifiers were used in conjunction with the
474 Field Techniques for Refraction Shooting
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
SURFACE
NHYDRITE
19,500'/s,
DEVONIAN
476
Field Techniques for Refraction Shooting
Shooting Through Screening Layers 477
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
-[06
R.EF.RAC.:T'ION
VELOCITY
PROFILE
::::ANHYDRIT :':';':':':';'
,' :::1
.:.:.:.:.:.:.;'///,'.',', ;,:,1,:.:,:/,,/,
...
::::::::::::19500
':..:.:.:.:.:.:.:,
:.:.,:.:.:.:,:.:.:
CA S T I L L E --:
o
I,-
ILl
0
0
o
o
o
o
o
q
0''
o
Shooting Through
Screening Layers
z
o
rn
',
8
'n
479
480 Field Techniques for Refraction Shooting
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
From
themtdel
andfieldwork,
criteria
were
established
fordistinguishing
the
deep refractions from arrivals associatedwith the shallow screeninglayers. These
include:
1. Frequency
The predominantfrequencyof arrivalsfrom deepmassiverefractorslying belowa
screeninglayer was generallylow, with 8 to 12 cpseventsnormal.
Under certain conditions,the predominant frequency increased.The controlling
factors were:
1. shothole environment,
2. layering of the screeningzone, and
3. velocity of the material between the screeninglayer and the deep refractor.
With a high-velocity shooting medium, with hole depths such that the surface
reflection reinforcesthe higher frequencies,and with a screeninglayer made up of
layers less than one normal wavelength in thickness,the recorded events could in-
creaseto 12 or even 20 cps. A highervelocity betweenthe screeninglayer and deep
refractor contributesto higher frequencyarrivals.
2. Attenuation rate
Where the deep refractor was essentiallya massivebed, the attenuation rate was
slower and was fairly consistentwith normal sphericaldivergence.
In somecasesthe sameinitial energy cyclesprevailed consistentlyfrom 50,000 ft
to 150,000ft.
3. Depth computations
The events identified as deep refractorsmust occur at the proper position on the
time-distance curve. Where well control was available, the time-distance curve was
predictedbeforefield shootingbegan.Thesepredictionsgave the interpreter a quick
check on refractor conformanceand made anomalousevents immediately evident.
4. Geologicallysound
The data gathered on the refractorsmust be geologicallysound.In the courseof
various surveys,mapping has includedknown geologicfeatures. The conformity of
the refraction data with known geologywas good.
5. Internally consistent
Internal checking of observed data was possible by changing shot-to-detector
positionswhile maintaining the same structural position on the refractor.
Other internal checksallowed good estimates on the relative effect of structural
complicationsbetweenthe shotpointand spread,as well as an evaluation of pene-
tration of the refractor. In the West Texas area theseproblemswere not considered
serious.
RESULTS
8 MILES
-4.7
4.8 -4.8
4.9 -4.9
5.0 5.0
-5.1
5.2
5.3- 5.$ :
5.4- -5.4 rn
o
z
5.5- 5.5 o
5.6
-5.7
5.8
5.9
-6.0
-6.1
ARC PROFILE
DELAWARE BASIN
ROBERT E. SHERIFF*
ABSTRACT
INTRODUCTION
Much refraction work at sea consistsof limited programscarried out with pro-
duction reflection crews.Plannersof offshorerefraction work may not realize that
operational factors peculiar to conventional marine work make both the conduct of
the work and its interpretation somewhat different from refraction work on land.
This paper is written to point out some of these differences.
The mostimportant step toward avoidingor solvingtheseproblemsis beingmade
aware,in sufficienttime, of the dangerwhich may arise.Hence,this paperis mainly
confinedto citing the problems,solutionsbeing suggestedonly occasionally.Solu-
tions for the special circumstancesof a particular survey are a problem for the
planners of the particular work.
About half of theseproblemsconcernuncertainty in location, and, hence,a divi-
sion into location and nonlocationproblemsis convenient. The use of record sections
to aid in solving some of these problems will be discussedlast.
LOCATION PROBLEMS
Shot-locationproblems
Usually only the recordingboat is equippedfor precisenavigation, and shotpoints
are locatedwith respectto presetbuoys.Such buoysare commonlyanchoredby
* The California Company,New Orleans,Louisiana.
482
Operational Problems in Marine Work 483
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
ropes of length three or four times the water depth, and, hence, the buoy may be
some distance from the predetermined location of its anchor. The line connecting
the buoy to its anchor cannot be shortened because,if it were, the bobbing of the
buoy in the waves might walk the anchor. Not uncommonly buoys drift, are pulled
under by the current, or otherwise disappear. Multiple buoys or color coded buoys
are frequently used to decreasethe possibility of erroneousbuoy location.
The location of the shotpoint with respectto the buoy alsointroducesuncertainty
inasmuch as the shot is not located right at the buoy (in order not to damage it by
the shot). The fact that shotpoint location in ordinary reflection work is not critical
and that the time of an offshoreseismicparty is expensive tends to make for care-
lessnessabout the location of the shot. It is often difficult to get a shooting crew to
write down exactly where the shot is located with respect to the buoy or the direc-
tion of the current. The fact that supervisory personnel usually are looking at the
recordson the recording boat a long distance off also encourages"unknowns" in the
shooting operations. The overall effect is frequently an uncertainty on the order of
300 ft in shot location. Supervisionon the shooting boat and detailed shot location
notes are worthwhile.
Where navigation equipment is available on the shooting boat, the shot can be
located by directing the boat through the water churned up by the shot and noting
its location.
Detector location
Where currents are present in the water, the geophonestreamer tends to drift
away from the desiredin-line position. This drifting is more seriousthan in reflection
work becausethe recording boat spendsmore time waiting, thus allowing the cable
more time in which to drift. It is desirable to have the shooting boat prepare to
shoot before the recording boat proceedsinto position in order to minimize cable
drift. The direction from the recordingboat to a buoy at the center or far end of the
cable may be determined for each shot to help determine the cable position.
Marine detector cables usually are less than two-thirds of a mile long because
longer cablescannot be handled easily. Hence, to cover a moderate range of offset
distancesrequires several shots.
5100
5000
4900
z
)
4700
=oo ,;oo eo- 00'
tion of a delay time involves the subtraction of a large term proportional to the
offset distance,significanterrorscan be introducedunlesssuch distancemeasure-
mentsare accuratelymade. This type of error tendsto be systematic.
Temperatureand salinity variations(particularlytemperatureand salinity layer-
ing in the wateror variationcausedby moderatelywell-defined
currents)affectthe
water breaks in another way also. The distribution of temperature and salinity
variations may bend the path of the water-bornewave and, thus, influencethe
amount of energycontainedin the water breaks.The nature and configurationof
the water bottom also affects the amount of energy in this wave. Sometimesthese
effectsare solarge that the direct wave doesnot carry enoughenergyto give a clear
arrival above the noise level. Part of the direct wave is reflected back and forth
between the bottom and the water surface. This tends to lengthen the direct wave
and make the onsetlesspronounced.This effectseemsto be accentuatedwhen the
Operational Problems in Marine Work 485
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
water depth decreasesin the direction in which the direct wave is traveling. Hence,
it is sometimesimpossibleto detect the direct energy from the explosionin the
direction of shallow water, whereas,it can be done easily in the opposite direction.
Since the water breaks arrive late on the records becauseof their relatively slow
velocity, care should be taken to run the records long enough to record these
events. As a further complication,thesewater arrivals may occur in the presenceof
higher amplitude wavesor noiseand, as a result, may not be recognized.
Most marine detectorsand amplifiers(exceptfor piezoelectriccrystal types) dis-
criminate against the high frequenciesof the water breaks so that it is better to use
separatehigh-frequencyhydrophonesand high-frequencyamplifiers to detect their
arrival. However, a complete high-frequencysystem is normally impractical, and
the most that is usually available is one high-frequencychannel, the detector being
hung from the recording boat. In addition, one may be able to record the output of
a few of the detectorsin the cable separately through high-frequencyamplifiers; in
spite of the discriminationof the detectorsagainst high frequencies,adequatedirect
arrivals can frequently be detected in this manner. One tries to get sharp water
breaks recorded for detectors at each end of the cable and one or two in between
so that the location of other detectors can be determined by interpolation from
these.
OTHER PROBLEMS
Shot bubbleproblem
The shotsin reflectionwork are locatednearthe surfaceof the water to prevent
the bubbleeffect,but sucha chargelocationcutsdownon the explosion efficiency
becausemuch of the energyis vented to the atmosphere.Conversely,refraction
chargesare usuallydetonatedon the sea bottom in order to get enoughseismic
energyto carry over the long distancesto the detectors.Hence, large bubbleeffects
Fro. 2. The shotpointis locatedover a relatively thick sectionof semiconsolidated Plioceneand Mio-
cene sandsand shales.With increasingoffset,an upfaulted block of Cretaceouslimestoneis crossedat
approximately 16,600-ft offset distance.Note the diffractionsoff the fault. The departure of the first
break event around 28,000-ft offsetdistanceis attributed to departure in cable position. Note that each
event has a parallel bubbleevent (which doesnot alwaysline up on adjacentrecords)which might be
misinterpretedas a secondaryarrival.
All of the recordswere 24-tracerecords,but duplicatetraceswere cut off to prepare this profile. The
horizontalscaleis not everywherethe samebecauseof cabledrift.
Operational Problems in Marine Work 487
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
::.
.............
.,.::::a.z;,
'* ................
:::
......
Use to determineoffsetdistance
The various records of a single refraction profile in marine work often form a
ragged set. The effective geophonespacing along a line from the shot may vary
becauseof cable drift or becausethe shot is not located on the line of the spread, and
overlap and offlap of adjacent recordsare common. Laying each record besideits
companionsand shifting the distance, or overlap, between records to achieve the
best overall fit of primary and secondaryalignments, including water breaks, is a
practical way of checking offset determinations. The use of variable-area or vari-
able-densitydisplays(seeFigures2 and 3) aids in this processbecauseit improves
the detectability of secondaryrefractions and various other types of arrivals which
are difficult to seeon conventional recordsbecauseof high amplitude. In the match-
ing processone needsto recognizeenergy which originated from bubblesrather than
from the shot itself, becausethe bubbles on different shots may have different time
and energy relationships.
Use to check time break
Matching adjacent records can also be used to check the time breaks since the
time breaks recordedat large offsetsare frequently very poor. When the time break
is recordedby the firing of a thyratron upon commandfrom a radio signal from the
shooting boat, false time breaks may occur, and one may have to chooseamong
several possibletime breaks.
CONCLUSIONS
ACKNOWLEDGMENTS
CASE
HISTORIES
OF
Section
7
REFRACTION
METHODS
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
Section 7
492
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
JOHN H. HOKE*
ABSTRACT
Refractionsurveyshave beencarriedout over a large portion of SaudiArabia during the past five
years. Refractionwork has proven to be both definitiveand rapid, and has considerablyexpandedthe
scopeand versatility of the geophysicalapproachin this area.
The interpretationprinciplethat has been utilized has been referred to by othersas the three-point
system,the ABC method, the reciprocalprocedure,two-way wavefronttargeting,and the plus-minus
method.It hasapparentlybeenwidely usedfor weatheringdeterminations,bedrock configurations,and
shallowbasementsurveys.A more universalapplicationof this theory for the interpretation of rela-
tively incompetentrefractorsat great depthsof burial has perhapsnot been as generallyrecognized.
The recordingand interpretationprocedurethat has been developedfor refractorswith a limited re-
cordingrangeis called the total time method. The constructioninvolved is not complicatedand pro-
videsa preciseresolutionof the refractiondata. The computationsare straightforward,and a program
has been developedfor an IBM 705 computerwhich will automaticallycalculatethe intercepttime,
refractorvelocity,refractionangle,vertical time, and depth to the refractorfor eachgeophonelocation.
INTRODUCTION
494
Case Histories
of Refraction
Methods
Prospecting in the Rub' AI Khali 495
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
INTERVAL VELOCITY
FEET/SEC.
GENERALIZED
DEPTH
SECTION o. o. o. o.o..o.o.
(FEET) I i i
NEOGENE
LI_I
RS-FORAION
IOOO Z
I.I_1
2000
;5000
4000 .......
5000
THANANA
GROUP
6000
9000
FIO. 2. Averageformationalvelocities
basedon velocitysurveysthroughoutthe ArabianPeninsula.
RANGE'37,000-48,000 FEET
SURFACE
CHARGE-
I000POUNDS
EOCENE ,JURASSIC
ARRIV.L ARRIVAL
.........
MONITOR
RECORD
T_,.J
'-tEJl
i''1
VXrIH
li,111 FILTERING 31-0
PLAYBACKRECORD
.FILTERING
,20-0
Fro.3. Jurassic
refraction
recordfromtheeastern
Rub'al Khali
with interference from the shallow Eocene refractor.
lb of powder. The pattern followedin arranging the stacksvaries with the nature
and configurationof the near-surfacematerial. The energycouplingis not very
effectivewhenlimestoneis exposedat the surface,and drilled holesare usedin these
areasto eliminatethe necessityof goingto very large surfacecharges.
OVERBURDEN VELOCITIES
FULL
GAIN
PLITUDE
RANGE50,000-61,000 FEET
SURFACE CHARGE- 4500 POUNDS
FILTERING 32-0
ONE GEOPHONE PER TRACE
Fro. 4. Jurassicrefractionrecordnear the Ghawar Field, Saudi Arabia.
SECONDS
rEPTi COPUTAOrS
= + - + + - x x 0.
The constant changesin V and V2 require that a unique "cosine factor" must be
computedfor eachgeophonelocation.Severalthousanddepth computationsmay be
involved on each line, and this number is later compoundedby each new control
point that modifies the previous overburden overlay. This presentsa formidable
volume of computingand recomputing,if the refraction data are to be continually
adjusted to all the available well control. For this reason, a program has been de-
velopedfor the IBM 705 computer,which can readily make all the necessarydepth
computations.
The input data for this program consistof the two arrival times from the opposed
shotpointsand the extrapolated total time for each reversal. The total times from
shotpoint to shotpoint may be projected on a standard time-distance plot, or tabu-
lated by adding together the successiveincrements of refractor traveltime on the
adjacent spreadsthat will make up a "total time." The surface elevation, weather-
ing referencedatum, and the computedweathering times for each geophonelocation
are also included to adjust the depth calculationsto a sea-level datum. The over-
burden velocities are introduced as a separate input and can be changed without
affecting the basic input data.
The refractor velocitiesare computedas in (2) with the distanceinterval being
equal to the spacing between geophones.The time differentials between adjacent
geophonesare statistically averaged until the time deviation between stations is
less than one ms. These smoothed-out time increments are then converted into a
refractor velocity for each geophonelocation. Intercept times from the surface are
computedby
ti = (TR - Try) -- TRr, (6)
502 Case Histories of Refraction Methods
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
where TR and TL are the two arrival times and TRL the total time between shot-
points.The refractionangleis computedthroughformula (4) and the intercepttimes
are converted into vertical times through
2tv- [t- wxco]/cos i. (7)
These times can then be used in conjunctionwith reflection work to provide inte-
grated interpretations.
Two depths are computed to help in evaluating the probable successof the
weatheringcorrectionsapplied in the area.
The equationsfor these depth calculationsare'
Z(s) = (2tv X V/2) - ERE (8)
and
WEATHERING CORRECTIONS
The large amount of relief and the very low near-surfacevelocitiesin someof the
sand dune areasof the Rub' al Khali make the computation of adequate weathering
correctionsan important aspectof the interpretation process.When the sand dunes
are very large or the water table is very deep, the weatheringcorrectionsare often a
significantportion of the raw intercept time. These weatheringanomaliesmay occur
as local featuresaffectingonly a portion of a spread,or as broad distortionsinfluenc-
ing areas large enoughfor them to attract possiblestructural interest. When the
errors in the weathering correctioncreate only a scattering in the depth values, they
can be largely eliminated by an averaging process.This procedurealso tends to
smooth out the random picking errors causedby noise or other interference.
Weathering correctionsare normally obtained at 10,000 ft intervals along the line
by shootinga short weatheringspread.These correctionsare not always satisfactory
due to the absenceof vertical upholecontroland the probability of velocity stratifica-
tion in the near-surfacebeds with potentially undetectable low-velocity layers.
Weathering correctionsfor the individual geophonesare determinedby establishing
a secondaryterrain weatheringcorrectionfor each geophonelocation, and applying
this value to the primary weatheringtimes that have been prorated betweenshot-
point locations. The terrain correction is obtained by first constructinga smooth
terrain datum, which is a plane that passesthrough the surface elevation at each
of the weatheringshotpoints.The interval betweenthis datum and the surfaceeleva-
tion at each geophonelocation is converted into a plus or minus time correction
based on an arbitrary near-surfacevelocity that may range from 1,200 to 3,000
ft/sec.
Prospecting in the Rub' AI Khali 503
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
SECONDS
1.200,
E
400
'!TOTALTIME"
RECORDIN6 SPREAD
SHOTPOINTS MAIN RECORDIN6
SPREAD SHOTPOINT N
FEET I'-'-I
', -. ,,'T.,,7,?,
g3,'..l'
V/=/0,000
Feet
cond
,,'
IOO0l , ,' , ,,y , ',, ,, I ---I000
,_ ,, ,' I,' ,,!,_,.z, ,,'!
2000 2000
=20 Feet
persecd /000
When a very competentrefractor is present the total time from shotpoint to shot-
point can be recordeddirectly at one of the shotpoints.The method is illustrated
in Figure 6. This procedureis normally used when the terrain is extremely rugged
or partially inaccessible.It can alsobe usedwhen it is considereddesirableto cover
an area more rapidly than would be possibleby continuousprofiling.
The shotpointscan be at any range from the main spread providing they are
beyond the critical distance and in-line. The total time is obtained from a second
spread at one of the shotpoints,and this time is recordedsimultaneouslywith the
shot into the main spread. Occasionalcontinuouscoverageon the refractor out to
the total time spreadis desirableto verify that the refraction arrivals on both spreads
are equivalent.
Refractor depths between Za and Z4 can be determined without any information
on the ranges,weathering corrections,or elevationsat the shotpoints.This can be
a distinct advantage in difficult terrain, where establishinghorizontal and vertical
control is a major problem. Depths at Za and Z4 are determined as in (3),
z = {[( + ) - ]/(2 coi)} x (0)
and
CONCLUSIONS
ACKNOWLEDGMENTS
The author would like to thank the Arabian American Oil Company for permission
to publish this paper, and expressappreciation for the encouragementand assist-
ance received from various members of the Exploration Department in the prepa-
ration of this article.
REFERENCES
Birch, F., 1960, The velocity of compressionalwavesin rocksto 10 kilobars,Part I: Jour. of Geophys
Res., v. 65, p. 1099.
de Segonzac,P. D., and Laherrere, J., 1959, Applicationof the continuousvelocity log to anisotropy
measurements in Northern Sahara; resultsand consequences: Geophys.Prosp., v. 7, p. 202-217.
Hagedoorn,J. G., 1959, The plus-minusmethod of interpretingseismicrefractionsections:Geophys.
Prosp.,v. 7, p. 158-182.
Hawkins, L. V., 1961, The reciprocalmethodof routine shallowseismicrefractioninvestigations:Geo-
physics,v. 26, p. 806-819.
Heiland, C. A., 1940, Geophysicalexploration:New York, PrenticeHall.
Schenck,F. L., 1962, Wavefront refraction solutions:Unpublishedmanuscript,R. H. Ray Geophysics,
Inc.
Stam, J. C., 1962, Modern developmentsin shallowseismicrefraction techniques:Geophysics,v. 27,
p. 198-212.
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
ABSTRACT
INTRODUCTION
The Heletz oil field (Figure 1), Israel'sfirst, wasdiscoveredin 1955 by the Lapi-
doth Israel Petroleum Company. The field has sincebeen fully developed,produc-
ing some2,000 barrels per day from about 20 wells.
Located on the southeastflank of a concealed,major coastal plane feature, the
accumulation is essentially stratigraphic, controlled by the updip reduction in
porosity of a few thin reservoirsands.Depth to the producingLower Cretaceous
sandsis 1,500 m (Tschopp, 1957).
The relatively small producingarea is located near the axis of a major gravity
anomaly which parallels the coast line and traversesa large part of the southern
Coastal Plain (Figure 2).
A number of earlier attempts at reflection shootingin the critical regionsof the
structure gave disappointingresults,which must be attributed to a combinationof
adversefactors: irregular topography; high-velocity beds, such as chalk closeto the
surface;dry and highly porouscalcareoussandstonebedsat surface;and thick low-
velocity layer, in placesover 100 m thick. But probably the most critical factor
influencingreflectionquality is the existenceof extremely cavernousdolomite for-
mations in the Cenomanian and Lower Cretaceous sections.
An experimentalrefraction survey, sponsoredby the Ministry of Development,
governmentof Israel, was carried out in the Heletz-Negba area during the period
from April to October,1957. The object of the investigationwas to test the efficacy
of the refraction method for structural mapping. Considerablegeologicalcontrol
was available for this area, which had been obtained from development and wildcat
drilling.
Two profiles were shot, one along the gravity axis, the secondat right angles.
The results obtained along the axis line Heletz-Negba are describedin this report
(location map, Figure 3).
kilometers
33
SCALE
T i beria
HAIFA lake
TEL-AVIV
32
32
JERU;
NEGBA
HELETZ
31
N E G E V
:30
30
}ULF OF AQABA
m. 1.1V[p of Tsre].
Heletz-Negba Area, Israel 507
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
..REVAHA,
'"
IlS
/ 7EL
I10
gO 105
LEGEND
OILWE REFRACTION
LINE
TEE o REFRACTI
SHPNT IOO
[LD TECIrIqUES
In orderto establish
the sequenceof formationvelocities,
a 14-kmreversedpro-
file wasshotfirst,shownon the shotpoint locationmapbetweenshotpoint 50 to
thenortheast andshotpoint 345to thesouthwest.Failuretodetectthedeeper
refrac-
torsfromshotpoint 50,alongthelineofseismometers fromkm7 to km 12.5,neces-
sitatedan additionalseriesof shotsfrom a newshotpoint281, alongthe sameline
of seismometers.A few shortreversed profileswerealsoshotpermittingclosersub-
surface
coverage
ontheshallow
refractors
(profiles
50-160,160-281,and281-345,
the latter wascoveredin onedirectiononly). More completesubsurface
coverageon
thedeeper
refractor
wassecured
by making
a number
ofrecordings
withtheseis-
508 Case Histories of Refraction Methods
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
117
HELETZ
NO23
HELETZ
NO
30 / sP.
2O6
HELETZ
NO.
22
o $P.
236
-- //
.-RUR
NO,
-- ,-LEGEND
DRY
TESTHOLE
OIL WELL
107
III 113 115 I 8 120 122
INTERPRETATION OF RESULTS
A study of the uphole times on the refraction recordsand of the traveltime curves
for the near spreadsconfirmed earlier findings of the coincidencebetween the base
LVL and the water level (Behr, 1958). In the caseof the Heletz-Negba line, its direc-
tion is such that the water level remains reasonably constant over the surveyed
area, at a depth of only a few meters above mean sealevel.2 It could thus be assumed
Table 1.
ShotpointElevation Depth
shot
of ObservedTheoretical
uphole time
vertical time
also that the base of the LVL remained constant, which meant that the effect of the
LVL could be eliminated in a simple and effective manner by referring all arrival
times to a datum at MSL. The average velocity from surface to base of LVL was
found to be remarkably constantat 1,000 m/sec. The validity of the abovemay be
judged by Table 1, which comparesobserveduphole times with theoreticalvertical
times, computedon a basisof the 1,000-m/secaveragevelocity to the datum and a
subweatheringvelocity of 2,000 m/sec.
Average LVL velocitiesand subweatheringvelocitieswere known from previous
refraction and reflection surveys in the area.
After correction to datum, all useable events appearing on the records were
plotted against distance from sourceon a time scale of one cm--50 m. Vertical
paths were assumedfor the LVL, the validity of which may be judged by the small
anglesof refractionfor the upper interfaces(Figure 4). Due to the inconvenientsize
of this graph, on which the overlappingof an unusually large number of events tend
to obscurethe more important data, the original is not presentedwith this report.
"Data on water levelsobtainedfrom HydrologicalService,Water Department, Ministry of Develop-
ment.
510 Case Histories of Refraction Methods
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
ooo .
'-,..... . 2089
oo
I ' ' ' I _ _ i '-.[, -
es NI. ' ' ' '' ' HI ' ' H2.H. ' '8 I. '
,.rfo=e
.... ""' ''"'' .......................................' ;;'.........................................................
_ ____Z'::::c:===:-
:,ooo
_
' /
.........
note
exoggemted
depth
I0_
000
m LEGEND
_ 200
m 2.) VELOCl
IN KM/SEC. REFRACTION
INTERFACES
) 21 SHOTPOiNT NO.AND DIRECTION
OF ROFILE folr OCCUFOcy
m 57 TIMEINMILLISECS. ,o vory)ng
degrees
of occurocy,
m .... owingto m uenceo uppermeroces
............... -- 000
FIG.4. Refraction
traveltime
curves
fortheHeletz-Negba
profilewith
corresponding
interpretationof velocityinterfaces.
Instead,Figure4 presents
onedrawnona greatlyreduced
scaleandonwhichmost
ofthelaterphases
ofearlyevents
andmanysecondaryonesareomitted.
Thisplot
isalsoa moreconvenient
formforidentifying
refractor
velocities
withcorrespond-
ingbranches
ofthetraveltime
curves.
However,
theoriginal
plotswereusedforthe
actualdepthand dip computationsof the interfaces.
Identification
ofvelocitybedsfromapparentvelocities
wasguided bycomparing
overalltimesforshotsin opposite
directions,
alongmutualrefraction
paths,and,
also,relating
intercept
timesat common shotpoints
forspreadsin opposite
direc-
tions.Theupperpartofthevelocitysequence
isknownfromearlier,shallow
refrac-
tion surveys.Comparison
of overalltimesis givenin Table2.
Heletz-Negba Area, Israel 511
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
Table 2.
Apparent
No.of Totaltravel- velocity
at Remarks
Shotpoint
endseis- timeinsec endofprofile
mometer inkm/sec
50 160 1.310 6.1 Arrival time via 6.0 km/sec equals time via
160 50 1.311 6.0 5.0 km/sec at this shot-detector distance.
50 142 1.134 4.6 Apparent velocities refer to 5.1 km/sec re-
142 50 1.141 5.2 fractor.
160 345 2. 075 (projected) 5.0 6.2 km/sec refersto dip in 5.1 km/sec.
345 160 2.089 (assumed) 5.0
50 345 3.100 6.2 Discrepancy in time possibly due to picking
345 50 3.140 6.0 of different phase.
of depth and dip of the interfaces,particularly when dealing with multiple, non-
conformable, and discontinuousinterfaces, associatedwith high velocities. Uncer-
tainties associatedwith computing the upper-velocity sequenceaccumulatein sub-
sequentcomputationsfor the underlyingvelocity beds.
Below the traveltime curves,the sequenceof velocitybedsis shown(Figure 4), as
interpreted from these data. A number of trajectories have been indicated in order
to relate branches of the various refractors with correspondingup- and downdip
segmentsof the traveltime curves.
DISCUSSION OF RESULTS
The 2.0-/2.8-km/secinterface
Position of this interface is based on separate depth calculationsin the vicinities
of shotpoints 50 and 160. The T-X segments connected with this interface are
limited to short lengths closeto the shotpoints, which means that for the interven-
ing and greater part of the sectionthere is no subsurfacecontrol at this level. Non-
detectionof the 2.8-km/sec interface in the vicinity of shotpoint 281 implies a very
thin bed or its complete absence.The interface as drawn between shotpoints160 and
281 must, consequently,be consideredpartly hypothetical.
Velocity range
Stratigraphy (km/sec)
Neogene;a marl, clay ...................................... 1.8 to 2.1
From this table it is noted that a velocity of 2.0 km/sec may refer to marl and clay
of the Neogeneas well as to the chalk and marl of the Eocene-Senonian.Similarly, a
3.0-km/secvelocity couldbe associatedwith a numberof parts of the stratigraphic
section, viz., the Eocene-Senonian when the facies is limestone and flint or the
Turonian-Cenomanian when the facies is marl and chalk. A 2.0-/3.0-km/sec inter-
face may, therefore, represent any of the following contacts: Neogene/Eocene/
Senonian, Neogene/Turonian-Cenomanian, Eocene-Senonian/Eocene-Senonian,
and Eocene-Senonian/Turonian-Cenomanian.Since both the base of the Neogene
and the Eocene-Senonianmark unconformitiesin this area, the 2.0/2.8 km/sec can-
not be considered a reliable structural marker.
Previous, detailed, shallow refraction studiesin the area have establishedgood
correlationbetweenthe 2.8-/3.9-km/secinterfaceand a marker on the electriclogs,
which coincideswith the top of the limestone-dolomitecomplexof the Cenomanian,
possiblyof structural significance(Tschopp, 1957; Grader, 1957; Behr, 1958).
The closestreliablestructural marker to the 3.9-/5.1-km/sec interface is a remark-
ably persistentshale bed of the Lower Cretaceous,often referred to as the "Shale
Break," as shown in Figure 5. Its occurrenceis a few-hundred meters deeper than
the seismic horizon. No correlation is at all possiblebetween the refractor and the
geologicalmarker.
On the same crosssectionit is immediately clear that the top of the 6.0-km/sec
bed bears no relation to the structure at that level, accepting that structural atti-
tude is reflectedby the "Heletz sand" of the Lower Cretaceousand the top of the
limestoneof the Jurassic.
In Figure 6, the depths of the 6.0-km/sec bed have been compared with electric
logs taken in the three holes, viz., Negba No 1, Heletz No 22, and Heletz No 23.
The first two holes are located on or near the refraction line, while Heletz No 23
lies two km northwest of Heletz No 3, the latter on the line (Figure 3).
At Heletz No 22, the depth of the refractor coincideswith the top of the lime-
stone seriesof the Jurassic,which is overlain by a sectionwhich is predominantly
shale. A velocity survey in this well (Figure 7) indicated interval velocities no
greater than 4.2 km/sec as deep as the Jurassiclimestone,followed by an interval
velocityof 6.35km/secfor theselimestbnes,
150metersin thickness.
Thereis, there-
fore, a remarkable agreement between the velocity of the refractor (6.0 km/sec)
and the measuredinterval velocity (6.35 km/sec) as well as the corresponding
a Neogeneis the term used to describethe undifferentiatedMiocene-Oligocene.
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
o
o
I
I
I
I
I
I
I
I
I
I
"-'1
/
- I
,.,,- J
I
I
-
o
_o
o gtJ.'Fi .
o
NI #ld](]
I
Heletz-Negba Area, Israel
o
515
516 Case Histories of Refraction Methods
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
1700
' 1800
TopJur,os
- m/s. refroor 1900
.u-'oSl?
206
/--assumed
2000
2 IOO
Depth of6000
rn/s. refroctor
SR345
-ztsoo to-$ooom 2200
).z8s
depths at which they were found, respectively, -1,900 and -1,870 m. Another
important observationhereis that the computedaveragevelocitydownto the refrac-
tor is of the samemagnitudeas that actually measured,respectively,3.7 km/sec
and 3.6 km/sec.
Identifying the 6.0-km/secrefractorwith the Jurassiclimestoneat Heletz 22 is,
therefore, basedon conclusiveevidence.
Heletz-Negba Area, Israel 517
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
-LITHOLOGY
INTERVAL
HARDLS LS. SH SH8SS SHtLS. LS. LS DOL. MARL
'1'
131
I-
rq
---I
METERS/SEC. o o o o o I%1
METERS
FROM
MSL. o
0
o
0
o
o
.= I
.LITHOLOOY LS. SH. LS SSI LS LS LS.8SH. DOLOMITE 8 LS. CHALKY LS. rq
INTERVAL
VELOCITY o
METERS/SEC. o o
o
o o
o o
o
SE.
SURFACE
Geological Section
ing the reducedgravity values from the Bouguer anomaly. From this contour map
a gravity profilewas constructedalongthe Heletz-Negba refractionline (Figure 9).
As this line runs roughly along the strike of the transgression,the influence of the
latter, comparedto the size of the total anomaly, can be consideredto be small and
of no consequencehere. On the same diagram, the 6.0-km/sec refractor has been
somewhatmodifiedin shapeto permit simple computation of the theoretical gravity
anomaly along the profile. A density contrast of 0.2 was assumedat the interface
representingthe lumped distribution of the shale and the underlying limestone.The
problem was treated as a two-dimensionalone on the assumptionthat the line of
profile was closeenough to the axis of a symmetrical feature.
Allowing for deviations inherent in the oversimplification of the problem there
appears to be a definite relationship between the gravity anomaly and the inter-
preted faciesboundary.It is evenpossiblethat the generalincreasinggravity north-
ward is mainly the manifestation of the thickening effect, in this direction, of the
deeperlimestonesection.
520 Case Histories of Refraction Methods
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
LEGEND
_j THEORETICAL
VERTICAL
COMPONENT
OF
-.,..,...,.
-...,.
,... .i,.,,. 6RAVITY
0.2 ATFORDENSITY
CONTRAST
I.TZR;ACE. OF
' .'. __ EXPERIMENTAL
REGIONAL
i BOUeUER
>,
m GRAVITY ANOMALY
>_ a
M SL. NEGBA
I HELETZ18 HELETZ22 BRURI S.P:345
...- 500
-. I000
1500
zm
ASSUMED
DENSITY
REPRESENTING
BOUN
SHALE--LIMESTONE LITHOFACES
IN
DAR 2000
2500
. LOWER
CRETACEOUS
AND
JURASSIC. OOO
CONCLUSIONS
ACKNOWLEDGMENTS
REFERENCES
CHARLES L. ELLIOT*
ABSTRACT
INTRODUCTION
* GeophysicalEngineer,Tucson,Arizona.
522
Applications of Refraction in Mining 523
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
Emergence-angle
raypath tracing
The use of emergenceanglesin raypath tracing as a means of solvingrefraction
problemshas been reported in the literature. This procedurehas been describedfor
the rapid graphical solutionsof multilayer problems using reversedprofiles (Slot-
nick, 1950) and for partial solutionsof unreversedprofiles (Dix, 1941, 1952).
With known velocitiesof the upper layer and the times of first arrivals at a known
separationbetweentwo detectorpoints, it is possibleto determinethe angleof the
emergentraypath within the plane of observation.Figure 1 showsthe relationships
for determination of the emergenceangle e so that
sin e = V, (1)
where/XT is the time differenceof the first arrivals at geophonesR and R2.
But/XT//XX is the reciprocalof the apparent velocity of the lowest medium and
is directly obtainable from the slope of the traveltime curve segment between
geophonesR and Ra.
Applications of Refraction in Mining 525
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
R R2
MEDIUM
V
This is solvedfor CD and yields an absolutelimit to the depth of the interface and,
hence,is a guide to the determination of trial raypaths.
Now trial raypaths can be constructed, such as SF and SGH, with as many
separatesegmentsas are required. The nature of convexitiesand concavitiesof the
bedrock surfacebetween the shotpoint and geophoneswill determine the segments
requiredas the raypath must remain in the high-velocitymedium. Actually, analysis
526 Case Histories of Refraction Methods
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
Reversedfan shooting
Fan shooting over piercement-typesalt domes in the Gulf Coast region was
actively and successfullypursuedin the early days of petroleum prospecting.Its
V2
T V$
V4
R8
R7
BODY R6
o R8
o R
Rs
VI R2
v,v, R,
Therefore, the raypaths that traverse the sulfide body will time lag the normal
raypaths expected in the absenceof the body. Using conventional practices of
shootingcrossedfans, the resultant time lags can be solved for approximate thick-
nessand position of the sulfide body.
An important and useful, additional condition arises in this case. For detectors
in the fanned spreadnear the extremitiesof the sulfidebody the first arrivals will be
along raypaths through the surroundinghigh-velocity rock and diffracted by the
edgesof the body. In Figure 3 the first arrival raypaths are depictedfor a hypotheti-
cal sulfidebody with velocity V. buried in a medium of velocity V. High-velocity
raypaths from the shotpointin medium V that circumventthe low-velocitysulfide
body are the diffracted first arrivals at geophonesR, R7, and Rs. The first-arrival
raypaths with time lag due to traversingthe sulfide body are indicatedat geophones
R3,R4, and Rs. At somepoint near the extremityof the body the traveltime of first
arrivals will be the same for both the circumventing high-velocity and traversing
time-laggedraypaths. Thesepoints are indicated at geophonesR.and R6.
528 Case Histories of Refraction Methods
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
For the high-velocity raypaths that circumvent the body, the emergence-angle
raypath tracing proceduresof the previoussectioncan be used to advantage.By
this meansthe edgepointsof the body can be ascertainedfrom the directionof the
emergenceray at the geophoneand the total traveltime, assumingthat velocity V
is known or can be independently determined. The derived edge point positions
can then be correlated with the body position as determined by the time lags as a
means of refinement of the interpretation and establishmentof the validity of the
derived body outline.
This method has all of the inherent limitations that are present in conventional
salt dome fan shootingwith someadditional ones.The surroundingrock formations
shouldbe homogeneous and/or uniform in velocity propertiesand their positions
and parametersknown before these techniquescan be applied with any degreeof
success.Any velocity variations in the surroundingrock could give rise to appreci-
able variations in first-arrival times and, hence,make possiblesolutionsambiguous
or nonexistent. Unfortunately, geologicconditionssurroundingore occurrencesare
usually complex,and consequentvelocity variations couldbe extremelydeleterious.
Ideal velocity conditionsof the surroundingrock formations would likely be the
exception rather than the general rule.
Also the sulfidebody must be very massiveand approach completereplacement
of the host rock in order that suffi'cientvelocity contrastswould exist for measurable
differencesin first-arrival times. Even tens of percent sulfide by volume would not
appreciably decreasethe velocity of the body from that of the host rock. Again,
extremelymassivesulfidebodiesin nature are the exceptionrather than the general
rule. Also, when only consideringexplorationfor massivesulfides,other geophysical
methods are more ideally suited to their direct detection, so that seismicrefraction
techniqueswould likely not materially offer any advantage.
Refinements to this basic technique are indeed possibleand might result in a
practical method for sulfideexploration.Suitable modificationsof the procedures
for accuratelydelineatingsalt domeboundariesusingaplanaticsurfaces(Gardner,
1949) or wavefront plotting (Musgrave et al, 1960) conceivablycould offer a new
and important approach.
FIELD RESULTS
tive and consistedof several shallow shafts and a small open pit from which about
100,000 tons of copper carbonate ore had been extracted. The mine lies in a broad,
mature valley between the Cerbat Mountains to the east and the Black Mountains
to the west, and about one mile from the foothills to the east.
In the valley, outcrops are sparse.Alluvium and gravels fanning out from the
mountain washes overlie Gila conglomerate which rests unconformably on the
Precambrianbasement.The bedrock consistsmainly of granitic gneissand feldspar
porphyry. The Gila conglomerate, in which the carbonate mineralization occurs,
and the Precambrian basement were exposedin the open pit.
The primary sourceof the copper was not known but was believed to lie beneath
the Gila conglomeratein the valley and likely within a mile of the openpit. Outcrop
control toward the foothills to the east ruled out the possibility in this direction.
In view of the lack of geologicalcontrol, explorationfor the primary sulfidesbecame
a geophysicalproblem. The area west of the open pit was extensivelycoveredon the
ground by magnetic, induced-polarization,and resistivity surveying. In addition,
two seismicrefraction profileswere run.
The purposeof the two seismicprofileswas primarily to determine the depth to
the Precambrian basement under the Gila conglomerate and alluvial cover. This
servedtwo purposes'to determineif sectionsof the basement,and, hence,any possi-
ble sulfide mineralization, was beyond the depth of detectability of the other geo-
physical methodsbeing employed; and to aid in depth correlationin the interpreta-
tion of the other geophysicaldata in areas where the basement was not excessivein
depth.
No geophysicalindications of extensive sulfide zones in the bedrock were ap-
parent, so that no follow-up drilling was done. Therefore, an accurate check of the
seismic-deduceddepths to bedrock is not available. However, depths to bedrock
deducedfrom resistivity data from Wenner array expandersgave excellent correla-
tion with the seismicdepths.
The equipment used in the survey was standard eight-channel, seismicrefraction
equipment manufactured by Century GeophysicalCorporation, and conventional
field procedureswere followedin the survey operations.Geophoneseparationsof 200
ft were used for the main long profiles and variable geophoneseparationsin the
short spreadsfor upper layer solutions.
Bothprofiles
wereshotfromthesame
shotpoint,
S,located
in bedro'k
in the
bottom of the open pit. Seismicline no. ! (Figure 4) bears S 25W from &, and
seismicline no. 2 (Figure 5) bearsN 55W from &.
Figures 4 and 5 show the results of the two seismiclines. Short spreadsat inter-
mediate points along the profile were shot to determine depths, dips, and velocities
of the upper layers. In addition, velocitiesfor the Precambrian bedrock and the Gila
conglomeratewere determined by shooting in shallow diamond-drill holes in the
open pit and vicinity. The traveltime curves shown on the figures have been cor-
rected for elevation before plotting. The resultant velocities shown are average
velocities for each layer determined from the traveltime curves and the drill-hole
velocity determinations.
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
530
o
o
I
o
o
I
SONOO3SI9911N NI 3111.1.
Case Histories of Refraction Methods
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
/
/
I
Applications
SONOO3SIqqlIN
I
of Refractionin Mining
NI 31NI.I.
I
531
532 Case Histories of Refraction Methods
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
The undulating bedrock surfacewas derived from the emergence-angle and ray-
path tracing using the depth, dip, and velocitiesof eachupper layer. Only selected-
raypaths have been plotted for illustration.
Total traveltimes from the seismic records were readable to within one ms.
For the surfaceweatheringlayer with an averagevelocity of 3,000 ft/sec and a
geophonespacingof 200 ft, calculationsof the emergenceangleswould be in error
by no more than onedegree.Therefore,with reasonablecontrolof the parametersof
the upper layers and the velocity of the bedrock, the emergence-angleraypath
tracing techniqueresulted in accurate derivation of the undulating bedrock surface.
In the interpretation, convexitiesand concavitiesof the bedrock surfaceinfluenced
the shortest time path in the high-velocity bedrock, and consequentlysome trial
and error and cross-checkingof the various raypaths were necessaryin order to
arrive at the final undulating surface.
United Verde Mine
The United Verde Mine is located in Jerome, Yavapai County, Arizona, and is
one of the largest known copper-pyrite ore bodiesin the world. From its discovery
in 1876 until it was finally shut down in 1952, in excessof 30,000,000 tons of ore,
averagingapproximately 5 percent copper with significantamounts of gold and sil-
ver, were produced.
The ore body is in Precambrian and consistsof nearly massiveiron and copper
sulfidesin a curved lenticular pipe that plunged very steeply to the north. Vertical
extent was in excessof 3,000 ft with a maximum crosssectionapproximately 800 ft
by 700 ft. The main mass was nearly pure pyrite with copper-richsulfidepockets
locatedalongthe footwall. The oxidezoneextended160 ft from surfaceand a second-
ary enrichmentzone below this to 600 ft. A massivediorite stock forms the hanging
wall and quartz porphyry the footwall. Locally, patchesof chloritic schistremain on
the footwall side. A network of small basic dikes cut the sulfide mass parallel to
and near the hanging wall side.
The seismicsurvey at the United Verde Mine was an experimental program to
determine if massive sulfide bodies could be detected by seismicrefraction tech-
niques. The sulfide body here offered an excellentopportunity for such an experi-
mental program. The sulfide zone and surroundinggeologywere well known and
fortuitously the major massof the sulfideson the lower levels was still in place and
relatively undisturbed.This was mainly a pyrite zone with low coppercontent, and,
hence,had not beenmined out. At the time of the survey, stoppingon the lower levels
was confinedto selectedzoneswithin the main sulfidemassand along the footwall
contact only.
Five geophonespreadswere used undergroundon the 2,550 level; the spreadlay-
outs are shownin Figure 6. For each spread two shotpointslocated on surfacewere
used, as indicated in Figure 6. Both shotpointswere in the hanging wall diorite and
belowthe surfaceweathering.The geophoneswereon the footwall sideof the massive
sulfidezone and implanted in either black chloritic schistor quartz porphyry.
Ideally, the shotpointsand geophonespreadsshouldbe located so that two ap-
proximatelyperpendicularfan patterns would result with the spreadsextendingbe-
Applications of Refraction in Mining 533
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
TO $,
1700 FT HORIZ.
2120 FT VERT.
'-':'
'""-_-_-zz
2550
DRIFTS
SULFIDE OUTLINE
GEOPHONE
POSITION -:
4-, SPREAD NO.--GEOPHONE NO.
HIGH
VELOCITY
RAY,
HORIZ.
PROJECTION
SCALE IN FEET
? ,7o ,o ,oo
I
TO s
Co
Bo
AO
TO Sz
/
?.
ee
SULFIDE OUTLINE
j DERIVEDOUTLINE
GEOPHONE POSITION 2550 LEVEL
SCALE IN FEET
,o o o
1-1 0 0 0 1 21 0
2 1.5 32 175 1.5 32 0
3 5 105 420 3.5 74 0
4 9 190 555 1.5 32 160
5 12 253 485 0.5 10 145
6 27 570 440 3.5 74 220
7 16 337 365 2.5 53 245
8 30.5 643 345 6.5 137 240
2-1 3 66 300 29.5 620 370
2 5.5 116 300 20.5 432 410
3 8.5 179 295 15 316 410
4 13 274 320 15.5 327 415
5 8.5 179 370 12.5 264 415
6 10 210 395 16 338 430
7 9 190 460 10 211 400
8 12 263 375 13 274 485
3-1 11 232 375 21 443 485
2 13 274 460 29 612 390
3 13 274 450 28.5 600 310
4 13 274 445 29.5 621 175
5 19.5 410 375 26 548 210
6 ? ? 300 21 443 140
7 13 274 325 34.5 726 240
8 17.5 370 320 17.5 369 260
4-1 2.5 53 320 17.5 369 260
2 5.5 116 340 15.5 327 330
3 9 190 325 6 127 285
4 11.5 242 430 ? ? 230
5 8.5 179 450 6 127 125
6 6 126 540 8 169 60
7 4 84 210 10.5 221 0
8 2.5 53 0 ! 21 0
5-1 13 274 445 21.5 453 175
2 12 263 450 ? ? 240
3 12 263 460 25 527 250
4 13 274 240 24 506 230
can be obtained. This was done for various levels using vertical and horizontal
projectionsof the thicknessesof the sulfidesas indicated by the two crossedfans.
The straight-line raypaths interceptedthe sulfidebody approximately at the 2,100 ft
level due to the vertical displacementof the shotpoints relative to the geophone
spreadson the 2,550 ft level.
The probable outline and position of the derived sulfide body are shown in com-
parison with the known sulfidebody in Figure 7. Under the circumstancesand pro-
ceduresused, the agreementis remarkably good. This method cannot be expected
to give accurate indications of sulfide position and extent due to the simplified
assumptionsusedin the analysis.However, it has indicated the presenceof the low-
velocity sulfideswith some measureof extent and position.
This would be sufficient justification for employment of this method in mining
exploration. Normally, detection and determination of approximate position and
extent of a sulfide body are all that is required. Measurement of time differences
Applications of Refraction in Mining 537
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
and, therefore, the resultingtime lags and emergenceangles can now be made to a
higherdegreeof accuracythan at the time of this experimentalprogram. Therefore,
significantimprovementsin techniqueare possible.Refinementsof the method and
more detailed work might resolvesomeof the ambiguity.
CONCLUSIONS
ACKNOWLEDGMENTS
REFERENCES
INTRODUCTION
540
Case Histories of Refraction Methods
Oco
oEc
Seismic and Gravity Profile, Utah 541
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
slightly from those originally postulated by Lum (1957, Figure 4) prior to the
seismicsurvey.
Except along its east end, the seismicprofile was establishedessentiallyto co-
incide (within 1 miles) with the gravity profile. Thus, it was hoped that the
combination of seismicand gravity data would indicate the geologicalstructure of
the valley and permit a better evaluationof the densitycontraststhat are applicable
in the interpretation of the structuresof the basinsin the Basin and Range province.
INSTRUMENTATION
PROCEDURE
Refractionprofiles
Two eastward-trending refraction profiles were established east of the base of
Little Mountain South (Figure 1): one extendedbetween shotpoints1 and 2, a
distanceof 3,200 ft; and the other extendedbetweenshotpoints6 and 7, a distance
of 15,600 ft. These two profiles were both reversed. Also east of Little Mountain
South, an additional, northward-trending,reversedrefraction profile was established
between shotpoints3 and 4 to confirm the velocitiesof the other two profiles.
The refractionprofileswere established(1) to follow the Precambrianrock that
crop out on Little Mountain South until the horizon was deep enoughto obtain
reflectionsfrom it, and (2) to obtain velocity control to be used in the reflection
work.
usedto fill in betweenthem. The shot was offsetfrom the spreadby a maximum of
1,900 ft, in line with the profile.
Reflectionprofile
After the refraction profileswere completed,shotswere detonated at shotpoints
8 and 9 to obtain reflectionresults(Figure 2). Becausepoor reflectionsresultedand
becauseof the populatedarea,the profilewasmoved! milesnorth, wheresuccessful
results were obtained. The overlapping of the reflection and refraction profiles en-
abled a correlation of the results of the two methods. The reflection work was then
continued acrossthe valley wherever sufficientspacewas available for the spreads.
Questionableareas were filled in as the survey progressed.The dot-dash line on
Figure ! showsthe positionof the reflectionprofile.
The short lines extendingfrom shotpoints8 through 28 in Figure 1 indicate the
subsurfacereflection coverage.The direction of the spreadsis given by the direc-
tion of the lines. Dip shotswere made where two lines extend from a station at an
angleother than 180 degrees.In general,a 1,100-ft spreadwasusedon both sidesof
the shotpointwith station intervals of 200 ft. Two reflectionseismometers
that were
separatedby 24 ft at each station were usedas a meansof enhancingthe signal.
For all shotpoints,a Mobile auger drill, which was mounted on a Jeep pickup
truck, wasusedto drill shotholeswhich rangedin depth between10 and 30 ft. Sixty
percenthigh velocitydynamiteand Herculesno. 7 capswereusedfor all shots.For
the refraction work, the maximum charge used was 10 lb of dynamite and the
average was 5 lb. For the reflection work, the maximum charge used was 5 lb of
dynamite, and the average size was between 3 and 4 lb. On the average, about
four shotsper hole were detonated for the reflection work.
The time instant for the reflection work was recordedin a conventionalmanner,
and the time instant for the refraction work was transmitted from the blaster to the
recordingtruck by radio. The time instant in all casescould be read to _+0.001sec
Gravity profile
The gravity data were obtained with a Worden gravimeter using conventional
field techniques.Horizontal and vertical controlwere obtainedwith U.S. Geological
Survey topographicquadrangle maps, which included bench marks and elevations
of road intersections.The gravity values were corrected for instrument drift, lati-
tude variation, and combinedelevation effect (including free-air and Bouguer
effects).Terrain correctionsout to and includingzoneI (outer radiusof 8.44 km) on
the zone charts of the U.S. Coast and Geodetic Survey correspondingto the
Bullard modificationof the Hayford-Bowie method (Swick, 1942, p. 68) were also
applied. An averagedensity of 2.67 g/cc was assumedfor the rocksin making the
Bouguer and terrain corrections.The station values were tied to U.S. Coast and
Geodetic Survey pendulum station no. 49 in Salt Lake City, for which the simple
Bouguer anomaly value was taken as --183 mgal (Cook and Berg, 1961, p. 76;
Duerksen, 1949, p. 8). The accuracyof the finally reducedgravity values is be-
lieved to be within 0.5 mgal for most of the stations, and within one to two mgal
Seismic and Gravity Profile, Utah 543
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
I J 1 IIIIII I II I I I III II II
:: II I II I I I I II II II
II II I II I I I I II II II
II IIIII I I I I II II II
II ;'11 I/ II I III I II I
/ II I I /I I II I I
i // / / // // / // / /
I .// / I i /-/ / .// / /
i II i I i ii / /I i I
/ //. / / //t// t /// // / I
I I / /// II I I I II
I / I / /// // / I I //
.11 I ' I / // I I I
., / /////// ///// / /. /
/ ,,,S /// // / II I I I I I
,,I 1 /// // / II III 1 I
/// /
I / ""/
/1 ///////
// ,, ,,,,--
tt - t t
,, / / /
//II ,/ /I /
/
at the Wasatch front and within OgdenCanyon. The gravity valuesfor the inter-
pretive geologiccrosssectionwerecomputedwith a two-dimensionalJunggraticule'
no end correctionswere applied.
INTERPRETATION
Seismic results
546
o
,
o
,
XI IAV19
XIIAV19 -IVn01sq 1
3n9noa
o
o
o
Case Histories of Refraction Methods
Seismic and Gravity Profile, Utah 547
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
eastern edge of the valley was suggestedby the occurrenceof shallow reflections
which were interpreted to be causedby the basementrock of Precambrian and/or
Paleozoicage. No deep reflectionscould be picked on the records.The location and
attitude of the fault as shownon the time sectionis uncertain, and additional work
is neededin the area to resolvethis uncertainty.
The dips computedfrom individual recordsusing the assumedvelocitiesshoweda
generaltwo- to three-degreedip toward the centerof the profile. Severalof the com-
puted dips are shownon the time crosssection.The erratic dips are believedto be
causedby subsurfaceirregularities.For example,a few small westdipsin the upper
west-centralportion of the time sectionmay be causedby slumpingin the vicinity
of the fault zone.
The interpretedgeologicsectionis shownat the bottom of Figure 2. The faulting
wasplacedby the aboveinterpretation.The depthsto the horizonswere computed
by taking one-halfof the two-way reflectiontime and multiplyingthis time by the
averagevelocity that was applicablefor that two-way reflectiontime as read from
the columnof averagevelocities.On the assumptionthat the last correlatablereflec-
tionswerefrom the bedrock,the total indicatedthicknessof the valley fill alongthe
sectionis about 6,200 ft; thus the bedrocksurfacewould be about 2,000 ft below sea
level. It shouldbe emphasized,however,that this indicatedthicknessis probablya
minimum depth becausethe observedreflectionsfrom any goodseismicreflector--
as, for example,a basaltlava flow--within the rocksof Cenozoicagewouldlead to
an erroneousinterpretation of the total depth to the true bedrock.
It shouldbe noted that our resultsfor the minimum depth to the basementrock
agreeswell with thosefound independentlyin the samegeneralarea by McDonald
and Wantland (1960, p. 21), who concludedfrom their seismicreflectiondata that
"the basementrock might be as much as 6,000 ft deepin the bottom of the trough,
which lies some 5 miles west of Ogden." A test well, which was drilled in this area
by the U.S. Bureauof Reclamationincident to their ground-waterstudies,pene-
trated rocksof Cenozoicageto a depth of 3,006 ft, whichwasthe total depth of the
hole.
Gravity results
Figure 3 showsthe observed,assumedregional,and residualgravity, as well as an
interpretive geologiccrosssection,along the gravity profile that extendseastward
from Little Mountain South,throughOgden,and up OgdenCanyonin the Wasatch
Range.The small amountof topographicrelief for the WasatchRangeindicatedon
the crosssectionresultsfrom the fact that the elevationsusedin the profile are the
elevationsof the gravity stations,whichweretakenin the bottomof OgdenCanyon,
rather than the elevationsat the high mountain surface.
For the interpretive geologicprofile shownin Figure 3, the computedgravity
valuesare shownwith appropriatesymbolson or near the residualgravity curvein
accordance
with whetherthe assumeddensitycontrastbetweenthe valley fill of
Cenozoicage and the bedrock along each of the three assumedbedrock surfacesis
0.4, 0.5, or 0.6 gm/cm, respectively.It shouldbe notedthat the total gravity relief
due to the grabenis about44 mgal. Althoughit is believedthat the geologicstru-
548 Case Histories of Refraction Methods
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
tures shown are reasonable,it must be emphasizedthat instead of the few faults
which are postulated, a seriesof smaller step faults will give similar resultsfor the
theoreticalgravity. It is believed,however,that the total displacementalonga fault-
block mountain often occursby large movements along a relatively few major faults
rather than the cumulative effect of small movementsalongmany minor faults.
The indicated thicknessof the valley fill for the largest assumedaveragedensity
contrast (0.6 gm/cma) betweenthe bedrockand valley fill is about 7,600 ft. At this
depth, the bedrock surfacewould be about 3,400 ft below sea level. The rather low
compressionalseismicvelocities that extend to considerabledepth (Figure 2) in-
dicatethat the assumedaveragedensitycontrastof 0.6 gm/cmacouldconceivablybe
too low. If so, the estimated thicknessof the valley fill from the gravity results
(7,600 ft) would be too great.
Of the two faults shownon the west side of the valley, the southwardprojection
of the west fault at point B coincideswith Hooper Hot Springs,just off the map in
Figure 2; and it is therefore suggestedthat these springsare possibly due to this
fault (Lum, 1957,p. 12). The fault shownat point C' on the gravity profile(Figure3)
is interpreted as the southwardextensionof the fault indicated at point C on the
seismicprofile (Figure 1).
The northward and southward extents of the graben and its relationship to the
Wasatch trench are treated in another paper (Cook et al., in preparationfor publica-
tion).
CONCLUSIONS
ACKNOWLEDGMENTS
ABSTRACT
INTRODUCTION
TRANSPORTATION
The sameDecca 9f Lambda Survey Chain usedin 1960 was employedfor regional
navigation in 1961. The Master station was at Cape Isachsen on Ellef Ringnes
Island, Red Slave on the north shoreof Borden Island, and Green Slave on Meighen
Island. Positive lane identification was not included on the instruments.
It has been concludedthat the accuracy of the Decca chain is not sufficient for
long-rangeseismicrefraction surveying.This is emphasizedparticularly if there is a
bend in the line of detectors, becausedistancesof shotpoint to detector cannot be
determined with sufficientaccuracy. A Tellurometer was used in the 1962 survey
tying into the existing Shoran and Tellurometer network. A sky-wave interference
reflected by the ionospheremay also add to the inaccuraciesof the Decca Survey
systemduring certain eveninghours. The authors hasten to point out however that
while the Decca Systemis a dependablenetwork for many navigational and survey
requirements,the Decca Companyitself hasnever claimedthat their systemis suffi-
ciently accurate for seismicsurveying.
Referencing of the Decca instruments to a known location should be done while
the aircraft is stationary. If the Decca equipment is mounted in a helicopter, the
rotor bladesshouldbe rotating as before take-off to balanceout instrumental phase
shift and so eliminate the proximity of a rotor blade being randomly near to or far
from the Decca antenna.
552 Case Histories of Refraction Methods
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
AREAS SURVEYED
AXEL
I
CAPE ISACHSEI
REINDEER
DEER
ELLEF AMUND
BORDEN
RINGNES
IS.
IACKENZIE
KING
)ELCHER
LOUGHEED IS, CHANNEL
40 MILES I
secondaryarrivals.
Onplayback thetapespeed wasincreased to 13in/secto com-
pensate,
in some measure, forincreasedattenuation of thelowfrequencies.Texas
Instruments'seven-cycle,
modelS-43geophones wereusedto detectrefractedsig-
nalsovera spreadwitha generaldetectorintervalof240ft. Thesignalwasrecorded
usingalow-cutfilterof"out"andahigh-cut filtervariable
from40to92cpsdepend-
entupontheresponse characteristics
of theareaof interest.Theinstruments were
notpermanently installed
in fabricated
racks butwerefreeforfloorinstallationin
either a tent or helicopter.
Thefirststepin thefieldprocedure
wasto house
theinstruments
in a tentat a
Decca-controlled
site.The S-55helicopter
then movedshooting
personnel,
explo-
sives,andequipment
to selected
shotpointlocations
moving
progressively
further
awayfromthegeophone spreaduntilallvelocities
downto,andincluding,
20,000
ftfsecwererecorded.
Whenthisvelocity wasobserved,
theshooting
andrecording
crews interchanged
positions.
The shooting crewthenremainedstationary
while
the instruments
and geophonespreadweremovedby the helicopter.Reversal
spreads
werelaidoutat suitable
locations
dictated primarily
bysegmentsof the
time-distance
plotsand,secondly,
bythesuitability
oftheicetopography
tolaying
outthespread.
Nocontinuousreflection
profiles,
air,orsurface
shots
wereattempted
during the 1961 program.
554 Case Histories of Refraction Methods
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
METHOD OF COMPUTATION
The geographicalcoordinatesof each shotpoint and the center and both ends of
the detectorspreadwere computedfrom the Decca coordinatestabulated in the field.
The straight-line distance of each detector from every shotpoint was readily ob-
served when these coordinateswere plotted graphically. Arrival times at each geo-
phone station were plotted against distance to determine the velocities indicated
on the spread.Figure 2 is an exampleof a time-distanceplot with the reverseveloci-
ties plotted.
Two corrections,using the velocities indicated on the time-distance plots, were
applied to the total times of eachset of arrivals on eachrecord.This method of com-
putation will apply only to first arrivals of energy sincelater arrivals require a dif-
ferent set of corrections.The correctionsapplied are as follows'
(1). One correction minimizes the effect of the variable depth to the first sedi-
mentary layer underlying the water. The variable depth to this sedimentary layer
was assumedto correlatewith observedwater depths.This first correction,/xtd,is
SPREAD
X -- SURFACE
% sP.
'
i
VI (WATER)
II WATER
AVERAGE
DEPTH
UNDEROF
la (-rE! //SHOTNS
__ _ __ .... -- --
v F
- /;a ....
V3
(SECOND
SEDIMENTARY
YER)
//// ,.
Vh
Then t = V2 fromSnell
Sin02r s
Law
and from the fundamental equation
AfA&
' = - adltan
Vn
e2r.+ 6dl
V2 cos 02rt
= Ad cos G2n
V2
Fro.3.Diagram
andderivation
ofcorrection
applied
forthevariable
depth
ofthefirstsedimentary
layer.
thecorrection
applied
to first-arrival
timesanditsderivation
isindicated
in Figure
3. It is similar to a weatheringcorrection.
Theentirewaterlayerof velocityV couldhavebeenreplaced
in thiswayby
materialof subbottom
velocityV.and,therefore,
onlyonecorrection
wouldhave
beenrequired.
However,V.mayvaryoverthelengthoftheprofile
and,therefore,
toreplace
daswellasAdbyanerroneous
velocity
V.wouldintroduce
errors
propor-
tional to AVe. and d.
(2). A second
correction
wasthenapplied.
Variations
inbottom
topography
were
corrected
to a valueofdusinga valueof V.determined
closeto thespread
andthe
shotposition
waselevated fromitsresulting
depth,d,to thesurface of theice,
(wheretheicesurface
isthedatum plane),
bya correctionwhichisindependent of
V2.Thiscorrection,
Ata,
isderived
asindicated
in Figure
4.It issimilar
toanuphole
correction.
The totalcorrection
appliedis thesumof thesetwocorrections.
Therefore,
/XT = /xta-5-/Xta
/Xd cos 0u d cos 0
= --
Vu V
556 Case Histories of Refraction Methods
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
All time-distance data are therefore referred to the common ice surface and the effect
of the undulating velocity discontinuityat the water-bottom interface tends to be
eliminated as much as possiblewithout detailed information on V.. These correc-
tions were applied without considerationfor dipping velocity layers, but the final
results would be modified only slightly by dips up to a few degrees.
The thicknessesof the velocity layers indicated on the time-distanceplots were
computedusingthe standardformula
x ()
V =
2d
cos
V
0
GEOLOGY
DISCUSSION OF RESULTS
SPREAD SURFACE
.____ s,).
...........
ii/! APPLYING
V4 ETC.
Then
SinOn= V_L
Vn
fromSnell's
Low
andatd=-dtanO,n+ d
Vn V cos
= d cos
V
These two locationsare on the broad rim of the Sverdrup Basin as indicated on
the crosssectionby the nearestapproachto surfaceby the 20,000-ft/seclayer. The
!7,500-ft/sec layer appearsto pinch out or becomeseismicallyundetectableon the
rim. The rim also marks the southern facies change of the Beaufort Formation and
probablyof the post-Jurassic
sediments.On shorenear theselocations,outcropis of
the Lower CretaceousIsachsen Formation or the Upper JurassicDeer Bay Forma-
tion. These may not be presentoffshore,and, consequently,Jurassicbeds may lie
immediately under the sea-floorsediments.
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
558
Case Histories
of Refraction
Methods
Sverdrup Basin, Canadian Arctic Islands 559
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
2 3 4 5 5 6 7
_ ..I/5,ooo', .5,ooo.
_. -_ I /5-,000
k',,_ 9,000J<,,
9,000J<,,
_2 oo'(,__. -" lo.ooo,, -,'qo,000,,
.200, /' ",12,500,
_:,] ,6.--./_.
- ........
....
,4,aoo,,
'u5.,00020,500
. 14,500,,
14,500Y,
,'"'"'""_...- _ .--,v.e . ,, --
15,000, ------ -,
)OUU II I ; --e I
:' ,, .o,ooo,,
\ /e,soox,
e,soox,
2O-
o,ooox,
?
40-
2o,oooX,
50
o 3o
I m . m I
6o
Thequestion
arises
whether
theupliftindicated
at locations
3 and4 occurred
at
the time of the suggestedunconformityor whether it was there before such an un-
conformity and thereby a sourceof the sedimentson both sidesof the Basin rim.
The unconformitymay, therefore,be an important feature in the developmentof the
Sverdrup Basin itself, for the solution of this question may have a bearing on
whether similar sedimentarymaterials and conditionsmay occurinside and outside
the Basin. This in turn will have an influenceon the probable extent of potential
petroleum provinces. More geophysicaland stratigraphic information is required
before the initial questioncan be answered.
In this general area, Hobson (1962) recordsthe presenceof a material of low
velocity, 6,100 ft/sec, immediatelyunder the water asrecentlydepositedsediments
and alsoa velocity of 22,600 ft/sec asa secondaryarrival. Neither of thesevelocities
was observedin 1961, so that a further evaluation of their presencecannot be at-
56O Case Histories of Refraction Methods
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
tempted. However, analysisof the results of the 1960 program indicates the pres-
ence of a 7,000-to-8,000-ft-thick layer at Isachsen below what was then termed
"basement."If this layer indicatedby the 22,600-ft/secvelocity be extrapolatedto
the Kristoffer Bay profile of 1961, there may be approximately 64,000 ft of unmeta-
morphosedsedimentswithin the Sverdrup Basin. This extensionof sediment thick-
nessfrom the 1960 estimate of 40,000 ft is made possibleonly through information
at hand from the Dome et al. Winter Harbour no. 1 well (personalcommunication).
The gamma-soniclogs run in this hole indicate generally high sectional velocities
with recognizedformational and lithologicalcontactscorrespondingto large velocity
discontinuities.The average interval velocitiesobservedin the Winter Harbour hole
correspondvery favorably with those indicated in Figure 6 of this paper. An ex-
tremely high interval velocity averaging 22,500 ft/sec was recordedin the Lower
Cape Phillips and/or Cornwallis carbonateformation of Ordovician age at depth in
the Winter Harbour hole. Based on this velocity information the thicknessof the
unmetamorphosedsedimentswithin the Sverdrup Basin is probably in excessof
60,000 ft of which only 40,000 ft may be a sourcefor petroleum.
Kristoer Bay--location 5
Location 5 appearsto be near the center of the Sverdrup Basin. This profile was
reversedover a spreadof approximately 170,200ft using500 lb of explosivesin 775
ft of water. Fair to poor first arrivals of first energy were recordedat this distance,
whereasgood to fair first breaks were observedout to about 136,200 ft usingonly
75 lb of explosives.Shooting this profile in reverseproduced no indication of dip
north or south lending further weight to the conclusionthat this profile is near the
center of the Basin.
Since Permian formations are present at the surface within the domal structures
east and south of Kristoffer Bay, Jurassicformations are probably presentat depth
and it is suggestedthat the 10,000-ft/seclayer shownon the sectionmay correspond
to theseJurassicbeds. There is no evidenceon which to correlate the 12,500-ft/sec
layer with a known geologicalformation, but it is suggestedthat this horizon is
probably Triassic in age.
year, and operating from a helicopterit was desirableto shoot from ice floeswhich
are seldom found where desired.
An interestingphenomenonis displayedon this profile by considerabledisplace-
ment of a segmentof the time-distancecurve. It can be interpreted as a fault, but,
usinga body velocity of 20,000 ft/sec, it could be an intruded dyke, about 2,000 ft
thick. This latter interpretation is reasonablesince such intrusions are known to
exist on Ellef RingnesIsland.
Location 7 is an unreversedprofile in Belcher Channel south of Ellef Ringnes
Island and very closeto location 6. In fact, the southernmost shotpoint of location
6 is the northernmostof location 7. If the time-distanceplot of location 7 is com-
pared with that of location 6 there is a distinct similarity. However, the data ac-
cumulatedfrom location7 are toosparseto allowa rigorousinterpretation.This pro-
file wasshotearly in June, 1961,and operationaldifficultiespreventedthe recording
of data from a sufficientnumber of locationsfor a meaningfulinterpretation. Profile
7 was unreversedover approximately 100,400 ft over water depthsranging from 350
to 845 ft. Extensive movement of the ice prevented the recording of sharp first
arrivals. A reflection velocity survey was attempted in this area but was unsuccess-
ful.
The crosssectionat locations6 and 7 indicatesa noticeabledecreasein velocity
within the 17,500-ft/seclayer. This may be due, as suggestedearlier, to a disappear-
ance of the suggestedunconformity at the top of the 17,500-ft/sec layer.
SUMMARY OF RESULTS
1. The 1961 seismicprogram has outlined the northern rim and the center of the
$verdrup Basin.
2. The thicknessof the unmetamorphosedsedimentswithin the $verdrup Basin,
from seismicrefraction data and interval velocity data from Dome et al. Winter
Harbour no. 1 well, may be approximately 64,000 ft with probably 40,000 ft of that
sectionbeing a potential sourceof petroleum. The lower sedimentsare probably too
dense to be reservoirs.
$. A crosssectionexpressedin seismicvelocitiespresentsthe results of the 1961
program acrossthe $verdrup Basin. No direct attempt has been made to correlate
seismicvelocitieswith lithology due to the scarcityof geologicaldata at depth within
the area.
ACKNOWLEDGMENTS
The data discussedin this paper were the direct result of a cooperativeeffort on
the part of the authorsand G. W. Sander,W. Tyrlik, and J. Parkhill. Their coopera-
tion in the field is gratefully acknowledged. Information by Dome Petroleum
Limited and their partners in the Winter Harbour hole is alsoacknowledged.E. F.
Roots, as coordinator of the Polar Continental Shelf Project, has an untiring inter-
est in the seismicprogram. R. Thorsteinsson,of the GeologicalSurvey of Canada,
advisedin the correlationof the seismicvelocitieswith lithology. Acknowledgments
are due H. A. MacAulay who checkedthe computations and many others who con-
tributed to the successof the program.
562 Case Histories of Refraction Methods
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
REFERENCES
BIBLIOGRAPHY
Section
8
OF REFRACTION
PAPERS
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
Section 8
564
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
A. SEG BIBLIOGRAPHY
1. Antoine, J., and Ewing, J. I., 1963, Seismicrefraction measurementson the marginsof the Gulf of
Mexico:Jour. Geophys.Research,v. 68, no. 7, p. 1975-1996. F
2. Antoine, John W., and Harding, JamesL., 1965, Structure beneathcontinentalshelf,northeastern
Gulf of Mexico:Am. Assoc.PetroleumGeologists Bull., v. 49, no. 2, p. 157-171. F
3. Antoine, John W., and Henry, Vernon J., Jr., 1965, Seismicrefractionstudy of shallowpart of
continentalshelf off Georgiacoast:Am. Assoc.Petroleum GeologistsBull., v. 49, no. 5, p. 601-
609. F
4. Behrendt, John C., 1963, Seismicmeasurementson the ice sheetof the Antarctic Peninsula:Jour.
Geophys.Research,v. 68, no. 21,p. 5973-5990. F
5. Berg, JosephW., Jr., 1963, Seismicrefractionprofiles,NarragansettBay, Rhode Island: Geol. Soc.
AmericaBull., v. 74, no. 11,p. 1305-1312. F
6. Berg, JosephW., Jr., Cook, Kenneth L., Narans, Harry D., Jr., and Leamer, Richard J., 1961,
Seismicprofilesin northwesternUtah: Pilot Range and GrouseCreek range area: Jour. Geophys.
Research,v. 66, no. 4, p. 1255-1263. D
7. Berger,J., Blanchard,J. E., Keen, M. J., McAllister, R. E., and Tsong,C. F., 1965,Geophysical
observationson sedimentsand basementstructure underlying Sable Island, Nova Scotia: Am.
Assoc.PetroleumGeologists Bull., v. 49, no. 7, p. 959-965. F
8. Birch, William B., and Dietz, Frank T., 1962, Seismicrefraction investigationsin selectedareas of
NarragansettBay, RhodeIsland:Jour. Geophys.Research,v. 67, no. 7, p. 2813-2821. F
9. Blaik, Maurice, Northrop, J., and Clay, C. S., 1959, Someseismicprofilesonshoreand offshoreLong
Island, New York: JournalGeophysicalResearch,v. 64, no. 2, p. 231-239.
10. Bonini, William E., and Hickok, Eugene A., 1958, Seismic-refractionmethod in ground-water ex-
ploration:Mining Engineering,v. 10,no. 4, p. 485-488. A
11. Bonini, William E., and Woollard, GeorgeP., 1950, Sub-surfacegeologyof North Carolina-South
Carolina coastalplain from seismicdata: Am. Assoc.Petroleum GeologistsBull., v. 44, no. 3, p.
298-315. F
12. Bouchon,R., Ortynski, H. I., de Lapparent, C., and Pommier,G., 1959,The developmentof seismic
refractionsurveyingin the geologicinterpretation of the northern Sahara--Its role in the discovery
and the study of the Hassi Messaoudfield: World Petroleum Cong., 5th, New York, Proc., sec. 1,
paper39, p. 729-744. F
13. Brown,M. Vertner, 1962,Seismicprofilein St. Andrew Bay (Florida): Jour. Geophys.Research,v.
67, no. 11,p. 4513-4515. F
14. Brown, M. V., Northrop, John, Frassetto,Roberto, and Grabner,L. H., 1961, Seismicrefraction
profileson the continentalshelfsouthof Bellport,LongIsland,New York: Geol.Soc.AmericaBull.,
v. 72,no. 11,p. 1693-1706. F
15. Bunce,Elizabeth T., and Fahlquist, Davis A., 1962, Geophysicalinvestigationsof the Puerto Rico
trenchand outerridge:Jour. Geophys.Research,v. 67, no. 10,p. 3955-3972. F
16. Bush, Boyd, and Schwarz,SigmundD., 1964, Report on geophysicalinvestigationfor reconstruc-
tion studies,Anchorage,Alaska, in Anchoragearea soil studies,Alaska: Seattle, Shannonand Wil-
son, Inc., App. F. p. F3-F4. F
17. Cagniard,Louis, 1962,Reflectionand refractionof progressiveseismicwaves: [translationand revi-
sionof Cagniard'soriginalbook (1939) by Edward A. Flinn and C. Hewitt Dix], New York, Mc-
Graw-Hill Book Company, Inc. 282 p. H
18. Conwell, C. L., 1952, Seismicinvestigationsof areas near Ephrata, Washington-ColumbiaBasin
project:U.S. Bur. Reclamation,GeologyRept. G-120, 23 p. F
19. Cook,John C., 1963,Seismicreconnaissance on an ice-coveredAntarctic Sea (with Frenchand Ger-
man abstracts):Jour. Glaciology,v. 4, no. 35, p. 559-568. F
20. Crary, A. P., 1961,Marine-sedimentthicknessin the easternRossSeaarea, Antarctica: Geol. Soc.
AmericaBull., v. 72,no. 5, p. 787-790. F
21. 1963, Seismicstudies,in Resultsof United States traversesin East Antarctica, 1958-1961:
Am. Geog.Soc.IGY World Data Center,A Glaciol.Rept., no. 7, chap.4, p. 39-67. D
22. Currier, L. W., 1960,The seismicmethodin subsurfaceexplorationof highwayand foundationsites
in Massachusetts:U.S. Geol. SurveyCirc. 426, 8 p. A
23. Dobrin, M. B., et al., 1946, Seismic-refractionsurvey of Bikini Atoll: Geol. Soc. America Bull.,
v. 57, no. 12,pt. 2, p. 1189. F
24. Dobrin, M. B., Perkins, Beauregard,Jr. and Snavely, B. L., 1949, Subsurfaceconstitutionof Bikini
Atoll as indicated by a seismic-refraction survey: Geol. Soc.America Bull., v. 60, no. 5, p. 807-
828. F
25. Donath, Fred A., and Kuo, John T., 1962, Seismic-refractionstudy of block faulting, south-central
Oregon:Geol.Soc.AmericaBull., v. 73,no.4, p. 429-434. F
26. Dow, Willard, 1960,A telemeteringhydrophone:Deep-SeaResearch,v. 7, no. 2, p. 142-147. C
27. Dudley, William W., Jr., and McGinnis, Lyle D., 1964, Seismic-refraction and earth-resistivity
investigationof hydrogeologicproblemsin Humboldt River Basin, Nevada: Nevada Univ. Desert
Research
Inst. Tech.Rept.,no.2, pt. 1,p. 7-23. F
Other U.S. Bibliography 569
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
28. Eby, J. B., and Harkins, T. I., 1942. Geophysicalhistory of Darrow Dome: Am. Inst. Min. Met.
Eng. Tech. Pub. 1495,8 p.
29. Ewing,JohnI., 1963,Elementarytheoryof seismicrefractionand reflectionmeasurements, in The
sea--Ideas and observationson progressin the study of the seas,v. 3, The earth beneaththe sea
(and) history:New York, IntersciencePublishers,Chap. 1, p. 3-19.
30. Ewing, John I., Antoine, J., and Ewing, Maurice, 1960, Geophysicalmeasurements in the western
Caribbeanandin the Gulf of Mexico:Jour.Geophys.Research,v. 65, no. 12,p. 4087-4126. F
31. Ewing,JohnI., and Tirey, G. B., 1961,Seismicprofiler:Jour. Geophys.Research,v. 66, no. 9, p.
2917-2927. C
32. Ewing, John I., Worzel, J. Lamar, and Ewing, Maurice, 1962, Sedimentsand oceanicstructural
historyof the Gulf of Mexico:Jour.Geophys.Research,v. 67, no. 6, p. 2509-2527. F
33. Ewing,Maurice,1938,Topographyof the buriedlandsurfaceof theformercontinentof Appalachia:
From seismicevidence:Geol.Soc.AmericaBull., v. 49, no. 12,part 2, p. 1948-1949. F
34.- 1947, Geophysicalinvestigationsof continentalborders:Tulsa Geol. Soc. Digest, v. 15,
p. 77. F
35. Ewing, Maurice, Crary, A. P., and Rutherford, H. M., 1936, Geophysicalinvestigationsin the
emergedand submergedAtlantic CoastalPlain; part I, Methods and results:Geol. Soc.America
Proc.,p. 75. F
36. Ewing, M., Ludwig, W. J., and Ewing, J. I., 1963, Geophysicalinvestigationsin the submerged
Argentinecoastalplain. Part I. BuenosAiresto PeninsulaValdez:Geol.Soc.AmericaBull., v. 74,
no. 3, p. 275-292. F
37. Ewing,Maurice, Sutton,G. H., and Officer,C. B., Jr., 1954,Seismicrefractionmeasurements in the
AtlanticOcean,Part VI: Typical deepstations,North Atlantic Basin:Seismol.Soc.AmericaBull.,
v. 44, no. 1, p. 21-38. F
38. Ewing,Maurice,Woollard,G. P., and Vine, A. C., 1938,Geophysicalinvestigationsin emergedand
submergedAtlantic CoastalPlain, part 4, Cape May, New Jersey,Section:Geol. Soc. America
Bull., v. 49, no. 12, part 2, p. 1880. F
39. 1939,Geophysical investigations
in the emergedandsubmerged AtlanticCoastalPlain,Part
3, BarnegatBay, New Jersey,section:Geol.Soc.AmericaBull., v. 50, no. 2, p. 257-296. F
40. 1940,Geophysical investigations
in the emergedandsubmerged AtlanticCoastalPlain, Part
4, CapeMay, N.J., section:Geol.Soc.AmericaBull., v. 51, no. 12,pt. 1, p. 1821-1840. F
41. Ewing, Maurice, Woollard,G. P., Vine, A. C., and Worzel, J. L., 1946,Recentresultsin submarine
geophysics: Geol.Soc.AmericaBull., v. 57, no. 10,p. 909-934. F
42. Fritz, AxelM., 1960,The MD engineering seismographandits applicationto highwayengineering:
11thAnnualSymposium
on HighwayEngineering
GeologyProceedings,
FloridaGeol.Survey,p.
55-65. A
43. Gaskell,T. F., 1956,Seismic
prospecting
in the Middle East:Oil Forum,v. 10,no. 7, p. 242-243,
254. F
44. Gaskell,T. F., and Swallow,J. C., 1952,Seismicrefractionexperiments
in the Pacific:Nature, v.
170,no.4337,p. 1010-1012. F
45. Geophysical InstrumentCo., Washington, D.C., Seismic prospecting
apparatus refractiontype:
Circular "C," 4 p. C
46. Gregory,J. N., 1960,Portabletimeroffersnewseismic
applications:
Oilweek,v. 11,no.28,p. 28-
30. C
47. Gutenberg,
Beno,andBuwalda,J.P., 1938,Geophysical
investigation
of YosemiteValley:Geol.
Soc.AmericaProc.for 1937,p. 240. F
48. Harris,Sidon,1947,Modernseismic
techniques
appliedto geophysical
exploration
in WestTexas:
Oil and Gas Jour., v. 46, no. 11, p. 60-63. F
49. 1948,Modernseismic refractiontechniques:
WorldPetroleum,v. 19,no.3, p. 72-76. G
50. Healy, JohnH., 1963,Crustalstructurealongthe coastof Californiafrom seismic-refraction
mea-
surements:Jour. Geophys.Research,v. 68, no. 20, p. 5777-5787. F
51. Healy, JohnH., Steinhart,JohnS., and Meyer, Robert P., 1965,Seismicrefractionstudiesin the
Transcontinental Geophysical Survey:Am. Geophys.UnionTrans.,v. 46, no. 2, p. 383-384. F
52. Hill, Maurice N., 1963, Single-shipseismicrefractionshooting,in The sea--Ideasand observa-
tionsonprogress in the studyof the seas,v. 3, The earthbeneaththesea(and)history:New York,
IntersciencePublishers,chap.3, p. 39-46. G
53. Houtz,R. E., andEwing,J. I., 1963,Detailedsedimentary
velocities
fromseismic
refraction
profiles
in thewesternNorth Atlantic:Jour.Geophys.Research,
v. 68,no. 18,p. 5233-5258. F
54. 1964,Sedimentary velocities
of the westernNorth Atlanticmargin:Seismol.Soc.America
Bull. v. 54, no. 3, p. 867-895. F
55. International
Geophysical
YearBulletin(No.33), 1960,Seismic
crustalstudies
duringtheIG, pt.
1:Marineprogram:Am. Geophys.
UnionTrans.,v. 41, no. 1, p. 107-113. A
56. International
Geophysical
YearBulletin(No.34), 1960,Seismic
crustalstudiesduringtheIGY, pt.
2: Continental
program:Am. Geophys.
UnionTrans.,v. 41, no.2, p. 351-355. A
57. Jameson,
M. H., 1937,Wavefrontsencountered
in weathering
determinations:
MinesMag.,'. 27,
no.9, p. 16-17. K
570 Bibliography of Refraction Papers
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
58. Janoschek,R. H., 1959, Oil explorationin the MolasseBasin of westernAustralia(with French
abstract):World PetroleumCong.,5th, New York, Proc.,sec.1, paper47, p. 849-859. F
59. John,A.M., andWesley,R. H., 1952,Applicationsof seismicmethodsto foundationexploration,in
Symposium onsurfaceandsubsurfacereconnaissance:Am. Soc.for TestingMaterials,SpecialTech.
Pub. 122,p. 142-150. A
60. Jones,J. H., 1938, The refractionmethod of seismicprospecting,in Scienceof Petroleum:v. 1,
p. 282-286,New York, OxfordUniv. Press. H
61. Katz, Samuel,Edwards,Richard S., and Press,Frank, 1953, Seismicrefractionprofileacrossthe
Gulf of Maine: Geol. Soc.AmericaBull., v. 64, no. 2, p. 249-252. F
62. Kovach, Robert L., Allen, ClarenceR., and Press,Frank, 1962, Geophysicalinvestigationsin the
Coloradodeltaregion:Jour.Geophys.Research,v. 67, no. 7, p. 2845-2871. F
63. Lee, F. W., Farnham,F. C., and Raspet,A., 1940,The seismicmethodfor determiningdepthsto
bedrockas appliedin the Lowell quadrangle,Massachusetts: Massachusetts Dept. Public Works,
U.S. Dept. Interior, Geol. Survey, Cooperativegeologicproject, Specialpaper 3, 46 p. F
64. Leet, L. D., 1938,Longitudinalvelocitiesin someweatheredand unweatheredCarboniferous rocks:
Seismol.Soc.AmericaBull., v. 28, no. 3, p. 163-168. I
65. Linehan, Daniel, 1952,Seismology appliedto shallowzoneresearchin Symposiumon surfaceand
subsurfacereconnaissance: AmericanSocietyfor testingMaterials SpecialTechnicalPublication
no. 122,p. 156-170. K
66. Ludwig,W. J., Ewing,Maurice,Ewing,JohnI., and Drake, C. L., 1962,Discussion of a paperby
C. It. Sayit,D. M. Blue,and J. G. Smith,Explorationseismictechniques appliedto oceaniccrustal
studies:Jour. Geophys.Research,v. 67, no. 12,p. 4946-4947.
Reply by authorsto the precedingdiscussion: ibid., p. 4948-4955. G
67. Ludwig, William J., Ewing, JohnI., and Ewing, Maurice, 1965,Seismic-refraction measurements in
the MagellanStraits:Jour.Geophys.Research,v. 70,no. 8, p. 1855-1876. F
68. Lyons,P. L., 1946,The low velocitylayer in seismicexploration,parts 1-2: Mines Mag., v. 36, no.
11,p. 501-508. K
69. Macelwane,J. B., 1937,Seismicprospecting: MissouriAcad. Sci. Proc.,v. 3, no. 4, p. 133. A
70. Macpherson,J. D., 1962, Seismicrefractionmeasurements of the bottom structureoff the north
shoreof PrinceEdwardIsland: Seismol.Soc.AmericaBull., v. 52, no. 2, p. 399-404. F
71. 1962,A newinterpretationof seismicrefractionprofilesobtainedin the Hartlen Point region
of the ScotianShelf:Seismol.Soc.AmericaBull., v. 52, no. 3, p. 693. F
72. McGinnis, Lyle D., 1964, Seismicstudiesof three areasin northern Nevada: Nevada Univ. Desert
ResearchInst. Tech. Rept., no. 2, pt. 2, p. 25-37. F
73. Milne, Allen R., 1964, Two seismicrefraction measurements:North PacificOceanBasin and Dixon
Entrance:Seismol.Soc.AmericaBull., v. 54, no. 1, p. 41-50. F
74. Mining Engineering,1959, Seismicanalysisaids in overburdenremoval: Mining Eng., v. 11, no.
8, p. 803-804. A
75. Mooney, Harold M., and Kaasa, Robert A., 1958, New refractionseismograph: Rev. Sci. Instru-
ments,v. 29, no. 4, p. 290-294. C
76. Nesbit, R. It., 1950,The work of the geologistin civil engineering: Ohio State Univ., Eng. Exper.
Sta. News, v. 22, no. 2, p. 4-8, 23-25. A
77. Northrop, John, and Ransome,Morris, 1962, Some seismicprofilesnear the western end of the
PuertoRicoTrench:Jour. of GeneralPhysiology,v. 45, no. 4, pt. 2, supplement,p. 243-251. F
78. O'Brien, P. N. S., 1959,The useof amplitudeand frequencyin explorationseismology(with French
abstract):World PetroleumCong.,5th, New York, Proc., sec.1, paper 34, p. 627-642. G
79. Officer,C. B., Jr., and Ewing, Maurice, 1954, Geophysicalinvestigationsin the emergedand sub-
mergedAtlantic CoastalPlain, Part VII: Continentalshelf,continentalslope,and continentalrise
southof Nova Scotia:Geol.Soc.AmericaBull., v. 65, no. 7, p. 653-670. F
80. Pakiser, Louis C., Press,Frank, and Kane, Martin F., 1960, Geophysicalinvestigationof Mono
Basin,California:Geol.Soc.Amer. Bull., v. 71, no. 4, p. 415-448. F
81. Parks,L. B., and McKay, A. E., 1947,Geophysical historyof Block 31 field, CraneCounty,Texas
(abstract):Oil and GasJour.,v. 45, no.47, p. 126. F
82. Pekeris,C. L., 1947,New methodsfor interpretationof seismicrecordsfrom underwaterexplosions:
Oil and GasJourn.,v. 45, no. 47, p. 126. G
83. Phillips,RichardP., 1964,Seismicrefractionstudiesin Gulf of California,in Marine geologyof the
Gulf of California--A symposium:Am. Assoc.PetroleumGeologists Mem. 3, p. 90-121. F
84. Pratt, J. G. D., 1960,SeismicsoundingsacrossAntarctica:Trans-AntarcticExped., 1955-1958,Sci.
Repts., no. 3, 69 p. F
85. Press,Frank, and Beckmann, Walter, 1954, Geophysicalinvestigationsin the emergedand sub-
mergedAtlantic coastalplain: Part VIII Grand Banks and adjacent shelves:Geol. Soc.America
Bull., v. 65, no. 3, p. 299-314. F
86. Raitt, RussellW., 1963,The crustalrocks,in The sea--Ideas and observationson progressin the
study of the seas,v. 3, The earth beneaththe sea(and) history:New York, IntersciencePublishers,
chap.6, p. 85-102. I
87. Robin, G. de Q., 1958, Seismicshootingand related investigations:GlaciologyIII, Norwegian-
British-Swedish AntarcticExpedition,1949-52' Sci.Results,v. 5, pt. 3, p. 134. F
Other U.S. Bibliography 571
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
63, 67, 70, 71, 72, 73, 77, 79, 80, 81, 83, 84, 85, 87, 88, 89, 91, 95, 97, 98, 99,
100, 101, 103, 104, 108, 112
G. Techniques: 6, 22, 23, 24, 32, 34, 47, 63, 65, 70, 83; 49, 52, 66, 78, 82, 90, 92, 96, 105,111,113
H. Theory: 3, 12, 13, 14, 20, 28, 35, 72, 79, 80, 84, 87, 93; 17, 29, 60, 64
I. Velocity: 67, 68, 74, 81, 89, 95; 64, 86, 93
J. Waves: 18, 29, 57, 85
K. Weathering: 5, 15, 16, 27, 57, 65, 68
1. Abdulayev, R. A., 1961, Comgositionof normal traveltime curves and determinationof mean
velocity to refracting boundarieswith the aid of nomograms:Razvedochnayai Promyslovaya
Geofizika,no. 40, p. 18-22.
2. 1960, Tectonicsof the deep horizonsof Lokbatan and the Khudat-Khachmasof the cis-
Caspianregionfrom seismicprospecting data: Vyssh.Ucheb.ZavedeniyIzv., Neft' i Gaz,no. 10, p.
9-12.
3. Adachi,Ryuzo,1957,A methodof analysisof seismicrefractionexploration:Butsuri-Tanko,v. 10,
no. 4, p. 184-190.
4. --- 1959,An experimentand its explorationconcerningseismicrefractionmethod:Kumamoto
Iour. Sci.ser.A, v. 4, no. 1, p. 99-102.
5. Xdfim,Oszkfir,
andKilnyi,lgva,1964,Determination
oftheapproximate
velocityfunction
from
refractiontraveltimecurves:Geofiz.K6zlemnyek,v. 12, nos.3-4, p. 67-78.
6. Akrabova,A., 1961,Applyingthe methodof refractedwavesto the areasof fault tectonicsin the
Kamchiaforedeep:Minno Delo i Metalurgija,v. 16,no. 12,p. 25-30.
7. Alfano,Luigi, 1959,On the formandamplitudeof the refractedimpulse'Quad.Geofisica Appl.,v.
20, p. 15-57.
8. Allen,Adrian, 1960,Seismicrefractioninvestigations of the preglacialvalley of the River Teifi near
Cardigan:Geol.Mag., v. 97, no. 4, p. 276-282.
9. Andreyev,N. S., andAndreyeva,K. S., 1958,Seismicexplorationof bauxitedeposits: Vses.nauchne-
issled.inst. metodikii tekhniki razvedkiTrudy, no. 1, p. 309-322.
10. Averbukh,A. G., 1964,Determinationof the positionof a refractingboundaryand wave velocities
from traveltimecurvesof discontinuous refractionprofiles:PrikladnayaGeofizika,no. 38, p. 58-68.
11. Aver'yanov,A. G., 1962, On the appearanceof refractedwavesin deep seismicsoundingin the
northwestpart of the OkhotskSea,in GlubinnoyeSeysmicheskoye ZondirovaniyaZemnoyKory v.
SSSR:Moscow,Akad. Nauk SSSR,p. 342-362.
12. Banerjee, S. L., and Venkateswarlu,V., 1963, Geophysicalinvestigationsat Sharavati Valley
HydroelectricProjectsites,Mysore,in Geophysical exploration,a symposium,Baroda,India, 1959:
New Delhi, India, Councilof Scientificand Industrial Research,p. 151-161.
13. Barenboym,M. I., 1953,Dynamiccharacteristics of the refractedwaveson the slopesof salt domes:
PrikladnayaGeofizika.,vypusk10,p. 84-96.
14. Behrens,Jorn, 1960, Seismicrefraction measurementsat sea (with English abstract): Zeitschr.
Geophysik,v. 26, no. 4, p. 161-176.
15. Beranek,Bretislav,and Zounkova,Milada, 1957(1958), Resultsof seismicrefractioninvestigation
in the RimavskaSobotaregion:Ceskoslovenska Akad. Ved Geofys.Ustav Prace, Geofys.Sbornik,
no. 66, p. 367-379.
16. 1959,Interpretationof seismicrefractionmeasurements in the RimavskaSobotaareausing
the zero-timerefractionmethod(with Russianand Germansummaries): PraceUstavupro Naftovy
Vyzkum,v. 13, no. 45, p. 73-86.
17. 1964,On the methodof seismic
workin thesouthpart of the Carpathianforedeep:Cesko-
slovenska
Akad. Ved Geofys.Sbornik1963,no. 184,p. 205-326.
18. Ber.roth,
A., 1937,Problems
ofthesurveying
engineer
in prospecting
forores,applied
geophysics
for
engineers:Allg. Vermessungs-Nachr., no. 49, p. 553 and 569, Berlin.
19. Berzon,I. S., 1948, On the determinationof seismicvelocitiesfrom the initial pointsof surficial
hodographs of Mintrop waves(in Russian):Akad.Nauk SSSRIzvestiia,Ser.Geog.i. Geofiz.,v. 12,
no. 6, p. 561-566.
20. 1949, Interpretation of travel time curvesof Mintrop's wavesfor a refractingboundaryof
arbitrary shape(in Russian):Akad. Nauk SSSRIzv., Ser.geobg.i. geofiz.,v. 13,no. 6, p. 545-555.
21. 1951,Determinationfrom experimental data of the powerexponentof the functionrepre-
sentingthe spreading of refractedwaves(in Russian):Akad. Nauk SSSRIzv., Ser. geofiz.,no.4, p.
1-30.
22. 1951,The revealingof dynamicpropertiesof seismicrecordsas influencedby differences in
the sensitivityof individualreceivingchannels:
Akad. Nauk SSSRIzv., Ser.geofiz.,no. 6, p. 5-10.
23. 1952,Seismicwavesin verticallystratifiedmedium:Akad. Nauk SSSRIzv., Ser.geofiz.,no.
3, p. 3-33.
24. 1956, Determinationof vertical discontinuitiesby using dynamic propertiesof seismic
refractedwaves:Akad. Nauk SSSRIzv. Ser.geofiz.,no. 11,p. 1294-1308.
Outside U.S. Bibliography 573
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
tion and refraction boundaries within the northern border of the Black Sea Basin: Akad. Nauk
Ukrayin. RSR Dopovidi, no. 3, p. 323-327.
227. O'Brien,P. N. S., 1963,A noteon the reflectionof seismicpulseswith applicationto secondevent
refraction shooting:Geophys.Prosp., v. 11, no. 1, p. 59-72.
228. Oelsner,Christian, 1965, Hammer-blow seismicinvestigationsundergroundfor the determination
of the physicalpropertiesof rock in place:FreibergerForschungshefte, C 178, 128 p.
229. Officer,Charles,B., 1959, On someoffshoreseismicrefractionprofilesin the Cook Strait, Tasman
Bay, and Golden Bay areas of New Zealand: New Zealand Jour. Geologyand Geophysics,v. 2,
no. 2, p. 350-354.
230. Oks, Meyer, 1938,Influenceof elasticanisotropyof layersof seismicrefractioncalculations:Bol.
Inform, Petrol., v. 15, no. 161, p. 109-118.
231. Oliveira, Geraldode, 1960,Analytic and graphicdeterminationof the refractinghorizonby means
of envelopeparabolas:Bol. Tecn. Petrobras,v. 3, no. 1, p. 5-19.
232. OverseasGeologicalSurveys, 1962, Geophysicalfield surveys---Bechuanaland 1961: Overseas
Geol. SurveysRept. for 1960-61, p. 12-14.
233. Paicu, Dumitru, and Patrichi, Constantin, 1961, Seismicresults obtained on metamorphicand
igneousformations:Acad. Romine, Problemede Geofizica,v. 1, p. 205-212.
234. Palmason,Gudmundur, 1963, Seismicrefractioninvestigationof the basalt lavas in northern and
easternIceland: Jokull, v. 13, p. 40-52.
235. Perrier, Guy, 1963. Seismicwaves registeredin the Forez Hills: Acad. Sci. (Paris) Comptes
Rendus,v. 257, no. 6, p. 1321-1322.
236. Petkow (Petkov), I. N., 1959, On an analytical method of determining the layer velocitiesof
seismicwaves: Geophysiku. Geologie,no. 1, p. 12-14.
237. Prodehl,C., 1964,Interpretation of refractionobservations in the BavarianAlpsforeland(quarry
explosions at Eschenlohe,1958-61)with regardto the depthof the basement:Zeitschr.Geophysik,
v. 30, no. 4, p. 161-181.
238. Puzyrev, N. N., 1960,Regardingapplicationof simplifiedproceduresof observationin the study
of the folded basementof the West SiberianLowland by the method of refractedwaves:Akad.
Nauk SSSR, SibirskoyeOtdeleniye, Geologiyai Geofizika, no. 11, p. 102-105.
239. 1963, Theory of interpretation of isolated seismic observations:Akad. Nauk SSSR
SibirskoyeOtdeleniye, Geologiyai Geofizika, no. 9, p. 66-82.
240. Pyzyrev, N. N., Krylov, S. V., Potap'yev, S. V., and Treskova, Yu. A., 1963, Seismicrefraction
soundingsfor regional geologicinvestigations: Akad. Nauk SSSR Sibirskoye Otdeleniye, Ge-
ologiyai Geofizika,no. 8, p. 55-67.
241. Quadir, M. A. 1959, A measurementand study of the velocity of seismicwavesin rocksof Quetta
and around: Pakistan Jour. Sci., v. 11, no. 4, p. 192-199.
242. Rabinovich, G., 1962, Determination of average velocity from compositetraveltime curves:
Razvedochnayai PromyslovayaGeofizika,no. 45, p. 42--47.
243. Radzhabov, M. M., 1958, Certain problemsof interpretations of singlelongitudinal traveltime
curvesof refracted waves: Akad. Nauk Turkmen. SSR Izv., no 4, p. 3-12.
244. 1958, The determination of the boundary velocitiesfrom transversetraveltime curvesof
refractedwaves,part 1: Akad. Nauk SSSRIzv. ser.geofiz.,no. 12, p. 1491-1503.
245. 1960,Selectionof meanvelocitiesin determinationof effectivedepthsfrom the individual
transverse traveltime curves of refracted waves: Akad. Nauk SSSR Izv. Ser. Geofiz., no. 6, p.
863-867.
246. Radzhabov,M. M., and Agranovskiy,L. Ye., 1960, Deternfination of depthsand relief of a dis-
continuity from individual transversetraveltime curvesof refractedwaves:Akad. Nauk SSSRIzv.
Ser. Geofiz.,no. 6, p. 854-862.
247. Raih, R. J., and Wyrobek,S., 1956,Improvementsin seismicprospecting: Nature, v. 177,no. 4522,
p. 1209-1210.
248. Redpath, Bruce B., 1961. Seismicoperations,in Jacobsen-McGillArctic researchexpeditionto
Axel HeibergIsland, PreliminaryReport 1959-1960:Montreal, McGill Univ., p. 101-107.
249. Reich,H., 1937,On the geologicinterpretationof seismicrefractionmeasurements: Oel und Kohle,
v. 13, no. 3, p. 53-59, Vienna.
250. 1937,Seismicrefractionmethodsin the geophysical surveyof Germany:Geophysik,v. 7,
no. 1, p. 1-16.
251. 1951, Results of seismicinvestigationsin the Peissenbergcoal mine: Erd61 und Kohle,
Jahrg.4, Heft 9, p. 538-542.
252. ---- 1953,On reflectedrefractionimpulses:Geol.Jahrb.,Band 67, p. 135-142.
253. -- 1960, Seismicinvestigationsof the Flysch trough at Lenggrieswest and east of the Isar:
Akad. Wiss.G6ttingenNachr., Math.-Phys. KI., no. 11, p. 205-255.
254. Reich, H., Schulze,G. A., and Fortsch,O., 1948, Geophysicalfindingsresultingfrom the Haslach
explosionin the southernBlack Forest:Geol. Rundschau,v. 36, p. 85-96.
255. Riabinkin,L. A., 1939,Seismicmethodsof prospecting(in Russian):RazvedkaNedr, v. 9, no. 1, p.
24-27, Moscow.
256. Richards,T. C., 1959,The evolutionof the geophone: Bur. Central Seismol.Internat. Pubs.,Ser.A,
Travaux Sci.,no. 10, p. 169-180.
Outside U.S. Bibliography 581
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
286. Sokolov,A. F., 1963,On the workingout of traveltime curvesof refractedwaves(in the first arrivals)
in reflectionsurveying:Razvedochnayai PromyslovayaGeofizika,no. 49, p. 32-38.
287. Solaini,Luigi, 1951,An exampleof seismicexplorationof a rocky foundation:Riv. Geofis.Appl, v.
12, no. 2, p. 73-82.
288. Sollogub,V. B., Chekunov,A. V., Kalyuzhnaya, L. T., and Khilinskiy, L. A., 1963, On the deep
structureof the Korostenpluton accordingto seismicdata: Akad. Nauk SSSR Doklady, v. 152, no.
5, p. 1212-1217.
289. Sollogub,V. B., Chekunov,A. V., Kalyuzhnaya, L. T., Khilinskiy, L. A., and Kharechke, G. Ye.,
1963, Internal structure of the crystallinebasementin the southwestpart of the Korostenbatholith
accordingto the seismicdata: Akad. Nauk Ukrain. SSR Inst. Geofiziki,Geofiz.Sbornik,no. 5 (7), p.
122-130.
290. Sollogub,V. B., Chekunov,A. V., Pavlenkova, N. I., and Khilinskiy, L. A., 1964, Nature of the
Novotsaritsa gravity anomaly in lowland Crimea accordingto data of seismicinvestigations:Akad.
Nauk Ukrain, SSR Inst. Geofiziki,Geofiz.Sbornik,no. 8(10), p. 3-12.
291. Spinicci,Gian Luigi, 1962,Aspectsand problemsof the refractionsurveyin the dunearea of the
Libyan Desert:Quadernidi GeofisicaApplicata,v. 23, p. 67-75.
292. Starodubrovskaya,S. P., 1957, Experiencesfrom an attempt to locatethe buried faults through the
dynamic characteristicsof refractedwaves:Akad. Nauk SSSRIzv. Ser. geofiz.,no. 5, p. 553-567.
293. Sterlyagova,G. V., 1964, Structure of the southeastpart of the CaspianSea accordingto data of
marineseismicinvestigations:Razvedochnayai PromyslovayaGeofizika,no. 51, p. 11-23.
294. Szenas,Gyorgy, 1960, Someproblemsof the developmentand applicationof the seismicmethod
(with Germanand Englishsummaries)'Geofiz.Kozlemenyek,v. 8, no. 4, p. 255-278.
295. Szenas,Gyorgy, and Geresben,Laszlo, 1955, Application of seismicrefraction measurementsto
bauxiteexploration:Magyar Allami EotvosLorand Geofiz.Intezet Geofiz.Kozlemenyek,kotet 4,
szam1, p. 67-74.
296. Taanila, P., 1963,A profilecalculationmethodin seismicrefractionsurveysbasedon the useof the
effectivevertical velocity: Geoexploration,no. 1, p. 36-49.
297. Tal'virskiy, D. B., 1959,Seismicprospectingof the basementin the southernpart of the Tobol zone
of the West SiberianLow]and: Prikladnaya Geofizika,no. 22, p. 3-23.
298. Tateishi, Tetsuo, 1951, Seismicprospectingat Toyosuto Coalfield,Hokkaido: Geol. Survey Japan
Bull., v. 2, no. 9, p. 29-34.
299. Telyakova, Z. Kh., 1961,Resultsof investigationsof the crustin the southpart of West Siberian
Lowland by the method of deep seismicsounding: Akad. Nauk SSSR, Sibirskoye Otdeleniye,
Geologiyai Geofizika,no. 7, p. 3-15.
300. T'eng, Chi-Wen, 1963, Dynamic features of seismicwaves propagatingin model mediumswith
right-anglefaults:Acta Geophys.Sinica,v. 12,no. 1, p. 75-92.
301. Teupser, Christian, and Olszak, Gerd, 1960, A contributionto the techniqueand applicationof
hammerblowseismicsurveying:Geophysiku. Geologie,no. 2, p. 84-88.
302. Teplitskiy, V. A., 1961,Someresultsof seismicprospectingoperationswithin the Bukharo-Gazly
and Chardzhou-Pitnyakzones:Vses.Nauchno-Issled.Geologorazved.Neft. Inst. Trudy, no. 30, p.
79-82.
303. Thomas,Erich, 1960,Structuralelementsin the northern German Democratic Republic according
to the resultsof seismicrefractionmeasurements: Internat. Geol. Cong., 21st, Copenhagen1960,
Proc., pt. 2, p. 189-199.
304. 1961, Structural elementsin the eastMecklenburgand northeastBrandenburgarea on the
basisof seismicrefractionsurveydata: Zeitschr.angew.Geologie,v. 7, no. 6, p. 274-281.
305. Timofeyeva,N.M., 1962,Investigationsof tectonicdislocations by seismicmethods:Akad. Nauk
Ukrain, SSSRInst. Geofiziki,Geofiz.Sbornik,no. 2(4), p. 84-92.
306. Tregub, F. S., and Pushkarev,N. K., 1963, A variation in the constructionof the tape-drawing
mechanismof the OS-60oscillographfor work by the correlationrefractionand deepseismicsound-
ing methods:Geofiz.Razvedka,no. 11, p. 85-89.
307. Trigubov, A. V., 1962,On a comparativeappraisalof the accuracyof the methodsof longitudinal
and convertedrefractedwaves (with English abstract): Akad. Nauk SSSR,SibirskoyeOtdeleniye,
Geologiyai Geofizika,no. 12,p. 87-95.
308. 1963, On the appearanceof loopsin traveltime curvesof head waves:Akad. Nauk SSSR
SibirskoyeOtdeleniye,Geologiyai Geofizika,no. 7, p. 103-109.
309. Troyanskiy,V. T., 1957,On the possibilityof the investigationof intrusiverocksby seismicmethods
of exploration:Razvedkai okhrananedr,no. 2, p. 32-41.
310. Troyanskiy,V. T., and Komissarchik,B. S., 1963, Use of the refractionmethodfor studyingthe
relief of the basementsurfaceunder the conditionsof the southernareas of the Rostov Region and
the westernpart of the Kalmyt A.S.S.R.:Geofiz.Razvedka,no. 14,p. 60-73.
311. Tseng, Chung Sheng,Ho, Chuan Da, and Kan, Jung Chu, 1961, Refracted wavesand multiple
wavesin Chaitamu Basin by low-frequencyseismicrefractionmethod:ScientiaSinica,v. 11, no. 1,
p. 100-116.Alsoin Acta Geophys.Sinicav. 10,no. 1, p. 39-53, 1961.
312. Tseng,Jung-Sheng,Kan, Yung-Chu,Ho, Chuan-Da,and Lee, Pang-Nian, 1960, A study of the
crystallinebasementin Chai-Da-Mu Basinby the low frequencyrefractioneimicmethod:Acta
Oeophys.
Sinica
v. 9no.2p. 155-168,
Outside U.S. Bibliography 583
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
343. Willmore, P. L., and Tolmie, R., 1956, Geophysicalobservationson the history and structureof
SableIsland,in OceanfloorsaroundCanada(a symposium):Royal Soc.CanadaTrans., ser.3, v. 50,
p. 13-20.
344. Wilson,C. D. V., 1959,Geophysical investigations
in the Vale of Clwyd:Liverpooland Manchester
Geol.Jour., v. 2, pt. 2, p. 253-270.
345. Winkler, H. A., 1963, Notes on the seismicrefractiontests on the Masinloc chromitedeposit,
Zambales,Philippines:PhilippineGeologist,v. 17, no. 4, p. 138-143.
346. Woollard,G. P., and Ewing, Maurice, 1939,Structuralgeologyof the BermudaIslands:Nature, v.
143,no. 3630, p. 898.
347. Wyrobck,S. M., 1964,Influenceof dipsin seismicrefraction:PrzegladGeol.,v. 12, no. 1, p. 3541.
348. Yamashita,Shiro,and Noritomi, Kazue, 1957,On the surveyof undergroundstructureby using
E.T.L. seismicequipment:Akita Univ. Mining CollegeResearchInst. UndergroundResources
Rept., no. 17,p. 18-23.
349. Yanez Correa,Amado, 1962, Geophysicalstudy in the Sigsbeedeep: Soc. Geol. Mexicana Bol., v.
25, no. 2, p. 95-119.
350. Yepinat'yeva,A.M., 1956,Experiences in recordingrefractedtransversewavesin seismicexplora-
tion: Akad. Nauk SSSRIzv. Ser. geofiz.,no. 11, p. 1309-1315.
351. --- 1959, Presentstatus of KMPV (correlationmethod of refractedwaves): Acta Geophys.
Sinica,v. 8, no. 1, p. 55-59.
352. -- 1963,On the dynamicrelationshipsof reflectedand refractedwavesin seismicsurveying,in
GlubinnoyeSeysmicheskoye ZondirovaniyeZemnoyKory vSSSR:Moscow,Akad. Nauk SSSR,p.
334-342.
353. Yepinat'yeva,A.M., Ivanova, L. A., Kun, V. V., and Shport,L. P., 1960,Someproblemsin the
methodof seismicexplorationof the Paleozoicbasementin westernSiberia:Akad, Nauk SSSRInst.
Fiziki ZemliTrudy, no. 12(179),p. 3-67.
354. Yoshikawa,Soji, 1959, Seismicexplorationin the vicinity of the crater of Nakadake, AsoVolcano:
Volcanol.Soc.JapanBull., ser.2, v. 4, no. 1, p. 20-32.
l'anne gdophysique
internationale(A.G.I.)' IzvestijaAkademiiNauk SSSR--Serijageofiziceskaja
(Moskva) no. 2, p. 169--84.
Aynard, C., 1956,Les mthodesg5ophysiques de prospection'La Revue ?dtroli[re--FrSvrier,p. 55-6.
Aynard, C., and ?ommier,G., 1959, Considerations sur l'aide que la gophysiquepeut apporteraux
probl[mesdes variationsde la composition de la srie stratigraphique:Bulletin de l'Association
FrangaisedesTechniciensdu Ptrole, Novembre,no. 138,p. 1141-3.
Barenboim,M. I., and Puzyrev,N. N., 1952,LesrSsultatsobtenuslorsde l'applicationde la mthodede
corrdlationdesondesrfractdesdansla dpressionprcaspienne:Prikladnaya[Geofizika,U.R.S.S., dd.
8, p. 191.
Barrabe,L., 1946,Les recherches de ptrole dansl'Allemagnedu Nord-Ouestau coursde la guerre:
Houilles,Minerais,Ptrole, France,no. 5-6, p. 139-54.
Barsh, O., and Reich,H., 1930,Ergebnisse seismicher Untersuchungen tiber den Schichtenaufbau yon
Nord Deutschland:Erginzungshefte ftir angewandteGeophysik,I, no. 2, p. 165-88.
Baumgarte,J., 1955,Constructiverepresentation of reflectionhorizonswith regardto refraction,three-
dimensionalcase'Geophys.Prosp.,v. 3, p. 126-62.
Baumgarte,J., and Krey, Th., 1961, Reflexionund Brechungbeim schrigenDurchgangebenerseis-
mischerWellen durchN planparalleleMedien' Geophys.Prosp.,v. 9, no. 2, p. 242-60.
Belooussov, U. G., Volvosky,B. S., Volvosky,I. S., and Riaboi, V. Z., 1962, Etudesexperimentales
basessurl'enregistrement desondesprofondes rfractdes:Izvestija AdademiiNauk. SSSR(Moskva),
serijageofiziceskaja,
no. 8, p. 1034-44.
Bentley,C. R., Pomeroy,P. W., andDorman,H. J., 1957,Seismic measurements on the Greenlandice
cap.PartI: Studies at 7659'N 5605'W:Annalesde Gophysique (Paris),Tome13,no.4, Oct.-Dc.,
p. 253-76.
Berckhemer, vonH., andOliver,J., 1955,ZurDeutungseismischer Einsitzemit parallelenLaufzeitkurven:
Zeitschriftftir Geophysik,no. 3, p. 152-63.
Bernard,P., 1956, InterprStationdes ondessismiquesdes explosions des Rochilles(Ao/tt-Septembre
1956):ComptesRendusHebdomadaires desSe/nces del'AcadmiedesSciences (Paris),p. 2115.
Berzon,I. S., 1942,Hodographes desondesrSflEchies, rflEchies-rEfractEes
et rEfractesrflEchies,
mul-
tiples:IzvestijaAkademiiNauk S.S.S.R.(Moskva),serijageofiziceskaja,
6, p. 275-89.
1945,Possibilit5cindmatiquede l'existencedesondesde Mintrop dansquelquesmilieuxde vitesse
continue:Izvestija Akademii Nauk S.S.S.R.,serijageofiziceskaja,2, p. 122-33.
1946,Solutiond'unproblmedirectdela mthodedesondesrfractespour desmilieuxcontinus
limitesde separation:Izvestija AkademiiNauk S.S.S.R.,serijageofiziceskaja,1, p. 71-89.
1953,DElimitationdesrdgionsd'enregistrement d'ondesr.flEchies
et rfractes d'aprs les indi-
catricescombin,es:Izvestija Akademii Nauk S.S.S.R.,serijageofiziceskaja,3, p. 209-14.
1957,DveloppementdesmEthodesde la prospectionsisrniqueen U.R.S.S.:Izvestija Akadm!ie
Nauk S.S.S.R.,serijageofiziceskaja,no. 11,p. 1347-58.
1958,Experimentationsurlesondesrfract{esconvertiesPSP: Izvestija AkademiiNauk S.S.S.R.
serijageofiziceskaja, no.6, p. 397-406.
Berzon,I. S., Vassil'ev,You. I., and Stardubrovskaya, S. ?., 1959, Sismiquerfraction sur dessables
aquifres.Iet II: IzvestijaAkademiiNauk S.S.S.R.,$erijageofiziceskaja, p. 17-29.
Bhat, H., and Rimbaut,M., 1957,Essaisd'application de la petiterfraction la dterminationdes
meilleuresprofondeurs de tir: E.A.E.G.mCongrsdeBordeaux--DEc.,6 p., 14pl. h. t.
Boccalery,M., 1949,Suggerimenti per una condottaregionaleed economicanella ricercadegli idro-
carburiin Italia: Atti del vi ConvegnoNazionaleper il Metano (Padova),p. 156-66.
1952,Aspetti practicinell'esplorazione geofisica:Atti del VII ConvegnoNazionaledel Metano e
del Petfolio (Taormina), I, p. 443.
Bornnann,G., 1959,Grundlagen und Auswerteverfahren der dynamischen Baugrundseismik: Freiberger
Forschungshefte, C 65, no. 163,"Geophysik"Allemagne--(Berlin),95 p.
Bortfeld,R., 1962,Exactsolutionof the reflectionand refractionof arbitrarysphericalcompressional
wavesat liquid-liquidinterfacesand at solid-solidinterfaceswith equal hear velocitiesand equal
densities:
Geophys. Prosp.,v. 10,no. 1, p. 35-67.
1962,Reflectionand refractionof sphericalcompressional wavesat arbitrary plan interfaces:
Geophys. Prosp.,v. 10,p. 517-38.
Bortfeld,R., andHurtgen,H., 1960,On the identificationand construction of reflectedrefractions:Geo-
phys.Prosp.v. 8, p. 12-24.
Bouasse, M. H., 1892,Rflexionet rfractiondansles milieuxisotropes, transparents et absorbants:
Th[seprOsentre la Facult desSciences de Parispourobtenirle gradede Docteurs Sciences Math-
matiques: GauthierVillarsParis,France.13Dcembre,158p.
Bouchon,R., 1954,L'applicationde la sismique-rdfraction au Sahara:II me Congr[sNational du
PdtroleFrangais,(Alger)Mai, p. 61-70.
----- 1957,Rfractionau Sahara:III [me Congr[sNational du Ptrole Frangais(AFPT), (Pau), Mai,
p. 91-103.
1961, Evolution de la sismiquerfraction au Sahara: PEtrole Informations (Fr.) no. 313,
p. 52.
Brand,J., 1925,Ein Beltragzum Studlureder Bodenbewegungen
nicht seismischen
Ursprungs:Zeit-
schriftftir Geophysik,
p. 348--59.
85 Bibliography of Refraction Papers
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
Brauch,
W., 1958,Onreflectedrefractionwaves: Geophys. Prosp., v. 6,no.4, p. 365-81.
Breusse,
J. J., 1959,Application
desmthodes gophysiques . la recherchedel'eau:"Association Inter-
nationaledesHydrogdologues--Congrs Gologique International.--Mdmoires--T. II. (Rdunionde
Liege1958)":Secretariat
deI'A.I.H., p. 107-12,5 fig.
Breusse,
J. J., andHuot,G., 1953,Recherches d'eauenpaysdsertiques (Mauritanie-Soudan) parles
m4thodes de prospection
41ectrique
et sismique rdfraction'Congresde l'EuropeanAssociation
ofEx-
plorationGeophysicists--(Paris),
Mai, 5 p.
Brewer,H. L., andHoltzscherer,
J., 1958,Results ofsubsurface investigations
usingseismic
detectorsand
deepboreholes:Geophys. Prosp.v. 6, no.2, p. 81-100.
Brockamp, B., andMothes,H., 1930,Seismische Untersuchungen auf demPasterzegletscher--I:
Zeit-
schriftfar Geophysik.no. 8, p. 482-500.
Bun'kov,A.M., 1958,Elevation
dela precision
deddtermination
dela vitesse
moyenne
d'aprsle point
d'intersection deshodographes
desondesrdfractdes:
Trudy Gorno-Geologiceskogo
Instituta (Moskva),
no. 34, p. 87-93.
Bygajlo,V. A., 1957,Descriptiond'unemdthodederestitution
desmiroirsquinesontpasplans:Geo-
fysikalni,Sbornik,
TrudyGorno-Geologiceskogo Instituta(Moskva),
30.,no.2,p. 142-45,
3 fig.
Cagniard,L., 1939,Rflexionet rdfraction
desondessismiquesprogressives:
Gauthier-Villars.
(Paris),
255p., 70 pl.
1941,Quclques
rdsultats
d'unetudethdorique
dela rflexionet dela rfractiondesondessdis-
miques
progressives:
Annales
de l'Institutde Physique
du Globe(Strasbourg),
3mepartie:Gdo-
physique,
p. 106-16.
Cleya,J. de.,Closs,
H., andDurbaum,
H., 1952,ZurBerficksichtigung
derBrechung
in derReflexions-
seismik:Erd61und Kohle,D4cembre,5, p. 765-68.
1953,Interpr.tation
desrdsultats
dela sismique
rflexionentenantcomptedel'effetde rfrac-
tion:Minerariay Metallurgia,Espagne,no. 243,mars,p. 22-25.
Caloi,P., 1953,La prospection
sismique
applique
la dtermination
del'6.paisseur
deglacier
(glacier
de
la Marmalada). (Principedesmdthodesde refraction):Bolletinodel ComitatoGlaciologico
Italiano
(Turin), no.4, p. 265-66.
Cassinis,
R., 1958,La sismicadi rifrazionenellaricercadegliidrocarburi:
Rivoluzione
Industriale
(Milano),nov.,6, p. 37-54.
1959,Applicazione
dellasismica di rifrazionenellericherche a piccolissima
profondita'
Rivistadi
Ingegneria
(Milano),no.1, 15p.,
1955Ricercasperimentale della seriesolfiferain Sicilia'Congresde l'EuropeanAssociation
of
Exploration
Geophysicists
(Londres),
7-9Ddcembre, 15p.
1956,Ricercasperimentale
dellaseriesolfiferain Sicila:Quaderni
di Geofisica
Applicata(Mi-
lano),XVII, no. 1-2, p. 3-8.
1960,Application of seismic
methods
to geothermal
energyexploration'
Quadernidi Geofisica
Applicata(Milano),
XXI, p. 28-37.
Cassinis,
R., andFolicaldi,
R., 1957-1958,
Preliminari
di unaricerca
di laboratorie
sulleffettedi strati
sottilinellasismicadi rifrazione:Quadernidi Geofisica
Applicata(Milano), 1957-1958,18-19,p. 71--8.
Cekine,
B. S.,1959,Surle spectre
desondes
rdfldchies
et rdractdes
dansunecouche:
IzvestijaAkademii
Nauk. S.S.S.R.,serijageofiziceskaja,
no. 11,p. 1279-87,13fig.,3 rdf.biblio.
1959,Reflexionet rdfractiondesondessismiques au niveaud'interfaces
peu rdfldchissantes.
IzvestijaAkademiiNaukS.S.S.R.(Moskva),no.1,janvier,p. 9-13.
Closs,
H., 1951,
Applied
geophysics
inWestGermany
during
thelastfiveyears'
3meCongrs
Mondial
du
Petrole(TheHague),section
I, p. 564-82.
1963,DeutschegeophysikalischeUntersuchungen:
OelZeitschriftfar dieMinera161wirtschaft,
no. 1, p. 24-27.
Closs,
VonH., andBehnke, CI.,1961,Fortschitte
derAnwendung seismischerMethodenin derErfor-
schungderErdkruste:Geologische
Rundschau (Stuttgart),v. 51,no.2,p. 315-30.
Closs,
H., deCaleya,
J.,andDiirbaum,H. J.,1952,ZurBeriicksichtigungderBrechunginderReflexions-
seismik:
Physikalische
Verhandlungen
(Karlsruhe),
3, no.6, p. 156.
Collignon,
M., 1928,Lesexplosions grandedistance.
Errat relatifs cettecommunication,
p. 452:
Comptes Rendus HebdomadairesdesSdancesdel'Acadmie
desSciences, p.357.
1928,Surl'auditiondesondesd'explosion
grandedistance: Comptes RendusHebdomadiarcs
desSancesde l'Acadmie desSciences,p. 1615.
Contini,C., 1949,Le richerche
geofisiche
dell'A.G.I.P.:
Atti delvi Convegno
Nazionale
peril Metano
(Padova),p. 167-82.
Contini,C., Maino,C., andRanucci,
M., 1958,Nuovicriteriperla prospezione
sismica
a rifrazione.
Annalidi Geofisica
(Roma),T. XI, no. 1,p. 15-54Annalidi Geofisica
(Roma),T. XI, no.3, p. 207-32.
Coyaud,
J., 1958,Procddds
destines
remplacer
lesexplosions
danslesmdthodes
sismiques
utilisdes
dans
lesprospections
gdophysiques:
BrevetNo.L.219.337
(P.V.18.295Bouches-du-Rh6ne),
24.12.1958G
01 c.
Davydova,
N. I., 1962,Etudesurlesmodules
dela dfpendance
descaract4ristiques
dynamiques
des
ondesde front longitudinales
de la puissancede la coucherffringente:Izvestija AkademiiNauk
S.S.S.R.,no. 1, p. 11-22.
Compagnie Generale de Geophysique Bibliographle 587
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
Petrucci,
G.,1952,
Applicabilittr
deimetodi
geofisici
allaricerca
degli
idrocarburi:
AttidelVIIConvegno
Nazionale
delMetanoedelPetrolio
(Taormina),
I, p. 465.
Podyapolsky,
G.S.,1961,
Coefficients
derdfraction
etderdflexion
desondes
lastiques
surunecouche'
Izvestija
Akademii
NaukS.S.S.R.,
no.4,p.520-33.
1959, Expression
approximative
pourledplacement
auxenvirons
dufrontfondamental
dansle
casdepetit angle
entre
lerayonetlasurfacedesparation.
Enseservant
delaformulegnlrale
on
considre
lecas
particulier
del'onde
rfracte:
Izvestija
Akademii
NaukS.S.S.R.,
no.12,p.1761-73.
1959,
Ausujet
d'une
formule
itablissant
larelation
entre
lescoefficients
des
ondes
detteetceux
delarflexion
etdelardfraction:
Izvestija
Akademii
NaukS.S.S.R.,
no.11,p.1560-9.
Posgay,K.,1956, Seismic
surveyinthe
coal-basin
ofEsztergom,1955'Geofizikai
Kozlemenyek,
Hongrie,
no. 4 p. 39.
Radzabov, M.M.,1958, Dtermination
desvitesses
limites
d'aprslesdromochroniques
transversales
des
ondes rfractes:
Izvestija
Akademii
NaukS.S.S.R.,
no.12,p. 1491-503.
1958, Ddtermination
desvitesses
limites
aumoyen desdromochroniques
d'ondes
rfractes:
Izvestija AkademiiNaukS.S.S.R.,
serija
geofiziceskaja,
no.12,p.870-7.
1959, Quelques
traitscaractristiques
spciauxdesdromochroniques
desondes'
Izvestija
Akademii
NaukS.S.S.R.,
serijageofiziceskaja,
no.7, p. 1046-51.
1959Precision
dela ddtermination
dela vitesse
d'interface
trpartird'unsy.s.t[me
dedromo-
chroniques transversales
d'ondes
rfractdes:
IzvestijaAkademii Nauk S.S.S.R.,
senjageofiziceskaja,
no. 3, p. 304-10.
1960, Choix des vitesses
moyennes pour ladtermination desprofondeursvraies,
aumoyen des
dromochroniques d'ondesrdfractdes:
Izvestija
Akademii NaukS.S.S.R., serijageofiziceskaja,
no.6,p.
p. 571-4.
Radzabov, M.M.,and Agranovsky,
L.E.,1960, Dtermination desprofondeursetdelaforme d'une
interface aumoyen des dromochroniques
transversalesd'ondesrfractes:
IzvestijaAkademii Nauk
S.S.S.R., serija
geofiziceskaja,
no.6,p.560-70.
Rajkher, L.D.,1957, Etudedtaille
des structures
failles:
Prikladnaya
Geofisiceskaja
(Leningrad),
no.
16, p. 98-113.
Reich, DrH.,1921, ElasticheBodenwellenalsHilfsmittelzurAufsuchung von Lagerstitten'
Stahl und
Eisen--Zeitschrift fardasDeutscheHtittenwesen,21.4,p.547-8.
1925, DieBedeutung derneun geophysikalischen UntersuchungsmethodenftirdieGeologie
nach denbisherigen Erfahrungen'
Zeitschrift
farGeophysik, p.120-1.
1931, DerUntergrund vonSchleswig-Holstein
nach denErgebnissenseismischer
Refraktions-
messungen' Pumpen und BrunnenbauBohrtechnik,Allemagne, no.24,26novembre,p.763-9.
1948, GeophysikalischeLagerst/tttenforschung:
Naturforschung undMedizin inDeutschland,
1939-46, (FiatReview ofGermanScience)17,Geophysik, I Wiesbaden.
1957, InSaddeutschland Seismisch
ermittelte
tiefeGrenzflachen
und ihre Geologische
Bedeutung:
Geologische Rundschau, Allemagne,Band 46,Heft1,p.1-17.
1960, ZurFrage derGeologischen
Deutung seismischerGrenzflichen
inden Alpen'Geologische
Rundschau, Allemagne, Band50,p.465-73.
Rempp, G.,Hee, Mme and Damman, Melle Y.,1925, Observationsfaitestrl'occasion
del'explosion
d'Hagondange parI'I.P.G.S.'Association
Francaise
pour L'Advancement des Sciences,
49[me session,
(Grenoble),p. 252-5.
Reyre, D.,and Vallet,J.M.,1958,Etude ded6mes desel auGabon' Geophys.Prosp.,
v.6,no.1,p.58-
74.
Riznichenko, Ju.V.,1940, Construction
d'une plate-formerfractante
ourflchissante
d'apr[s
l'hodo-
graphe des ondes transormdes
rflchies
oupassag[res: Izvestija
Akademii Nauk S.S.S.R.,serija
geofiziceskaja,
5,p. 751-8.
Riznichenko, Ju.V.,andShamina, O.G.,1960, Ondes multiples
rfl4chies
ettransmises'Izvestija
AkademiiNauk S.S.S.R.,no. 12,p. 1129-39.
Riznichenko, Ju.V.,Shamina, O.G.,and Khanutina, R.V.,1961, Ondes41astiques
sepropageantdans
des
mod[les
bidimensionnels
avecunevitesse
uniforme'
Izvestija
Akademii
NaukS.S.S.R.,
no.4,
p. 321-34.
Roberts,F.A.,andDennisson, A.T.,1953, Adevice overcoming theeffects
ofstaticonseismic
shot
signals'
Geophys. Prosp.,v. 1, p. 192-4.
Robin,G.deQ.,1958, Glaciology III (Seismic
shooting
and relatedinvestigations),
inNoregian-British-
Swedish Antarctic Expedition 1949-52, Scientific
Results.v. 5' Oslo,
Nors?olarinstitut.
Rocard,Y.,1960, Surlessignaux sismiquesdelongue pdriode obtenuslorsdutestnuclaire
dereg-
ganele13.2.1960: Comptes Rendus HebdomadairesdesSances del'Acadmie
desSciences
p.2244.
Rogozza,I. O.,1961, Particularit'sdynamiquesdesondesrfractes
etdiffractes'
Prikladnaya
Geofizika,
U.R.S.S.,ed.29, p. 3-11.
Rothe,E.,1924, Comparaison entre l'explosion
d'OppauetcelledelaCourtine'
Comptes Rendus
Hebdo-
madaires desSances del'Acadmie desSciences,
179,8.9.
1924, Principe d'une mdthode deddterminationprecisedelapropagation
des ondes
sismiques:
Comptes Rendus Hebdomadaires desSancesdel'Acadmie desSciences,
179,
8.9,p.507-9.
1952,
Lesmthodes
deprospection
dusous-sol.
Tome I: Gauthier-Villars,
392p.
Compagnie Generale de Geophysique Bibliographie 593
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
Rothe, E., Lacoste,J., and Bois, Ch., 1925, Observationssismologiques faites & l'occasiond'une forte
explosion'ComptesRendusHebdomadaires desSancesde l'AcadmiedesSciences, 180,29.6,p. 2053-4
Rothe, E., Lacoste,J., Bois, Ch., Dammann, Melle, and Hee, Mme, 1924, Etude de la propagationde
l'branlementdesexplosions dela Courtine(Part I); Comparaison entrel'explosiond'Oppauet celledu
23 mai & la Courtine (Part II): ComptesRendusHebdomadairesdes Sancesde l'Acaddmiedes Sci-
ences,p. 339 (Part I), p. 507 (Part II).
Rothe, J.P., Peterschmitt,E., and Stahl, P., 1948,Les ondessismiquesdesexplosionsd'Haslach (Fort
Noire): ComptesRendusHebdomadairesdesSancesdel'Acadmie desSciences, p. 354.
Sandner,A., 1956, Das Problem der seismischen Grenzschichtwellebei der Behandlungder Wellengleich-
ung:FreibergerForschungshefte C 26 "Geophysik"(Berlin), 45 p.
Schmidt, O. von, 1928, AngewandteSeismik (Zusammenfassender Bericht tiber Arbeiter von 1921 bis
1928):Zeitschriftftir Geophysik,p. 134-46.
1928,
Angewandte
Seismik:
Zeitschrift
ftir
Geophysik,
1928,
v.4,i134.
1931,Theorieder3 SchichtenSeismik:Zeitschriftftir Geophysik, 37-56.
-- 1932,B..rechungsgesetz
odersenkrechten Strahl?Zeitschriftftir Geophysik,p. 376-96.
--1954, Uber den Energietransportbei der Sprengseismik: Zeitschrift ftir Geophysik,v. 10, p.
378-85.
Schlich,R., 1958,Interpr6tation d'un proill sismiquede rfraction & la station Charcot en Terre Ad61ie:
ComptesRendusHebdomadaires desSancesdel'AcadmiedesSciences, p. 1761.
Schneerson, M. B., 1958, Carottagesismiquepar ondesrdfractes:RazvedocnajaI PromyslovajaGeo-
fizika, U.R.S.S., no. 26, p. 3.
1959, Estimation de la precisiondu procd d'interprtation diffrentielledes rdsultatsobtenus
par correlationdes ondesrfractes dans les conditionsde la plate-forme russe:RazvedocnajaI
PromyslovajaGeofizika,U.R.S.S.,no. 27, p. 3-14.
1960,Interpr6tation deshodographes desondesr6fract6esau coursde la prospectiondes struc-
turesdesplate-formesen pentedouce:Prikladnaya Geofizika,U.R.S.S., ed. 25, p. 3-19.
--, Determination des vitessesde la propagationdes ondeslastiques d'apr[s les hodographesde
diff4rencesdes ondesrflchies: Prikladnaya Geofizika,U.R.S.S., ed. 31, p. 109-15.
Schulz,G., 1960,lbereinRefraktionsproblem:
Zeitschritftir Geophysik,
v. 26,no.5, p. 235-45.
Schuster,K., 1959,Geophysikalische Erkundungim Sudan:FreibergerForschungshefte C 50"Geophysik"
(Berlin), Mai, 55 p.
Schweydar,W., and Reich, H., 1926, Aufzeichnungenyon ktinstlichen Erdbeben: Zeitschrift ftir Geo-
physik, v. 2, p. 350-1.
----1927, Ktinstliche elastiche Bodenwellen als Hilfsmittel geologischenForschung: Beitrige
(Gerland's)zur Geophysik,t. 17,p. 121.
Seya,K., 1960,On an influenceof anisotropyin a seismicrefractionmethod:Butsuri-Tanko.(Kanagawa-
ken), Japan,v. 13, no. 2, p. 89-95.
Shima, E., 1957, Note on the depth calculationby the seismicrefraction method: Butsuri-Tanko
(Kanagawa-ken),Japan,v. 10, no. 1, p. 16-8.
Solaini,L., 1959, Sui rilievi geofisicia Pian di Fedaia: Energia Elettrica, Iralie, v. XXXVI, fasc.5, 6 p.
Stenz,E., 1954,Wstep do geofizyki.Warszawa(Varsovie),PWN, S. 179,ryc. 75.
Susstrunk,A., 1948,Techniqued'valuation et limite d'applicationdesmesuresde rfraction sismique
dansles couchesirrgulires:Bulletin de L'AssociationSuissedesSciences Naturelies,128,p. 111-8.
Tabuteau, F., 1953,Travaux de sismologiede la premi[re expeditionen Terre Adlie, 1950.Installation
et fonctionnement;Rapportsscientifiques desexpeditionspolairesfransaisesPaul-EmileVictor. S IV 3:
Annalesde Gophysique,t. 9, p. 277-83.
Tal'virskij, D. B., 1959, Particularit,s dessismogrammes et des dromochroniques de tirs de rfraction
effeztusdansleszonesd'effondrementde SibrieOccidentale:Prosp.Geoph.Ind., no. 21, p. 8.
Tanaka, A., and Ninagawa, S., 1960, Geophysicalprospectingin the Eastern part of Kushiro City,
Hokkaido:Bulletin of the GeologicalSurveyof Japan,v. 2, p. 15-30.
Tardi, P., 1956,M. PierreTardi rendcompted'unesrie d'expriencesgophysiques qui ont eu lieu du 24
Ao6t au 6 Septembredans les Alpes Fransaises:ComptesRendus HebdomadairesdesSancesde
l'Acadmie desSciences, p. 1089.
1957, Exp4riencessismiquesdans les Alpes occidentalesen 1956, rsultats obtenus par le
"Groupe d'Etudes et des ExplosionsAlpines": ComptesRendus Hebdomadairesdes Sancesde
L'Acadmie desSciences,p. 1114.
1959,Enregistrements sismiques de grandesexplosionsdansles Alpes occidentales en Septembre
1958:ComptesRendusHebdomadaires desSancesdel'AcademiedesSciences (Paris),p. 1069.
Tarrant, L. M., 1956,A rapid methodof determiningthe form of seismicrefractionfrom line profilere-
sults:Geophys.Prosp., v. 4, no. 2, p. 131-9.
Tateischi,T., 1951,Seismicprospecting at the TokyosaveCoalfield:Bull. Geol. Surv.Japan,no. 12, p.
23-9.
1954,Seismicprospectingon KusanoPlain: Bull. Geol.Surv.Japan,no. 12,p. 39-44.
Tateishi, T., and Hirasawa, K., 1957, Seismicprospectingin the vicinity of YokoshibaTown-Chiba
Prefecture--Seismicprospectingin Kwanto, District II (Part I), Seismicprospectingin the vicinity of
YokoshibaTown, Chiba Prefecture(Part II): Bull. Geol.Surv.Japan,v. 8, no. 9, p. 41-50 (Part I);
v. 8, no. 11,p. 41-50 (Part II).
594 Bibliography of Refraction Papers
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
INDEX
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
Index
The Index was made by requestingthe authors to contribute an index for each
article, which was then compiledand supplementedby the Editor.
The Index is divided into two parts, "Name and Place Index" and "Subject
Index."
The "Name and Place Index" includes both names and locations felt more
important or pertinent to the Indexers. The "Subject Index" includes important
words and phrasesthat occurredin each individual paper. Page numbers are listed
usuallyfor the first or more important occurrenceof the word or phrase.Multi-word
phrasesare sometimeslisted under several important words.
,596
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
SUBJECT INDEX
Curvedpathrefraction
charts,320 Emergence-angle
raypathtracing,524
Curvedray, 217 Emergentoverburden
wavefront,420
Curvedraypaths,294 Emergentpoint, 298, 340
Curvedtime-distanceplot, 217 Emergentwavefronts, 416
Curved-ray
problem,nomogram,
296 Energy,partitionof, 168
Cycloidalraypath,430 releaseof seismic,161, 165
shingling,469
Data averaging,501 Energyconversions,
compressional-to-shear,
33
Data errors, 459 Eocene, 493
Datum plane, 348 Eocene-Senonian,514
Decay, theoretical,149 Epoxy resin, 79
DeccaLambda Surveychain,551 Equipment,special,485
Deepcontinuousrefractor,
171 Equivalentparabolic
methodfor linearvelocity,
386
Deeprefractions,
identification
of,480
DeHavilland Otter, 551 Equivalent
radiator,
153
Delaytime,172,338,348,483 Erosinalsurfaces,buried, 199
absolute, 343 Errors, buoy location,483
definition,172 due to incorrectassumptions,
463
versusdepth,356 in data, 459
Delay-time
curve,average
relative,
343 in overburdenvelocity, 463
determinationof, 342, 343 refractor velocity, 423
final absolute, 343 refractor-velocitydeterminations,419
relative, 343 Explorationphase,426
Delaysin continuous
profiling,
174 Explosion
cavity,radiusof, 156
Depthanddipcomputations,510 Explosion
efficiency,
486
Depth chart, 341 Explosions,
surface,167
Depthpointdeterminations,
524 underground,152
Depth to refractor,
beneathVz--Vo+kZ,324 underwater, 164
beneathVz= Vo(t-t-kz)
', 326
Depthversus
delaytime,341 F factor chart, 318
Depth-distance
andtime-distance
plots,272 Facies change, 518
Desiredin-line position,483 Family of velocitydepth-functions,
290
Detectability
of secondary
refractions,
489 Fan shooting,5, 267, 526
Detector location,483 Fault example,282
Development phase,426 Faulting, 343
Diagonalplot, 272 Basin and Range, 539
Differencesignal,67 Fermat's principle, 339
Diffraction,401,513,527 Field procedure,proposed,81
to checkoverburdenvelocity, 411 Filter, phase-velocity,
68
Dilatation, 139 velocity,67
Dilatation head wave, 150 Final absolutedelay-timecurve,343
Dilatation potential, 140 First arrival, 340, 353
Dimensional scaling, 155 time lag technique,524
Dip anddepthcomputations,
510 First breaks, 544
Dip angle,268 First motionapproximation,22
Dip determinations,
relative,340 Flatten an event, 490
Dip errors,251 Flexiblevelocityfunction,217
Directed wavefront, 365
Forerunner,476
Discontinuous
profiling,503 Fossildrainagechannels,197
Domalmass,outlining,456 Fossilweatheredlayers, 198
Dominantperiod,71
Fourierspectrum,140
Drainage
channels,
fossil,197 Frequencies,
predominant,163
Drainage
ofwater-saturated
sandlayers
related
to Fresh sea water, 483
velocitychanges,
207
Marker velocity,assigned,
341 Offset, 351
Maximum
propagation
velocity,
217 versusdepth,356
Mean relativecurve,176 Offsetdistance,351,489
Metal miningapplications,
523 versusdelaytime, 341
MiddlePrecambrian
Farmington
Canyon,539 One-layer
refractions,
nomograph
for, 308
Midpointmethod,260 One-wayrefractions,
378
Minimum trajectory,419 One-waytargeting,420
Miningexploration,
522,523 Optimum
distances,
274
Mobile augerdrill, 542 Ordovicianage,560
Model,analog,15, 74 Outlining,
saltmassby recording
in a deephole,
449
geologic,130
interpretation,
115 shalemasses
by recording
in a deephole,452
the domal mass, 456
plate, 74
Overburden,thickness,523
plexiglass
andaluminum,
469 Overburdenvelocity, 107, 498
steel and brass,471
check with diffractions, 411
two-dimensional,469
two-dimensionalanalog, 15 complex,383
errors in, 463
Modelseismogram, 17
selectionof, 408
Model studies,281 Overburden wavefront, 420
Modes,antisymmetric,
143
attenuated, 145
Pr mode, 62
symmetric,143
unattenuated, 143 Papermonitor
records,
489
Parabolic method, 367
Modifications
to standardrefraction
equations
re- Parabolicvelocityconstants
fromrefractions,
323
quired
because
of depthof shot,200 Particularapparentvelocity,489
(or)modifications
toequations
duetoshot
depth Partition of energy, 168
Moenkopiformation,198 Peakpressure,153
Monotonicvelocityincreases,
290 Period,dominant,71
Motor toboggans,
551 Permafrost lenses, 198
Moveout time, 296 Permian formations, 560
Multilayerproblem,281 Phase corrected,272
Multiple buoys,483 Phasedistortion,95
Multiplerefractions,
31, 511 Phase velocities, 139
Phase-velocityfilter, 68
n-layer
problem,analytic
solution
of,260 Plate model, 74
n layers,
refractions
through,
307,315,316 Plexiglass
andaluminum
model,469
Narrowpassband-producedringing,544 Point determinationsof depths,524
Navigationequipment,
483 Point source,365
Near-frontapproximation,
119 Post-glacial
erosion:
effectuponvelocities,
198
Near-surfacecorrections,354 Precambrianbasementrocks,527
Near-surfacevelocity variations,331 Precambrian rocks, 539
Neogene,514 Predominant frequencies,163
Nomogram,296, 336 Primaryphenomenon
relativeto the amplitude
for curved-rayproblem,296 variations, 180
for refractioncomputations,
209 Primaryrefractions,
511
Nomograph
forrefractions
through
onelayer,308 Principle
of reciprocity,
178,374
Nondispersive
propagation,
146 Prism-protractor
slopemeter,
230
Nonlinearity
ofrefracted
traveltime
graphs,
203 Profile, in-line, 340
Nonplane
refractor
beneath
linearvelocity
over- unreversed,332, 524
burden,382 Profile shooting, 340
Nonreversedshots, 174 Propagation,
nondispersive,
146
Normal knee case, 232 Propagation
characteristics,
147
Normal line, 272 Propagation
constant,
complex,
139
Normal mode equation, 142 Propagation
velocity,139
Normal sea water, 483 maximum, 217
Normal time-distancecurve, 7
Northward-trendinggraben,539 Quaternaryfill, 198
Radial refraction shooting, 11
Oblique
shooting
intoa wellsurvey,
464
802 Subject Index
Downloaded 06/25/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/