古海洋学及古气候替代性指标.docx

古海洋学及古气候替代性指标.docx

《古海洋学及古气候替代性指标.docx》由会员分享，可在线阅读，更多相关《古海洋学及古气候替代性指标.docx（32页珍藏版）》请在冰豆网上搜索。

古海洋学及古气候替代性指标

Paleoceanography

Introduction

Paleoceanographyisthestudyofthehistoryoftheoceans.Itencompassesaspectsofoceanography,climatology,biology,chemistryandgeology.Themainsourcesofinformationarebiogenicandinorganicmarinesediments,aswellascorals.Biogenicsedimentincludesplanktonicandbenthicfossilswhereasinorganicsedimentincludesice-rafteddebrisanddust.Onland,paleo-shorelinesanderosionalfeaturesaswellasoutcroppingsofpaleomarinesedimentsaretheprincipalsourcesofproxydata.Glaciologicalrecordscanalsogiveindirectinformationaboutpaleoceans.Theocean'shighheatcapacityanditsabilitytotransportenergyandtosequesterandreleasegreenhousegasesgiveitanimportantroleinhelpingtodeterminethestateoftheplanet'sclimate.Thus,paleoceanographicresearchisalsointimatelylinkedtopaleoclimatology.

Methods

Thereconstructionofpaleoceancharacteristicsanddynamicsrequiresclimaticdetectivework.Itinvolvesthedatingandinterpretationofpaleoclimaticrecordsaswellasthedefinitionofphysicalanddynamicalconstraintswhichspecifypossiblecirculationpatternsandcharacteristics.

Reconstructions（ProxyData）

Directmeasurementsofthequantitiesofinteresttooceanographersextendonlyintotherelativelyrecentpastandinmostcasesdonotgofurtherbackthanthemid-nineteenthcentury.Tostudytheoceanduringperiodsforwhichtherearenodirectmeasurementsonemustrelyonindirectevidence.Historicaldocumentscanbeusedassourcesofdata.Shiplogbooksandsailingtimesacrossfrequentlytraveledrouteshaveprovidedestimatesofthedirectionsandstrengthsofpastprevailingwinds（Brazdiletal.,2005）.ThefrequencyandintensityofElNiñoeventssincethe1500shavebeenreconstructedbasedpartiallyonhistoricalaccountsoflargefloodsandcroplosses（QuinnandNeal,1987）.Thistypeofanalysisfurnishesqualitativedescriptionsofthepast.

Quantitativereconstructionsarepossiblebyproxy,whereaquantitywhichispreservedinanaturalarchiveandcanbemeasured,standsasasurrogateoftheparameterofinterest.Abasicrequirementisthattherelationshipbetweentheproxyparameterandthequantityofinteresthastobeknown.Whenthisisthecase,thehistoryoftheproxyvariablecanbeconvertedintothehistoryofthevariableofinterestbytheuseofmathematicalexpressionsofthetype:

（

（1））

whichstatethattheparameterofinterest,Int,isafunctionoftheproxyquantity,Prx.Thetindexreferstotime.Equationsofthiskindarecommonlycalledtransferfunctions.TheconfidencewithwhichIntcanbeestimatedwilldependonaseriesoffactors,startingwiththequalityoftheproxymeasurements.Alsorelevant,andacommonsourceofuncertainty,ishowwellfrepresentstherelationshipbetweenPrxandInt.

Inmostcases,transferfunctionsareobtainedempiricallybycomparingdirectlymeasuredvaluesofthequantityofinteresttoapertinentsetofproxydata.Apotentialsourceoferroristhatthefunctionobtainedbythisproceduremightnotbegeneral,butinfactcouldrepresentarelationshipbetweenPrxandIntthatispeculiartothedatasetsusedtogeneratef.Thisproblemcanbeminimizedbyexpandingthespatialandtemporalcoverageofthedatausedtoestablishthetransferfunction.Still,evenassumingthatfisaperfectrepresentationofhowtheproxyandthequantityofinterestareconnectedtoeachotherinthepresent,thereisnoguaranteethattherelationshipbetweenthemwasthesameinthepast.

Anothersourceoferrorcanbeeasilyunderstoodbyre-writingEquation

（1）sothatitexpressestheproxyquantityasafunctionofthevariableofinterest.Itisreasonabletoassume（andinmanycasesithasbeendemonstrated）thatIntisnottheonlyfactorcontrollingPrx,sothatinfactweendupwith:

（

（2））

whereE1,E2,…,Enrepresentenvironmentalparametersthatalsoinfluencetheproxyvariablebutareindependentofthequantityofinterest.AnimmediateconclusionisthatreconstructionsofIntbasedonPrxwillbe"contaminated"byotherfactorssothatpartofthevariabilityobservedintheproxyquantityisnotrelatedtochangesintheparameterofinterest.Comprehensiveanalysisoftherelationshipsbetweenproxiesandaseriesofobservedparameterscanoffersomeinsightintohowtoremovepartoftheundesiredinfluenceofotherfactorsfromthereconstruction.

Giventhecomplexityinvolvedindevelopingskillfultransferfunctionsaswellasinidentifyingandcorrectingforpotentialsourcesoferror,acommonstrategyistoreconstructthesameparameterofinterestusingdifferentproxies.Suchanalysesareknownasmulti-proxyreconstructions（FischerandWefer,1999）.

Typesofproxies:

Thesystematicuseofproxiesinquantitativereconstructionsofpastoceanicenvironmentsoriginatedinthesecondhalfofthetwentiethcentury.Sincethen,alargenumberofproxytechniqueshavebeenestablishedandmoreareconstantlybeingdeveloped.Proxiescanbegroupedinsixbroadcategories,basedonthetypeofdirectmeasurement（FischerandWefer,1999）.Thesearelistedbelow,togetherwithbriefdescriptionsofthemainvariablesofinterestassociatedwitheachproxy.Thefollowingchaptersonpaleotemperature,paleoproductivityandpaleocirculationpresentinmoredetailtheproxiesrelevanttoeachofthesefields.Comprehensivediscussionsofoceanographicproxiescanbefoundintheliterature（Bradley,1999;FischerandWefer,1999;Henderson,2002）.

∙Microfossilassemblages.Therelativeabundanceofplanktonicandbenthicspeciesofforaminifera,coccoliths,radiolaria,diatomsandotherorganismscanbeusedtoestimatepastoceantemperature,productivityandseaicedistribution.ThisproxytypewasusedfortheCLIMAPproject,whichproducedthefirstglobaldistributionofseasurfacetemperaturefortheLastGlacialMaximum（CLIMAP-ProjectMembers,1976）.

∙Stableisotopesarebasedontheratiobetweendifferentisotopesofanelement.Theratiosareusuallystandardizedbyareferencevalueandnamedaftertheheavierisotope.Theratiobetween18Oand16O,forexample,isrepresentedbyδ18O.Isotopereadingsareretrievedmainlyfromforaminiferaskeletons（tests）,organicmatterorothersources（e.g.,watermoleculesincontinentalicesheets）.Theamountof18Oincorporatedbyorganismslikeforaminiferaandcoralsincreasesastemperaturedecreases.Continentaliceisrelativelydepletedin18Ocomparedtoseawater.Thismakesδ18Oaproxyforbothtemperatureandtheextentofcontinentalicesheets.δ11BisusedasaproxyforpH.Productivity,nutrientconcentrationandpastcirculationcanbereconstructedfromδ15Nandδ13C（12Cistakenupwithslightpreferenceto13Cduringphotosynthesis）.Togetherwithmicrofossilassemblages,δ18Oandδ13Carethepaleoceanographicproxieswiththemostwidespreaduse.

∙Radiogenicisotopes.Thedifferentsolubilitiesofuraniumandtwoproductsofitsnaturallyoccurringdecay,thorium（Th）andprotactinium（Pa）canbeusedtoestimatetherateofdeepwaterflowandthefluxofparticlesfromthewatercolumntothesediments.Thisfluxcanalsobeusedasaproductivityestimate.14Cpreservedinorganicmatterisusedtoestimatetheagedifferencebetweennearsurfaceanddeepwatersandhence,ventilationrates.

∙Biogeniccompounds.Theconcentrationsofsomecompounds,mainlyorganiccarbon,calciumcarbonateandopal,areusedasestimatesofpastproductivity.Calciumcarbonateisalsoanindicatorofthecalcitecompensationdepth.Alkenonesarelongchainedorganicmoleculesresistanttodegradation.Thealkenonesproducedbysomecoccolithophorscanhavetwoorthreedoublebondsintheirstructure.Theratiobetweenmoleculeswithtwoandthreedoublebondsreflectthetemperatureatthetimeofsynthesis.

∙Elements.Theconcentrationsandratiosofcertainelementsinthesediment,organicremains,testsandcoralsarealsousedasproxies.Theratiosofstrontiumtocalcium（Sr/Ca）andmagnesiumtocalcium（Mg/Ca）inbiologicallyprecipitatedmarinecarbonatesaretemperaturedependent.Thecadmiumtocalcium（Cd/Ca）ratioisusedfornutrientreconstructions.BariumconcentrationandtheBa/Caratioareproxiesforproductivityandalkalinity,respectively.

∙Sedimentology.Grainsizedistributioncanprovidequalitativeinformationaboutbottomcurrentspeedsandactasanindicatoroficerafteddebris.Informationaboutpasttidescanbeinferredfromlayeredsedimentscalledrhythmites.Themineralogyofthesedimentscanbeusedtoestablishsourceareasanddirectionoftransportforbothwaterandwindbornesediments.

Conspicuouslyabsentfromthequantitiesofinterestlistedaboveissalinity,afundamentalparameterwhichinfluencesmanyaspectsoftheoceanenvironment.Thereis,atthemoment,noindependentproxyforthisquantity.Assalinityalsoinfluencesδ18O,anindirectmeasurementcanbeobtainedbyusingindependentestimatesoftemperature（alkenones,forexample）toremovethetemperaturesignalfromexistingδ18Oseries.Attemptshavealsobeenmadetoinfersalinityfrommicrofossilassemblages.Unfortunately,bothapproachesgenerateerrorsof∼1psu,verylargecomparedtotherangeofsalinityvariabilityintheoceans（FischerandWefer,1999）.

Reconstructions（Models）

Thetheoreticalapproachtopaleoceanographyusesquantitativeoceanandclimatemodelstoreconstructpaleoconditionsoftheoceanandinterpretobservations.Anumberofdifferentoceanandcoupledclimatemodelshavebeenusedinthepastandwewillgiveaveryshortoverviewofthehierarchyanduseofthesemodelsinthebroadresearchareaofpaleoceanography.