thehighamplitudere?ectorsindicatespossiblelowconcentrationsofgashydrate(Fig.2andTable1).Zone2appearstobedividedroughlyintotwohalvesbyfaulting(Fig.2),withstrongtroughamplitude(yellowinFigure2andTable1)ontheleftsideoftheseismiclinewhichisinterpretedtobeagaszonewithafewthickhighlyconcentratedgashydratelayersinthemiddlepartofthestructure(theleftsideofZone2inFigure12)andstrongpeakamplitude(aquainFigure2andTable1)ontherightsideoftheseismiclinewhichisinterpretedtobeathickhydratelayerwithoutgasbelowit(therightsideofZone2inFigure12).Well-log-derivedgashydratesaturationsareestimatedtobemorethan50%oftheporespacebetween410mbsfand450mbsfintheGC955-Hwell(Guerinetal.,2009).Zone4hastwohighpeakamplitudeanom-alies(aquainFigure2andTable1)andZone6showsastrongtroughamplitudere?ector(yellowinFigure2andTable1)withmoderatepeakamplitudeeventsabove(blueinFigure2andTable1).Figure12showsZone4ispredictedtobeahighlyconcentratedgashydratelayerandZone6ispredictedtobeafreegaszonewithathingashydratelayerabove.Lowconcentrationorthinhighlyconcentratedhydrate-bearinglayersareinterpretedinmostareasofZone3andZone5whichareassociatedwithhightroughamplitudere?ectorswithmoderate-to-highpeakampli-tudeabove(Fig.2andTable1).Hightroughamplitudes(yellowinFigure2andTable1)areinterpretedtobefreegaszones.Highlyconcentratedgashydrateisinterpretedtooccurinthemiddleofthetwozones(Zones3and5inFigure12),suggestingthathydratesaturationsorthicknessofthehydratelayerarehorizontallychanginginthesandlayers.

4.2.Thechemicalandphysicalcontrols

Wehaveshownagashydrateandfreegaslayeredsystemwithhighlyvariablecontactrelationshipsbetweentheoverlyinggashydrateandunderlyingfreegas.Thevariationsinthelayersofgashydrateandfreegasverticallyandlaterallysuggestdifferentlocalgashydratestabilityconditions.Fluidandgasmigrationsarethoughttobecurrentlyactiveinthestudyarea.Thisactive?owmayimpacttheporewatersalinityandsubsurfacetemperatureconditions(Woodetal.,2002;DevandMcMechan,2010).Thegasesinvariousstructures(zones)couldalsohavedifferentcomposi-tions,suchasbiogenicgaswithhighmethanecontentandther-mogenicgascontainingmoreethane,propane,andbutanes,whicharefoundinGulfofMexico(Brooksetal.,1986).

Localvariabilityingashydratestabilityconditionscanalsobeattributedtolocalizedstructuralcontrolsand?uid?owdynamics.DifferencesinphysicalpropertiesofsedimentsareanotherpossiblereasontointerpretgashydrateandfreegasatdifferentdepthsacrossGC955.AsupwardmigratingwaterandgasentersthevariousfaultblocksalongthecrestofGC955,the?uidcanaffectthevarioushydrate-bearingsandreservoirsdifferentlydependingonlocalstructuraland?uidmigrationcontrols(e.g.thickhighlyconcentratedgashydrateinthezone2andzone4).Itispossiblethattheformationofhydrateinaclosedstructurecouldaffectlocalporepressureconditions,whichinturnwillimpactgashydratestabilityconditions(Ruppel,1997;Kraemeretal.,2000;Torresetal.,2004).Ithasbeensuggestedthatgashydrateformationcanalso“clog”largeporespaceconduitscausinglocalizedover-pressuring(Eckeretal.,1998);thus,gasaccumulatesbelowhydrates(e.g.freegasoverlainbythickhighlyconcentratedgashydrateinthemiddleofthezones1,3,and5).5.Conclusions

Anapproachintegratingrockphysicsmodeling,amplitudeanalysis,andspectraldecompositionimprovesourabilityto

132Z.Zhangetal./MarineandPetroleumGeology34(2012)119e133

identifyandcharacterizethepresenceofgashydrateandfreegasintheGreenCanyon955(GC955)area.Theinteractionofvaryinggashydratesaturations,reservoirthicknesses,andfreegasoccurrencegeneratesacomplexexpressioninseismicpro?lesacrossGC955.Ingeneral,lowamplitudepeakeventsoccurinthelow-saturationhydrate-bearingsandsandhighamplitudepeakeventsoccurinhighlysaturatedhydrate-bearingsands.Theamplituderesponse,however,ismorecomplicatedinthemixed-layergashydrateandfree-gassystem.Highsaturationgashydrateoverfreegasinsandswilldriveamplitudeslightlylowerthanforagassandwithnooverlyinggashydrate.Sandswiththickgashydratedepositsunderlainbyfreegasmaywellhavestrongtrough-dominatedwaveforms.Furthermore,amplitudeisaffectedbybedthick-nessesinallthingashydratedeposits,butespeciallysowhereunderlainbyfreegas.Lowamplitudeeventsoccurinverythinhighlyconcentratedhydrate-bearinglayersandhighamplitudeeventsoccurinrelativelythick,butnotthethickest,highlyconcentratedhydrate-bearinglayers,ifthereisnotfreegasbelow.Thus,thisworksuggeststhepossibilitythatthehighlysaturatedgas-hydrate-bearingsandsdrilledinthe2009JIPLegIIcampaignmaynothavebeenthethickestinthiscomplexsystem.

Inthisstudyweinterprettheoccurrenceofgashydrateandfreegasfromwedgemodelsbuiltonlocalgeologicandgeophysicalcriteria.Themodelspresentedprovideimportantexamplesofgashydrateandfreegasaccumulationsandrelationships,butnotacompleteviewofthefullrangeofpossibleconditions.Wehaveconcentratedoncompressional-wavepost-stackamplitudeanal-ysis,whileneglectingattenuation,scattering,andanisotropy.WehavenotdiscussedtheprestackAVOeffectthatcanshowampli-tudechangeindifferentoffsetswithrespecttovariablehydratesaturationandgassaturation(e.g.,CarcioneandTinivella,2000).Wealsoassumethatgashydrateisatthetopofthemodelandinferredsandlayersandthein-situfreegasisuniformlydistributedinthicklayers.Wedidnotinterprettheamplitudeanomaliesatthebottomofeachzone,astheymayintroduceerroneousorambig-uousinterpretations.Forexample:astrongpeakamplitudeanomalycouldbeassociatedwiththebaseoffreegaslayerorthetopofgashydratelayer.Toidentifythecauseoftheseanomalies,itispreferabletohaveexplicitcalibrationdatafromawell.

Finally,thisworkshowsthatspectraldecompositionofconventionalexplorationseismicdata,guidingbytheproperrockphysicsmodel,maybeaneffective,quantitativeprospectingtooltoidentifythick,highsaturatedgas-hydrate-bearingsandlayers.Theresultsofthischaracterizationofacombinedgashydrateandfree-gassystemareusefulexamplesthatallowinterpreterstoidentifygashydratereservoirsfromseismicdatainsimilarenvi-ronments.Additionally,thestudyofshallowfree-gasusingthesetechniquescanbeapplicabletoimprovethequantitativeidenti?-cationofpotentialshallowgasaccumulationsinpre-drillinggeo-hazardassessments.Acknowledgments

TheauthorsthankEmrysJones,whoguidedtheJIPfromitsstartthroughtheJIPLegIIexpedition,JohnBalczewski,Chevron,RayBoswell,NationalEnergyTechnologyLaboratories,theExecutiveBoard,andthemanyscientiststhatworkedwiththeGulfofMexicoGasHydrateJointIndustryProjectthatmadethisstudypossible.WegratefullyacknowledgeTimCollett,RayBoswell,andScottSingleton,whometiculouslyeditedthemanuscripttoimprovethetechnicalcontentandreadability.Constructivecommentsbythemandananonymousrevieweraremuchappreciated.WeespeciallythankJIPmembers,WesternGeco,forpermissiontoshowthedata.WethankAOAGeophysicsInc.forthesupportofthestudy.Finally,theauthorsthankAdrianDigby,AOAGeophysicsInc,forhishelpful

commentsanddiscussionoftheLWDdata.Preliminaryresultsoftheprojectwerepresentedatthe2010OffshoreTechnologyConference(OTC20659)andFireinTheIce(January2011).References

Adriansyah,A.,McMechan,G.A.,2002.Analysisandinterpretationofseismicdata

fromthinreservoirs:northwestJavaBasin,Indonesia.Geophysics67,14e26.Andreassen,K.,Hogstad,K.,Berteussen,K.A.,1990.Gashydrateinthesouthern

BarentsSeaindicatedbyashallowseismicanomaly.FirstBreak8,235e245.Avseth,P.,Mukerji,T.,Mavko,G.,2005.QuantitativeSeismicInterpretation:

ApplyingRockPhysicsToolstoReduceInterpretationRisk.CambridgeUniversityPress.

Boswell,R.,2009.IsgashydrateenergywithinReach?Science325,957e958.

Boswell,R.,Shelander,D.,Lee,M.Y.,Latham,T.,Collett,T.,Guerin,G.,Moridis,G.,

Reagan,M.,Goldberg,D.,2009a.OccurrenceofgashydrateinOligoceneFriosand:alaminosCanyonblock818:northernGulfofMexico.MarineandPetroleumGeology26,1499e1512.

Boswell,R.,Collett,T.S.,Frye,M.,McConnell,D.,Shedd,W.,Mrozewski,S.,Guerin,G.,

Cook,A.,2009b.GulfofMexicogashydrateJointindustryprojectLegIIdtechnicalSummary.In:ProceedingsoftheDrillingandScienti?cResultsofthe2009GulfofMexicoGasHydrateJointIndustryProjectLegII.http://www.netl.doe.gov/technologies/oil-gas/publications/Hydrates/2009Reports/TechSum.pdf.

Boswell,R.,Collett,T.S.,Frye,M.,Shedd,McConnell,D.,Shelander,D.,2012a.

SubsurfacegashydratesinthenorthernGulfofMexico.JournalofMarineandPetroleumGeology34,3e40.

Boswell,R.,Frye,M.,Shelander,D.,Shedd,W.,McConnell,D.,2012b.Architectureof