Milkov-2005-Gas hydrate systems at Hydrate Rid(2)

FourSites(1247,1248,1249,1250)arelocatednearorontheSouthernSummit,twosites(1251,1252)occurintheslopebasineastofHydrateRidge,andthreesites(1244,1245,1246)weredrilledasatransectjustnorthoftheSouthernSummit(Figs.1and2).Sites1248to1250arelocatedintheareaofrelativelyhighgas?uxfromthesubsurfacetothesea?oor,whereasSites1244to1247areintheareaofrelativelylowgas?ux(Milkovetal.,2003,2004b,2004c;Tréhuetal.,2003,2004).NogashydratesampleswererecoveredatSites1251and1252intheslopebasin,andtheseSitesarenotdiscussedhere.Thelocation,waterdepthandothercharacteristicsofSites1244to1250arelistedinTable1.AllthesesitesweredrilledintheareawhereaBSRoccursatseismicallyestimateddepthsof114to134mbsf(Tréhuetal.,2003)andgenerallyliesabovethedepthscalculatedfrommeasuredtemperaturesforthebaseoftheGHSZ(Table1).

3.PREVIOUSSTUDIESOFHYDRATE-BOUNDGASES

ATHYDRATERIDGE

GashydrateshavebeensampledatHydrateRidgebeforeLeg204.Hovlandetal.(1995),WhiticarandHovland(1995),andWhiticaretal.(1995)reportthecompositionofagashydratesamplefrom3.67mbsfatSite892nearthe

Northern

GashydratesystemsatHydrateRidgeoffshoreOregon1009

Fig.2.Schematiccross-sections(interpretedfrom3DseismiclinesinTréhuetal.,2003)showingthegeologicalsettingofSites1244to1246(a),1245,1247,and1248(b)and1249,1250(c)(seeFig.1forthelocationofthelines).Shadedareashighlightthewipe-outzonesandzonesofhighre?ectivityonseismicsections.Mainseismiceventssuchasthesea?oor,BSR,BSR2,HorizonsA,Y,andY=,andthetopofaccretionarycomplexsedimentsareindicated.Diamondscorrespondtotheintervalsfromwhichsamplesofgashydrateswererecoveredandanalyzed.

SummitofHydrateRidge(Fig.1).Gasfromhydratedecom-positionwasdominatedbymethane(C1)withlesserhydrogensul?de(H2S)andminorcarbondioxide(CO2),ethane(C2),propane(C3),isobutane(i-C4)andnormalbutane(n-C4)(Table2).Basedontheisotopiccompositionofcarbon,C1inthegashydratewasinferredtohaveamicrobialorigin.AssociatedH2Swasassumedtobefrommicrobialreductionofthedissolvedsulfate.Whiticaretal.(1995)concludedthatnear-surface(2–19mbsf)gashydratesatSite892wereformedfromdiage-

neticinsitugases.Althoughnogashydratesampleswererecoveredfromdeepsediments,theauthorsusedthecomposi-tionofvoidgasestosuggestthatthedeepergashydratesneartheBSRmayhaveasigni?cantcomponentofallochthonousthermogenicgas.

TheothergashydratesampleatSite892wasrecoveredfrom?17mbsf(Kastneretal.,1998).ThesamplecontainedhighconcentrationofC1andH2SwithminorCO2andC2(Table2).Hydrate-boundC1wasinferredtobelargelymicrobialbasedontheisotopicproperties.Incontrasttothepreviousconclu-sionsofWhiticaretal.(1995),Kastneretal.(1998)proposedthatC1withsomeC2advectedintothesulfate-reductionzone(0–20mbsf)alongafaultfromsedimentsatintermediateburialdepth.GashydratethatinitiallywasdominatedbyC1,incor-poratedH2Sfromsulfatereductionovertime.OxidationofC1mayaccountfortheminorvolumeofCO2ingashydrate(Kastneretal.,1998).

Manysamplesofmassivegashydratewererecoveredfromthesea?ooratboththeSouthernandNorthernsummitsofHydrateRidge(Suessetal.,1999,2002).However,gasprop-ertiesweremeasuredonlyinafewsamples.Suessetal.(1999)foundthatC1dominatesinthehydrate-boundgases,andsmallamountsofH2S,C2,C3andCO2werealsopresent.TheconcentrationsofH2Swerefoundtobesigni?cantlylower(byanorderofmagnitude)thanthosefoundatSite892.ThiswasattributedtohigherconcentrationsofH2SatgreaterdepthspenetratedatSite892.C1/C2?ratiosandcarbonisotopiccom-positionofC1indicatethatmicrobialmethaneisthemajorcomponentofgashydrates.

Winckleretal.(2002)providedmoremeasurementsoftheisotopiccompositionofC1fromsea?oorgashydratesattheSouthernSummitofHydrateRidge(Table2).Thesedatacon?rmedthatC1hasmicrobialoriginandformedviareduc-tionofCO2.Moreover,methanelackedmeaningful14Csug-gestingthattheorganicmatterfromwhichthemethaneorigi-natedisolderthan?50ka,andthatmethaneinsea?oorgashydratesisnotfromcurrentlygeneratinginsitusources.Winckleretal.(2002)alsofoundthatgashydratesamplescontainsigni?cantamountsofargon,krypton,andxenon,butheliumandneonaretracecomponents.

Tosummarize,severalgashydratesampleswererecoveredfrom0to18mbsfatHydrateRidgepriorLeg204.ItwasestablishedthatmicrobialC1isthemaincomponentofhydrate-boundgases.However,theC1mostlikelywasnotgeneratedinsitubutwassuppliedfromunde?neddepthsinsediments.SmallvolumesofC2?hydrocarbongasesfoundingashydrateswereassumedtobeallochthonousandofthermogenicorigin.MixedC1-H2Sgashydratesweresuggestedtobemostcom-mon.Itwasproposedthatthesea?oorgashydratesrepresenta“secondaryphase”becausetheyformedfromgassuppliedfromgashydratesdecomposedatthebaseoftheGHSZandwerere-stabilizedbyadditionofH2Sinsedimentsnearthesea?oor(Suessetal.,1999).InthefollowingsectionswewilltestthepreviousconclusionsbasedonthenewgasdatafromLeg204.

4.SAMPLING,MATERIALANDMETHODS

Inthispaper,weconcentrateontwotypesofgassamples,namelyhydrate-boundgasandvoidgas.Gashydrate

samples

1010A.V.Milkovetal.

Table1.LocationandmaincharacteristicsofSites1244to1250drilledonLeg204(compiledfromdatapresentedbyTréhuetal.,2003).

Site

Characteristic

NumberofholesLatitude(N)Longitude(W)Waterdepth(mbsf)

Bottomwatertemperature(°C)Geothermalgradient(°C/km)Maximumpenetrationdepth(mbsf)

DepthofBSR(mbsf)

DepthofHorizonA(mbsf)

PorewatersalinitynearthebaseofGHSZ(g/kg)

CalculatedbaseofGHSZa(mbsf)

12446(A–F)?44°35.17=?125°7.18=?895

4.0560380127–129Notpresent

33137

12455(A–E)?44°35.17=?125°8.94=?870

3.9855540.3

13418035149

12462(A,B)?44°35.16=?125°8.13=?850

4.5461180114Notpresent

36115

12472(A,B)?44°34.66=?125°9.01=?835

3.7552270129–134150–16535156

12483(A–C)?44°34.45=?125°9.15=?830

4.1457194115127–12934131

124912(A–L)?44°34.22=?125°8.84=?775

4.54990115(notpenetrated)

NodataNodata

136

12506(A–F)?44°34.11=?125°9.00=?792

4.2253210114148–15235133

ThethicknessoftheGHSZiscalculatedinthisstudyusingCSMHYDsoftware(Sloan,1998)andassumingthegivenbottomwatertemperaturesandgeothermalgradient,hydrostaticpressuregradient,puremethanegas,andsalinityasmeasuredneartheBSR.

incoreswerelocatedusinganinfraredcamerathathelped?ndcoldspotswheregashydratewasdecomposing(Tréhuetal.,2003,2004).Gashydratewasrecoveredfromthesediments,cleaned,andfullydecomposedinsidea20-cm3syringe.Al-thoughsigni?canteffortwasmadetocollectasmanygashydratesamplesaspossible,manysedimentintervalswithcoldspotswerepreservedinliquidnitrogenimmediatelyafterre-coveryandnotmadeavailableforanalysis.Atotalof55gashydratesampleswereobtainedandanalyzed(Table3).Gasvoidsforminthecorelinerupontransfertosurfaceconditions(PaullandUssler,2001)whentheconcentrationofdissolvedC1exceeds10to20mmol/L(Tréhuetal.,2003).Thevoidsweresampledbypuncturingtheplasticcorelinerwithasteelpenetrationtoolconnectedthroughavalvetoasyringe.Atotalof494gasvoidsweresampledandanalyzedonLeg204(Tréhuetal.,2003;seealsoTableEA-1intheElectronicAnnex).AveragefrequencyofgasvoidsamplingatLeg204wasonevoidperevery?6mofrecoveredsediments,withmuchgreaterfrequencywithintheGHSZ(Milkovetal.,2004c).Analiquotofsampledhydrate-boundgasesandvoidgaseswasanalyzedonboardoftheJOIDESResolutionwithin?30minofsampling.AtSites1244and1247,concentrationsofC1,C2andC3wereanalyzedusinganHP6890gaschromatographequippedwitha?ameionizationdetector(FID)setat250°C(PimmelandClaypool,2001).AtSites1245,1246,and1248to1250,methanethroughpentanes(C1-C5)hydrocarbongasesandCO2,O2,N2,andH2SgaseswereanalyzedonboardusinganHP5890AgaschromatographequippedwithbothFIDandathermalconductivitydetector(TCD).Calibrationofinstru-mentdifferenceslimitedtheprecisionwithwhichsmallcon-centrations(severalppm)ofC3-C5gasescouldbereliablymeasuredbyTCD.Aftershipboardchemicalanalysis,theremaininggaswaspreservedinvacutainersforpostcruiseresearch.Ingeneral,nosigni?cantamountofsedimentorporewaterwasallowedtoenterthevacutainers,anditisnotlikelythatthemolecularandisotopicpropertiesofgasesweremod-i?edbymicrobialactivityduringthestorage.

Althoughallsampleswerecontaminatedbyair,concentra-tionscalculatedaspercentageofthesumofnaturalgases(hydrocarbongases,CO2andH2S)arediscussedbelow.The

measuredconcentrationsofgases(aspartspermillionbyvolumeoftotalgas)aretabulatedinTréhuetal.(2003).Voidgasesusuallyhaverelativelylowaircontamination(?10%)whilesomehydrate-boundgasesarecontaminatedbysigni?-cantamountofair(usually?50%butupto98%insomesamples).Theaircontaminationlikelydidnotaffectthemo-lecularandisotopicpropertiesofhydrocarbongases.However,theCO2ingashydratesandinvoidgaseswaslikelydilutedbyCO2fromtheair,whichcomplicatedtheinterpretationofthepropertiesandoriginofthisgas.