numseasonalwinds
seasonalwinds 时间:2021-01-14 阅读:()
Keywordspaleoseismology–AltynTaghFault–strike-slipfaults–India-Eurasiacollision1.
IntroductionThesinistralAltynTaghFault(ATF)travers-esthenorthernboundaryofTibetfor>1500kmandisamajorstructureintheIndia-Eurasiacon-PaleoseismologyoftheXorxolSegmentoftheCentralAltynTaghFault,Xinjiang,ChinaZackWashburn(1),JRamónArrowsmith(2),GuillaumeDupont-Nivet(3),WangXiaoFeng(4),ZhangYuQiao(4)andChenZhengle(4)(1)GeomatrixConsultantsInc.
,CostaMesa,CA,U.
S.
A.
(2)DepartmentofGeologicalSciences,ArizonaStateUniversity,Tempe,AZ,U.
S.
A.
(3)PaleomagneticLaboratoryFortHoofdijk,FacultyofEarthSciences,UtrechtUniversity,Netherlands(4)InstituteofGeomechanics,ChineseAcademyofGeologicalSciences,Beijing,ChinaAbstractAlthoughtheAltynTaghFault(ATF)isthoughttoplayakeyroleinaccommodatingIndia-Eurasianconver-gence,littleisknownaboutitsearthquakehistory.
Studiesofthisstrike-slipfaultareimportantforinterpreta-tionoftheroleoffaultingversusdistributeddeformationintheaccommodationoftheIndia-Eurasiacollision.
Inaddition,the>1200kmlongfaultrepresentsoneofthemostimportantandexemplaryintracontinentalstrike-slipfaultsintheworld.
WemappedfaulttracegeometryandinterpretedpaleoseismictrenchexposurestocharacterizetheseismogenicbehavioroftheATF.
Weidentified2geometricsegmentboundariesina270kmlongreachofthecentralATF.
TheseboundariesdefinethewesternmostWuzhunxiao,theCentralPingding,andtheeasternmostXorxol(alsowrittenasSuekuliorSuo'erkuli)segments.
Inthispaper,wepresenttheresultsfromtheCamelpaleoseismicsitealongtheXorxolSegmentat91.
759°E,38.
919°N.
Thereevidenceforthelasttwoearthquakesisclearand14Cdatesfromlayersexposedintheexcavationbrackettheirages.
Themostrecentearthquakeoccurredbetween1456and1775calA.
D.
andthepenultimateeventwasbetween60and980calA.
D.
CombiningtheCamelinterpretationswithourpublishedresultsforthecentralATF,weconcludethatmul-tipleearthquakeswithshorterrupturelengths(50km)ratherthancompleteruptureoftheXorxolSegmentbet-terexplainthepaleoseismicdata.
Wefound2-3earthquakesinthelast2-3kyr.
Whencoupledwithtypicalamountsofslipperevent(5-10m),therecurrencetimesaretentativelyconsistentwith1-2cm/yrsliprates.
Thisresultfavorsmodelsthatconsiderthebroaderdistributionofcollisionaldeformation,ratherthanthosewithnorthwardmotionofIndiaintoAsiaabsorbedalongafewfaultsboundingrigidblocks.
tinentalcollision(e.
g.
,TapponnierandMolnar,1977;Peltzeretal.
,1989;YinandHarrison,2000;Tapponnieretal.
,2001;fig.
1).
Twocom-petingend-membermodelshavedevelopedthatdescribehownorthwardconvergenceofIndiaisaccommodatedbyTibetandSouthernAsia.
Thefirstmodeldividesthisregionintoafewrigidblocksboundedbymajorfaults(i.
e.
AltynTaghandKunlunfaults,fig.
1)(e.
g.
,AvouacandTap-ponnier,1993;Tapponnier,etal.
,2001).
Thesecondmodelhasconvergenceaccommodat-edbycrustalthickeningandalongnumerousminorandafewmajorfaults(e.
g.
,EnglandandMolnar,1997).
Modelsofthefirsttypearecon-1015ANNALSOFGEOPHYSICS,VOL.
46,N.
5,October2003Mailingaddress:Dr.
JRamónArrowsmith,DepartmentofGeologicalSciences,ArizonaStateUniversity,Tempe,AZ85287-1404U.
S.
A.
;e-mail:ramon.
arrowsmith@asu.
edu1016ZackWashburn,JRamónArrowsmith,GuillaumeDupont-Nivet,WangXiaoFeng,ZhangYuQiaoandChenZhengleChinaTibetAltynTaghFault80°100°120°40°20°KunlunfaultRedRiverFaultHimalayanFrontalThrustsIndiaFig.
1.
OverviewmapshowingmajorstructuresandtopographyintheIndia-Eurasiacontinentalcollisionzone.
Insetshowslocationoffig.
2.
FiguregeneratedfromGTOPO301kmdigitalelevationdatasetusingGMT(WesselandSmith,1995).
sistentwithhighersliprates(3cm/yrfortheATF)whilelowerslipratesareconsistentwiththesecondtypebecausedeformationisdistrib-utedinsteadoflocalizedalongasmallnumberofdiscretefaultplanes.
Globalseismicitydatabases(CNSS)andtheChineseCatalogofHistoricStrongEarthquakes(2300B.
C.
to1911A.
D.
)recordnomajorearthquakesalongtheATF.
However,weknowthatlargeearthquakeshaveoccurredbecauseofmoletracksseenonsatelliteimagery,surfacebreaksshownbyGeetal.
(1992),andlimitedfieldinvestigationbywesternscientists(Molnaretal.
,1987).
Despitetheclearevidenceofactivefaulting,slipratedeterminationsarediscrepant.
Overtimescalesof10s-100sofkyr,slipratesareinferredtobe2-3cm/yr.
Peltzeretal.
(1989)inferred3cm/yrratesintheareaof84°E-92°EbasedonanassumptionofinitialHoloceneageforoffsetlandforms.
RecentworkhasfocusedondeterminingtheslipratebydatingoffsetQuaternarylandformsusingcosmogenicmeth-ods.
Meriauxetal.
(1999,2000)determined3cm/yrbetween85°Eand89°Eand2-3cm/yrnear90°E.
Merieauxetal.
(2003)determinedaminimumof20±3mm/yrat94°E.
Chineseearthquakegeologystudies(mapping,paleoseis-mictrenching,and14C)inferredaminimumrateof5mm/yralongtheentireATF(Geetal.
,1992).
StudiesofdeformationintheIndia-Eur-asiancollisionovergeodetictimescales(5-10years;Bendicketal.
,2000;Shenetal.
,2001;Wangetal.
,2001)indicatethatdistributeddeformationinTibetisanimportantprocessbecausefaultssuchastheATFhaveslipratesontheorderof10mm/yr(9±5mm/yr,Ben-dicketal.
,2000;9±2mm/yr,Shenetal.
,2001).
Holtetal.
(2000)alsousedarateof10±8mm/yralongtheATFintheiranalysisofthevelocityfieldinAsia.
Twomajorstrike-slipearthquakeshaverecentlyoccurredinNorthernTibet.
The1997Mw7.
6Manyiearthquakerupturedforadis-tanceofabout170kmwithpeakleft-lateralslipofabout7m,apparentlyreactivatingaQuaternaryfault(fig.
2;Peltzeretal.
,1999).
The2001Mw7.
8KokoxiliorCentralKunlunearthquakehada300-400kmlongrupturezoneandsinistralslipupto16.
3m(Linetal.
,2002;Klingeretal.
,2003).
ItrupturedtheeasternportionoftheKunlunFault(vanderWoerdetal.
,2002).
Theseearthquakesdrawattentiontothestyleofearthquakerupturealongthegreatstrike-slipfaultsofAsiaandalsototheirresponsetothegeodynamicsetting.
Becausethehistoricrecordinthisregionisonlycom-pleteforaboutthelast100years,paleoseismo-logicalmethodsareimportantforinvestigatingthetimingoflargerupturesliketheseandthosealongneighboringfaults.
Earthquakedisplacements,surfacerupturelengths,andrecurrenceintervalsalongmajorstrike-slipfaultssuchastheATFhaveimplica-tionsforwhichgeodynamicmodelsbetterexplaincontinentaldeformationandforregion-alearthquakesequencesincontinentalsettings.
Therefore,weconductedastudyoftheearth-quakegeologyoftheATFtocharacterizetheseismogenicbehaviorofthismajorstrike-slipfault.
Inthispaper,wefirsthighlightthegeo-PaleoseismologyoftheXorxolSegmentoftheCentralAltynTaghFault,Xinjiang,China808590953540KunlunFaultTibet678White2500yearsBP1997Mw7.
6ManyiearthquakeTarimBasinQaidamBasinQimenTaghMilanRuoqiangNATFHuatagoLapequanAltynTaghFault1924M>7earthquakes1951~M6earthquakeAltynTaghAltynTaghFaultAkatoTaghMinfengSubeiSulamuTaghTula2001Mw7.
8KokoxiliearthquakeFig.
2.
OverviewmapshowingtheCentralAltynTaghFault(ATF),locationofstudyarea(rectangularbox,fig.
3),instrumentalseismicity(CounciloftheNationalSeismicSystem(CNSS);http://quake.
geo.
berkeley.
edu/cnss/cata-log-search.
html),historicearthquakes(ChineseCatalogofHistoricStrongEarthquakes2300B.
C.
toA.
D.
,1991),paleoearthquakes(StateSeismologicalBureau(SSB);Geetal.
,1992),andmajortowns.
TheGeetal.
(1992)dataareplottedatthereportedlocationswiththeirinferredmagnitudeandcolor-codedbyagerange.
TheeventsmostrelevantforthisstudyareM6.
TheactivetracetraversesalluvialflatsaroundLakeWuzhunxiaoandtheBitterSeaandbedrockintheArkatengandPingdingmountainranges.
Twogeometricboundariesareidentified:the7kmwideArkatengRestrainingBendandthe4kmwidePingdingreleasingstep.
ThesegeometricboundariesdefinethewesternmostWuzhunxiao,theCentralPingding,andtheeasternmostXorxolsegments.
Theboundariesserveasaframeworkforinterpretingoffsetdataandesti-matingruptureextentbecausesuchdiscontinuitiescommonlyactasrupturebarriers.
1018over110km(figs.
2and3).
Proterozoicintru-siverocksandsliversofPaleozoic-Mesozoicgranitoid,ultramafic,marinecarbonate,andophioliticrockscomprisethemountainsinthestudyarea(Cowgilletal.
,2003).
ThickNeoge-nebouldergravelstosiltsencirclethePro-terozoic-Mesozoicmountainsandfilladjacentbasins.
IntensefoldingofthesestrataintheQaidamBasinandthroughoutthefieldareaindicateacontractionalregime(regionally)fol-lowingdepositionoftheNeogeneunits.
RelativelythinQuaternarygravelsandsilts(pied-montdepositsandprimaryandreworkedloess)overlietheNeogenestrata.
3.
FaulttracemappingandpaleoseismicsiteidentificationInthefield,wemappedthefaulttracesthatcomprisetheindividualfaultsthathavehadrecentrupturebasedongeomorphicageofthescarps,disruptionsofHolocenedepositsandlandforms.
Theactivetracedesignationgener-allyreferstosurfacebreakscreatedbythelast1-3earthquakes.
WehaveidentifiedtheactivetracesandmostQuaternaryfaults,butmaynotshowalloftheearlyQuaternaryfaultsbecauseweareprimarilyconcernedwiththegeometryandearthquakehistoryforthelastfewkyr.
Infig.
3,faultslabeledQmaresimilarinagetoorslightlyyoungerthansurfacesanddepositsofLatePleistoceneandEarlyHoloceneinferredage.
TheseweremappedasunitQmandarethemostextensiveoftheQuaternarydepositsandformbroadfloorsthatdepositionallyabutadja-centhillsidesandoftendefineasharpbreakinslope.
Theygenerallyhaveafewto10sofmincisionandmaybecoveredwithvarnishedclasts.
TectoniclineamentsrefertosuspiciousgeomorphicalignmentsthataremostlikelyfaultsolderthanQm.
OurmappingapproachwastoinspecttheprincipalactivetracesasidentifiedonCORONAimagery.
Wemappedthemainfaulttraces,theQuaternaryunitsdiscussedabove,andocca-sionallyobtainedbeddingorotherstructuralattitudes.
Wetypicallycovereda5kmwidezonecenteredoverandparalleltotheactivetracetoidentifyrelevantstructuresandgeolog-icrelations.
Wemademanysmallexcavationswithshovelstoclarifystratigraphicrelation-shipsandexposebeddingsurfaces.
Thesesmall-erexcavationshelpedaddconfidencetoourgeo-morphicandstratigraphic-basedpaleoseismicsitedecisions.
Weloggednaturalandhand-dugexposures,anddocumentedoffset(bothverticalandhorizontal)landformsandstrata.
Whilemappingthefaulttrace,welookedforsiteswherelandformsindicatedthatsteadyaccumulationofdistinctivesedimentswouldpreserveevidenceforpastruptures.
Inanum-berofexcavationsatsuchsites,brokenandtilt-edbedsoverlainbycontinuousbedsandupwardterminationsoffaulttracesidentifypastearth-quakesinexcavations.
Disruptionseenatthesamestratigraphiclevelatseverallocationsandonbothtrenchwallsqualifiesasadistinctevent(afterGrantandSieh,1994).
Organicmatterwithinthestratigraphywasdatedusing14Candfinesiltwasdatedbyinfraredstimulatedlumi-nescence(IRSL)(Forman,1999).
Weusethe2mcalibratedsampleagesandOxCalv3.
5(Ram-sey,2000)fortimingconstraints.
WealsouseaMonteCarloprogramthatusesorderingcon-straintstotrimprobabilitydistributionsofra-diocarbonsamplesandthenyieldsprobabilitiesforevents(HilleyandYoung,2003).
Thisapproachallowsformoreobjectiveeventcorre-lationbetweenpaleoseismicsites.
4.
FaulttracegeometryAsmall-scalerepresentationoftheactivetracesthatwemappedisshowninfig.
3.
Forconvenience,werefertolocationsalongthemappedtracesintermsofdistancefromthewesternendofthestudyarea.
Washburnetal.
(2001)describedthewestern150kmoftheareashowninfig.
3andestablishedthegeo-metricsegmentation.
Oursubsequentmap-pingaddedanother120kmtowardtheeast(Washburn,2001).
Aviewalongthestrikeoftheactivetraceinfig.
3showsaprominentrightbend(65to85km)andleftstep(97km)inthegenerallystraightCentralAltynTaghFault.
WefollowedtheconventionsofdePoloetal.
(1991)andKnuepfer(1989)andidenti-fiedthesediscontinuitiesasgeometricbound-1019PaleoseismologyoftheXorxolSegmentoftheCentralAltynTaghFault,Xinjiang,ChinaarieswhichinturndefinethewesternmostWuz-hunxiaoSegment,theCentralPingdingSeg-ment,andtheEasternXorxolSegment.
Thesesegmentsarenotnecessarilyrupturesegments;however,thisapproachisusefulforcharacterizingfaultzonesbecausegeometricboundariesmaybelocationsofsignificantchangeinslipormomentrelease(e.
g.
,Wesnousky,1989;HarrisandDay,1993,1999).
Historicalrupturedatashowthatrup-tureendpointscommonlycoincidewithgeometricboundariesandtheseboundariesappeartoretardrupture(Knuepfer,1989).
ThemajorgeometricboundaryinthestudyareaisthePingdingReleasingStep.
Here,theactivetracejumpsfromthesouthernsideofPingdingShan4kmnorthtotheactivetracethatcontinuestotheXorxolValley(fig.
3;Wash-burn,2001;Washburnetal.
,2001).
The3-4mverticaloffsetofHolocene-agefanonthesout-heastsideofPingdingShan,aprominentscarpontheeastfaceofPingdingShan,andthepres-enceofrecentlandslidesindicatethatthenor-malfaultthatlinkstheboundingstrike-slipfaultsrupturedrecently.
Ourmappingstopsat268kminastraightandsimpleportionoftheXorxolSegment(fig.
3).
This070°strikingsimplegeometrycontinueseastward,butjustofftheeasternedgeoffig.
3,around300km,aprominentnorthernstrandde-velops(fig.
2).
WeinferthattheXorxolgeomet-ricsegmentisatleast200kmlongandremark-ablystraight(Washburn,2001).
5.
XorxolSegmentpaleoseismicinvestigationOverviewSeismogenicpropertiessuchasfaultstraight-nessandmaximumdisplacementduringthelastearthquakesatspecificsitesalongthecentralATFwerepresentedaboveandinWashburnetal.
(2001)andWashburn(2001).
Thenextstepsincharacterizingtheseismogenicbehav-ioroftheCentralAltynTaghFaultaredetermi-nationsofrecentearthquaketimingandesti-matesofmagnitude.
Inordertoestablishthetimingofpastearthquakes,wemade10exca-ZackWashburn,JRamónArrowsmith,GuillaumeDupont-Nivet,WangXiaoFeng,ZhangYuQiaoandChenZhengleTableI.
14CdatafortheCamelpaleoseismicsite.
AllanalysesarestandardAMSdatesfromCAMSfacilityatLLNL.
b.
13CvaluesaretheassumedvaluesaccordingtoStuiverandPolach(1977).
Valuesmeasuredforthematerialitselfaregivenwithasingledecimalplace.
Thequotedageisin14CyearsusingtheLibbyhalflifeof5568yearsandfollowingtheconventionsofStuiverandPolach(1977).
14Cconcentrationisgivenasfractionmodern,D14C,andconventional14Cage.
FieldsampleCAMS#b.
13CFraction±D14C±14.
Cage±2mcalibratednamemodernage(*)(calA.
D.
)C169687–250.
89230.
0035–107.
73.
5920401024-1209C769688–250.
95920.
005–40.
85330501456-1650C869689–250.
98550.
0048–14.
54.
8120401676-1939C10A69690–250.
79210.
0034–207.
93.
418704060-240C1669691–250.
89120.
0043–108.
84.
3930401022-1191C2069692–250.
86390.
0047–136.
14.
7118050717-980C2269693–250.
98030.
0039–19.
73.
9160401662-1950C2469694–250.
73760.
0042–262.
44.
2244050762-403B.
C.
C2569695–250.
98740.
0039–12.
63.
9100401678-1937C2669696–250.
89290.
0045–107.
14.
5910501022-1221C3269697–250.
62120.
003–378.
833820402404-2140B.
C.
C4569698–250.
57450.
0027–425.
52.
74450403337-2924B.
C.
SLrabbit69699–251.
12450.
0054124.
55.
4>Modern>1950SLcamel69700–251.
20740.
0058207.
45.
8>Modern>1950SLstick69701–251.
18980.
0052189.
85.
2>Modern>1950(*)DeterminedwithCalib4.
2(Stuiveretal.
,1998).
Doesnotincludecurveinterceptscomprising30mlengthofthetrench.
Therefore,thelim-iteddeformation,seenasonly3faults,indi-catethatthisintermediateeventwasdifferentandhadlowermagnitudethantheMREandpenultimateevents.
Wefavoraninterpretationofupwarddieoutand/orplungingtiplineofMREfracturesbecausethefaultingrelation-shipsarenotthesameonbothtrenchwallsandthemuchsmallerdegreeofdeformationcom-paredtorupturesseeninothertrenchesinthestudyarea.
Moresubstantialevidenceforapre-penulti-mateeventisseen1mnorthofandinthebot-tomoffaultzone5.
Here,faultedandtiltedstrataareoverlainbyflatlyingbeddedsandandfinegravelofunit85.
Theserelationshipsaresuggestiveofanearthquakebeforethepenulti-mateevent,howeverthetrenchwasnotdeepenoughtoconfirmthisevent.
Faultzone6preservesevidenceofthefail-ureoflargeblocksofthelocalNeogenebase-mentinduratedsiltsintomostlyeoliansands.
Theblocksarestronglyfracturedandincombi-nationwiththewindblownsandsrepresentaslightlydifferentenvironmentthantheeoliansandandalluvialfanmuddysandsandgravelstothenorth,furthermakingcorrelationofe-ventswiththerestofthetrenchdifficult.
5.
5.
Camel14CdatingWecollected57samplesoforganicmate-rialfromthetrenchanddated12,aswellas3fromamodernstrandline.
BecausemuchoftheorganicmaterialintheCameltrenchwasrepresentedinmodernstrandlines,andoftendepositedasstrandlines,wedatedthemodernonestoassesstheirabilitytodefinethedepo-sitionalage.
The3strandlinesamplesareallmodern.
Thus,the14Cdatesinthetrenchprob-ablydonotincludeasignificantpre-deposi-tionalage.
The14CdataarepresentedintableIandtheprobabilitydistributionsfromcalibra-tionsusingOxCal(Ramsey,2000)areshowninfig.
7a,b.
An8by5by3mmpieceofwoodwithintactbark(C8)fromunit150providesanupperboundof1676-1939calA.
D.
,whilearabbitdropping(C7)yieldsalowerboundof1456-1650,forthelastearthquakeinthisarea.
Likewise,anelongaterabbitdropping(C20)placesanupperboundsof766-980calA.
D.
and10smallpiecesofwoodydebrisweremixed(C10)toprovidealowerboundsof60-240calA.
D.
forthepenultimateevent.
Theoldeststrataexposedinthetrencharebe-tween3337-2924calB.
C.
basedonaradio-carbondatefrompieceofwood(C45).
Becauseofthecompletestratigraphica-greementwithinall12samples,weusedOxCal1026ZackWashburn,JRamónArrowsmith,GuillaumeDupont-Nivet,WangXiaoFeng,ZhangYuQiaoandChenZhengle1027PaleoseismologyoftheXorxolSegmentoftheCentralAltynTaghFault,Xinjiang,Chinatorefinetheconstraintsoneventtiming(Ram-sey,2000).
Thisprogramusesstratigraph-icorderingconstraintstoeliminateoverlap-pingageprobabilities,whichrefinesourearthquaketimingbrackets.
Weused1950A.
D.
asanuppereventboundarybecausetheworldwideseismicnetworkwasabletodetectmoderatesizedearthquakesbythistime.
OxCalismostusefulforsamplesthatarecloselyspacedstratigraphicallyandyieldsimilarradiocarbonages.
C22islocatednearthetopofunit160whileC25inthelowerpartofthisunitandC8islocatedinunit150.
Thesesamplesallyieldsimilarcalendarages(tableI),howevertheirstratigraphicpositionallowsOxCaltocutoffoverlappingportionsoftheirprobabil-itycalendaragedistributions.
Thisisusefulbe-causethispartofthecalibrationcurveresultsinawideagerange(e.
g.
,1676-1939forC8).
AfterrunningOxCal,C8yieldsanarrowup-perboundof1676-1775fortheCamelMRE.
Thelowerboundof1456-1650cal.
A.
D.
remainsunchangedbecausethenextlowestsample(C261022-1221calA.
D.
)ismucholderandthereforeprobabilitydistributionsdonotoverlap.
Thelowerboundonthepe-nultimateearthquakeisalsounchangedforthesamereason.
Abovethepenultimateeventhorizon,unit125contains2radiocarbonsam-ples,howevertheyarespacedoveran8mhorizontaldistanceandsitattheapproximatesamestratigraphiclevel.
Thereforewecannotapplyorderingconstraintstothesesamples,althoughC26inunit127doesshifttheiragedistributions30yearsolder.
Figure8a,bshowstheearthquaketimingconstraintsatanumberofsitesalongthecen-tralATF.
Theagesdisplayedinfig.
8aarecal-ibratedbutotherwisenotrefinedwithstrati-050100150200250distance(km)CalendarYears-2500-1500-5005001500Kulesayi2-6-172-6-18CamelExplanation14MRExxxmostrecenteventyoungerthanMREolderthanMRE/youngerthanpen.
olderthanpenultimateeventearthquakefromActiveAltunFaultmonograph(Geetal.
)2sigmacalibratedageofC(solidline)andIRSL(dottedline)samplesolderthanearliesteventnosym.
-2500-1500-5005001500CamelProbability0.
010.
0CalendarYearsExplanationMREPenultimateeventFig.
8a,b.
a)Earthquaketimingconstraintsfromstratigraphicrelationshipsand14CandInfraredStim-ulatedLuminescence(IRSL)samples(fromthisstudyandWashburnetal.
,2001).
NotetheactivetracemapaboveplotAforreference.
ThesearethecalibratedageswithoutOxCalorotherrefinementsfromstratigraphicordering.
b)MonteCarloproba-bilityplotfortheCamelpaleoseismicsiteearthquaketimingdata(HilleyandYoung,2003).
Theareaunderthecurvesrepresentstheprobabilitydistributionoftheageofpastearthquakes.
Sharperpeaksrepresentwell-constrainedeventsandshowtherefinementoftheCameleventages.
abZackWashburn,JRamónArrowsmith,GuillaumeDupont-Nivet,WangXiaoFeng,ZhangYuQiaoandChenZhengle1028Fig.
9.
TrenchlogsfortheCamelpaleoseismicsiteat198km.
Mostfaultsterminateattheupperblueline,theeventhorizonforthelastearthquakeatthissite.
Thepenultimateearthquakeisidentifiedbyupwardfaultter-minationsandfoldedbeddingoverlainbycontinuousstrata,whichismarkedbythelowerblueline.
Thispenul-timateeventhorizoncoincideswithasmallunconformity(yellowline)identifiedbydifferentlevelsofindura-tionandanirregularcontactwithcmdeeprillsinplaces.
StratigraphiccolumnofCameltrenchatlowerleftshowsrelativeunitthickness,unitnumber,majorunconformities,andstratigraphiclocationofradiocarbonsam-ples.
Seediscussionintextforfurtherinterpretationandtimingconstraintsonearthquakesseeninthistrench.
N2xxxxxC221662-1950A.
D.
C251678-1937A.
D.
C71456-1650A.
D.
C261022-1221A.
D.
C161022-1191A.
D.
C20717-980A.
D.
C11024-1193A.
D.
C1060-240A.
D.
C24762-403B.
C.
C322404-2140B.
C.
C453337-2924B.
C.
uncfuncfuncf170160150140130125123127115110105100908580C81676-1939A.
D.
StratigraphiccolumnUnit#SamplecalibrateddateyoungerthanMREolderthanMRE/youngerthanpen.
olderthanpenultimateeventxRadiocarbonsamplesuncfunconformity115unitnumberKeyExplanationmatrixsupportedmassivesandgravelclastsupportedgravelbeddedsandeoliansandsiltsandysiltprominentstrandlineeventhorizonunconformity14Csamplefault,fractureNeogenebedrockeastwallC32(2404-2360,2354-2140calB.
C.
)C26(1022-1221calA.
D.
)C25(1803-1907,1678-1744calA.
D.
)C16(1022-1191calA.
D.
)C1(1024-1193calA.
D.
)westwallfaultzone1faultzone2faultzone3faultzone4faultzone5faultzone6C24(762-678,671-608,599-403calB.
C.
)C7(1456-1650calA.
D.
)C8(1802-1939,1676-1765calA.
D.
)C20(766-980calA.
D)C22(1662-1711,1717-1886,1911-1950calA.
D.
)C45(3337-3208,3194-3150,3140-3008,2987-2924calB.
C.
)C10(60-240calA.
D.
)05mprominentstrandline14CsampleExplanationeventhorizonunconformityfault,fracturematrixsupportedclastsupportedbeddedsandmassivesandgravelgraveleoliansandsiltsandysiltNeogenebedrock120130SOUTHNORTHUnitnumbershUnitnumbers17016015514013012311510510017016013011511015014090120130100105170160155140130123115105100Unitnumbers8090100110160170Fig.
6.
(continued)1029PaleoseismologyoftheXorxolSegmentoftheCentralAltynTaghFault,Xinjiang,ChinagraphicorderingconstraintsusingOxCal.
Hil-leyandYoung(2003)havedevelopeda14CagerefinementmethodthatissimilartoOxCalandproducesprobabilitycurvesforeventswithaMonteCarlo-basedanalysisof14Cagecalibrationandstratigraphicordering.
Weshowtheprobabilitycurvesforthelasttwowell-definedCameleventsdeterminedusingtheHilleyandYoung(2003)methodinfig.
8b.
6.
Discussion6.
1.
EarthquakerupturesalongtheCentralAltynTaghFaultCorrelationofearthquakesbetweenpaleo-seismicsitescanleadtomisinterpretationbe-causenotallstrandsinwidefaultzonesmaybeactivatedduringanearthquake.
Wehaveelimi-natedmostofthepotentialforthisscenariobythoroughevaluationofpaleoseismicsitesandexcavationacrossonlythesharpest,mostactivetraces.
However,dogtailruptureterminationsinthemiddleofgeometricsegmentsandmaynotrupturethroughallexcavationsinquestion(e.
g.
,Ward,1997).
Thereforetrenchesshowingoverlappinggeochronologicearthquakecon-straintsthatareonly10sofkilometersapartmaynotrecordthesameearthquake.
Wethere-foreutilizeotherobservationstohelpclarifytherupturehistory.
Weknowthepeaksurfacedisplacementatagivensitefromtheoffsetdata(Washburn,2001;Washburnetal.
,2001)andwehavecon-strainedearthquaketimingatanumberofpaleo-seismicsites.
Withthesedata,wecandeterminecruderecurrenceintervals,becausewehaveonly1-3eventsineachtrench.
Wecanalsomakemagnitudeestimatesusingthedisplace-mentdata.
Betterestimatesofmomentmagni-tude(Mw)canbeachievedbyusingSurfaceRuptureLength(SRL)(WellsandCoppersmith,1994).
Weestimatesurfacerupturelengthbyutilizinggeometricconstraints,offsetdistribu-tion,surfacerupturemorphology,timingcon-straintsfromthetrenchsites.
Figure8a,bshowsthatthelastearthquakeatKulesayi,2-6-17,2-6-18,andtheCamelsitealloccurredattheapproximatesametime,be-tween1215-1775calA.
D.
(datafromWashburnetal.
,2001,andthispaper).
However,itisunlikelythatonelargeearthquakerupturedtheentirefieldareabetween1215-1775calA.
D.
Insuchacase,theMREwouldhavejumpedtwogeometricboundariesandrupturedfromtheKulesayitrench(23km)toatleasttheCameltrench(199km)andlikelyfartherbasedontheoffsetdistributionandsharpsurfacerup-turethatcontinuesto>km240.
ThisscenarioyieldsaSRLof>220kmandMw>7.
8.
However,considerationofdetailsoftheagecontrolfromtheexcavationsandthegeometricboundariesmakesitunlikely,andtherelativelyhighslipim-plied(10m)isnotconsistentwithourinfer-redoffsetdistribution(maximumoffset7m;Washburn,2001;Washburnetal.
,2001).
Itispossiblethatthe4kmPingdingleftstepbarredrupture.
TheArkatengRestrainingBendandassociatedcomplexitiesaroundDayangShanarealsopotentialrupturebarriersfig.
3.
Furthermore,thediscontinuousanddegradedsurfacerupturearoundtheseboundariesarguesforsmallerearthquakesthatdidnotrupturethewholefieldarea.
ThereforeweseparateWuzhunxiaoandPingdingeventsfromXorxoleventsbasedonsurfacerupturemorphologyandgeometricboundaries.
Consequently,thelastearthquakeatKulesayioccurredbetween1215and1750A.
D.
andmayhaveproducedthe3-5moffsetsonthewestsideofLakeWuzhunxiao(fig.
8a,b;Washburnetal.
,2001).
AlongtheXorxolSegment,thetimingcon-straintsimplythatthelastearthquakeruptureditswesternportionbeforetheeasternportionbroke.
Thelastearthquakeinthewestruptured1270and1430calA.
D.
(fromBitterSeapublishedresultsofWashburnetal.
,2001;fig.
8a,b).
ThisrupturemaygofromwesternendoftheXorxolSegmenttoaroundtheBitterSea(50kmSRL)withpeakdisplacementof7m(Washburn,2001;Washburnetal.
,2001).
TheyoungerCamelMREoccurredbetween1456and1775calA.
D.
(fig.
8a,b).
ItrupturedfromsomewhereeastoftheBitterSeathroughtheCameltrenchandlikelyfartherbasedontheroundedandsemicontinuousmoletracksbe-tween247kmtotheedgeofthestudyarea(268km,fig.
3).
ThedegradednatureofthesurfacerupturesuggeststhatthisreachiseithertheeasternterminusoftheXorxolMREorthatero-sionrates,whicharedetectablyhigherinthisarea,ledtothemoredegradedsurfacerupture.
HighererosionratesareplausiblebecauseofthepositionoftheEasternXorxolValleyrelativetothemouthoftheTarimBasinonly7kmnorthoftheplayaat265kmonfig.
3.
ThemouthallowsmorefrequentstormsandintensedailywindstoblowthroughEasternXorxolValleyandmoreanimalaccessandpotentialdegradation.
Insummary,thelastearthquakesalongtheCentralAltynTaghFault:1)occurredbetween1215and1775calA.
D.
;2)hadmaximumdisplacementof5montheWuzhunxiaoandPingdingsegmentsand7montheXorxolSe-gment(Washburn,2001;Washburnetal.
,2001),and3)hadsurfacerupturelengthsbetween50and>150kmwithcorrespondingmagnitudesoflowMw7stohighMw7s,respectively.
6.
2.
Implicationsforslipratebasedonthesizeandfrequencyofearthquakesinthelast3kyrAlthoughtheearthquakedatadonottightlyconstrainthesurfacerupturelengthorMwoftheMRE,therecurrencetimesfromtrenchesshowgoodagreement.
Wefind2earthquakesinthelast0.
8-2.
0kyrattheCamelsite,3eventsinthelast2.
4-3.
0kyrattheBitterSea,and2Kulesayieventsinthelast0.
8-2.
2kyr(thispaper;Washburn,2001;Washburnetal.
,2001).
Althoughwehavealimitednumberofeventsforrobustrecurrencecalculations,takingaveragesfromthesedatagiverepeattimesof0.
7±0.
3kyrfortheCamelsite,0.
9±0.
1kyrfortheBitterSeatrenches,and0.
75±0.
35kyrfortheKule-sayisite.
Thesefrequenciesareconsistentwiththe2earthquakesinthepast3kyrreportedbyGeetal.
(1992)inthestudyareaandaregen-erallysimilartothe800-1000yearrecurrencetimesestimatedfortheKunlunFault(vanderWoerdetal.
,1998).
Giventhelimitednumberofearthquakesinthisstudy,weemphasizethetentativeresultsderivedfromrelatingearthquakerecurrencetosliprates.
However,systematicunderinterpre-tationorrupturelocationvariabilitymustbeinvokedforsignificantlyhigherrecurrenceratesorslip/event.
Ifalargeearthquakeoc-curredinthenearfuture,ourrecurrencerateswouldbemaximaandanyinferredslipratewouldincrease.
Tocomparethesepaleo-earthquakedatawithotherfindingsalongtheATFwecastsliprateintermsofaveragerecurrenceintervalsbyusing5and10mofslippereventbecausethesevaluesbracketthepeakoffsetsrecordedfortheMRE(s)(Washburnetal.
,2001).
Using:1)10mm/yrdeterminedfromrepeatGPSsurveys(Bendicketal.
,2000;Shenetal.
,2001)yieldsrecurrenceintervalsof0.
5and1.
0kyrfor5and10mofslipperevent,respectively;2)3cm/yrdeterminedbylargescalepostglacialoffsetsandoffsetterracesandmorainesinareaswestofthestudyarea(Peltzeretal.
1989;Meriauxetal.
,2000)yieldsrecurrencetimesof0.
2and0.
3kyr.
Figure10showsacompilationofthevariousslipratesthathavebeeninferredforthecentralATF.
Comparisonofthedifferentrecur-renceintervalsshowthatthepaleoearthquakedatamostcloselyagreewitha1-2cm/yrslipratebecausewewouldexpecttosee2-3timesthenumberofearthquakesinourpaleoseismicinvestigationsandthefaultingandsedimentaryrecordinthemisnotconsistentwithsomanymissingearthquakes.
AnimportantissueinslipratecomparisonsistheobservedeastwarddecreaseinslipratealongtheATF(fig.
10).
Peltzeretal.
(1989)in-ferred3cm/yrratesintheareaof84°-92°E.
Meriauxetal.
(1999,2000)determined3cm/yrbetween85°and89°Eand2-3cm/yrnear90°E.
Ryersonetal.
(2003)reportawellconstrain-ed18mm/yratthePingdingShan(90.
5°E).
Meriauxetal.
(2003)determinedaminimumof20±3mm/yrat94°E.
Ryersonetal.
(2003)alsoreport12.
5-15mm/yrat95°E.
Meyeretal.
(1996)concludedthatsliprateseastof96°were4±2mm/yr,consistentwithdecreasingslipratealongtheATFandaccommodationoftheslipgradientbythrustingalongthesouthsideintheQaidamBasin(alsosuggestedbyPeltzeretal.
,1989;Chenetal.
,1999).
Meyeretal.
(1998)observedthatthethrustsbranchfromtheATFandarethuskinematicallylinked;andvanderWoerdetal.
(2001)showedthedecreaseinsliprateandgeometryofthefaultjunctionattheAltynTaghFault-TanghenanShanThrusttriplejunction.
1030ZackWashburn,JRamónArrowsmith,GuillaumeDupont-Nivet,WangXiaoFeng,ZhangYuQiaoandChenZhengleThisstudycoverstheportionoftheATFfrom89.
6°Eto92.
5°E.
ItiswithintheareacoveredbythehigherratesinPeltzeretal.
(1989)andspanstheMeriauxetal.
(1999)andRyersonetal.
(2003)sitenearthePingdingShan(18mm/yr).
Ourbestearthquakerecur-rencedatacomefromtheCamelsiteisatabout91.
75°E,sowemaybegintoobserveadecreaseinslipandthusrecurrencerateduetointerac-tionbetweentheATFandtheadjacentactivethrustsoftheQaidamBasin(forexample,notethattheactivenorthweststrikingYousha-ShanthrustfaultofChenetal.
(1999)approachestheATFneartheGobiling).
Althoughthere-sultsofRyersonetal.
(2003)andourinferredslipratefromnearthePingdingShanoverlap(fig.
10),the20±3mm/yrminimumrateat94°E(Meriauxetal.
,2003)impliesahigherearthquakerecurrenceratethanweinferredfortheXorxolSegment.
Thegeodeticinvestigationsarenotsuffi-cientlydensetoimagevariationsinslipratealongthecentralATF.
Giventheirgreatbreadth(100sofkmperpendiculartotheATF),nearfaultlockingandseismiccyclevariationinvelocitycannotexplaintheslipratediscrepancy(fig.
10;Thatcher,1990;Shenetal.
,2001).
Shenetal.
(2001)alsopointoutthatbecausethereisnomajorconvergenceparalleltotheATFonitssouthsideacrosstheQimenTagh1031PaleoseismologyoftheXorxolSegmentoftheCentralAltynTaghFault,Xinjiang,China848688909294969800.
511.
522.
533.
54Longitude(°E)ATFparallelsliprate(cm)Geetal.
(1992)Shenetal.
(2001)Bendicketal.
(2000)Meriauxetal.
(1999,2000)Peltzeretal.
(1989)Meriauxetal.
(2003)Ryersonetal.
(2003)ThisstudyRyersonetal.
(2003)Meyeretal.
(1996)Fig.
10.
InferredAltynTaghFault-parallelslipratesfromlong-termstudies(blue),geodesy(red),andearth-quakegeology(black)compiledfromtheliteratureandthisstudy.
Theeastwarddecreaseinsliprateindicatedbythelong-termstudiesisconsistentwiththeactivityofnumerousthrustfaultssplayingfromandtothesouthoftheATF(e.
g.
,Meyeretal.
,1996,1998;Tapponnieretal.
,2001;althoughnotconsistentwiththegeodeticresultsfromShenetal.
,2001).
Ingeneral,thegeodeticresultsshowlowerslipratesrelativetothelong-termstudies.
Ourearthquakegeologyresultsspanthetwoothermethodsinboththetimescaleofmeasurements(afewkyrversusdecadaland>10kyr)andtheresultinginferredrates.
MountainsatthesouthernmarginoftheQaidamBasin,theslipratealongtheATFmustnotdecreasesignificantlybetween80°and92°E.
Oneexplanationforthediscrepancyinrecur-rencetimesfromGPSsitevelocitiesandtheearthquakerecordandgeologyisthatamajorearthquakealongthecentralATFisoverdue.
AnotherintriguingexplanationisthatthelongtermloadingratealongtheATFhasdecreasedinthepast100kyrandstrainisnowaccommodat-edbyotherstructuresinCentralAsia.
Forexam-ple,thefaultsthatproducedfourM8earth-quakesinMongoliainthepastcenturyreleasedsignificantstraininanareaofrelativelylowerinferredlongtermdeformationrate(Baljinnyametal.
,1993).
Earthquakes,suchasthe1997Mw7.
6Manyiearthquakethatruptured170kmalongtheKunlunFault(Peltzeretal.
,1999)andthe2001Mw7.
8Kokoxiliearthquake(vanderWoerdetal.
,2002;Linetal.
,2002)mayindi-catethatthewesternportionoftheKunlunFaultandothercentralandnorthTibetanstrike-slipfaultsmaybemoreactivethanpreviouslythought.
Finally,shorteningwithintheQimenTaghandothercontractionalstructuresinNorthernTibetmaynowaccommodateahigherpercentageofIndia-Eurasiaconvergence(YinandHarrison,2000;Cowgill,personalcommu-nication).
7.
ConclusionsInthispaper,wehavecharacterizedthefaulttracegeometryandpaleoseismologyoftheCentralAltynTaghFault,emphasizingits230kmlongandremarkablystraightXorxolgeometricsegment.
Tracegeometry,timingofpaleoearthquakes,andinferencesofrupturemagnitudescontributetotheglobalcatalogueofstrike-slipfaultsystemsandprovideinforma-tionaboutitsgeodynamicroleintheIndia-Eurasiancollision.
ThetracegeometryofthecentralATFisdominatedbytheArkatengRestrainingBend,PingdingReleasingStep,andthenotablystraightXorxolSegment.
Thegeometricbound-ariesareprobableearthquakerupturebound-ariesbasedonsurfacerupturecontinuity,distri-butionofsmallgeomorphicoffsets,andhistor-icaldata.
MajorHolocenedeformationalongtheAltynTaghFaultSystemisconfinedtoanarrowzonelessthan3kmwide.
Thegeometricboundariesandtimingcon-straintsprobablyprecludeonelargeearthquakerupturingall3segments(270km).
Further-more,multipleearthquakeswithshorterrupturelengths(50km)ratherthancompleteruptureoftheXorxolSegmentbetterexplainthepaleo-seismicdata.
Wefound2-3earthquakesinthelast2-3kyr.
Whencoupledwithtypicala-mountsofslipperevent(5-10m),therecur-rencetimesaretentativelyconsistentwith1-2cm/yrsliprates.
Thisresultfavorsmodelsthatconsiderthebroaderdistributionofcollisionaldeformation,ratherthanthosewithnorthwardmotionofIndiaintoAsiaabsorbedalongafewfaultsboundingrigidblocks.
AcknowledgementsWethankP.
Molnar,R.
Bilham,P.
Tap-ponnier,M.
Hamburger,R.
Bürgmann,A.
Yin,S.
Graham,andJ.
Freymuellerforearlyreviewsofthisresearch,ourcolleaguesattheInstituteofGeomechanicsinBeijing,andS.
Holloway,M.
Baillie,S.
Selkirk,E.
Catlos,C.
Schrein,andS.
Zhangforhelpinthelab,field,andtheoffice.
ConstructivereviewsbyK.
OkamuraandF.
GaladiniandtheguidanceofDanielaPantostiforAnnalsofGeophysicswerealsoveryhelpful.
R.
Ryersonarrangedour14Cdat-ingatLawrenceLivermoreNationalLabora-tory/UniversityofCaliforniaCenterforAcce-leratorMassSpectrometrywherewewereguidedbyG.
Sietz,M.
Caffee,J.
Southon,andtheircolleagues.
G.
Hilleyprovidedimpor-tantadviceoninterpretationof14Cdata.
TheNorthernCaliforniaEarthquakeDataCenter(NCEDC),andthemembernetworksoftheCounciloftheNationalSeismicSystem(CNSS)providedtheseismicitydatashowninfig.
2.
WeareparticularlygratefultoAnYinforinvitingustojointheprojectandtoEricCowgillforintroducingustofieldworkinXinjiang.
ThisworkispartofacollaborativeresearcheffortsupportedbytheU.
S.
NationalScienceFoundationContinentalDynamicsProgram(grantEAR-9725780).
1032ZackWashburn,JRamónArrowsmith,GuillaumeDupont-Nivet,WangXiaoFeng,ZhangYuQiaoandChenZhengleREFERENCESAVOUAC,P.
J.
andP.
TAPPONNIER(1993):KinematicmodelofactivedeformationinCentralAsia,Geophys.
Res.
Lett.
,20,895-898.
BALJINNYAM,I.
,A.
BAYASGALAN,B.
A.
BORISOV,A.
CISTERNAS,M.
G.
DEM'YANOVICH,L.
GANBAATAR,V.
M.
KOCHETOV,R.
A.
KURUSHIN,P.
MOLNAR,H.
PHILIPandY.
Y.
VASHILOV(1993):RupturesofmajorearthquakesandactivedeformationinMongoliaanditssurround-ings,Geol.
Soc.
Am.
Mem.
,181,pp.
62.
BENDICK,R.
,R.
BILHAM,J.
FREYMUELLER,K.
LARSON,G.
PELTZERandG.
YIN(2000):GeodeticevidenceforalowsliprateontheAltynTaghFaultandconstraintsonthedeformationofAsia,Nature,404,69-72.
CHEN,W-P.
,C-Y.
CHENandJ.
L.
NBELEK(1999):Present-daydeformationoftheQaidambasinwithimplicationsforintra-continentaltectonics,Tectonophysics,305,165-181.
COWGILL,E.
,A.
YIN,X.
WANGandQ.
ZHANG(2000):IstheNorthAltynFaultpartofastrike-slipduplexalongtheAltynTaghFaultSystem,Geology,28,255-258.
COWGILL,E.
,A.
YIN,J.
R.
ARROWSMITH,W.
XIAOFENGandS.
ZHANG(2003):TheAkatoTaghbendalongtheAltynTaghFault,NWTibet,1.
Cenozoicstructure,smooth-ingbyvertical-axisrotation,andtheeffectoftopo-graphicstressesonborderlandfaulting,Geol.
Soc.
Am.
(inreview).
DEPOLO,C.
M.
,D.
CLARK,B.
SLEMMONSandA.
RAMELLI(1991):HistoricalsurfacefaultingintheWesternBasinandRangeProvince,WesternNorthAmerica:impli-cationsforfaultsegmentation,J.
Struct.
Geol.
,13,122-136.
ENGLAND,P.
C.
andP.
MOLNAR(1997):ThefieldofcrustalvelocityinAsiacalculatedfromQuaternaryratesofsliponfaults,Geophys.
J.
Int.
,130,551-582.
FORMAN,S.
L.
(1999):Infraredandredstimulatedlumines-cencedatingofLateQuaternarynearshoresedimentsfromSpitsbergen,Svalbard,Arctic,Antarct.
AlpineRes.
,31,34-49.
GE,S.
and34OTHERS(1992):ActiveAltunFaultZoneMonograph(StateSeismologicalBureau,SeismologicalPress,Beijing,China),1-319.
GRANT,L.
B.
andK.
SIEH(1994):PaleoseismicevidenceofclusteredearthquakesontheSanAndreasFaultintheCarrizoPlain,California,J.
Geophys.
Res.
,99,6819-6841.
HARRIS,R.
A.
andS.
M.
DAY(1993):Dynamicsandfaultinteraction:parallelstrike-slipfaults,J.
Geophys.
Res.
,98,4461-4472.
HARRIS,R.
A.
andS.
M.
DAY(1999):Dynamic3Dsimula-tionsofearthquakesonenéchelonfaults,Geophys.
Res.
Lett.
,206,2089-2092.
HILLEY,G.
E.
andJ.
J.
YOUNG(2003):Determiningevenmagnitudeandtimingfrompaleoseismicandgeomor-phicdataalongtheCentralSanAndreasFault,California,Bull.
Seismol.
Soc.
Am.
(inreview).
HOLT,W.
E.
,N.
CHAMOT-ROOKE,X.
LEPICHON,A.
J.
HAINES,B.
SHEN-TUandJ.
REN(2000):VelocityfieldinAsiainferredfromQuaternaryfaultslipratesandGlobalPositioningSystemobservations,J.
Geophys.
Res.
,105,19,185-19,209.
KLINGER,Y.
,J.
VANDERWOERD,P.
TAPPONNIER,X.
W.
XU,G.
KING,W.
B.
CHEN,W.
T.
MAandD.
BOWMAN(2003):DetailedstripmapoftheKokoxiliearthquakerupture(MW7.
8,14/11/01)fromspace,Geophys.
Res.
Abstr.
,5,08487.
KNUEPFER,P.
L.
K.
(1989):Implicationsofthecharacteris-ticsofend-pointsofhistoricalsurfacefaultrupturesforthenatureoffaultsegmentation,inFaultSegmentationandControlsofRuptureInitiationandTermination,editedbyD.
SCHWARTZandR.
SIBSON,U.
S.
Geol.
Surv.
Open-FileRep.
89-315,193-228.
LIN,A.
,B.
FU,J.
GUO,Q.
ZENG,G.
DANG,W.
HEandY.
ZHAO(2002):Co-seismicstrike-slipandrupturelengthproducebythe2001Ms8.
1CentralKunlunearth-quake,Science,296,2015-2017.
MRIAUX,A.
S.
,P.
TAPPONNIER,F.
J.
RYERSON,J.
VANDERWOERD,C.
LASSERRE,X.
XU,R.
FINKELandM.
CAFFEE(1999):Large-scalestrainpatterns,greatearthquakebreaks,andLatePleistoceneslip-ratealongtheAltynTaghFault(China),EuropeanUnionofGeosciencesConferenceAbstracts,EUG10,TerraAbstracts,11,p.
55.
MRIAUX,A.
S.
,F.
J.
RYERSON,P.
TAPPONNIER,J.
VANDERWOERD,R.
C.
FINKEL,M.
W.
CAFFEE,C.
LASSERRE,X.
XU,H.
LIANDZ.
XU(2000):FastextrusionoftheTibetPlateau:a3cm/yr,100kyrslip-rateontheAltynTaghFault,EosTrans.
Am.
Geophys.
Un.
,81(48),p.
1137.
MRIAUX,A.
S.
,P.
TAPPONNIER,F.
J.
RYERSON,X.
XU,J.
VANDERWOERD,G.
KING,R.
C.
FINKEL,LIHAIBING,M.
W.
CAFFEEandX.
ZHIQIN(2003):Post-glacialsliprateontheAksaysegmentoftheNorthernAltynTaghFault,derivedfromcosmogenicradionuclidedatingofmorphologicaloffsetfeatures,Geophys.
Res.
Abstr.
,5,08062.
MEYER,B.
,P.
TAPPONNIER,Y.
GAUDEMER,G.
PELTZER,G.
SHUNMINandC.
ZHITAI(1996):Rateofleft-lateralmovementalongtheeasternmostsegmentoftheAltynTaghFault,eastof96°E(China),Geophys.
J.
Int.
,124,29-44.
MEYER,B.
,P.
TAPPONNIER,L.
BOURJOT,F.
MTEVIER,Y.
GAUDEMER,G.
PELTZER,G.
SHUNMINandC.
ZHITAI(1998):CrustalthickeninginGansu-Qinghai,lithos-phericmantlesubduction,andoblique,strike-slipcon-trolledgrowthoftheTibetplateau,Geophys.
J.
Int.
,135,1-47.
MOLNAR,P.
,C.
BURCHFIEL,L.
K'UANGYIandZ.
ZIYUN(1987):GeomorphicevidenceforactivefaultingintheAltynTaghandNorthernTibetandqualitativeesti-matesofitscontributiontotheconvergenceofIndiaandEurasia,Geology,15,249-253.
PELTZER,G.
,P.
TAPPONNIERandR.
ARMIJO(1989):MagnitudeofLateQuaternaryleft-lateraldisplace-mentsalongthenorthedgeofTibet,Science,246,1285-1289.
PELTZER,G.
,F.
CRAMPandG.
KING(1999):EvidenceofnonlinearelasticityofthecrustfromtheMw7.
6Manyi(Tibet)earthquake,Science,286,272-276.
RAMSEY,C.
B.
(2000):OxCalv3.
5softwarethatusesstrati-graphicconstraintstotrimcalendarageprobabilitydis-tributions,http://www.
rlaha.
ox.
ac.
uk/oxcal/oxcal.
htm.
RYERSONF.
J.
,A-S.
MRIAUX,P.
TAPPONNIER,J.
VANDERWOERD,XUXIWEI,LIHAIBINGandXUZHIQIN(2003):NortheastwardsdecreaseintheLatePleistocene-Holoceneslip-rateandpropagationoftheAltynTaghFault(China),Geophys.
Res.
Abs.
,5,08070.
1033PaleoseismologyoftheXorxolSegmentoftheCentralAltynTaghFault,Xinjiang,ChinaSHEN,Z-K.
,M.
WANG,Y.
LI,D.
D.
JACKSON,A.
YIN,D.
DONGandP.
FANG(2001):CrustaldeformationalongtheAltynTaghFaultSystem,WesternChina,fromGPS,J.
Geophys.
Res.
,106,30,607-30,621.
STUIVER,M.
andH.
A.
POLACH(1977):Discussion,report-ingof14Cdata,Radiocarbon,19,355-363.
STUIVER,M.
,P.
J.
REIMERandE.
BARD(1998):INTCAL98radiocarbonagecalibration,24.
000-0calBP,Radio-carbon,40,1041-1081.
TAPPONNIER,P.
andP.
MOLNAR(1977):ActivefaultingandtectonicsinChina,J.
Geophys.
Res.
,82,2905-2930.
TAPPONNIER,P.
,X.
ZHIQIN,F.
ROGER,B.
MEYER,N.
ARNAUD,G.
WITTLINGERandY.
JINGSUI(2001):ObliquestepwiseriseandgrowthoftheTibetplateau,Science,294,1671-1677.
THATCHER,W.
(1990):Present-daycrustalmovementsandthemechanicsofcyclicdeformation,inTheSanAndreasFaultSystem,California,editedbyR.
E.
WALLACE,USGSProf.
Paper1515,189-205.
VANDERWOERD,J.
,F.
J.
RYERSON,P.
TAPPONNIER,Y.
GAUDEMER,R.
FINKEL,A.
S.
MERIAUX,M.
W.
CAFFEE,Z.
GUOGUANGandH.
QUNLU(1998):Holoceneleft-slipratedeterminedbycosmogenicsurfacedatingontheXidatanSegmentoftheKunlunFault(Qinghai,China),Geology,26,695-698.
VANDERWOERD,J.
,X.
XU,H.
LI,P.
TAPPONIER,B.
MEYER,F.
J.
RYERSON,A.
S.
MRIAUXandX.
ZHIQIN(2001):RapidactivethrustingalongthenorthwesternrangefrontoftheTangheNanShan(WesternGansu,China),J.
Geophys.
Res.
,106,30,475-30,504.
VANDERWOERD,J.
,A.
MRIAUX,Y.
KLINGER,F.
J.
RYERSON,Y.
GAUDEMERandP.
TAPPONNIER(2002):The14No-vember2001,Mw=7.
8KokoxiliearthquakeinNort-hernTibet(QuinghaiProvince,China),Seismol.
Res.
Lett.
,73,125-135.
WANG,Q.
,P-Z.
ZHANG,J.
T.
FREYMULLER,R.
BILHAM,K.
M.
LARSON,X.
LAI,X.
YOU,Z.
NIU,J.
WU,Y.
LI,J.
LIU,Z.
YANGandQ.
CHEN(2001):Present-daycrustaldefor-mationinChinaconstrainedbyGlobalPositioningSystemmeasurements,Science,294,574-577.
WARD,S.
N.
(1997):Dogtailsversusrainbows:syntheticearth-quakerupturemodelsasanaidininterpretinggeologicaldata,Bull.
Seismol.
Soc.
Am.
,87,1422-1441.
WASHBURN,Z.
(2001):QuaternarytectonicsandearthquakegeologyoftheCentralAltynTaghFault,Xinjiang,China:implicationsfortectonicprocessalongthenorthernmarginofTibet,M.
S.
Thesis,Tempe,ArizonaStateUniversity,pp.
179.
WASHBURN,Z.
,J.
R.
ARROWSMITH,S.
FORMAN,E.
COWGILL,X.
F.
WANG,Y.
ZHANGandC.
ZHENGLE(2001):LateHoloceneearthquakehistoryoftheCentralAltynTaghFault,China,Geology,29,1051-1054.
WELLS,D.
L.
andK.
COPPERSMITH(1994):Newempiricalrelationshipsamongmagnitude,rupturelength,rupturewidth,rupturearea,andsurfacedisplacement,Seismol.
Soc.
Am.
Bull.
,84,974-1002.
WESNOUSKY,S.
G.
(1989):Seismicityandthestructuralevo-lutionofstrike-slipfaults,inFaultSegmentationandControlsofRuptureInitiationandTermination,editedbyD.
SCHWARTZandR.
SIBSON,U.
S.
Geol.
Surv.
Open-FileRep.
89-315,409-431.
WESSEL,P.
andW.
H.
F.
SMITH(1995):NewversionoftheGenericMappingToolsreleased,Eos,Trans.
Am.
Geo-phys.
Un.
,76(33),p.
329.
YIN,A.
andM.
HARRISON(2000):GeologicevolutionoftheHimalayan-Tibetanorogen,Annu.
Rev.
EarthPlanet.
Sci.
,28,211-280.
1034ZackWashburn,JRamónArrowsmith,GuillaumeDupont-Nivet,WangXiaoFeng,ZhangYuQiaoandChenZhengle
IntroductionThesinistralAltynTaghFault(ATF)travers-esthenorthernboundaryofTibetfor>1500kmandisamajorstructureintheIndia-Eurasiacon-PaleoseismologyoftheXorxolSegmentoftheCentralAltynTaghFault,Xinjiang,ChinaZackWashburn(1),JRamónArrowsmith(2),GuillaumeDupont-Nivet(3),WangXiaoFeng(4),ZhangYuQiao(4)andChenZhengle(4)(1)GeomatrixConsultantsInc.
,CostaMesa,CA,U.
S.
A.
(2)DepartmentofGeologicalSciences,ArizonaStateUniversity,Tempe,AZ,U.
S.
A.
(3)PaleomagneticLaboratoryFortHoofdijk,FacultyofEarthSciences,UtrechtUniversity,Netherlands(4)InstituteofGeomechanics,ChineseAcademyofGeologicalSciences,Beijing,ChinaAbstractAlthoughtheAltynTaghFault(ATF)isthoughttoplayakeyroleinaccommodatingIndia-Eurasianconver-gence,littleisknownaboutitsearthquakehistory.
Studiesofthisstrike-slipfaultareimportantforinterpreta-tionoftheroleoffaultingversusdistributeddeformationintheaccommodationoftheIndia-Eurasiacollision.
Inaddition,the>1200kmlongfaultrepresentsoneofthemostimportantandexemplaryintracontinentalstrike-slipfaultsintheworld.
WemappedfaulttracegeometryandinterpretedpaleoseismictrenchexposurestocharacterizetheseismogenicbehavioroftheATF.
Weidentified2geometricsegmentboundariesina270kmlongreachofthecentralATF.
TheseboundariesdefinethewesternmostWuzhunxiao,theCentralPingding,andtheeasternmostXorxol(alsowrittenasSuekuliorSuo'erkuli)segments.
Inthispaper,wepresenttheresultsfromtheCamelpaleoseismicsitealongtheXorxolSegmentat91.
759°E,38.
919°N.
Thereevidenceforthelasttwoearthquakesisclearand14Cdatesfromlayersexposedintheexcavationbrackettheirages.
Themostrecentearthquakeoccurredbetween1456and1775calA.
D.
andthepenultimateeventwasbetween60and980calA.
D.
CombiningtheCamelinterpretationswithourpublishedresultsforthecentralATF,weconcludethatmul-tipleearthquakeswithshorterrupturelengths(50km)ratherthancompleteruptureoftheXorxolSegmentbet-terexplainthepaleoseismicdata.
Wefound2-3earthquakesinthelast2-3kyr.
Whencoupledwithtypicalamountsofslipperevent(5-10m),therecurrencetimesaretentativelyconsistentwith1-2cm/yrsliprates.
Thisresultfavorsmodelsthatconsiderthebroaderdistributionofcollisionaldeformation,ratherthanthosewithnorthwardmotionofIndiaintoAsiaabsorbedalongafewfaultsboundingrigidblocks.
tinentalcollision(e.
g.
,TapponnierandMolnar,1977;Peltzeretal.
,1989;YinandHarrison,2000;Tapponnieretal.
,2001;fig.
1).
Twocom-petingend-membermodelshavedevelopedthatdescribehownorthwardconvergenceofIndiaisaccommodatedbyTibetandSouthernAsia.
Thefirstmodeldividesthisregionintoafewrigidblocksboundedbymajorfaults(i.
e.
AltynTaghandKunlunfaults,fig.
1)(e.
g.
,AvouacandTap-ponnier,1993;Tapponnier,etal.
,2001).
Thesecondmodelhasconvergenceaccommodat-edbycrustalthickeningandalongnumerousminorandafewmajorfaults(e.
g.
,EnglandandMolnar,1997).
Modelsofthefirsttypearecon-1015ANNALSOFGEOPHYSICS,VOL.
46,N.
5,October2003Mailingaddress:Dr.
JRamónArrowsmith,DepartmentofGeologicalSciences,ArizonaStateUniversity,Tempe,AZ85287-1404U.
S.
A.
;e-mail:ramon.
arrowsmith@asu.
edu1016ZackWashburn,JRamónArrowsmith,GuillaumeDupont-Nivet,WangXiaoFeng,ZhangYuQiaoandChenZhengleChinaTibetAltynTaghFault80°100°120°40°20°KunlunfaultRedRiverFaultHimalayanFrontalThrustsIndiaFig.
1.
OverviewmapshowingmajorstructuresandtopographyintheIndia-Eurasiacontinentalcollisionzone.
Insetshowslocationoffig.
2.
FiguregeneratedfromGTOPO301kmdigitalelevationdatasetusingGMT(WesselandSmith,1995).
sistentwithhighersliprates(3cm/yrfortheATF)whilelowerslipratesareconsistentwiththesecondtypebecausedeformationisdistrib-utedinsteadoflocalizedalongasmallnumberofdiscretefaultplanes.
Globalseismicitydatabases(CNSS)andtheChineseCatalogofHistoricStrongEarthquakes(2300B.
C.
to1911A.
D.
)recordnomajorearthquakesalongtheATF.
However,weknowthatlargeearthquakeshaveoccurredbecauseofmoletracksseenonsatelliteimagery,surfacebreaksshownbyGeetal.
(1992),andlimitedfieldinvestigationbywesternscientists(Molnaretal.
,1987).
Despitetheclearevidenceofactivefaulting,slipratedeterminationsarediscrepant.
Overtimescalesof10s-100sofkyr,slipratesareinferredtobe2-3cm/yr.
Peltzeretal.
(1989)inferred3cm/yrratesintheareaof84°E-92°EbasedonanassumptionofinitialHoloceneageforoffsetlandforms.
RecentworkhasfocusedondeterminingtheslipratebydatingoffsetQuaternarylandformsusingcosmogenicmeth-ods.
Meriauxetal.
(1999,2000)determined3cm/yrbetween85°Eand89°Eand2-3cm/yrnear90°E.
Merieauxetal.
(2003)determinedaminimumof20±3mm/yrat94°E.
Chineseearthquakegeologystudies(mapping,paleoseis-mictrenching,and14C)inferredaminimumrateof5mm/yralongtheentireATF(Geetal.
,1992).
StudiesofdeformationintheIndia-Eur-asiancollisionovergeodetictimescales(5-10years;Bendicketal.
,2000;Shenetal.
,2001;Wangetal.
,2001)indicatethatdistributeddeformationinTibetisanimportantprocessbecausefaultssuchastheATFhaveslipratesontheorderof10mm/yr(9±5mm/yr,Ben-dicketal.
,2000;9±2mm/yr,Shenetal.
,2001).
Holtetal.
(2000)alsousedarateof10±8mm/yralongtheATFintheiranalysisofthevelocityfieldinAsia.
Twomajorstrike-slipearthquakeshaverecentlyoccurredinNorthernTibet.
The1997Mw7.
6Manyiearthquakerupturedforadis-tanceofabout170kmwithpeakleft-lateralslipofabout7m,apparentlyreactivatingaQuaternaryfault(fig.
2;Peltzeretal.
,1999).
The2001Mw7.
8KokoxiliorCentralKunlunearthquakehada300-400kmlongrupturezoneandsinistralslipupto16.
3m(Linetal.
,2002;Klingeretal.
,2003).
ItrupturedtheeasternportionoftheKunlunFault(vanderWoerdetal.
,2002).
Theseearthquakesdrawattentiontothestyleofearthquakerupturealongthegreatstrike-slipfaultsofAsiaandalsototheirresponsetothegeodynamicsetting.
Becausethehistoricrecordinthisregionisonlycom-pleteforaboutthelast100years,paleoseismo-logicalmethodsareimportantforinvestigatingthetimingoflargerupturesliketheseandthosealongneighboringfaults.
Earthquakedisplacements,surfacerupturelengths,andrecurrenceintervalsalongmajorstrike-slipfaultssuchastheATFhaveimplica-tionsforwhichgeodynamicmodelsbetterexplaincontinentaldeformationandforregion-alearthquakesequencesincontinentalsettings.
Therefore,weconductedastudyoftheearth-quakegeologyoftheATFtocharacterizetheseismogenicbehaviorofthismajorstrike-slipfault.
Inthispaper,wefirsthighlightthegeo-PaleoseismologyoftheXorxolSegmentoftheCentralAltynTaghFault,Xinjiang,China808590953540KunlunFaultTibet678White2500yearsBP1997Mw7.
6ManyiearthquakeTarimBasinQaidamBasinQimenTaghMilanRuoqiangNATFHuatagoLapequanAltynTaghFault1924M>7earthquakes1951~M6earthquakeAltynTaghAltynTaghFaultAkatoTaghMinfengSubeiSulamuTaghTula2001Mw7.
8KokoxiliearthquakeFig.
2.
OverviewmapshowingtheCentralAltynTaghFault(ATF),locationofstudyarea(rectangularbox,fig.
3),instrumentalseismicity(CounciloftheNationalSeismicSystem(CNSS);http://quake.
geo.
berkeley.
edu/cnss/cata-log-search.
html),historicearthquakes(ChineseCatalogofHistoricStrongEarthquakes2300B.
C.
toA.
D.
,1991),paleoearthquakes(StateSeismologicalBureau(SSB);Geetal.
,1992),andmajortowns.
TheGeetal.
(1992)dataareplottedatthereportedlocationswiththeirinferredmagnitudeandcolor-codedbyagerange.
TheeventsmostrelevantforthisstudyareM6.
TheactivetracetraversesalluvialflatsaroundLakeWuzhunxiaoandtheBitterSeaandbedrockintheArkatengandPingdingmountainranges.
Twogeometricboundariesareidentified:the7kmwideArkatengRestrainingBendandthe4kmwidePingdingreleasingstep.
ThesegeometricboundariesdefinethewesternmostWuzhunxiao,theCentralPingding,andtheeasternmostXorxolsegments.
Theboundariesserveasaframeworkforinterpretingoffsetdataandesti-matingruptureextentbecausesuchdiscontinuitiescommonlyactasrupturebarriers.
1018over110km(figs.
2and3).
Proterozoicintru-siverocksandsliversofPaleozoic-Mesozoicgranitoid,ultramafic,marinecarbonate,andophioliticrockscomprisethemountainsinthestudyarea(Cowgilletal.
,2003).
ThickNeoge-nebouldergravelstosiltsencirclethePro-terozoic-Mesozoicmountainsandfilladjacentbasins.
IntensefoldingofthesestrataintheQaidamBasinandthroughoutthefieldareaindicateacontractionalregime(regionally)fol-lowingdepositionoftheNeogeneunits.
RelativelythinQuaternarygravelsandsilts(pied-montdepositsandprimaryandreworkedloess)overlietheNeogenestrata.
3.
FaulttracemappingandpaleoseismicsiteidentificationInthefield,wemappedthefaulttracesthatcomprisetheindividualfaultsthathavehadrecentrupturebasedongeomorphicageofthescarps,disruptionsofHolocenedepositsandlandforms.
Theactivetracedesignationgener-allyreferstosurfacebreakscreatedbythelast1-3earthquakes.
WehaveidentifiedtheactivetracesandmostQuaternaryfaults,butmaynotshowalloftheearlyQuaternaryfaultsbecauseweareprimarilyconcernedwiththegeometryandearthquakehistoryforthelastfewkyr.
Infig.
3,faultslabeledQmaresimilarinagetoorslightlyyoungerthansurfacesanddepositsofLatePleistoceneandEarlyHoloceneinferredage.
TheseweremappedasunitQmandarethemostextensiveoftheQuaternarydepositsandformbroadfloorsthatdepositionallyabutadja-centhillsidesandoftendefineasharpbreakinslope.
Theygenerallyhaveafewto10sofmincisionandmaybecoveredwithvarnishedclasts.
TectoniclineamentsrefertosuspiciousgeomorphicalignmentsthataremostlikelyfaultsolderthanQm.
OurmappingapproachwastoinspecttheprincipalactivetracesasidentifiedonCORONAimagery.
Wemappedthemainfaulttraces,theQuaternaryunitsdiscussedabove,andocca-sionallyobtainedbeddingorotherstructuralattitudes.
Wetypicallycovereda5kmwidezonecenteredoverandparalleltotheactivetracetoidentifyrelevantstructuresandgeolog-icrelations.
Wemademanysmallexcavationswithshovelstoclarifystratigraphicrelation-shipsandexposebeddingsurfaces.
Thesesmall-erexcavationshelpedaddconfidencetoourgeo-morphicandstratigraphic-basedpaleoseismicsitedecisions.
Weloggednaturalandhand-dugexposures,anddocumentedoffset(bothverticalandhorizontal)landformsandstrata.
Whilemappingthefaulttrace,welookedforsiteswherelandformsindicatedthatsteadyaccumulationofdistinctivesedimentswouldpreserveevidenceforpastruptures.
Inanum-berofexcavationsatsuchsites,brokenandtilt-edbedsoverlainbycontinuousbedsandupwardterminationsoffaulttracesidentifypastearth-quakesinexcavations.
Disruptionseenatthesamestratigraphiclevelatseverallocationsandonbothtrenchwallsqualifiesasadistinctevent(afterGrantandSieh,1994).
Organicmatterwithinthestratigraphywasdatedusing14Candfinesiltwasdatedbyinfraredstimulatedlumi-nescence(IRSL)(Forman,1999).
Weusethe2mcalibratedsampleagesandOxCalv3.
5(Ram-sey,2000)fortimingconstraints.
WealsouseaMonteCarloprogramthatusesorderingcon-straintstotrimprobabilitydistributionsofra-diocarbonsamplesandthenyieldsprobabilitiesforevents(HilleyandYoung,2003).
Thisapproachallowsformoreobjectiveeventcorre-lationbetweenpaleoseismicsites.
4.
FaulttracegeometryAsmall-scalerepresentationoftheactivetracesthatwemappedisshowninfig.
3.
Forconvenience,werefertolocationsalongthemappedtracesintermsofdistancefromthewesternendofthestudyarea.
Washburnetal.
(2001)describedthewestern150kmoftheareashowninfig.
3andestablishedthegeo-metricsegmentation.
Oursubsequentmap-pingaddedanother120kmtowardtheeast(Washburn,2001).
Aviewalongthestrikeoftheactivetraceinfig.
3showsaprominentrightbend(65to85km)andleftstep(97km)inthegenerallystraightCentralAltynTaghFault.
WefollowedtheconventionsofdePoloetal.
(1991)andKnuepfer(1989)andidenti-fiedthesediscontinuitiesasgeometricbound-1019PaleoseismologyoftheXorxolSegmentoftheCentralAltynTaghFault,Xinjiang,ChinaarieswhichinturndefinethewesternmostWuz-hunxiaoSegment,theCentralPingdingSeg-ment,andtheEasternXorxolSegment.
Thesesegmentsarenotnecessarilyrupturesegments;however,thisapproachisusefulforcharacterizingfaultzonesbecausegeometricboundariesmaybelocationsofsignificantchangeinslipormomentrelease(e.
g.
,Wesnousky,1989;HarrisandDay,1993,1999).
Historicalrupturedatashowthatrup-tureendpointscommonlycoincidewithgeometricboundariesandtheseboundariesappeartoretardrupture(Knuepfer,1989).
ThemajorgeometricboundaryinthestudyareaisthePingdingReleasingStep.
Here,theactivetracejumpsfromthesouthernsideofPingdingShan4kmnorthtotheactivetracethatcontinuestotheXorxolValley(fig.
3;Wash-burn,2001;Washburnetal.
,2001).
The3-4mverticaloffsetofHolocene-agefanonthesout-heastsideofPingdingShan,aprominentscarpontheeastfaceofPingdingShan,andthepres-enceofrecentlandslidesindicatethatthenor-malfaultthatlinkstheboundingstrike-slipfaultsrupturedrecently.
Ourmappingstopsat268kminastraightandsimpleportionoftheXorxolSegment(fig.
3).
This070°strikingsimplegeometrycontinueseastward,butjustofftheeasternedgeoffig.
3,around300km,aprominentnorthernstrandde-velops(fig.
2).
WeinferthattheXorxolgeomet-ricsegmentisatleast200kmlongandremark-ablystraight(Washburn,2001).
5.
XorxolSegmentpaleoseismicinvestigationOverviewSeismogenicpropertiessuchasfaultstraight-nessandmaximumdisplacementduringthelastearthquakesatspecificsitesalongthecentralATFwerepresentedaboveandinWashburnetal.
(2001)andWashburn(2001).
Thenextstepsincharacterizingtheseismogenicbehav-ioroftheCentralAltynTaghFaultaredetermi-nationsofrecentearthquaketimingandesti-matesofmagnitude.
Inordertoestablishthetimingofpastearthquakes,wemade10exca-ZackWashburn,JRamónArrowsmith,GuillaumeDupont-Nivet,WangXiaoFeng,ZhangYuQiaoandChenZhengleTableI.
14CdatafortheCamelpaleoseismicsite.
AllanalysesarestandardAMSdatesfromCAMSfacilityatLLNL.
b.
13CvaluesaretheassumedvaluesaccordingtoStuiverandPolach(1977).
Valuesmeasuredforthematerialitselfaregivenwithasingledecimalplace.
Thequotedageisin14CyearsusingtheLibbyhalflifeof5568yearsandfollowingtheconventionsofStuiverandPolach(1977).
14Cconcentrationisgivenasfractionmodern,D14C,andconventional14Cage.
FieldsampleCAMS#b.
13CFraction±D14C±14.
Cage±2mcalibratednamemodernage(*)(calA.
D.
)C169687–250.
89230.
0035–107.
73.
5920401024-1209C769688–250.
95920.
005–40.
85330501456-1650C869689–250.
98550.
0048–14.
54.
8120401676-1939C10A69690–250.
79210.
0034–207.
93.
418704060-240C1669691–250.
89120.
0043–108.
84.
3930401022-1191C2069692–250.
86390.
0047–136.
14.
7118050717-980C2269693–250.
98030.
0039–19.
73.
9160401662-1950C2469694–250.
73760.
0042–262.
44.
2244050762-403B.
C.
C2569695–250.
98740.
0039–12.
63.
9100401678-1937C2669696–250.
89290.
0045–107.
14.
5910501022-1221C3269697–250.
62120.
003–378.
833820402404-2140B.
C.
C4569698–250.
57450.
0027–425.
52.
74450403337-2924B.
C.
SLrabbit69699–251.
12450.
0054124.
55.
4>Modern>1950SLcamel69700–251.
20740.
0058207.
45.
8>Modern>1950SLstick69701–251.
18980.
0052189.
85.
2>Modern>1950(*)DeterminedwithCalib4.
2(Stuiveretal.
,1998).
Doesnotincludecurveinterceptscomprising30mlengthofthetrench.
Therefore,thelim-iteddeformation,seenasonly3faults,indi-catethatthisintermediateeventwasdifferentandhadlowermagnitudethantheMREandpenultimateevents.
Wefavoraninterpretationofupwarddieoutand/orplungingtiplineofMREfracturesbecausethefaultingrelation-shipsarenotthesameonbothtrenchwallsandthemuchsmallerdegreeofdeformationcom-paredtorupturesseeninothertrenchesinthestudyarea.
Moresubstantialevidenceforapre-penulti-mateeventisseen1mnorthofandinthebot-tomoffaultzone5.
Here,faultedandtiltedstrataareoverlainbyflatlyingbeddedsandandfinegravelofunit85.
Theserelationshipsaresuggestiveofanearthquakebeforethepenulti-mateevent,howeverthetrenchwasnotdeepenoughtoconfirmthisevent.
Faultzone6preservesevidenceofthefail-ureoflargeblocksofthelocalNeogenebase-mentinduratedsiltsintomostlyeoliansands.
Theblocksarestronglyfracturedandincombi-nationwiththewindblownsandsrepresentaslightlydifferentenvironmentthantheeoliansandandalluvialfanmuddysandsandgravelstothenorth,furthermakingcorrelationofe-ventswiththerestofthetrenchdifficult.
5.
5.
Camel14CdatingWecollected57samplesoforganicmate-rialfromthetrenchanddated12,aswellas3fromamodernstrandline.
BecausemuchoftheorganicmaterialintheCameltrenchwasrepresentedinmodernstrandlines,andoftendepositedasstrandlines,wedatedthemodernonestoassesstheirabilitytodefinethedepo-sitionalage.
The3strandlinesamplesareallmodern.
Thus,the14Cdatesinthetrenchprob-ablydonotincludeasignificantpre-deposi-tionalage.
The14CdataarepresentedintableIandtheprobabilitydistributionsfromcalibra-tionsusingOxCal(Ramsey,2000)areshowninfig.
7a,b.
An8by5by3mmpieceofwoodwithintactbark(C8)fromunit150providesanupperboundof1676-1939calA.
D.
,whilearabbitdropping(C7)yieldsalowerboundof1456-1650,forthelastearthquakeinthisarea.
Likewise,anelongaterabbitdropping(C20)placesanupperboundsof766-980calA.
D.
and10smallpiecesofwoodydebrisweremixed(C10)toprovidealowerboundsof60-240calA.
D.
forthepenultimateevent.
Theoldeststrataexposedinthetrencharebe-tween3337-2924calB.
C.
basedonaradio-carbondatefrompieceofwood(C45).
Becauseofthecompletestratigraphica-greementwithinall12samples,weusedOxCal1026ZackWashburn,JRamónArrowsmith,GuillaumeDupont-Nivet,WangXiaoFeng,ZhangYuQiaoandChenZhengle1027PaleoseismologyoftheXorxolSegmentoftheCentralAltynTaghFault,Xinjiang,Chinatorefinetheconstraintsoneventtiming(Ram-sey,2000).
Thisprogramusesstratigraph-icorderingconstraintstoeliminateoverlap-pingageprobabilities,whichrefinesourearthquaketimingbrackets.
Weused1950A.
D.
asanuppereventboundarybecausetheworldwideseismicnetworkwasabletodetectmoderatesizedearthquakesbythistime.
OxCalismostusefulforsamplesthatarecloselyspacedstratigraphicallyandyieldsimilarradiocarbonages.
C22islocatednearthetopofunit160whileC25inthelowerpartofthisunitandC8islocatedinunit150.
Thesesamplesallyieldsimilarcalendarages(tableI),howevertheirstratigraphicpositionallowsOxCaltocutoffoverlappingportionsoftheirprobabil-itycalendaragedistributions.
Thisisusefulbe-causethispartofthecalibrationcurveresultsinawideagerange(e.
g.
,1676-1939forC8).
AfterrunningOxCal,C8yieldsanarrowup-perboundof1676-1775fortheCamelMRE.
Thelowerboundof1456-1650cal.
A.
D.
remainsunchangedbecausethenextlowestsample(C261022-1221calA.
D.
)ismucholderandthereforeprobabilitydistributionsdonotoverlap.
Thelowerboundonthepe-nultimateearthquakeisalsounchangedforthesamereason.
Abovethepenultimateeventhorizon,unit125contains2radiocarbonsam-ples,howevertheyarespacedoveran8mhorizontaldistanceandsitattheapproximatesamestratigraphiclevel.
Thereforewecannotapplyorderingconstraintstothesesamples,althoughC26inunit127doesshifttheiragedistributions30yearsolder.
Figure8a,bshowstheearthquaketimingconstraintsatanumberofsitesalongthecen-tralATF.
Theagesdisplayedinfig.
8aarecal-ibratedbutotherwisenotrefinedwithstrati-050100150200250distance(km)CalendarYears-2500-1500-5005001500Kulesayi2-6-172-6-18CamelExplanation14MRExxxmostrecenteventyoungerthanMREolderthanMRE/youngerthanpen.
olderthanpenultimateeventearthquakefromActiveAltunFaultmonograph(Geetal.
)2sigmacalibratedageofC(solidline)andIRSL(dottedline)samplesolderthanearliesteventnosym.
-2500-1500-5005001500CamelProbability0.
010.
0CalendarYearsExplanationMREPenultimateeventFig.
8a,b.
a)Earthquaketimingconstraintsfromstratigraphicrelationshipsand14CandInfraredStim-ulatedLuminescence(IRSL)samples(fromthisstudyandWashburnetal.
,2001).
NotetheactivetracemapaboveplotAforreference.
ThesearethecalibratedageswithoutOxCalorotherrefinementsfromstratigraphicordering.
b)MonteCarloproba-bilityplotfortheCamelpaleoseismicsiteearthquaketimingdata(HilleyandYoung,2003).
Theareaunderthecurvesrepresentstheprobabilitydistributionoftheageofpastearthquakes.
Sharperpeaksrepresentwell-constrainedeventsandshowtherefinementoftheCameleventages.
abZackWashburn,JRamónArrowsmith,GuillaumeDupont-Nivet,WangXiaoFeng,ZhangYuQiaoandChenZhengle1028Fig.
9.
TrenchlogsfortheCamelpaleoseismicsiteat198km.
Mostfaultsterminateattheupperblueline,theeventhorizonforthelastearthquakeatthissite.
Thepenultimateearthquakeisidentifiedbyupwardfaultter-minationsandfoldedbeddingoverlainbycontinuousstrata,whichismarkedbythelowerblueline.
Thispenul-timateeventhorizoncoincideswithasmallunconformity(yellowline)identifiedbydifferentlevelsofindura-tionandanirregularcontactwithcmdeeprillsinplaces.
StratigraphiccolumnofCameltrenchatlowerleftshowsrelativeunitthickness,unitnumber,majorunconformities,andstratigraphiclocationofradiocarbonsam-ples.
Seediscussionintextforfurtherinterpretationandtimingconstraintsonearthquakesseeninthistrench.
N2xxxxxC221662-1950A.
D.
C251678-1937A.
D.
C71456-1650A.
D.
C261022-1221A.
D.
C161022-1191A.
D.
C20717-980A.
D.
C11024-1193A.
D.
C1060-240A.
D.
C24762-403B.
C.
C322404-2140B.
C.
C453337-2924B.
C.
uncfuncfuncf170160150140130125123127115110105100908580C81676-1939A.
D.
StratigraphiccolumnUnit#SamplecalibrateddateyoungerthanMREolderthanMRE/youngerthanpen.
olderthanpenultimateeventxRadiocarbonsamplesuncfunconformity115unitnumberKeyExplanationmatrixsupportedmassivesandgravelclastsupportedgravelbeddedsandeoliansandsiltsandysiltprominentstrandlineeventhorizonunconformity14Csamplefault,fractureNeogenebedrockeastwallC32(2404-2360,2354-2140calB.
C.
)C26(1022-1221calA.
D.
)C25(1803-1907,1678-1744calA.
D.
)C16(1022-1191calA.
D.
)C1(1024-1193calA.
D.
)westwallfaultzone1faultzone2faultzone3faultzone4faultzone5faultzone6C24(762-678,671-608,599-403calB.
C.
)C7(1456-1650calA.
D.
)C8(1802-1939,1676-1765calA.
D.
)C20(766-980calA.
D)C22(1662-1711,1717-1886,1911-1950calA.
D.
)C45(3337-3208,3194-3150,3140-3008,2987-2924calB.
C.
)C10(60-240calA.
D.
)05mprominentstrandline14CsampleExplanationeventhorizonunconformityfault,fracturematrixsupportedclastsupportedbeddedsandmassivesandgravelgraveleoliansandsiltsandysiltNeogenebedrock120130SOUTHNORTHUnitnumbershUnitnumbers17016015514013012311510510017016013011511015014090120130100105170160155140130123115105100Unitnumbers8090100110160170Fig.
6.
(continued)1029PaleoseismologyoftheXorxolSegmentoftheCentralAltynTaghFault,Xinjiang,ChinagraphicorderingconstraintsusingOxCal.
Hil-leyandYoung(2003)havedevelopeda14CagerefinementmethodthatissimilartoOxCalandproducesprobabilitycurvesforeventswithaMonteCarlo-basedanalysisof14Cagecalibrationandstratigraphicordering.
Weshowtheprobabilitycurvesforthelasttwowell-definedCameleventsdeterminedusingtheHilleyandYoung(2003)methodinfig.
8b.
6.
Discussion6.
1.
EarthquakerupturesalongtheCentralAltynTaghFaultCorrelationofearthquakesbetweenpaleo-seismicsitescanleadtomisinterpretationbe-causenotallstrandsinwidefaultzonesmaybeactivatedduringanearthquake.
Wehaveelimi-natedmostofthepotentialforthisscenariobythoroughevaluationofpaleoseismicsitesandexcavationacrossonlythesharpest,mostactivetraces.
However,dogtailruptureterminationsinthemiddleofgeometricsegmentsandmaynotrupturethroughallexcavationsinquestion(e.
g.
,Ward,1997).
Thereforetrenchesshowingoverlappinggeochronologicearthquakecon-straintsthatareonly10sofkilometersapartmaynotrecordthesameearthquake.
Wethere-foreutilizeotherobservationstohelpclarifytherupturehistory.
Weknowthepeaksurfacedisplacementatagivensitefromtheoffsetdata(Washburn,2001;Washburnetal.
,2001)andwehavecon-strainedearthquaketimingatanumberofpaleo-seismicsites.
Withthesedata,wecandeterminecruderecurrenceintervals,becausewehaveonly1-3eventsineachtrench.
Wecanalsomakemagnitudeestimatesusingthedisplace-mentdata.
Betterestimatesofmomentmagni-tude(Mw)canbeachievedbyusingSurfaceRuptureLength(SRL)(WellsandCoppersmith,1994).
Weestimatesurfacerupturelengthbyutilizinggeometricconstraints,offsetdistribu-tion,surfacerupturemorphology,timingcon-straintsfromthetrenchsites.
Figure8a,bshowsthatthelastearthquakeatKulesayi,2-6-17,2-6-18,andtheCamelsitealloccurredattheapproximatesametime,be-tween1215-1775calA.
D.
(datafromWashburnetal.
,2001,andthispaper).
However,itisunlikelythatonelargeearthquakerupturedtheentirefieldareabetween1215-1775calA.
D.
Insuchacase,theMREwouldhavejumpedtwogeometricboundariesandrupturedfromtheKulesayitrench(23km)toatleasttheCameltrench(199km)andlikelyfartherbasedontheoffsetdistributionandsharpsurfacerup-turethatcontinuesto>km240.
ThisscenarioyieldsaSRLof>220kmandMw>7.
8.
However,considerationofdetailsoftheagecontrolfromtheexcavationsandthegeometricboundariesmakesitunlikely,andtherelativelyhighslipim-plied(10m)isnotconsistentwithourinfer-redoffsetdistribution(maximumoffset7m;Washburn,2001;Washburnetal.
,2001).
Itispossiblethatthe4kmPingdingleftstepbarredrupture.
TheArkatengRestrainingBendandassociatedcomplexitiesaroundDayangShanarealsopotentialrupturebarriersfig.
3.
Furthermore,thediscontinuousanddegradedsurfacerupturearoundtheseboundariesarguesforsmallerearthquakesthatdidnotrupturethewholefieldarea.
ThereforeweseparateWuzhunxiaoandPingdingeventsfromXorxoleventsbasedonsurfacerupturemorphologyandgeometricboundaries.
Consequently,thelastearthquakeatKulesayioccurredbetween1215and1750A.
D.
andmayhaveproducedthe3-5moffsetsonthewestsideofLakeWuzhunxiao(fig.
8a,b;Washburnetal.
,2001).
AlongtheXorxolSegment,thetimingcon-straintsimplythatthelastearthquakeruptureditswesternportionbeforetheeasternportionbroke.
Thelastearthquakeinthewestruptured1270and1430calA.
D.
(fromBitterSeapublishedresultsofWashburnetal.
,2001;fig.
8a,b).
ThisrupturemaygofromwesternendoftheXorxolSegmenttoaroundtheBitterSea(50kmSRL)withpeakdisplacementof7m(Washburn,2001;Washburnetal.
,2001).
TheyoungerCamelMREoccurredbetween1456and1775calA.
D.
(fig.
8a,b).
ItrupturedfromsomewhereeastoftheBitterSeathroughtheCameltrenchandlikelyfartherbasedontheroundedandsemicontinuousmoletracksbe-tween247kmtotheedgeofthestudyarea(268km,fig.
3).
ThedegradednatureofthesurfacerupturesuggeststhatthisreachiseithertheeasternterminusoftheXorxolMREorthatero-sionrates,whicharedetectablyhigherinthisarea,ledtothemoredegradedsurfacerupture.
HighererosionratesareplausiblebecauseofthepositionoftheEasternXorxolValleyrelativetothemouthoftheTarimBasinonly7kmnorthoftheplayaat265kmonfig.
3.
ThemouthallowsmorefrequentstormsandintensedailywindstoblowthroughEasternXorxolValleyandmoreanimalaccessandpotentialdegradation.
Insummary,thelastearthquakesalongtheCentralAltynTaghFault:1)occurredbetween1215and1775calA.
D.
;2)hadmaximumdisplacementof5montheWuzhunxiaoandPingdingsegmentsand7montheXorxolSe-gment(Washburn,2001;Washburnetal.
,2001),and3)hadsurfacerupturelengthsbetween50and>150kmwithcorrespondingmagnitudesoflowMw7stohighMw7s,respectively.
6.
2.
Implicationsforslipratebasedonthesizeandfrequencyofearthquakesinthelast3kyrAlthoughtheearthquakedatadonottightlyconstrainthesurfacerupturelengthorMwoftheMRE,therecurrencetimesfromtrenchesshowgoodagreement.
Wefind2earthquakesinthelast0.
8-2.
0kyrattheCamelsite,3eventsinthelast2.
4-3.
0kyrattheBitterSea,and2Kulesayieventsinthelast0.
8-2.
2kyr(thispaper;Washburn,2001;Washburnetal.
,2001).
Althoughwehavealimitednumberofeventsforrobustrecurrencecalculations,takingaveragesfromthesedatagiverepeattimesof0.
7±0.
3kyrfortheCamelsite,0.
9±0.
1kyrfortheBitterSeatrenches,and0.
75±0.
35kyrfortheKule-sayisite.
Thesefrequenciesareconsistentwiththe2earthquakesinthepast3kyrreportedbyGeetal.
(1992)inthestudyareaandaregen-erallysimilartothe800-1000yearrecurrencetimesestimatedfortheKunlunFault(vanderWoerdetal.
,1998).
Giventhelimitednumberofearthquakesinthisstudy,weemphasizethetentativeresultsderivedfromrelatingearthquakerecurrencetosliprates.
However,systematicunderinterpre-tationorrupturelocationvariabilitymustbeinvokedforsignificantlyhigherrecurrenceratesorslip/event.
Ifalargeearthquakeoc-curredinthenearfuture,ourrecurrencerateswouldbemaximaandanyinferredslipratewouldincrease.
Tocomparethesepaleo-earthquakedatawithotherfindingsalongtheATFwecastsliprateintermsofaveragerecurrenceintervalsbyusing5and10mofslippereventbecausethesevaluesbracketthepeakoffsetsrecordedfortheMRE(s)(Washburnetal.
,2001).
Using:1)10mm/yrdeterminedfromrepeatGPSsurveys(Bendicketal.
,2000;Shenetal.
,2001)yieldsrecurrenceintervalsof0.
5and1.
0kyrfor5and10mofslipperevent,respectively;2)3cm/yrdeterminedbylargescalepostglacialoffsetsandoffsetterracesandmorainesinareaswestofthestudyarea(Peltzeretal.
1989;Meriauxetal.
,2000)yieldsrecurrencetimesof0.
2and0.
3kyr.
Figure10showsacompilationofthevariousslipratesthathavebeeninferredforthecentralATF.
Comparisonofthedifferentrecur-renceintervalsshowthatthepaleoearthquakedatamostcloselyagreewitha1-2cm/yrslipratebecausewewouldexpecttosee2-3timesthenumberofearthquakesinourpaleoseismicinvestigationsandthefaultingandsedimentaryrecordinthemisnotconsistentwithsomanymissingearthquakes.
AnimportantissueinslipratecomparisonsistheobservedeastwarddecreaseinslipratealongtheATF(fig.
10).
Peltzeretal.
(1989)in-ferred3cm/yrratesintheareaof84°-92°E.
Meriauxetal.
(1999,2000)determined3cm/yrbetween85°and89°Eand2-3cm/yrnear90°E.
Ryersonetal.
(2003)reportawellconstrain-ed18mm/yratthePingdingShan(90.
5°E).
Meriauxetal.
(2003)determinedaminimumof20±3mm/yrat94°E.
Ryersonetal.
(2003)alsoreport12.
5-15mm/yrat95°E.
Meyeretal.
(1996)concludedthatsliprateseastof96°were4±2mm/yr,consistentwithdecreasingslipratealongtheATFandaccommodationoftheslipgradientbythrustingalongthesouthsideintheQaidamBasin(alsosuggestedbyPeltzeretal.
,1989;Chenetal.
,1999).
Meyeretal.
(1998)observedthatthethrustsbranchfromtheATFandarethuskinematicallylinked;andvanderWoerdetal.
(2001)showedthedecreaseinsliprateandgeometryofthefaultjunctionattheAltynTaghFault-TanghenanShanThrusttriplejunction.
1030ZackWashburn,JRamónArrowsmith,GuillaumeDupont-Nivet,WangXiaoFeng,ZhangYuQiaoandChenZhengleThisstudycoverstheportionoftheATFfrom89.
6°Eto92.
5°E.
ItiswithintheareacoveredbythehigherratesinPeltzeretal.
(1989)andspanstheMeriauxetal.
(1999)andRyersonetal.
(2003)sitenearthePingdingShan(18mm/yr).
Ourbestearthquakerecur-rencedatacomefromtheCamelsiteisatabout91.
75°E,sowemaybegintoobserveadecreaseinslipandthusrecurrencerateduetointerac-tionbetweentheATFandtheadjacentactivethrustsoftheQaidamBasin(forexample,notethattheactivenorthweststrikingYousha-ShanthrustfaultofChenetal.
(1999)approachestheATFneartheGobiling).
Althoughthere-sultsofRyersonetal.
(2003)andourinferredslipratefromnearthePingdingShanoverlap(fig.
10),the20±3mm/yrminimumrateat94°E(Meriauxetal.
,2003)impliesahigherearthquakerecurrenceratethanweinferredfortheXorxolSegment.
Thegeodeticinvestigationsarenotsuffi-cientlydensetoimagevariationsinslipratealongthecentralATF.
Giventheirgreatbreadth(100sofkmperpendiculartotheATF),nearfaultlockingandseismiccyclevariationinvelocitycannotexplaintheslipratediscrepancy(fig.
10;Thatcher,1990;Shenetal.
,2001).
Shenetal.
(2001)alsopointoutthatbecausethereisnomajorconvergenceparalleltotheATFonitssouthsideacrosstheQimenTagh1031PaleoseismologyoftheXorxolSegmentoftheCentralAltynTaghFault,Xinjiang,China848688909294969800.
511.
522.
533.
54Longitude(°E)ATFparallelsliprate(cm)Geetal.
(1992)Shenetal.
(2001)Bendicketal.
(2000)Meriauxetal.
(1999,2000)Peltzeretal.
(1989)Meriauxetal.
(2003)Ryersonetal.
(2003)ThisstudyRyersonetal.
(2003)Meyeretal.
(1996)Fig.
10.
InferredAltynTaghFault-parallelslipratesfromlong-termstudies(blue),geodesy(red),andearth-quakegeology(black)compiledfromtheliteratureandthisstudy.
Theeastwarddecreaseinsliprateindicatedbythelong-termstudiesisconsistentwiththeactivityofnumerousthrustfaultssplayingfromandtothesouthoftheATF(e.
g.
,Meyeretal.
,1996,1998;Tapponnieretal.
,2001;althoughnotconsistentwiththegeodeticresultsfromShenetal.
,2001).
Ingeneral,thegeodeticresultsshowlowerslipratesrelativetothelong-termstudies.
Ourearthquakegeologyresultsspanthetwoothermethodsinboththetimescaleofmeasurements(afewkyrversusdecadaland>10kyr)andtheresultinginferredrates.
MountainsatthesouthernmarginoftheQaidamBasin,theslipratealongtheATFmustnotdecreasesignificantlybetween80°and92°E.
Oneexplanationforthediscrepancyinrecur-rencetimesfromGPSsitevelocitiesandtheearthquakerecordandgeologyisthatamajorearthquakealongthecentralATFisoverdue.
AnotherintriguingexplanationisthatthelongtermloadingratealongtheATFhasdecreasedinthepast100kyrandstrainisnowaccommodat-edbyotherstructuresinCentralAsia.
Forexam-ple,thefaultsthatproducedfourM8earth-quakesinMongoliainthepastcenturyreleasedsignificantstraininanareaofrelativelylowerinferredlongtermdeformationrate(Baljinnyametal.
,1993).
Earthquakes,suchasthe1997Mw7.
6Manyiearthquakethatruptured170kmalongtheKunlunFault(Peltzeretal.
,1999)andthe2001Mw7.
8Kokoxiliearthquake(vanderWoerdetal.
,2002;Linetal.
,2002)mayindi-catethatthewesternportionoftheKunlunFaultandothercentralandnorthTibetanstrike-slipfaultsmaybemoreactivethanpreviouslythought.
Finally,shorteningwithintheQimenTaghandothercontractionalstructuresinNorthernTibetmaynowaccommodateahigherpercentageofIndia-Eurasiaconvergence(YinandHarrison,2000;Cowgill,personalcommu-nication).
7.
ConclusionsInthispaper,wehavecharacterizedthefaulttracegeometryandpaleoseismologyoftheCentralAltynTaghFault,emphasizingits230kmlongandremarkablystraightXorxolgeometricsegment.
Tracegeometry,timingofpaleoearthquakes,andinferencesofrupturemagnitudescontributetotheglobalcatalogueofstrike-slipfaultsystemsandprovideinforma-tionaboutitsgeodynamicroleintheIndia-Eurasiancollision.
ThetracegeometryofthecentralATFisdominatedbytheArkatengRestrainingBend,PingdingReleasingStep,andthenotablystraightXorxolSegment.
Thegeometricbound-ariesareprobableearthquakerupturebound-ariesbasedonsurfacerupturecontinuity,distri-butionofsmallgeomorphicoffsets,andhistor-icaldata.
MajorHolocenedeformationalongtheAltynTaghFaultSystemisconfinedtoanarrowzonelessthan3kmwide.
Thegeometricboundariesandtimingcon-straintsprobablyprecludeonelargeearthquakerupturingall3segments(270km).
Further-more,multipleearthquakeswithshorterrupturelengths(50km)ratherthancompleteruptureoftheXorxolSegmentbetterexplainthepaleo-seismicdata.
Wefound2-3earthquakesinthelast2-3kyr.
Whencoupledwithtypicala-mountsofslipperevent(5-10m),therecur-rencetimesaretentativelyconsistentwith1-2cm/yrsliprates.
Thisresultfavorsmodelsthatconsiderthebroaderdistributionofcollisionaldeformation,ratherthanthosewithnorthwardmotionofIndiaintoAsiaabsorbedalongafewfaultsboundingrigidblocks.
AcknowledgementsWethankP.
Molnar,R.
Bilham,P.
Tap-ponnier,M.
Hamburger,R.
Bürgmann,A.
Yin,S.
Graham,andJ.
Freymuellerforearlyreviewsofthisresearch,ourcolleaguesattheInstituteofGeomechanicsinBeijing,andS.
Holloway,M.
Baillie,S.
Selkirk,E.
Catlos,C.
Schrein,andS.
Zhangforhelpinthelab,field,andtheoffice.
ConstructivereviewsbyK.
OkamuraandF.
GaladiniandtheguidanceofDanielaPantostiforAnnalsofGeophysicswerealsoveryhelpful.
R.
Ryersonarrangedour14Cdat-ingatLawrenceLivermoreNationalLabora-tory/UniversityofCaliforniaCenterforAcce-leratorMassSpectrometrywherewewereguidedbyG.
Sietz,M.
Caffee,J.
Southon,andtheircolleagues.
G.
Hilleyprovidedimpor-tantadviceoninterpretationof14Cdata.
TheNorthernCaliforniaEarthquakeDataCenter(NCEDC),andthemembernetworksoftheCounciloftheNationalSeismicSystem(CNSS)providedtheseismicitydatashowninfig.
2.
WeareparticularlygratefultoAnYinforinvitingustojointheprojectandtoEricCowgillforintroducingustofieldworkinXinjiang.
ThisworkispartofacollaborativeresearcheffortsupportedbytheU.
S.
NationalScienceFoundationContinentalDynamicsProgram(grantEAR-9725780).
1032ZackWashburn,JRamónArrowsmith,GuillaumeDupont-Nivet,WangXiaoFeng,ZhangYuQiaoandChenZhengleREFERENCESAVOUAC,P.
J.
andP.
TAPPONNIER(1993):KinematicmodelofactivedeformationinCentralAsia,Geophys.
Res.
Lett.
,20,895-898.
BALJINNYAM,I.
,A.
BAYASGALAN,B.
A.
BORISOV,A.
CISTERNAS,M.
G.
DEM'YANOVICH,L.
GANBAATAR,V.
M.
KOCHETOV,R.
A.
KURUSHIN,P.
MOLNAR,H.
PHILIPandY.
Y.
VASHILOV(1993):RupturesofmajorearthquakesandactivedeformationinMongoliaanditssurround-ings,Geol.
Soc.
Am.
Mem.
,181,pp.
62.
BENDICK,R.
,R.
BILHAM,J.
FREYMUELLER,K.
LARSON,G.
PELTZERandG.
YIN(2000):GeodeticevidenceforalowsliprateontheAltynTaghFaultandconstraintsonthedeformationofAsia,Nature,404,69-72.
CHEN,W-P.
,C-Y.
CHENandJ.
L.
NBELEK(1999):Present-daydeformationoftheQaidambasinwithimplicationsforintra-continentaltectonics,Tectonophysics,305,165-181.
COWGILL,E.
,A.
YIN,X.
WANGandQ.
ZHANG(2000):IstheNorthAltynFaultpartofastrike-slipduplexalongtheAltynTaghFaultSystem,Geology,28,255-258.
COWGILL,E.
,A.
YIN,J.
R.
ARROWSMITH,W.
XIAOFENGandS.
ZHANG(2003):TheAkatoTaghbendalongtheAltynTaghFault,NWTibet,1.
Cenozoicstructure,smooth-ingbyvertical-axisrotation,andtheeffectoftopo-graphicstressesonborderlandfaulting,Geol.
Soc.
Am.
(inreview).
DEPOLO,C.
M.
,D.
CLARK,B.
SLEMMONSandA.
RAMELLI(1991):HistoricalsurfacefaultingintheWesternBasinandRangeProvince,WesternNorthAmerica:impli-cationsforfaultsegmentation,J.
Struct.
Geol.
,13,122-136.
ENGLAND,P.
C.
andP.
MOLNAR(1997):ThefieldofcrustalvelocityinAsiacalculatedfromQuaternaryratesofsliponfaults,Geophys.
J.
Int.
,130,551-582.
FORMAN,S.
L.
(1999):Infraredandredstimulatedlumines-cencedatingofLateQuaternarynearshoresedimentsfromSpitsbergen,Svalbard,Arctic,Antarct.
AlpineRes.
,31,34-49.
GE,S.
and34OTHERS(1992):ActiveAltunFaultZoneMonograph(StateSeismologicalBureau,SeismologicalPress,Beijing,China),1-319.
GRANT,L.
B.
andK.
SIEH(1994):PaleoseismicevidenceofclusteredearthquakesontheSanAndreasFaultintheCarrizoPlain,California,J.
Geophys.
Res.
,99,6819-6841.
HARRIS,R.
A.
andS.
M.
DAY(1993):Dynamicsandfaultinteraction:parallelstrike-slipfaults,J.
Geophys.
Res.
,98,4461-4472.
HARRIS,R.
A.
andS.
M.
DAY(1999):Dynamic3Dsimula-tionsofearthquakesonenéchelonfaults,Geophys.
Res.
Lett.
,206,2089-2092.
HILLEY,G.
E.
andJ.
J.
YOUNG(2003):Determiningevenmagnitudeandtimingfrompaleoseismicandgeomor-phicdataalongtheCentralSanAndreasFault,California,Bull.
Seismol.
Soc.
Am.
(inreview).
HOLT,W.
E.
,N.
CHAMOT-ROOKE,X.
LEPICHON,A.
J.
HAINES,B.
SHEN-TUandJ.
REN(2000):VelocityfieldinAsiainferredfromQuaternaryfaultslipratesandGlobalPositioningSystemobservations,J.
Geophys.
Res.
,105,19,185-19,209.
KLINGER,Y.
,J.
VANDERWOERD,P.
TAPPONNIER,X.
W.
XU,G.
KING,W.
B.
CHEN,W.
T.
MAandD.
BOWMAN(2003):DetailedstripmapoftheKokoxiliearthquakerupture(MW7.
8,14/11/01)fromspace,Geophys.
Res.
Abstr.
,5,08487.
KNUEPFER,P.
L.
K.
(1989):Implicationsofthecharacteris-ticsofend-pointsofhistoricalsurfacefaultrupturesforthenatureoffaultsegmentation,inFaultSegmentationandControlsofRuptureInitiationandTermination,editedbyD.
SCHWARTZandR.
SIBSON,U.
S.
Geol.
Surv.
Open-FileRep.
89-315,193-228.
LIN,A.
,B.
FU,J.
GUO,Q.
ZENG,G.
DANG,W.
HEandY.
ZHAO(2002):Co-seismicstrike-slipandrupturelengthproducebythe2001Ms8.
1CentralKunlunearth-quake,Science,296,2015-2017.
MRIAUX,A.
S.
,P.
TAPPONNIER,F.
J.
RYERSON,J.
VANDERWOERD,C.
LASSERRE,X.
XU,R.
FINKELandM.
CAFFEE(1999):Large-scalestrainpatterns,greatearthquakebreaks,andLatePleistoceneslip-ratealongtheAltynTaghFault(China),EuropeanUnionofGeosciencesConferenceAbstracts,EUG10,TerraAbstracts,11,p.
55.
MRIAUX,A.
S.
,F.
J.
RYERSON,P.
TAPPONNIER,J.
VANDERWOERD,R.
C.
FINKEL,M.
W.
CAFFEE,C.
LASSERRE,X.
XU,H.
LIANDZ.
XU(2000):FastextrusionoftheTibetPlateau:a3cm/yr,100kyrslip-rateontheAltynTaghFault,EosTrans.
Am.
Geophys.
Un.
,81(48),p.
1137.
MRIAUX,A.
S.
,P.
TAPPONNIER,F.
J.
RYERSON,X.
XU,J.
VANDERWOERD,G.
KING,R.
C.
FINKEL,LIHAIBING,M.
W.
CAFFEEandX.
ZHIQIN(2003):Post-glacialsliprateontheAksaysegmentoftheNorthernAltynTaghFault,derivedfromcosmogenicradionuclidedatingofmorphologicaloffsetfeatures,Geophys.
Res.
Abstr.
,5,08062.
MEYER,B.
,P.
TAPPONNIER,Y.
GAUDEMER,G.
PELTZER,G.
SHUNMINandC.
ZHITAI(1996):Rateofleft-lateralmovementalongtheeasternmostsegmentoftheAltynTaghFault,eastof96°E(China),Geophys.
J.
Int.
,124,29-44.
MEYER,B.
,P.
TAPPONNIER,L.
BOURJOT,F.
MTEVIER,Y.
GAUDEMER,G.
PELTZER,G.
SHUNMINandC.
ZHITAI(1998):CrustalthickeninginGansu-Qinghai,lithos-phericmantlesubduction,andoblique,strike-slipcon-trolledgrowthoftheTibetplateau,Geophys.
J.
Int.
,135,1-47.
MOLNAR,P.
,C.
BURCHFIEL,L.
K'UANGYIandZ.
ZIYUN(1987):GeomorphicevidenceforactivefaultingintheAltynTaghandNorthernTibetandqualitativeesti-matesofitscontributiontotheconvergenceofIndiaandEurasia,Geology,15,249-253.
PELTZER,G.
,P.
TAPPONNIERandR.
ARMIJO(1989):MagnitudeofLateQuaternaryleft-lateraldisplace-mentsalongthenorthedgeofTibet,Science,246,1285-1289.
PELTZER,G.
,F.
CRAMPandG.
KING(1999):EvidenceofnonlinearelasticityofthecrustfromtheMw7.
6Manyi(Tibet)earthquake,Science,286,272-276.
RAMSEY,C.
B.
(2000):OxCalv3.
5softwarethatusesstrati-graphicconstraintstotrimcalendarageprobabilitydis-tributions,http://www.
rlaha.
ox.
ac.
uk/oxcal/oxcal.
htm.
RYERSONF.
J.
,A-S.
MRIAUX,P.
TAPPONNIER,J.
VANDERWOERD,XUXIWEI,LIHAIBINGandXUZHIQIN(2003):NortheastwardsdecreaseintheLatePleistocene-Holoceneslip-rateandpropagationoftheAltynTaghFault(China),Geophys.
Res.
Abs.
,5,08070.
1033PaleoseismologyoftheXorxolSegmentoftheCentralAltynTaghFault,Xinjiang,ChinaSHEN,Z-K.
,M.
WANG,Y.
LI,D.
D.
JACKSON,A.
YIN,D.
DONGandP.
FANG(2001):CrustaldeformationalongtheAltynTaghFaultSystem,WesternChina,fromGPS,J.
Geophys.
Res.
,106,30,607-30,621.
STUIVER,M.
andH.
A.
POLACH(1977):Discussion,report-ingof14Cdata,Radiocarbon,19,355-363.
STUIVER,M.
,P.
J.
REIMERandE.
BARD(1998):INTCAL98radiocarbonagecalibration,24.
000-0calBP,Radio-carbon,40,1041-1081.
TAPPONNIER,P.
andP.
MOLNAR(1977):ActivefaultingandtectonicsinChina,J.
Geophys.
Res.
,82,2905-2930.
TAPPONNIER,P.
,X.
ZHIQIN,F.
ROGER,B.
MEYER,N.
ARNAUD,G.
WITTLINGERandY.
JINGSUI(2001):ObliquestepwiseriseandgrowthoftheTibetplateau,Science,294,1671-1677.
THATCHER,W.
(1990):Present-daycrustalmovementsandthemechanicsofcyclicdeformation,inTheSanAndreasFaultSystem,California,editedbyR.
E.
WALLACE,USGSProf.
Paper1515,189-205.
VANDERWOERD,J.
,F.
J.
RYERSON,P.
TAPPONNIER,Y.
GAUDEMER,R.
FINKEL,A.
S.
MERIAUX,M.
W.
CAFFEE,Z.
GUOGUANGandH.
QUNLU(1998):Holoceneleft-slipratedeterminedbycosmogenicsurfacedatingontheXidatanSegmentoftheKunlunFault(Qinghai,China),Geology,26,695-698.
VANDERWOERD,J.
,X.
XU,H.
LI,P.
TAPPONIER,B.
MEYER,F.
J.
RYERSON,A.
S.
MRIAUXandX.
ZHIQIN(2001):RapidactivethrustingalongthenorthwesternrangefrontoftheTangheNanShan(WesternGansu,China),J.
Geophys.
Res.
,106,30,475-30,504.
VANDERWOERD,J.
,A.
MRIAUX,Y.
KLINGER,F.
J.
RYERSON,Y.
GAUDEMERandP.
TAPPONNIER(2002):The14No-vember2001,Mw=7.
8KokoxiliearthquakeinNort-hernTibet(QuinghaiProvince,China),Seismol.
Res.
Lett.
,73,125-135.
WANG,Q.
,P-Z.
ZHANG,J.
T.
FREYMULLER,R.
BILHAM,K.
M.
LARSON,X.
LAI,X.
YOU,Z.
NIU,J.
WU,Y.
LI,J.
LIU,Z.
YANGandQ.
CHEN(2001):Present-daycrustaldefor-mationinChinaconstrainedbyGlobalPositioningSystemmeasurements,Science,294,574-577.
WARD,S.
N.
(1997):Dogtailsversusrainbows:syntheticearth-quakerupturemodelsasanaidininterpretinggeologicaldata,Bull.
Seismol.
Soc.
Am.
,87,1422-1441.
WASHBURN,Z.
(2001):QuaternarytectonicsandearthquakegeologyoftheCentralAltynTaghFault,Xinjiang,China:implicationsfortectonicprocessalongthenorthernmarginofTibet,M.
S.
Thesis,Tempe,ArizonaStateUniversity,pp.
179.
WASHBURN,Z.
,J.
R.
ARROWSMITH,S.
FORMAN,E.
COWGILL,X.
F.
WANG,Y.
ZHANGandC.
ZHENGLE(2001):LateHoloceneearthquakehistoryoftheCentralAltynTaghFault,China,Geology,29,1051-1054.
WELLS,D.
L.
andK.
COPPERSMITH(1994):Newempiricalrelationshipsamongmagnitude,rupturelength,rupturewidth,rupturearea,andsurfacedisplacement,Seismol.
Soc.
Am.
Bull.
,84,974-1002.
WESNOUSKY,S.
G.
(1989):Seismicityandthestructuralevo-lutionofstrike-slipfaults,inFaultSegmentationandControlsofRuptureInitiationandTermination,editedbyD.
SCHWARTZandR.
SIBSON,U.
S.
Geol.
Surv.
Open-FileRep.
89-315,409-431.
WESSEL,P.
andW.
H.
F.
SMITH(1995):NewversionoftheGenericMappingToolsreleased,Eos,Trans.
Am.
Geo-phys.
Un.
,76(33),p.
329.
YIN,A.
andM.
HARRISON(2000):GeologicevolutionoftheHimalayan-Tibetanorogen,Annu.
Rev.
EarthPlanet.
Sci.
,28,211-280.
1034ZackWashburn,JRamónArrowsmith,GuillaumeDupont-Nivet,WangXiaoFeng,ZhangYuQiaoandChenZhengle
- numseasonalwinds相关文档
- readseasonalwinds
- 改良seasonalwinds
- Julyseasonalwinds
- yearsseasonalwinds
- resultsseasonalwinds
- originalseasonalwinds
美国Cera 2核4G 20元/45天 香港CN2 E5 20M物理机服务器 150元 日本CN2 E5 20M物理机服务器 150元 提速啦
提速啦 成立于2012年，作为互联网老兵我们一直为用户提供 稳定 高速 高质量的产品。成立至今一直深受用户的喜爱 荣获 “2021年赣州安全大赛第三名” “2020创新企业入围奖” 等殊荣。目前我司在美国拥有4.6万G总内存云服务器资源，香港拥有2.2万G总内存云服务器资源，阿里云香港机房拥有8000G总内存云服务器资源，国内多地区拥有1.6万G总内存云服务器资源，绝非1 2台宿主机的小商家可比。...
raksmart：香港机房服务器实测评数据分享，告诉你raksmart服务器怎么样
raksmart作为一家老牌美国机房总是被很多人问到raksmart香港服务器怎么样、raksmart好不好？其实，这也好理解。香港服务器离大陆最近、理论上是不需要备案的服务器里面速度最快的，被过多关注也就在情理之中了。本着为大家趟雷就是本站的光荣这一理念，拿了一台raksmart的香港独立服务器，简单做个测评，分享下实测的数据，仅供参考！官方网站：https://www.raksmart.com...
gcorelabs：美国GPU服务器，8张RTX2080Ti，2*Silver-4214/256G内存/1T SSD/
gcorelabs提供美国阿什本数据中心的GPU服务器（显卡服务器），默认给8路RTX2080Ti，服务器网卡支持2*10Gbps（ANX），CPU为双路Silver-4214（24核48线程），256G内存，1Gbps独享带宽仅需150欧元、10bps带宽仅需600欧元，不限流量随便跑吧。 官方网站 ：https://gcorelabs.com/hosting/dedicated/gpu/ ...
seasonalwinds为你推荐
-
虚拟空间什么是网络虚拟空间？cm域名注册.Cm是什么域名 网址尾部是.CM的是哪里的网址？哪可以注册？免费国外空间免费国外空间ip代理地址ip代理有什么用？香港虚拟空间最好的香港虚拟主机是哪家？免备案虚拟空间免备案的虚拟主机空间，买了以后会强制备案不？100m虚拟主机100元虚拟主机重庆虚拟主机万网M3型虚拟主机怎么样？速度如何？虚拟主机mysql虚拟主机支持mysql数据库，还需要额外购买mysql吗？河南虚拟主机新网的虚拟主机怎么样？