High-resolution seismic imaging in deep sea from a joint deep-towed/OBH reflection experiment

20 Pages·2017·2.21 MB·English

High-resolution seismic imaging in deep sea from a joint
deep-towed/OBH reflection experiment: application to
a Mass Transport Complex offshore Nigeria
Stephan Ker, Bruno Marsset, S. Garziglia, Yves Le Gonidec, Dominique
Gibert, Michel Voisset, J. Adamy
To cite this version:
StephanKer,BrunoMarsset,S.Garziglia,YvesLeGonidec,DominiqueGibert,etal.. High-resolution
seismicimagingindeepseafromajointdeep-towed/OBHreflectionexperiment: applicationtoaMass
Transport Complex offshore Nigeria. Geophysical Journal International, 2010, 182 (3), pp.1524-1542.
￿10.1111/j.1365-246X.2010.04700.x￿. ￿insu-00609554￿
HAL Id: insu-00609554
https://hal-insu.archives-ouvertes.fr/insu-00609554
Submitted on 17 May 2017
HAL is a multi-disciplinary open access L’archive ouverte pluridisciplinaire HAL, est
archive for the deposit and dissemination of sci- destinée au dépôt et à la diffusion de documents
entific research documents, whether they are pub- scientifiques de niveau recherche, publiés ou non,
lished or not. The documents may come from émanant des établissements d’enseignement et de
teaching and research institutions in France or recherche français ou étrangers, des laboratoires
abroad, or from public or private research centers. publics ou privés.
Geophysical Journal International
Geophys.J.Int. (2010)182,1524–1542 doi:10.1111/j.1365-246X.2010.04700.x
High-resolution seismic imaging in deep sea from a joint
deep-towed/OBH reflection experiment: application to a Mass
Transport Complex offshore Nigeria
S. Ker,1,2 B. Marsset,1 S. Garziglia,1 Y. Le Gonidec,3 D. Gibert,2 M. Voisset1
and J. Adamy4
1De´partementGe´osciencesmarines,IFREMER,France.E-mail:[email protected]
2InstitutdePhysiqueduGlobedeParis(CNRSUMR7154),4PlaceJussieu,75252ParisCedex,France
3Ge´osciencesRennes(CNRSUMR6118),Universite´Rennes1,Baˆt.15CampusdeBeaulieu,35042Rennescedex,France
4TOTAL,France
Accepted2010June14.Received2010June8;inoriginalform2010January4
SUMMARY
Weassessthefeasibilityofhigh-resolution seismicdepth imaging indeep water based ona
new geophysical approach involving the joint use of a deep-towed seismic device (SYSIF)
and ocean bottom hydrophones (OBHs). Source signature measurement enables signature
deconvolution to be used to improve the vertical resolution and signal-to-noise ratio. The
source signature was also used to precisely determine direct traveltimes that were inverted
to relocate source and receiver positions. The very high accuracy of the positioning that
was obtained enabled depth imaging and a stack of the OBH data to be performed. The
determinationoftheP-wavevelocitydistributionwasrealizedbytheadaptationofaniterative
focusingapproachtothespecificacquisitiongeometry.Thisinnovativeexperimentcombined
withadvancedprocessingsucceededinreachinglateralandverticalresolution(2.5and1m)
inaccordancewiththeobjectivesofimagingfinescalestructuresandcorrelationwithinsitu
measurements. To illustrate the technological and processing advances of the approach, we
present a first application performed during the ERIG3D cruise offshore Nigeria with the
y
g seismic data acquired over NG1, a buried Mass Transport Complex (MTC) interpreted as
o
ol a debris flow by conventional data. Evidence for a slide nature of a part of the MTC was
m
provided by the high resolution of the OBH depth images. Rigid behaviour may be inferred
s
ei frommovementofcoherentmaterialinsidetheMTCandthruststructuresatthebaseofthe
S
MTC.Furthermore,asiltlayerthatwasdisruptedduringemplacementbuthasmaintainedits
I
J
G stratigraphicpositionsupportsashorttransportdistance.
Keywords: Controlledsourceseismology;Seismictomography;Acousticproperties;Sub-
marinelandslides;AtlanticOcean.
1 INTRODUCTION
Geophysicalandgeologicalexplorationindeep-waterenvironmentshasattractedconsiderableresearchefforttoassessgeohazardssuchas
landslides, shallow gas and gas hydrates (Kvalstad 2007). Undoubtedly, the perspectives provided by 3-D seismic imaging
techniqueshaveyieldedsignificantinsightsintothelarge-scaleidentificationofsuchcomplexgeologicalfeatures(Heggland2004;Davies&
Clark2006;Cartwright2007;Calve`setal.2008;Bulletal.2009).However,marinegeohazardsrequirequantitativeseismiccharacterization
atsmallerscalesforcalibrationwithdetailedmeasurementsofgeotechnicalpropertiesandsamplingofthesubseabed(groundtruthing).The
objectiveistoachieveseismicimageryconsistentwiththeaccuracyofinsitumeasurements,thatis,high-resolutionseismicinvestigationat
frequencieshigherthan200Hz,withconsistentspatialsampling.Performinghigh-resolutionimagingindeepwater(500–6000m)encounters
technological,physicalandmethodologicalissues,particularlywhendealingwithshallowsedimentlayersinthefirsthundredmetresbelow
theseabed(Woodetal.2003).
High-resolution seismic imaging of shallow sediments in deep-sea environments cannot be achieved with conventional near-surface
seismicsystemsbecauseoftheoreticalandphysicallimitations.Thefirstlimitationistechnological.Thesourcehastobesufficientlypowerful
toovercometransmissionlossinthegreaterwaterdepth:thesphericaldivergenceoftheacousticwaveinawaterdepthof1000mimplies
alossof33dBoftheeffectivesourcelevel.Forconventionalsourceswithafrequencycontentsufficientlyhighforhigh-resolutionimaging
1524 (cid:2)C 2010TheAuthors
Journalcompilation(cid:2)C 2010RAS
High-resolutionseismicimagingindeepsea 1525
such as mini air-gun, piezoelectric transducer or sparker, the source level lies in the range 190–220 dB ref. 1 μPa at 1 m and to avoid
destructivecavitation,itcannotbeincreased.Thesecondlimitationconcernsvelocityanalysis,theaccuracyofwhichincreaseswiththerange
ofreflectionanglesoroffsets.Itispossibletoincreasethisrangeofanglesbyreducingthedistancebetweenthesourceandthereflectoras
withadeep-towsystem,orbyincreasingtheacquisitionstreamerlength,astypicallyachievedwithanear-surfacesystem.Thethirdl(cid:2)imitation
isoneofhorizontalresolution,controlledbytheradiusofthefirstFresnelzoneonthesurfaceinsonifiedbytheacousticwave,R= (hλ)/2
atnormalincidencewhereλisthewavelengthofthemonochromaticwaveandhisthereflectordepth.Forexample,forawaterdepthh=
1000m, R (cid:3)40mwhenusinga500-Hzsurfaceseismicdevice.Thelateralresolutioncanbereduceddowntothemeansignalwavelength
λ(cid:3)3musingmultichannelseismicacquisitionwithmigrationtechniques(Claerbout1985;Scales1997).Limitedstreamerlength,however,
does not usually allow the required precision in velocity determination to perform sufficiently accurate imaging for this resolution to be
achieved.
To overcome these limitations in high-resolution seismic imaging in deep water, one has to use (1) seismic sources/receivers with
frequency bandwidths consistent with the scale of in situ measurements and (2) multichannel acquisition and processing techniques to
increaselateralresolution.Hybridsystemsbasedonaconventionalsourceattheseasurfaceandreceiversclosetotheseafloor,forexample,
thoseofBowen(1984)andNouze´ etal.(1997),haveshownimprovementsinthespatialresolutionbutthisapproachisnotappropriatefor
accuratevelocityanalysis.Abettersolutionisafulldeep-toweddevicewheretheentiresystem,boththesourceandthereceivers,operates
close to the seafloor (Fagot 1983), simultaneously providing vertical resolution improvements and better penetration into the sediment.
Thefirstfulldeep-towedsystem,calledtheDTAGSforDeepTowedAcousticsandGeophysicsSystem,wasdevelopedbytheU.S.Naval
ResearchLaboratory(Gettrustetal.1991).ThefirstversionoftheDTAGShadtworeceiverarraysof24channels(respectively,2.1and
21mgroupspacing)andasolid-stateHelmholtzresonatorsourcethatemittedsignalsinthefrequencyrange(250–650Hz)(Wood&Gettrust
2001;Chapmanetal.2002).IFREMER(FrenchResearchInstituteforExploitationoftheSea)hasdevelopeditsowndeep-towedseismic
system,calledSYSIF(Syste`meSismiqueFond),thelatestversionofwhichhastwoacousticsourcesoperatinginthefrequencyranges(220–
1050Hz)and(580–2200Hz),respectively.SYSIFisthereforeabletorecordthesignalsreflectedfromsedimentlayersovertworesolution
scales,but,todate,doesnothaveamultichannelcapability(Keretal.2008;Marssetetal.2010).
DuringtheERIG3Dcruise(2008)ongeohazardsintheeasternNigerDelta,wedeterminedthesubseabedP-wavevelocityfield,avoiding
theacquisitioncomplexityofdeep-towedstreamergeometry,bydeployingocean-bottomhydrophones(OBHs)ontheseaflooratadepthof
800mandtowingaseismicsource150mabovethem.Thisconfigurationeliminatedtheproblemofcontinualchangesintherelativepositions
oftheshotandreceivercausedbythereceiver’smotion,simplifyingtheproblemtothatofthegeophysicalissueofdepthimaging.Aspart
ofacollaborationprojectbetweenIFREMERandtheFrenchoilcompanyTOTAL,thesurveywastoassessthefeasibilityofthisapproach
and its value for the characterization of a large-scale buried Mass Transport Complex (MTC), initially identified in 3-D seismic data. In
Section2ofthispaperdedicatedtotheintroductionoftheERIG3Dexperiments,wepresentthegeologicalcontextbasedonexistingdata,(i.e.
seismicdataandinsitumeasurements)andtheothergeophysicalsystemsusedduringthecruise.Section3coversthejointdeep-towed/OBH
experimentperformedduringthecruisetoinvestigatetheburiedMTC.Section4dealswiththeprocessingsequenceofthestreamerandOBH
data,includingnavigation,signalprocessingandvelocityanalysis.Theacquisitiongeometryandtheimprovementsintheseprocessespermit
depthimagingoftheOBHdatathroughaniterativefocusingapproachdescribedinSection5.InSection6,theresolutionprovidedbythe
differentgeophysicaldevices[hull-mountedsub-bottomprofiler,AutonomousUnderwaterVehicle(AUV)sub-bottomprofiler,deep-towed
systemandOBH]iscomparedandthedata,combinedwiththatobtainedbythejointdeep-towed/OBHexperiment,areusedtostudythe
buriedMTC.TheseOBHdatainterpretedinconjunctionwithsedimentologicalandgeotechnicaldataprovidenewinsightsintotheMTC;
evidenceofashorttransportdistanceandrigidbehaviourmayindicateaslidenatureofapartoftheMTC.
2 THE ERIG3D EXPERIMENTS
2.1 Geologicalbackground
TheNigerDeltaisa10–12-kmthicksedimentwedgethathasprogradedontotheAfricancontinentalmarginandadjacentoceaniccrust
sincetheearlyTertiary(Evamyetal.1978;Haacketal.2000).GravityslidingassociatedwithdifferentialloadingofearlytomiddleTertiary
marineshalesbytheprogradingdeltahasamajorimpactonitsstructuralstyleandevolution(Knox&Omatsola1989;Hooperetal.2002;
Morley2003;McClayetal.2003;Corredoretal.2005).Thedifferentialloadassociatedwithunevendeltaicsedimentationisthoughtto
havegeneratedoverpressureinmarineshalesandsubsequentdevelopmentofmultipledetachmentlevelslinkingupdipextension,downdip
contraction and compensating translation of the intervening overburden (Morley 2003; McClay et al. 2003; Corredor et al. 2005; Briggs
etal.2006).Extendingover(cid:3)200km2 fromtheshelfedge/upperslope((cid:3)200mbsl)tothemid-slope((cid:3)700–900mbsl)theMTCnamed
NG1liesbetweentheextensionalandtranslationaldomainoftheeasternNigerDelta(Garzigliaetal.2010).Extensionontheupperslope
isaccommodatedbynormalgrowthfaultsaffectingtheproximalpartofNG1.Itsdistalpartrestsontheflankofaprominentdome-shaped
bathymetrichighformedbyashale-coredanticlinewhoserelativelyrecentactivityismanifestedbycollapsenormalfaultsdiscernibleonthe
seabed(Fig.1).ThethicknessofthesedimentoverlyingNG1isupto120montheupperslopeanddecreasesprogressivelyto(cid:3)15–30mat
mid-slope,in800mwaterdepth,wherethisstudyfocuses.
(cid:2)C 2010TheAuthors,GJI,182,1524–1542
Journalcompilation(cid:2)C 2010RAS
1526 S.Keretal.
Figure1. (a)SeabeddipmapshowingingreythearealextensionofthedistalpartofNG1.Locationoftheseismicdata,coresandinsitumeasurements
collectedduringtheERIG3Dcruise.(b)Close-upofthelocationmapshowingthedatausedinthisstudy.
2.2 Seismicandinsitumeasurements
Theoverallgeologicalsettinghasbeenintensivelyinvestigatedforoilprospecting.3-Dseismicsurveysaswellasexplorationandgeotechnical
boreholeshavebeendrilledprovidingdetailedregionalconstraints.A3-DseismicvolumewascourteouslyprovidedbyTOTALforthisstudy.
Thisconventional3-DdatasethasbeenreprocessedbyTOTALtoenhancetheresolutionoftheupperpartofthevolume.Theprocessing
sequenceincludedshort-offsetalgorithmsandpre-stacktimemigrationwithinlineandcrosslinespacingof6.25and12.5m,respectively,
andthefinalverticalandhorizontalresolutions,7and25m.The3-Ddatasetprovidesthelarge-scaleimagingoftheMTCusedinthisstudy
toevaluatethehigh-resolutionimagingmethoddevelopedinthispaper.
OneoftheobjectivesoftheERIG3Dproject,wastofocusonsmall-scaleinvestigationoftheMTC.Duringthecruise,ontheFrench
R/VPourquoiPas?,asuiteofgeophysicalinstrumentsweredeployed,including:
(i) Twosub-bottomprofilers(SBP):one,aconventionalhull-mountedSBPandtheotheraSBPmountedonanAUV.Bothsub-bottom
profilersoperateatthesamefrequencyrange(1800–5000Hz),givingaverticalresolutioncloseto20cm.Theirlateralresolutionisconstrained
bythewaterdepthbelowtheinstrument.WhereastheAUVwasdeployedattheconstantaltitude(abovetheseafloor)of80m,givingalateral
resolutionof5m,thewaterdepthof800mrestrictedthelateralresolutionofthehull-mountedSBPto15m.
(ii) TheSYSIFdeep-towedseismicdevice:thelowerfrequencyrangesoftheseacousticsources(220–1050Hzand580–2200Hz)allowed
seismicimagesofthesub-bottomtobeacquiredwithametricresolution.However,thelowerfrequencybandwidthofSYSIFcomparedto
SBP,allowsdeeperpenetrationinthesubstrata.
(iii) Anoriginalhigh-resolutionseismicapproachinvolvingboththeSYSIFdevice(220–1050Hz)andOBHs:theaimofthisexperiment
istoprovidethemulti-offsetrecordingrequiredforvelocityanalysisanddepthimagingalgorithms.Itconstitutesthecentralsubjectofthis
paperandisthereforedescribedindetailinthefollowingsections,frombothatechnologicalandscientificpointofview.
InsitumeasurementscollectedduringtheERIG3Dcruisewereusedtogroundtruththegeophysicaldata.Inaddition,thisstudyuses
lithologicalandbulkdensitylogsdeterminedfromtheanalysisofcoresectionsofa60-mdeepboreholecourteouslyprovidedbyTOTAL.
Bulkdensitiesweredeterminedfrommeasurementsofmassandvolumeofrandomlysampledsediments.Insitumeasurementsincluded
conepenetrationtestsandP-wavevelocitymeasurements(at500MHz).Thesewerecarriedoutdownto30mbelowtheseabedusingthe
PENFELDpenetrometer(Sultanetal.2007)developedbyIFREMER(seeFig.1forlocationofthemeasurements).Themethodusedto
interpretsedimenttypesfromthePENFELDmeasurementsfollowedtheapproachproposedbyRamsey(2002).
(cid:2)C 2010TheAuthors,GJI,182,1524–1542
Journalcompilation(cid:2)C 2010RAS
High-resolutionseismicimagingindeepsea 1527
Figure2. PictureoftheSYSIFheavy-towedvehicleduringaJH250-6000deployment.
3 SYSIF DEEP-TOWED DEVICE AND OBH EXPERIMENTS
3.1 Descriptionoftheseismicdevices
The SYSIF deep-towed device, designed by IFREMER and illustrated in Fig. 2, consists of a seismic source and a streamer. This heavy
vehicle,weighing2.4tonsinair,istowedbythevesselwithanarmouredelectro-opticalcabledelivering1000VACpower,andbi-directional
telemetryfortheseismicpayloadandsafetycontrols.Thenavigationisachievedthroughfoursystems:the120kHzaltimetermeasuresthe
verticaldistancetotheseabed,thequartzpressuresensorcalculatesthedepthfromthesealevel,theminiatureattitudeandheadingreference
systemmeasuresthestabilityofthevehiclewhilethe14–16kHzultra-shortbaselineprovidestherelativepositionofthevehiclethrough
acousticpositioning.
TheSYSIFseismicsourcehastowithstandhighhydrostaticpressure;thesolutionwastoadaptthetechnologyofaJanus–Helmholtz
acoustic transducer, initially designed for low-frequency active sonars, to the needs of seismic surveying. A Janus–Helmholtz trans-
ducer consists of a piezoelectric ceramic stack inserted between two similar head-masses (Le Gall 1999). This structure, called a Janus
driver, is mounted inside a vented rigid cylindrical housing, providing a Helmholtz cavity. The coupling of mechanical resonance and
fluid resonance permits a large frequency bandwidth greater than two octaves. With this performance, the seismic source is able to
emit long-duration frequency-modulated acoustic signals, called Chirp signals, well adapted to increasing both resolution and signal-
to-noise ratio using specific processing algorithms. The amplitude variations of the output signal due to the Transmitted Voltage Re-
sponse of the transducer are taken into account through amplitude modulation. Based on this mature technology, two seismic sources
have been designed: the JH250-6000 and the JH650-6000. The source JH250-6000 operates between 220 and 1050 Hz, and is 112 cm
high,72cmdiameterandweighs450kg.TheJH650-6000forveryhigh-resolutionsurveysoperatesbetween580and2200Hz,is61cm
high,45cmdiameterandweighs90kg.Theoutputlevelof196dB(ref.1μPaat1m)overthewholefrequencyrangeisachievedusinga
single6.5kVAD-classpoweramplifier.
TheSYSIFstreamerisadualchannelantennamadeofTUBA6000hydrophones.Thesehydrophonesarepiezoelectricceramiccylinders
whosehighsensitivitytechnologyof−193dB(ref.1V/μPa)withstandshighhydrostaticpressurewithoutalossofsensitivity(1dB/600
bars). The first channel of the streamer is a single hydrophone with an offset of 10 m from the seismic source; this trace is used in the
experiment to process the recorded signal amplitude. The second channel has an offset of 15 m from the source and is made up of six
hydrophones,30cmapartandparallel-mountedtoincreasethesignal-to-noiseratio.Topreventsaturationfromthedirectwave,therecording
usesanalogueelectronicsincludingabandpassfilterof18dBperoctaveintherange100–3000Hzanda26.3dBpre-amplifier.Analogueto
digitalconversionisthenachievedat10kHzthrougha26bitanalogue-to-digitalconverter.
Becausethedevicelacksmultichanneltechnologytoachievedepthimaging,OBHsweredeployedontheseafloorduringtheERIG3D
cruisetorecordoffsetdata.TheOBHsareautonomousrecordinginstrumentsthatallowthedigitizationofacousticmeasurementsofthe
hydrophone. Their synchronization is achieved through a GPS clock compensated for long time drift. The storage capacity of the OBH
exceeds1Gbyte,whichallowsfourdaysofacquisitionusinga24-bitsigma-deltaADCat5kHZ(Auffretetal.2004;Westbrooketal.2008).
(cid:2)C 2010TheAuthors,GJI,182,1524–1542
Journalcompilation(cid:2)C 2010RAS
1528 S.Keretal.
Table1. GeophysicaltoolsusedduringtheERIG3Dcruiseandacquisitionparameters.
Seismicdevice Frequency/duration Altitude Shootingrate Shipspeed SL Linename
(Hz)/(ms) (m) (s) (Kn) (dBμPa@1m)
SysifJH650-6000 [580-2200]/50 60 2.4 2 196 SY13-VHR-Pr01
SysifJH250-6000 [220-1050]/100 150 2.6 1 196 SY10-HR-Pr01
Sysif/OBH [220-1050]/100 150 2.6 1 196 SY10-HR-Pr01
Chirp(Hull) [1800-5000]/50 Surface 1 5 210 CH18
Chirp(AUV) [1800-5000]/50 80 1.3 3 190 AUV03-Pr16
3.2 Deep-towedandjointdeep-towed/OBHseismicacquisition
The main objectives of the joint deep-towed/OBH experiments were to investigate the MTC NG1 both qualitatively (morphology and
subsurfacestructure)andquantitatively(acousticpropertiesandP-wavevelocitiesinparticular).Toreachthesescientificobjectives,accurate
positioning of both the deep-towed system in the water column and the OBH sensors on the seafloor is a critical issue for the frequency
bandwidth.Thepositioningofsuchsystems,thatcanbeaccuratelycontrolledonlandorinmarinesurface-towedsurveys,stillpresentsmany
difficultiesinunderwaterenvironments(Yoergeretal.2007).
DuringtheERIG3Dexperiment,fiveOBHswereloweredontoaflatseaflooratawaterdepthof800m(Fig.1b).Thedeploymentwas
controlledbyanacousticallynavigatedcoringwireabout10mabovetheseabedusinganultra-shortbaseline,thatis,anacoustictransponder
measuringthetraveltimeofsoundbetweenthedeviceandthevessel.Thesoundvelocitythroughthewatercolumnthatisrequiredforthe
timetodepthconversionwasmeasuredbyaconductivitytemperaturedensity(CTD)probeimmediatelybeforedeployment.Thenominal
receiverspacingwas25mbetweentheOBHswithatheoreticalpositioningerrorof1percentofthewaterdepth,thatis,roughly8maround
thenominalposition.
Identically to the OBHs, SYSIF was positioned by an ultra-short baseline navigation system. However the time measurements were
lessprecisebecausetheacquisitiongeometryinvolvedobliqueanglesinsteadofstraightpaths.ThedepthandthealtitudeofSYSIFwere
measuredsimultaneouslybyapressuresensorandanechosounder,withrelativeprecisionof6and12cm,respectively.
The shot intervals for the acquisition were calculated to avoid recording the high energy multiples related to the sea surface, that
is, source ghosts and seafloor-related multiples; such artefacts are impossible to remove once acquired on the SYSIF section. Multiple
removaltechniquesbasedondifferenceinmoveoutcannotbeusedhereandtechniquesbasedonthemultiple’sshape,includingpredictive
deconvolution,alsofailduetotheperturbationofthewaveformafterreflectionontheroughseasurface.Weuseddynamicraytracingand
computedthemultipleamplitudesandtraveltimesuptothefourthordertoadapttheshotintervaltoourspecificacquisitiongeometry.
TheinvestigationoftheMTCNG1wascarriedoutusingthewholesuiteofgeophysicaltoolsavailableonboardtheR/VPourquoiPas?.
Thereferencelinewasthereforesurveyedatdifferentfrequenciesthatprovidedauniquedatasetforthestudyofmarinegeohazards.The
differentacquisitionparametersarepresentedinTable1.AgeologicalinterpretationofthisdatasetisproposedinSection6.
4 STREAMER AND OBH DATA PROCESSING
4.1 ProcessingoftheSYSIFnavigationfromjointOBHdata
Dataacquisitionusingatowedsystemintroducesspecificvariationsinthefishpositionthattendtoincreasewiththelengthofthedeployed
(unfurled) cable. Initially, the acquisition system is towed at a constant altitude to track the bathymetry of the proposed line. Using this
informationthedepthofthesystemcanbecontrolledduringactualacquisitiontosmoothlyfollowthisbathymetryataconstantaltitude
andavoidsuddenchangesduetoabruptvariationsintheseafloor.Othervariationsarerelatedtothetransmissionofthemotionoftheship
tothedeep-towedsystemthroughthecable,andthehydrodynamicdragofthecable.Headingandlateralmotionsaffectthetrackofthe
deep-towedsystembyintroducinghigh-frequencyvariationsincomparisontothesmoothprogrammedvariationsinimmersiondepth.These
high-frequencyvariationsinthedeep-towednavigationappearastime-shiftsintheseismicdatameasuredonthefixedOBHs,asillustrated
on Fig. 3(a) for the OBH3. These undesirable variations also affect the streamer data, as reported by Rowe & Gettrust (1993) on their
DTAGSsystem.AmethodwasproposedbyWalia&Hannay(1999)toremovethis‘jitter’;theoverallmotionwasinterpretedasavertical
displacementandevaluatedusingtherecordedsea-surfacereflectiontimetocorrecttheseismicdatatoaconstantdepth.Takingadvantage
ofthefixedpositionoftheOBHs,weproposeamoremeticulousapproachtocorrecttheacquisitiongeometryandprocessthejitterartefact
byconsideringdirectSYSIF-OBHtraveltimesintheevaluationoflateralandverticaldisplacements.
RawnavigationdatafortheSYSIFcomefromanultra-shortbaselinepositioningsystem.Thesedatarelyonthedirecttraveltimesof
an8–17.5kHzsignaltothetowingvessel,andareunevenlyspaced.Theaccuracyofsuchpositioningdataisgivenasapercentageofthe
verticaldistancebetweenthetowedvehicleandthetowingvessel(1percent).Otherindividualsensors,thatis,thealtimeterandthepressure
gauge,areusedforreal-timemonitoringofthedeep-towedvehicle;theyproduceoversampledinformation(0.5Hz).Mergingthenavigation
datafromthedifferentindividualsensorsisconsequentlyapreliminaryprocessingsteptoincreasetheaccuracyinpositioning.
ForamoreaccuratenavigationoftheSYSIFdeep-towedvehicle,weincludedarelocationalgorithmbasedontraveltimeinversionof
theseismicsignalsrecordedattheOBHs.ThedirectarrivalsoftheOBHtracesaretimewindowedbyasinetapertoisolatethedirectarrival
(cid:2)C 2010TheAuthors,GJI,182,1524–1542
Journalcompilation(cid:2)C 2010RAS
High-resolutionseismicimagingindeepsea 1529
Figure3. OBH3before(left-handpanel)andafter(right-handpanel)thecorrectionofthejitterrelatedtothesourcepositionvariations.
waveforms and then cross-correlated with the source signature (the measurement of the far field signature of the different transducers is
describedinthenextsection).Thetimeofmaximumcross-correlationcorrespondstotheaccuratedirectarrivaltimeusedfortheinversion.
WeassessthevalidityofthisapproachwithaquantitativeestimatordefinedbyPearson’scorrelationcoefficient(Taylor1982;Molyneux&
Schmitt1999)expressedas
(cid:3) (cid:3) (cid:3)
n t=nx (t)x(t+τ)− t=nx (t) t=nx(t+τ)
r(τ)= (cid:4)(cid:5) (cid:3) t=0(cid:6)(cid:3)0 (cid:7) (cid:8)(cid:5) (cid:3)t=0 0 t=0 (cid:6)(cid:3) (cid:7) (cid:8), (1)
n t=nx2(t)− t=nx (t) 2 n t=nx2(t+τ)− t=nx(t+τ) 2
t=0 0 t=0 0 t=0 t=0
wherex (t)andx(t)aretwowaveformsandτ isthetimeofthecross-correlationmaximum.
0
Thisestimatorisapositiveparameterwhosemaximumvalueof1correspondstoaperfectmatchbetweenbothsignals.Wecalculated
thisestimatorforalltheshotsofourexperimentandthenexpressedtheresultsasafunctionoftheincidenceanglebetweenthesourceand
thereceiver(Fig.4a).DuringtheacquisitionoftheSY10-HR-PR01seismicline(Table1),theanglerangeis(−65◦,+65◦)withpositive
valueswhenthesourceistotheeastofthereceiver.Fromthisresult,weobservegoodsymmetricalbehaviourofthesystemcharacterizedby
constantcorrelationcoefficientsbetterthan0.9forangles−40◦to+40◦andastrongdecreasedownto0.6forhigherangles.Wecanassociate
theseanglerangestothesourcedirectivityestimatedfromthepowerspectraldensityofthedirectwaveformsrecordedbyOBHs(Fig.4b).
Foranglessmallerthan40◦,thetransducerdirectivityisnegligible,correspondingtotheconstantcorrelationcoefficients.Forhigherangles,
variationsinthedirectivityareinaccordancewithadecreaseinthecorrelationcoefficient.Thedirectwaveformwasconsideredtoodifferent
fromthesourcesignatureforacorrelationcoefficient lessthan0.7;thisvalue alsobeingatrade-offbetween anaccurate estimateofthe
traveltimeandarepresentativenumberofshotstoconsiderintherelocationprocess.Thevalueof0.7correspondstoamaximumincidence
angleof(cid:3)60◦oramaximumoffsetof(cid:3)250m.
Therelocationprocessthenconsistsoffindingthepositionsofboththe300shotsandthefiveOBHsutilizingthesignalsthatsatisfy
thecorrelationcoefficientcriterion.Theinverseproblemsearchesforthecalculateddirectarrivaltime‘tcal’whichminimizes,intheleast-
squaressense,thedifferencetotheobservedtime‘tobs’.Tosolvesuchanon-linearminimizationproblem,weapplytheTrust-Regionmethod
(Steihaug1983;Byrdetal.1988).
The sound velocity in the water column and the immersion depth of both the SYSIF seismic source and the OBH hydrophones are
assumedtobeaccurateenoughanddonotneedtobeinvertedintherelocationprocess.Therefore,therelocationprocedurefocusessolely
onvariationsinpositioninalateralsense.
The residuals from the inverse problem, defined as the mean values of ‘tobs − tcal’, are reported in Table 2 for each OBH. When
consideringonlypreliminaryprocessednavigationdata,twostepsarefollowedtoreducethemeanresidualvaluehigherthan1ms.
(i) instep1,theinitialshotpositionsarekeptconstanttoestimatetheOBHcoordinatesusingtheinverseproblem,
(ii) instep2,therevisedOBHlateralpositionsarethenusedtocorrecttheseismicshotcoordinates.
(cid:2)C 2010TheAuthors,GJI,182,1524–1542
Journalcompilation(cid:2)C 2010RAS
1530 S.Keretal.
Figure4. (a)CorrelationcoefficientforOBHdatacalculatedwithfarfieldsignature.(b)PowerspectraldensityofthedirectwaveformforOBH3.
Table2. Residuals(meanvaluesinms)obtainedaftereachstepoftherelocation
procedureforthefiveOBHs.
Step OBH1 OBH2 OBH3 OBH4 OBH5
Initial 1.69 2.12 1.92 0.70 1.14
Step1 0.32 0.34 0.36 0.65 0.93
Step2 0.31 0.23 0.32 0.31 0.6
Asaresult(Table2),therelocationprocessproducesaminimizationofthemeanresidualdownto(cid:3)350μs(exceptOBH5:600μs).
ThefinaldisplacementoftheshotsandOBHcoordinatesiswithinaradiusoflessthan10mandisthereforeconsistentwiththeresolution
oftheprimarynavigation.
Thederivedlateraldisplacementswerethenappliedtotheprimarynavigationdatatoevaluatetheoriginalprocessingfromthejoint
deep-towed/OBHacquisition.TheapproachconsistedoffixingadatumplanefortheSYSIFsourcedepth,computingtheassociateddirect
arrivaltimes‘tnew’andthenevaluatingthedifferences‘tobs−tnew’.Thistimecorrectionisgivenbytheexpression
(cid:9)(cid:10) (cid:10) (cid:13)
1 (cid:11) (cid:12)
(cid:4)t = cw (xs−xr)2− xs,z=z0−xr 2 , (2)
withcw thesoundvelocityinthewatercolumngivenbytheCTDprobe,xs andxr thesourceandreceiverpositions,respectively;xs,z=z0
standsforthesourcecoordinateswithaconstantimmersiondepthofz .TheresultsofapplyingthisrelocationcorrectionontheOBHtraces
0
areshowninFig.3.Theimprovedlateralcoherencyofthereflectedwavesclearlydemonstratesthevalidityoftherelocationprocedure.
4.2 Chirpsignalprocessingfromrealsourcesignature
Theconventionalsignalprocessingmethodusedinmarinesub-bottomprofiling(Quinnetal.1998)istocross-correlatetherecordedtraces
withatheoreticalpilotsignalorwiththedirectwaverecordedatthenearesthydrophone(Chapmanetal.2002;Heetal.2009).Theeffectis
tocompressthesignalatthereflectedhorizons.
Thelimitofthisapproachisbasedontheassumptionthatthepilotsignalmatchestherealoutputsourcesignal,thatis,thetransfer
function of the seismic source including the power amplifier, the impedance matching unit, the transducer, etc., can be considered to be
negligible, which is obviously not true.A more meticulous signal processingconsists of considering the in situfar field measurement of
thesourcetoperformsignaturedeconvolution(Ziolkowski1991);thissolutionisnotstraightforwardandrequiresspecificsourcesignature
measurementsatsea.
To evaluate this approach, we designed and performed sea trials in 2009 May to record the far field seismic signatures of both the
JH250-6000andJH650-6000transducers.Adedicatedsignaturehydrophone(HighTechHTI90hydrophone),connectedtoanautonomous
analoguerecorder,wasanchoredatadistanceof12mbelowtheSYSIFdeep-towedvehicle(thevalueof12misconsistentwithfarfield
signatureconstraintsasourlowerfrequencyis220Hz,i.e.6.7mwavelength).Toavoidinterferencefromechoesfromeithertheseabottom
ortheseasurface,themeasurementstookplacein2000mofwaterwiththeSYSIFbeingloweredto1000mbelowthestationaryvessel.
Themeasurementsunderlinedtheexcellentrepeatabilityofbothsources.Thedifferencesbetweenthefarfieldsignatureandthepilot
signalareclearandreflectnotonlythetransferfunctionofthesourcesbutalsoastrongresonancefrequencyassociatedwiththeHelmholtz
(cid:2)C 2010TheAuthors,GJI,182,1524–1542
Journalcompilation(cid:2)C 2010RAS
High-resolutionseismicimagingindeepsea 1531
Figure5. (a)Meansourcesignaturerecordedinfarfieldcondition.(b)Powerspectraldensityofthesourcesignature.
cavityat300Hz(Fig.5).Thisconfirmsthatasourcesignaturedeconvolutionapproachispreferabletoconventionalcorrelationprocessing
whichwouldhaveblurredthestructuresinthefinalseismicprocessing(Brittleetal.2001).
Thedeconvolutionprocedureisbasedontheconvolutionalmodelwhichstatestheseismictraceistheconvolutionbetweenthesource
signatures(t)andthegeologicalreflectivityr(t).Inthepresenceofnoisen(t),theseismictraceisexpressedas
x(t)=s(t)r(t)+n(t). (3)
AsourcesignaturedeconvolutionisthendefinedbyadivisionintheFourierdomain(Yilmaz1987).
X(ω) R(ω)S(ω) N(ω)
= + , (4)
S(ω) S(ω) S(ω)
where X(ω) isthedeconvolvedtraceintheFourierdomainwith X(ω),S(ω),R(ω)andN(ω)aretheFourierTransformof,respectively,
S(ω)
thedata,thesourcesignature,thegeologicalreflectivityandtheadditivenoise.Inpractice,anoisefactorhastobeaddedtothesourcebefore
thedivisiontopreventanydivisionbyzeroandtoreducetheamplificationofnoise.
OBH data present excellent signal-to-noise ratio exceeding 36 dB and therefore, on this particular data set, the noise factor is not
necessarytoensurethestabilityofthedivision.Thestreamerdataisalsocharacterizedbyahighsignal-to-noiseratio(27dB)whichmay
berelatedtotheperformancesofthereceiver,tothe‘peaceful’deep-waterenvironmentandtothelowspeedoftheacquisition(1–2knots).
Asaqualitativeresult,acomparisonbetweentheconventionalcorrelationandthesignaturedeconvolutionisshowninFig.6:thedifferences
inthelower partof thesectionwhere thesignal-to-noiseratioisminimum areclear. A350–1100 Hzbandpass frequency filter has been
appliedinthecorrelationsection(Fig.6a)toavoidtheHelmholtzresonancefrequencywhereasthefilterbandwidthwas200–1100Hzinthe
deconvolutionsection(Fig.6b)toattenuatefrequenciesoutsidetheemittedbandwidth.
As a quantitative result, we can estimate a vertical resolution of the imaging process between 0.5 and 1 m, thanks to both the large
frequencybandwidthofthesourceandthesuccessfulsignaturedeconvolution.Moreover,thelowaltitudeofthedeep-towedvehicleabove
theseafloor(150m)allowsagoodlateralresolution,between15mand20mdependingonthereflectordepthtobeachieved.
(cid:2)C 2010TheAuthors,GJI,182,1524–1542
Journalcompilation(cid:2)C 2010RAS
1532 S.Keretal.
Figure6. Left-handpanel:correlationprocessing(filter350–1100Hz).Right-handpanel:deconvolutionprocessing(filter200–1100Hz).
4.3 Analysisofthesubsurfacevelocityprofileinthetimedomain
Unlikesurfaceseismicrecords,thelocusofreflectionpointsforeachsource–receiverpairinourOBHsurveychangeswithreflectordepth.
In a horizontally layered medium, this locus tends towards the source–receiver midpoint with increasing depth. Therefore, the OBH data
cannotbesortedintocommonmid-pointgathersandthenormalmove-outvelocityanalysishastobecarriedoutonreceivergathers.The
geometryoftheacquisitionisasymmetricalduetothedifferenceinaltitudebetweenthesourceandthereceiver;thus,wehavetoredefine
thetraveltimeequationofreflectedeventsandtoadapttheNMOapproachtothisparticularcase.Forthesakeofsimplicity,weconsidera
strictly2-Dacquisitiongeometry(whereboththereceiverandthesourceareonthesameplane).Thereceiverislocatedontheseabedandif
weassumeflatunderlyinglayers,thearrivaltimeoftheP-wavereflectedwaveattheinterfacenisthusexpressedas
x2
t2≈t2 + , (5)
0,n 2
V
n
wheret0,nisthezerooffsetarrivaltimeoftheP-wavereflectionattheinterfacen,xisthehorizontaldistancebetweenthesourceandthe
receiver(offset)andVn isthermsvelocity.t0,nandVn canbeexpressedinfunctionofvk anddk whicharerespectivelytheP-waveinterval
velocityandthicknessofthek-thlayer.Thefirstlayeristhewatercolumn.
⎡ ⎤
2
t2 =⎣(cid:16)k=n dk +(cid:16)j=n dj⎦ (6)
0,n v v
k=1 k j=2 j
andV thequadraticvelocityexpressedas
n
(cid:3) (cid:3)
1 = (cid:3) kk==n1 dvkk +(cid:3)jj==n2 dvjj . (7)
V2n kk==n1vkdk+ jj==n2vjdj
Thisformulationofthetraveltimeissimilartotheoneusedinoffsetverticalseismicprofileprocessingwhichsharesitsasymmetrical
acquisitiongeometrywithourOBHexperiment(Dillon&Thomson1984).
ToperformtheP-wavevelocityanalysis,negativeandpositiveoffsets,correspondingtoOBHpositionsontheeastandwestsideof
theSYSIFseismicsourcewereconsideredseparately.ThesemblancepanelsforOBH3areshowninFig.7.AnaverageP-wavevelocityof
between1480and1490ms−1wouldbeinterpretedforthesedimentarycoverlocatedinthefirst200msbelowtheseabed.Thisvalueappears
(cid:2)C 2010TheAuthors,GJI,182,1524–1542
Journalcompilation(cid:2)C 2010RAS