MonitoringtheAtlanticmeridionaloverturningcirculationDarrenRayner,Joel - PDF document

MonitoringtheAtlanticmeridionaloverturningcirculationDarrenRayner,JoelJ.-M.Hirschi,TorstenKanzow,WilliamE.Johns,PaulG.WrightEleanorFrajka-Williams,HarryL.Bryden Correspondingauthor.E-mailaddress:dr400@noc.soton.ac.uk(D.Rayner). Deep-SeaResearchII58(2011)1744…1753 intheearlywarmingperiodattheendoftheiceage…wouldleadtotherapidcoolingofthenorthernAtlantic,andswitchingontheMOCwouldleadtotherapidwarmingfoundinpaleorecords.TheEarthsclimatehasbeenremarkablystableforthepast8000yearsandthisstableclimatehascoincidedwith,andarguablycontributedto,thedevelopmentofmoderncivilisationfromprehistoricnomadictribestomodernindustrialsociety.Wearenowperforminga“niteamplitudeperturbationexperimentontheclimatesystembydoublingtheamountofCOintheatmosphere.Willweperturbtheclimateoutofitsstablestate?Becauseofitsroleinpastrapidclimatechangeevents,theAtlanticmeridionaloverturningcirculationisafocusforunder-standingpresentandfutureclimatechanges.Coupledocean-atmosphereclimatemodelsareinagreementthattheAtlanticMOCwilldecreaseasCObuildsupintheatmosphere(Cubaschetal.,2001;IPCC,2007).ElevencoupledmodelsrunundergreenhousegasforcingwhereCOevery70years,allshowadecreaseintheAtlanticMOCby10…50%over140years(Gregoryetal.,2005).TheslowdownintheMOCisgradual,however,nomodelexhibitsasuddenchangeintheoverturningcirculation.Ontheotherhand,experimentswithcoupledmodelsusinga“xedCOlevel,wheredeepwaterformationisturnedoffbyaddingfreshwatertothesurfacewatersofthenorthernAtlantic,exhibitarapidshutdownoftheMOCresultingin4…8Creductionsinairtemperaturesoverthenorth-ernAtlanticandnorthwestEurope(VellingaandWood,2002Thus,ourbestmodelspredictaweakeningoftheAtlanticMOCunderanincreaseinCOintheatmosphereandsuggestthatiftheMOCabruptlyshutsdowntherewouldbeseverecoolingoverthenorthernAtlantic.ThereisconfusionabouthowtheNorthAtlanticcirculationischanging.SomeoceanobservationssuggestaslowdownoftheMOCbyasmuchas30%overthelast50yearswiththechangeinstructuresothatlessNorthAtlanticDeepWater”owssouthward,andmoreupperwatersarecirculatedinthesubtropicalgyreBrydenetal.,2005).ApossibleweakeningoftheMOCissupportedbyobservationsofacessationintheformationoflowerNorthAtlanticDeepWater(ØsterhusandGammelsrod,),adecreaseintheamountofcolddenseover”owwatersthroughtheFaroeBankChannel(Hansenetal.,2001),reducednorthward”owofupperwatersthroughthesubpolargyreLherminieretal.,2006),andafresheningofnorthernAtlanticsurfaceanddeepwaters(Curryetal.,2003,Dicksonetal.,2002HoweverotherobservationsdonotsupportaweakeningMOC.Olsenetal.(2008)foundthattherewasnotrendintheFaroeBankChannelover”ow,andHollidayetal.(2008)reportareversalofthepreviouslyobservedfresheningtrendofthenortheastNorthAtlanticandNordicSeaspossiblycausedbyareducedcontributionofwaterfromthesubpolargyre(etal.,2005;HakkinenandRhines,2004)orbysurfacewatersfromtheGulfStreamreachingtheRockallTroughthroughthesubtropicalgyre(akkinenandRhines,2009).Formorediscus-sionontheevidenceforthechangingMOC,seeCunninghamandMarsh(2010)NorthAtlanticsurfacetemperaturesareincreasing,evenfasterthantemperaturesintheNorthPaci“c,forexample.HowcanwereconcilewarmerseasurfacetemperatureswithaslowdownintheMOC?Analysesofnaturalvariabilityincoupledclimatemodelsrunover1000-yearperiodsindicatethatwarmerAtlantictemperaturesaresigni“cantlycorrelatedwithastrongerMOC.WarmerAtlanticseasurfacetemperatures(SSTs)couldbeadirectresultofradiativeheatingduetothegreenhouseeffectassociatedwithincreasedCOintheatmosphere(Cubaschetal.,2001;IPCC,).However,warmerSSTscouldalsobeassociatedwithhighNorthAtlanticOscillation(NAO)andAtlanticMultidecadalOscil-lation(AMO)indices.TheNAOindexmeasuresthestrengthofthewesterlywindsoverthenorthernAtlantic.Thisindexhasincreasedsubstantiallyfromthe1970stothe1990sandthelowfrequencyresponsetosustainedNAOforcingisexpectedtobewarmerSSTs(Visbecketal.,2003).TheAMOisanaturaloscillationof50…100-yearperiodintheAtlantic,whichcombinesastrongMOCandwarmSSTs(DelworthandMann,2000).ThewarmerSSTsinthepast25yearsarethentakenasevidencethattheAtlanticMOCispresentlyinastrongphase(Knightetal.,).Forcomparison,the11coupledclimatemodelrunswithdoublingatmosphericCOover70yearsallshowslowingoftheAtlanticMOCbutanoverallwarmingofAtlantictemperature;thecoolingassociatedwithagradualdecreaseoftheMOCissmallerthanthewarmingassociatedwithdirectradiativeforcing,soeverywhereintheAtlanticSSTsincreasewithtime(Gregoryetal.,).Thus,therearemanyviewsonthereasonforanincreaseinSSTsintheNorthAtlantic,butitispossiblethatacoolingtrendassociatedwithaslowdownintheAtlanticMOCisbeingmaskedbydirectradiativeheatingduetoincreasedatmosphericCOHowever,eveninthiscase,itisnotyetunderstoodwhytheNorthAtlanticregionwarmsfasterthan,e.g.theNorthPaci“c.ThereisaclearneedforobservationsoftheAtlanticMOCandhowitischangingovertime.TheMOCiscentraltoourunder-standingofthepresentclimateandhowitmightchangeinthefuture.Thus,itisessentialtoestablishabaselinemeasureoftheMOCstrengthanditsseasonaltointerannualvariabilitytoputwide-rangingandlongertimeseriesofAtlanticobservationsintoanoverallcontextofAtlantic(andglobal)climatechange.ComparisonofatimeseriesoftheobservedstrengthoftheMOCwithSSTs,NAOindexandAMOindexwouldclarifytherelation-shipbetweenpresentlydistinctphenomena.Finally,wewanttoknowifthereisatrendinthestrengthoftheMOCandwhetheritisasigni“cantdecrease(orincrease)intheoverturningInthispaperwedescribeanarraydeployedat26.5NtocontinuouslymonitortheMOC,withadescriptionoftheobserva-tionalstrategyandthepre-deploymentdesignvalidationwithnumericalmodels(Section2).ThearrayhasevolvedovertimeandthesechangesaresummarisedinSection2.5.Asummaryofthemainscienti“c“ndingsisgiveninSection3andinSection4weprovideadiscussionoftheseresultsandplacetheminthewidercontextofmeasuringtheMOC.2.MonitoringprogrammeWeinitiatedamonitoringprojectin2004withtimeseriesmeasurementsofthebasin-scaleoverturningcirculationatFig.1Marotzkeetal.,2002).Theprojectwasfundedfrom2004to2008intheframeworkoftheRapidClimateChange(RAPID)thematicprogrammeoftheNaturalEnvironmentResearchCouncil(NERC),theNationalScienceFoundation(NSF)MeridionalCirculationandHeatFluxArray(MOCHA)andbytheNOAAOf“ceofClimateObservations.FundinghassincebeenextendedbyNERC,NSFandNOAAtill2014aspartoftheRAPID-WATCH(WilltheAtlanticThermohalineCirculationHalt?)pro-grammetoprovideadecadelongtimeseriesofmeasurements.Themainaimoftheprojectisthedevelopmentofanoperational,cost-ef“cientobservationsystemthatcontinuouslymonitorsthestrengthandverticalstructureoftheAtlanticMOCat26.52.1.RationaleforobservingtheMOCat26.5Thelatitudeof26.5Nforthemonitoringarraywaschosenforthreereasons:itisclosetothepeaknorthwardheattransport;itisthelatitudeoffourmodernhydrographicsections(“veinclud-ingonein2004immediatelyafterthearraywasdeployed);D.Rayneretal./Deep-SeaResearchII58(2011)1744…1753 thewesternboundarycurrent(”owthroughtheFloridaStrait)hasalonghistoryofmeasurementsandcanbemeasuredrelativelystraightforwardlybycableandregularcalibrationcruises(Larsen,1992,BaringerandLarsen,2001).Additionally,Nhastheadvantageofhavingcomparativelysteepbasinboundariescomparedtootherlatitudes.2.2.ObservationalstrategyTheAtlanticMOCisdecomposedintothreecomponentsthatcanbemeasuredseparately:(1)TheGulfStreamtransportthroughtheFloridaStraits(BaringerandLarsen,2001),(2)thenear-surfacewinddrivenEkmancomponentand(3)themid-ocean”owbetweentheBahamasandAfrica(Fig.2At26.5NtheGulfStream”owsthroughthenarrow(80km),shallow(800m)FloridaStraitsbetweenFloridaandtheBahamas.TheGulfStreamtransporthasbeenestimatedformorethan25yearsbyrecordingtheinducedvoltageonsubmergedtelephonecablesbetweenWestPalmBeachandGrandBahamaIslandBaringerandLarsen,2001).Aconductor(inthiscaseseawater)passingthroughtheEarthsmagnetic“eldwillinduceanelectric“eldperpendiculartothemotionoftheconductor.Thisinducedelectric“eldvariesinrelationtotherateof”owoftheconductorandcanbedetectedbyvoltagechangesonthetelephonecablerelativetoanEarthground.ThevoltagevariationsarecalibratedwithdirectestimatesoftheGulfStreamtransportfromthevelocitypro“lesfromLoweredAcousticDopplerCurrentPro“ler(LADCP)sectionsandmeanverticalcurrentpro“lescollectedbyDropsonde”oats(Larsen,1992)togive.DailymeantransportestimatesareusedinthecalculationoftheMOC.isderivedfromsatellitebasedmeasurementsofthewindstress,andisintegratedfromwesttoeastacrosstheAtlanticaccordingto isthezonalcomponentofthewindstress,iswaterdensityandtheCoriolisparameter.isinferredfromQuickSCATscatterometermeasurementsoftheroughnessoftheseasurface(Grafetal.,1998).Windstressestimatesareavailableatadailyresolution.Kanzowetal.(2010)estimatethepossible4-yearmeanbiasintobe0.5Sv.Wedecomposeintothreecomponentsthatareobservedbyatrans-AtlanticarrayofmooringsbetweentheBahamasandthecoastofAfrica.istheinternalgeostrophic”ow,isthe Fig.1.Toppanel:BathymetryoftheNorthAtlanticsubtropicalgyreregion.TheredlinefromBahamastoAfricarepresentsthetrackofthe2004hydrographicCunningham,2005b)whosetemperaturedistributionisshowninthethreepanelsbelow.Themiddlelefthandpanelshowsthetemperaturedistributionofthenorthward”owingGulfStreamintheFloridaStrait.Themiddlepanelshowsthetemperaturedistributionoftheupper1000mtheisothermsslopinguptotheeastareindicativeofthesouthward”owofthermoclinewater.Thelowerpanelshowsthetemperaturedistributionofthelower5000mofthewatercolumn,whichisgenerallymovingsouthward.MagentabarsbelowthetoppaneldenotetheregionsoftheRAPIDmooringssub-arrays.(Forinterpretationofthereferencestocolorinthis“gurelegend,thereaderisreferredtothewebversionofthisarticle.) Fig.2.SchematicoftheMOCmonitoringarrayat26N.TheMOCisdecomposedintothreecomponents:(1)GulfStreamtransportthroughtheFloridaStraits(redarrow),(2)thenear-surfacewinddrivenEkmantransport(greenarrow)arisingfromthezonalwindstressand(3)geostrophic(thermalwind)contribu-(lightbluearrows)calculatedbetweenadjacentpairsofmoorings(verticallines).Yellowarrowsindicateaspatiallyconstantvelocitycorrectionthatensuresmassbalanceacrossthesection.(Forinterpretationofthereferencestocolorinthis“gurelegend,thereaderisreferredtothewebversionofthisarticle.)D.Rayneretal./Deep-SeaResearchII58(2011)1744…1753 zonallyintegratedreference-levelcontributionofthegeostrophic”owandisthemeridionaltransportoverthecontinentalshelfwestofthemooringsWB1andWB2referredtoasthewesternboundarywedge.iscalculatedfromthedifferenceinfulldepthdensitypro“lesoneithersideoftheAtlanticbasinwiththepro“lesderivedfromtemperature,conductivityandpressuremeasure-mentsatdiscretelevelshereafterreferredtoasgeostrophicmoorings.InthewestthecontinentalshelfslopeissteepandthetallgeostrophicmooringWB2isplacedclosetotheBahamasescarpment.Intheeasttheshelfslopeismuchmoregradualsoaseriesofshortermooringsisdistributeduptheslopetominimisetheeffectsofbottomtriangleswherehorizontalinterpolationatdepthwouldmisssigni“cantsectionsoftheocean(e.g.WhitworthandPeterson,1985).Thisseriesofmooringsismergedtoproduceasingleverticaldensitypro“leattheeasternbound-ary.Theeasternandwesterndensitypro“lesgivethegeostrophicinternaltransport()relativetothereferencelevel()usingistheEarthsaccelerationofgravity,thedensityintheeastandthedensityinthewest.Weuseareferencelevelof4740m.Bottompressurerecordersareusedtocomputethetime-varyingreferencelevelmeridionalgeostrophicvelocities.Fromthesetheverticallyintegratedexternaltransport”uctuation)integratedbetweentwosites+1canbeobtainedisthewatercolumnheightandisthebottompressureateachsite.Asisdifferentforeachsitethedeeperofthetwoisadjustedtotheshalloweronebyremovingthedensity”uctuationsbelowthedepthoftheshallowersitefoundfromthegeostrophicmooringnearesttothedeepersite.Thezonaltrans-portintegralisfoundbysummingthecontributionsbetweenadjacentmooringssothatThewesternboundarywedgecomponent(isobtainedfrominterpolatingandintegratingdirectvelocitymeasurementsfromcurrentmetresonmooringsinshoreofthewesternmostdensitymooring(Johnsetal.,2007).Currentsinthisregionconsistoftheupper-oceannorthwardAntillescurrentandtheupperandinshorefractionsoftheDeepWesternBoundaryCurrent(DWBC).Ifweassumethereisnonetmasstransportacross26.5Nthenwecansimplifythiswiththeverticalintegralofequallingtheverticalintegralofthesumof.Thismethoddoesnotandinsteadaddsabarotropictransportpro“letomaintainmassbalanceateachtimestep.Kanzowetal.(2007)showthevalidityofthisapproach.ThetimeseriesoftheMOCde“nedasthemaximumnorth-wardupper-oceantransportisproducedbysummingtransports,whereistheuppermid-oceantransportfoundbyverticallyintegratingdowntothedeepestnorth-wardvelocity(1100m)oneachday.2.3.ArraydesigntestsinnumericalmodelsNumericaloceanmodelsareavaluabletoolfortestingtheobservationalstrategy.PriortothedeploymentoftheRAPID-MOCarraywehaveusedmodelstocomparetheMOCcalculatedfromthefullmeridionalvelocitiestoreconstructionsbasedonsub-sampledmodeldata(Hirschietal.,2003,Baehretal.,2004,HirschiandMarotzke,2007).WeobtainthereconstructionsbysummingthemodelledtransportthroughtheFloridaStraits,theEkmantransportcalculatedfromthezonalwindstressusedtoforcethemodelatthesurfaceandthegeostrophictransportobtainedfrommooringsdeployedinthenumericalmodel.Additionally,weaddaspatiallyconstantvelocitycorrectionsothatthereisnonetmasstransportacrossthelongitude-depthsectionat26.5Fig.2).ThehighlevelofagreementfoundbetweenthesimulatedMOCanditsestimatebasedontheproposedobservingstrategyprovidesa“rstindicationforthesoundnessoftheapproach(Fig.3Hirschietal.,2003,Baehretal.,2004,HirschiandMarotzke,20072.4.ImplementationanddeploymentofthearrayThemooringarrayconsistsofthreesub-arrays:oneatthewesternboundary(eastoftheBahamas),oneattheeasternboundary(westofMorocco)andonewithmooringsoneithersideoftheMid-AtlanticRidge.Thewesternboundaryandeasternboundarymooringsprovideendpointdensitypro“lesusedduringthecalculationoftheocean-widezonallyintegratedgeostrophic”ow.TheMid-AtlanticRidgemooringsallowthecontributiontotheMOCfromtheeasternandwesternbasinstobedistinguished.Amooringtoobtainaverticaldensitypro“letimeseriescomprisesaseriesofself-loggingconductivity,temperatureanddepthinstruments(CTDs)verticallydistributedonananchoredmooringwiresupportedbydistributedbuoyancy.Verticalinstru-mentresolutionincreasesatshallowerdepthswherethehigherverticaldensitygradientrequiresmorereferencepointsforanaccurateinterpolationofthedensitypro“le.Theinstrumentsaretypicallysettorecorddataat15…30minintervalswithdatasubsequentlylow-pass-“lteredtoremovetides.MooringsareservicedannuallywiththewesternboundaryservicedinSpringandtheeasternboundaryandMid-AtlanticRidgesub-arraysservicedinAutumn. Fig.3.TestingtheMOCmonitoringstrategyinaneddy-permittingoceanmodelWebb1996).Thehorizontalresolutionis0.25inbothlongitudeandlatitude.Top:placementofmoorings(verticallines)andareawherevelocitycanbecalculatedbasedonzonaldensitydifferences(bluearea).Bottom:MOCatNand1100mdepth(blue),reconstruction(red,seemaintext)andEkmancontribution(green).(Forinterpretationofthereferencestocolorinthis“gurelegend,thereaderisreferredtothewebversionofthisarticle.)D.Rayneretal./Deep-SeaResearchII58(2011)1744…1753 Thewesternboundarysub-arrayisthemostimportant:thelargest”uctuationsintheAtlanticMOCoccurherecomparedtotherestoftheoceanbasin.OurkeymooringWB2,measuringbetween50and3800mdepth,isdeployedascloseaspossible3km)tothewallofthecontinentalshelf.WB3(50…4800m;27kmoffshorefromWB2)andWB1(50…1400m;7kminshorefromWB2)canbeusedasbackupstoprovidethedensitypro“leifWB2islost.WB1,WB0andWBAalsousecurrentmeterstodirectlymeasuretheDWBCandtheshallowerAntillescurrent,allowingthe”owinshoreofthegeostrophicarraytobemea-sured.WB4andWB5aredeployedoffshorefromtheprincipalmooringstomonitortheoffshoreextentoftheDWBC,thuscapturingthermalwindshearacrosstheentireboundarycurrent.Intheeast,tominimiseleakagethroughbottomtriangles,aseriesofshortermoorings(EBH1…EBH5)weredeployeduptheslope.AsthearrayevolvedthisserieswasextendedwithEBHthedeeperendandaseriesofstillsmallermini-moorings,EBM1…EBM7,addedattheinshoreendtoreduceriskofdatalossthrough“shingactivity.Theseriesofmooringsintheeastismergedtocreateasinglepro“leasthecounterparttoWB2.ThisisachangefromourinitialstrategywherewehadthetallmooringEB2deployedindeepwithitslocationchosenasacompromisebetweenthedesireforfullwaterdepthandthenearnesstotheshelfbreak.ThecontributiontoMOCvariabilityfromtheeasternandwesternbasinscanbedistinguishedbytheMid-AtlanticRidgemoorings.MAR1(upto50mdepth)andMAR2(upto1100m)aredeployedonthewestern”ankoftheridge,withMAR3(uptotheridgecrestat2500m)andMAR4(upto50m)initiallydeployedontheeastern”ank.2.5.EvolutionofthemooringarrayThearrayas“rstdeployedin2004consistedof22mooringsFig.4)withtheprimarygeostrophicmooringsbeingWB2inthewestandEB2intheeast.EB3wasdeployed10kmfromEB2asabackup.ThebackuptoWB2wasWB3,24kmfurtheroffshore.Theverticalresolutionofdensitymeasurementswas14discretelevelsatthe3900mdeepWB2siteand13discretelevelsatthe3500mdeepEB2site.Thearraycon“gurationhasbeenprogres-sivelymodi“edduringsubsequentdeployments,withthecurrentcon“gurationshowninFig.5WB2isstillourprimarydensitymooringinthewestbuttheinstrumentverticalresolutionhasbeenincreasedto22(whenmergingtheupper1400mwithWB1)toallowbetterinterpola-tionofthedensitypro“le.AttheeasternmarginEB2wasrelocatedoffshoretothesiteofEB1in5100mdepthfollowingdamagetothemooringduringthe“rstyearsdeploymentwasextendedto50mdepthtoactasthebackupandEB3wasremovedfromthearray.SubsequentlytheworkbyKanzowetal.demonstratedtheimportanceofthecontinentalsloperegionforcapturingseasonalvariabilityintheMOCsothefocusfortheeasternboundarydensitypro“leisnowtheseriesofshortmooringsthatstepsuptheslope.AssuchtheEB2siteislessimportantandsotheEB1backupmooringisnotrequired.Theseriesofshortermooringshaschangedslightlyasthearrayhasevolvedtotrytominimiseriskoflossthrough“shingactivityonthecontinentalslope.Thecurrentdesignhasaseriesofmini-moorings(EBM1,4,5and6)attheinshoreend,whicheachconsistofasingleinstrument,therebyspreadingtheriskoflosingalloftheshallowdatarecords.Experiencehasshownthat Fig.4.SchematicofthethreeRAPIDmooringsub-arraysasdeployedinMarchandApril2004(Cunningham,2005a).Mooringsaretheverticalredlinesandinstrumentssymbolsarede“nedinthekeyontherighthandpanel(CTDconductivity,temperature,depthrecorder;currentmeterdirectvelocitymeasuringinstrument;bottompressurerecordermeasuringthehydrostaticweightofwater,ADCPanacousticDopplercurrentpro“lerandMMPlimitisthepro“lingrangeofapro“lingselfpropelledCTD).Distributionofpotentialtemperaturewasobtainedonatrans-Atlantichydrographictransectin2004followingthedeploymentofthemooringarraysCunningham,2005b).(Forinterpretationofthereferencestocolorinthis“gurelegend,thereaderisreferredtothewebversionofthisarticle.) Fig.5.Fig.4butformooringdeploymentsfrom2009to2010.D.Rayneretal./Deep-SeaResearchII58(2011)1744…1753 theMid-AtlanticRidgemooringsarerelativelysafesothenumberofgeostrophicmooringsdeployedherehasbeenreducedfromfourtothree,withMAR4beingremoved.Thepressuregradientacrosstheridgeismonitoredbythemooringspro“linguptotheridgecrest,withMAR1providingtheupperwatercolumnpro“leforbothsidesoftheridge.Inthewesternboundarysub-arrayWBH1andWBH2wereremovedfromthearraywithWBH2subsequentlybeingrein-statedbutwithadifferentdesigntoincludecurrentmetersforbetterhorizontalinterpolationofdirectvelocitymeasurementsoftheDeepWesternBoundaryCurrent.Thebottompressurerecorders(BPRs)deployedinthe“rstyearwereattachedtodropoffmechanismstothebottomofthemooringsbymagnesiumbolts.WhentheseboltscorrodeafteracoupleofhourstheBPRsaredroppedtotheseabedtodecouplethemfrommooringmotion.TheBPRremainedattachedtothemooringbyashortlengthofropesothatwhenthemainmooringwasrecoveredtheBPRwasrecoveredtoo.Duetothelargeandsomewhatunpredictabledriftthatpressuresensorscanexhibitthe1-yeartimeseriesisoftennotenoughtoremovethedriftsatisfactorily.TheBPRswereremovedfromthemooringsanddeployedontheirowncustommoorings…termedlanders…thatmounttheBPRonastableframeontheseabed.Thesearenowdeployedfor2yearsatatimewithoverlappingrecordsof1yearsothatthesecondhalfoftherecord(whichislessaffectedbydrift)canbeusedforthecalculationsofAnotherchangethathastakenplaceinthearraydesignisthedeploymentofmooringsMAR0onthewestern”ankoftheMid-AtlanticRidgeandWB6640kmoffshoreoftheBahamasinthewesternboundarysub-array.Thesemooringsarebothatadepthof5500mandhavebeendeployedtostudythecontributiontotheMOCvariabilityfromAntarcticBottomWater.2.6.TheuseofglidersinthearrayTwoautonomousglidermissionshavebeenundertakentoassessthecontributionthatautonomousgliderscouldmakeinmonitoringtheMOC,withaspeci“cfocusontheiruseasasubstituteformooringsattheeasternboundary.ThispartoftheRAPIDarrayhassufferedlossesofinstrumentsanddataduetosuspected“shingactivityonthecontinentalslope.Furthermore,the“ndingsofKanzowetal.(2010)meanthatthedatafromthisareaaremoreimportantthanthe“rstthought.Itisexpectedthatgliderswillbelesssusceptibletolossby“shing(inparticulartrawling)thanthemooredinstrumentsandhenceimprovedatareturnfromthisregion.Anotheradvantageofglidersisthatdatamayberetrievedinreal-timeviairidiumsatellitecommunications,thusfurtherreducingtheriskofdataloss.Theseglidermissionstookplacebetween15September…24November2008and21May…21July2009,betweentheCanaryIslandsandthecoastofMorocco.The“ndingsarebeingpreparedforasubsequentpaper(Smeedetal.,20103.ResultsInthefollowingwearesummarizingthemostimportantscienti“c“ndingsofthe“rst4yearsofcontinuousMOCobserva-tionsat26.5NintheAtlantic.PriortoRAPIDthevalidityoftheRAPIDmonitoringapproachfortheMOCwasbasedontestsperformedwithnumericaloceanmodels.OnemainassumptionthatwasmadeisthatthesumofandINTiscompensatedbyazonallyconstant,barotropic”owacrossthesection.Thenet(toptobottomintegrated)meridional”owacross26.5Nshouldbesmallbecausethereisonlyasmall(1Sv,1Sv)net”owthroughtheAtlanticduetoPaci“ctoAtlantic”owthroughtheBeringStraitandanetinputoffreshwaterfromtheatmosphere,riversandicenorthwardof26.5N.Thus,thedifferenttransportcomponentsweobserve(Fig.6A)shouldcompensateforeachother,suchthatanoverallmassbalanceisachieved.Kanzowetal.(2007)demonstratethatthismassbalanceexistsatperiodslongerthan10daysFig.6B),withthesumofGulfStreamandEkmantransports”uctuatinginanti-phasewiththegeostrophic”owbetweentheBahamasandtheAfricancoast.Kanzowetal.(2007)interpretthecompensationbetweenEXTandINTasobservationalevidencethatthemonitoringapproachtakenisvalidandthattheRAPIDsystemisworkingreliablysinceEXTisequivalenttothesimplecompensationassumedinournumericaltests.Basedontheabovetransportobservations,ayear-longtime-seriesofthestrengthoftheMOCwasderivedbyetal.(2007),de“nedasthemaximumnorthwardupperoceantransportforeachdaybetweenApril2004andMarch2005.The”owisfoundtobenorthwardbetweentheseasurfaceandroughlythe1100mdepthlevel…asaconsequenceofthenorthward”owoftheGulfStream,theAntillesCurrentandtheEkman”ow…andiscompensatedbyasouthward”owbelowthat,concentratedmostlywithintheDeepWesternBoundaryCurrent(Johnsetal.,2007),astreamthat”owsalongthecontinentalslopeoftheAmericasandexportstheNorthAtlanticDeepWatersintotheotheroceanbasins.Aremarkablefeatureemergingfromthe“rstdeploymentyearisthelargevariabilityfoundforthemaximumMOCat26.5evenonsubannualWehavenowextendedtheMOCtimeseriesto4yearsfrom2April2004to10April2008.At26.5NthestrengthoftheMOC(10-daylow-pass“ltered)(Fig.7)hasameanof18.7Svandvariesby4.8Sv(onestandarddeviation)overthe4-yearperiod Fig.6.Toppanel:”uctuationsofverticallyintegratedinternal(,red),external,blue),westernboundarywedge(,orange),Ekman(,green)andGulfStream(,magenta)transports.Thetransportisobtainedfrommeasure-mentsofthebottompressureandisequivalenttothedepth-independenttransportcorrectionusedinthenumericaltests(seeFig.2).Thereisatwo-monthgapinbetween31/08/2004and29/09/2004.Alltimeserieshavebeen2-daylow-pass“lteredandsub-sampledonahalf-dailygrid.Theinitialsamplingrateswere15minfortheunderlyingdensityandcurrentmeasurementsand10minforbottompressure.Bottompanel:15-daylow-pass“ltered”uctuationsofverticallyintegratedmid-ocean()andboundarytransportsBOUND)asblackandgreylines,respectively.Thedashedpartofthegreylinedenotestheperiodwhencouldnotbemeasured.Alinearregressionisusedto“llthisgap(Kanzowetal.,2007).(Forinterpretationofthereferencestocolorinthis“gurelegend,thereaderisreferredtothewebversionofthisarticle.)D.Rayneretal./Deep-SeaResearchII58(2011)1744…1753 ofobservations,occupyingarangeofvaluesbetween3.2and32.1Sv.TheMOCchangesinstrengthonseasonaltimescalesbutalsoatperiodsasshortasweekstomonths.Allcomponents()contributeaboutequallytothetotalMOCvaria-bility(Fig.7).WhereasitwaspreviouslyknownthattheEkmanandGulfStreamtransportscanexhibitalargevariabilityonsubannualtointerannualtimescales,ourobservationswerethe“rsttoshowthatasimilarvariabilityisfoundforthemid-oceanThenatureoftheMOCvariability(inparticularthecontribu-tionsfrom)observedat26.5Nisfarfromfullyunderstood.Possiblesourcesofvariabilityincludeinternalwaves(RossbyandKelvinwaves)andeddies.However,theimprintofwavesand/oreddiesontheMOCisdif“culttoquantify.ResultsfromanumericalmodelsuggestthattransportanomaliestendtopropagatewestwardwithavelocitysimilartothatexpectedforRossbywavesoreddies(Hirschietal.,2007).BotheddiesandRossbywaveshaveasignatureintheseasurfaceheight.There-fore,ifinternalwavesoreddieswerethemaincauseoftheobservedvariabilityin,onewouldexpectto“ndasigni“-cantcorrelationbetweenandtheseasurfaceheight(SSH)variability.Surprisingly,thermsseasurfaceheightvariabilityobservedinbothdynamicheightsfrominsitudensitymeasure-mentsandaltimetricheightsreducesbyafactorofthree,approachingthewesternboundaryoveradistanceof100kmFig.8).Asaconsequenceofthissuppressionofvariabilityrightatthewesternboundarythevariabilityofisonly3.0Svrms.CorrelationsoftheSSHandarealsosmallclosetothewesternboundary.Thislackofarelationshipbetweensurfaceheightandupperoceantransportisrelatedtoanincreaseinimportanceofhigherorderverticalmodesofhorizontalvelocityrightatthewesternboundary.ThisdeteriorationoftheSSH-correlationisfoundinboththeRAPIDobservations(etal.,2009)andnumericaloceanmodels(Hirschietal.,2009However,ahighcorrelationisobservedbetweenSSHandthemid-oceantransportintegratedfromtheAfricanmargintothewesternmooringsthatarelocatedoffshoreawayfromthewesternboundary(e.g.WB5)(Fig.9).ItseemsthatSSHisunlikelytobeausefulpredictoronsubannualtointerannualtimescalesofMOCvariabilityat26.5N,butthemid-oceanvariabilityoffshorefromthewesternboundary…thewinddrivensubtropicalgyre…maybemonitoredbySSHvariability.The“rst4yearsofMOCobservationsat26.5Nalsosuggestthepresenceofaseasonalcycle,whichpartlyre”ectstheseasonal Fig.8.RMSamplitudeofseasurfaceheight()along26.5NfortheintervalsOctober1992…January2008(solidblackline),April2004…October2006(reddashedline)andOctober1995…June1997(bluedashedline).Alsoshownistheamplitudeofrmsdynamicheight”uctuations(dyn.cm;i.e.geopotentialanomalydividedbytheEarthsgravitationalacceleration)at200mdeterminedfromthemooringdensitymeasurementsatWB2,WB3andWB5(redcrosses).ThebluecrossdenotesdynamicheightcomputedfromdensityatmooringCoftheACCP-3experiment(Johnsetal.,2005).Thegreenlineshowsthermsamplitudealong26.5Nfor2-yearlow-pass“ltereddatafortheintervalOctober1992…January2008.(Forinterpretationofthereferencestocolorinthis“gurelegend,thereaderisreferredtothewebversionofthisarticle.) Fig.9.Uppermid-oceannorthwardtransport”uctuationsinSvshallowerthan1000minblackandeastwardofmooringsWB2(green),WB3(red)andWB5(blue)totheeasternboundaryoffMorocco.Transportsareoffsetby25Svbetweeneachcurve.Fluctuationsofseasurfaceheightincmoverthefullmid-oceansectionandtoeachmooringlocationareshowninorange.(Forinterpreta-tionofthereferencestocolorinthis“gurelegend,thereaderisreferredtothewebversionofthisarticle.) Fig.7.DailytimeseriesofGulfStreamtransport(blue),Ekmantransport(black),uppermid-oceantransport(magenta)andoverturningtransport(red)fortheperiod2April2004…10April2008.GulfStreamtransportisbasedonelectro-magneticcablemeasurementsintheFloridaStraits.Agapinthetimeseriesofapproximatelytwomonthsfrom4Septemberto28October2004isduetoHurricaneFrances,whichdestroyedthefacilityrecordingthevoltage.Herelinearinterpolationischosento“llthegap.EkmantransportisbasedonQuikSCATwinds.Theuppermid-oceantransportisbasedontheRAPIDarraymeasurementsandistheverticalintegralofthetransportperunitdepthdowntothedeepestnorthwardvelocity(1100m)oneachday.OverturningtransportisthenthesumoftheGulfStream,Ekmananduppermid-oceantransportsandrepresentsthemaximumnorthwardtransportofupperlayerwatersoneachday.(Forinter-pretationofthereferencestocolorinthis“gurelegend,thereaderisreferredtothewebversionofthisarticle.)D.Rayneretal./Deep-SeaResearchII58(2011)1744…1753 cycleobservedfortheuppermid-oceantransport.Recentworkhasshownthatforthisseasonalvariabilityhasitsoriginatboththeeasternandwesternmargins.Theslightlylargercontributionoriginatesfromtheeasternmarginandcanbeexplainedbytheheavingofisopycnalslinkedtotheseasonalcycleofthewind-stresscurlattheeasternmargin(Kanzowetal.,2010;Chidichimoetal.,2010Brydenetal.(2009)showthatat4000mdepthatthewesternboundaryoffAbaco,bottompressure”uctuationscompensateinstantaneouslyforbaroclinic”uctuationsinthestrengthandstructureoftheDeepWesternBoundaryCurrent.Therefore,baroclinic”uctuationsintheDeepWesternBoundaryCurrentarecompensatedlocallybybottompressure”uctuationsandsothereisnomid-ocean”owresultingfrom”uctuationsintheDeepWesternBoundaryCurrent.Residualbottompressure”uctuationsatthewesternboundary(bottompressure”uctuationsminusbottompressure,whichaccountforbaroclinicvariabilityoftheDeepWesternBoundaryCurrent)compensatefor”uctuationsinFloridaCurrenttransport.Thus”uctuationsinboththeFloridaCurrentandDeepWesternBoundaryCurrentsarecompensatedbarotropicallyveryclosetothewesternboundary.4.DiscussionandsummaryThe4yearsofMOCobservationshavealreadyprovidedanunprecedentedinsightintotheMOCvariability.WiththeinitialmeasurementswewerealsoabletodeterminethatthehistoricestimatesofthestrengthoftheMOC,basedonsynopticship-boardexpeditions(Brydenetal.,2005),werewithintherangeofsubannualvariabilityoftheMOC(Cunninghametal.,2007OneaspectthatneedstobebetterunderstoodandwhichisthesubjectofongoingresearchistheclimaticrelevanceoftheMOCobservations.AquestionofparticularinterestiswhetherwecanusetheRAPIDdatatoimproveclimatepredictions(onseasonaltoperhapsdecadaltimescales).OnthewaytoaddressthisquestionwewillneedtobeabletoputthelocalMOCobservationsfrom26.5NintoawiderspatialcontextandtrytoestablishthemeridionalcoherenceoftheobservedMOCvaria-bility.Doesthemeridionalcoherencedependonthefrequency(i.e.aresubannualsignalsmainlylocalto26.5Nwhileinter-annualandlongersignalsre”ectprocessesaffectingalargefractionoftheNorthAtlanticbasin(e.g.Binghametal.,2007Toaddressthesepointswewillneedtomakeuseofobserva-tionsfromotherlocationsandnumericalmodels.NumericalstudiessuggestthatfastpropagatingboundarywavescanleadtomeridionallycoherentMOCchanges.Thiswasfoundforidealisedmodelsetups(e.g.Kawase,1987,JohnsonandMarshall,2002)andinmorerealisticmodels(e.g.Binghametal.,2007,Biastochetal.,2008,Zhang,2008).However,modelresultsalsosuggestthatlocally,largehigh-frequencyMOCvariabilitycouldmaskthecoherence(e.g.Hirschietal.,2007).ToassesswhethermeridionallycoherentMOCchangescanbeobserved,theMOCtransportfrom26.5Nneedstobeconsideredalongsidedatafromotherobservingsystems.From2000to2009theMeridionalOverturningVariabilityExperiment(MOVE)providedNADWobservationsat16NintheAtlantic(e.g.Kanzowetal.,2006,2008).Additionally,continuousobservationshavebeenmadeatthewesternboundaryat40since2004intheframeworkoftheRAPIDfundedWesternAtlanticVariabilityExperiment(WAVE,http://www.pol.ac.uk/home/research/theme10/rapidII.phpHughesetal.,2002).Bottompres-suremeasurementsareavailableat26.5N,aswellasatthelocationsofMOVEandWAVEandcanbeusedtotestonwhattimescaleswe“ndcoherentsignalsbetweenthedifferentobser-vingsystems.ModelstudiessuggestacloselinkbetweenbottompressureandMOC”uctuations(e.g.Roussenovetal.,2008).Itwouldalsobeinstructiveifwecouldcomparetransports,e.g.ofNADWat26.5Nand16N,intermsoftransportsinisopycnalcoordinatesasthiscouldallowustoinferwatermasschangesbetweendifferentlatitudes.However,sincethetransportsatNand16Nareobtainedfromdensityobservationsatonlyafewlongitudes(endpointmethod),thefullzonaldensitystructureisnotavailable,whichmeansthataprojectionoftransportsontodensitycoordinatesisnotobvious.OnepossiblewaytoovercometheinabilityofoceanmodelstoreproducetheobservedoceancirculationandtheinevitablegapsinobservationsistoassimilatetheobservedMOCandotherobservationaloceandataintonumericalmodels.Therearedifferentdataassimilationschemes(e.g.WunschandStammer,1998,KohlandStammer,2008,SmithandHaines,2008,Balmasedaetal.,2007)thatassimilatedatafromhydrographicsections,Argo”oatsorfromsatelliteswiththeaimtoproduceoceanstatesthatareascloseaspossibletotherealocean(oceananalyses).Apartfromprovidingglobal,physicallyconsistentoceanstatesthatareusefulforstudyingoceanprocessesohl,2005Cabanesetal.,2008),thevalueoftheseoceananalysesliesintheirpotentialuseforimprovingclimateSmithetal.(2007)showedthattheassimilationofoceanobservationsintotheirdecadalpredictionsystem(DePreSys)improvedtheforecastqualityinasetof10-yearhindcasts.ResearchdoneintheframeworkofRAPID-WATCHwillestablishthevalueoftheRAPID-MOCdatafrom26.5N,whenitisusedasanadditionalconstraintinoceanmodelsandforecastingsystemslikeDePreSys(Smithetal.,2010;Baehr,2010TheRAPID-MOCmonitoringsystemisfundedbyNERC,NOAAandNSFforatotalof10yearsthroughto2014andshoulddocumentthesizeandstructureofthesubannualtointerannualvariabilityintheAtlanticMOC.Froma10-yearrecord,wecancomparetheinterannualvariationsintheMOCwithAtlanticseasurfacetemperaturevariationsandwiththeNorthAtlanticOscillationindexandstarttounderstandlinksbetweentheAtlanticmeridionaloverturningcirculationandclimate.TheobservationalestimatesofMOCvariabilitywillalsoserveasanewbenchmarkagainstwhichthevariabilityincoupledclimatemodels(whichexhibitsubstantiallydifferentamplitudeandstructureinMOCinterannualvariability)canbecomparedandvalidated.Witha10-yearrecordofMOCstrengthandstructureandbyconsideringoceanobservationsfromotherlocationsintheNorthAtlantic(e.g.intheframeworkoftheEUfundedThermohalineOverturningatRisk?(THOR)project),wecanalsostarttoassesswhetherthereisastatisticallysigni“canttrendinthestrengthoftheMOCabovethesubannualandinterannualvariabilityandwecanbuildthegroundworkforpredictingthecourseofAtlanticclimatechangeoverthenext50years.DataavailabilityDatafromtheRAPIDprojectareloggedwiththeBritishOceanographicDataCentre(BODC)onacquisition.FollowingtheNERCdatapolicyforRAPID-WATCH(http://www.bodc.ac.uk/projects/uk/rapid/data_policy/),dataaremadefreelyavailablefromtheBODCwebsite().Timeseriesoftheoverturningandcomponenttransports,alongwithgriddedmooringdata,areavailablefromtheprojectwebpage(www.noc.soton.ac.uk/rapidmocTheFloridaCurrentcabledataaremadefreelyavailablebytheAtlanticOceanographicandMeteorologicalLaboratoryD.Rayneretal./Deep-SeaResearchII58(2011)1744…1753 http://www.aoml.noaa.gov/phod/”oridacurrent/)andarefundedbytheNOAAOf“ceofClimateObservations.ThesurfacelayerorEkmancontributiontotheMOCiscalculatedfromwindsobtainedbytheQuickSCATsatellitescat-terometer(SeaWindsonQuickSCAT.Mission,http://winds.jpl.nasa.gov/missions/quickscat/index.cfmBaehr,J.,Hirschi,J.,Beismann,J.-O.,Marotzke,J.,2004.MonitoringthemeridionaloverturningcirculationintheNorthAtlantic:amodel-basedarraydesignstudy.JournalofMarineResearch62,283…312.Baehr,J.,2010.In”uenceofthe26NRAPID/MOCHAarrayandFloridaCurrentcableobservationsontheECCO-GODAEstateestimate.JournalofPhysicalOceanography40,865…879.Balmaseda,M.,Smith,G.,Haines,K.,Anderson,D.,Palmer,T.,Vidard,A.,2007.HistoricalreconstructionoftheAtlanticMeridionalOverturningCirculationfromtheECMWFoperationaloceanreanalysis.GeophysicalResearchLetters34,L23615.Baringer,M.O.,Larsen,J.C.,2001.SixteenyearsofFloridaCurrenttransportatN.GeophysicalResearchLetters28,3182…3197.Biastoch,A.,Boning,C.W.,Lutjeharms,J.R.E.,2008.Agulhasleakagedynamicsaffectsdecadalvariabilityinoverturningcirculation.Nature456.Bingham,R.J.,Hughes,C.W.,Roussenov,V.,Williams,R.G.,2007.MeridionalcoherenceoftheNorthAtlanticmeridionaloverturningcirculation.Geophy-sicalResearchLetters2007(34),L23606.Broecker,W.S.,Denton,G.H.,1989.Theroleofocean-atmospherereorganisationsinglacialcycles.GeochimicaetCosmochimicaActa53,2465…2501.Bryden,H.L.,Imawaki,S.,2001.Oceanheattransport.In:Siedler,G.,Church,J.,Gould,J.(Eds.),OceanCirculationandClimate.AcademicPress,pp.455…474.Bryden,H.L.,Longworth,H.L.,Cunningham,S.A.,2005.SlowingoftheAtlanticMeridionalOverturningCirculationat25N.Nature438,655…657.Bryden,H.L.,Mujahid,A.,Cunningham,S.A.,Kanzow,T.,2009.Adjustmentofthebasin-scalecirculationat26NtovariationsinGulfStream,DeepWesternBoundaryCurrentandEkmantransportsasobservedbytheRAPIDarray.OceanScience5,421…433.Cabanes,C.,Lee,T.,Fu,L.-T.,2008.MechanismsofinterannualvariationsofthemeridionaloverturningcirculationoftheNorthAtlanticOcean.JournalofPhysicalOceanography38,467…480.Chidichimo,M.P.,Kanzow,T.,Cunningham,S.A.,Marotzke,J.,2010.Thecontribu-tionofeastern-boundarydensityvariationstotheAtlanticmeridionalover-turningcirculationat26.5N.OceanScience6,475…490.Cubasch,U.,Meehl,G.A.,Boer,G.J.,Stouffer,R.J.,Dix,M.,Noda,A.,Senior,C.A.,Raper,S.,Yap,K.S.,2001.Projectionsoffutureclimatechange,Chapter9.In:Houghton,J.T.(Ed.),ClimateChange2001:TheScienti“cBasis.CambridgeUniversityPress,pp.525…582.Cunningham,S.A.,Kanzow,T.O.,Rayner,D.,Barringer,M.O.,Johns,W.E.,Marotzke,J.,Longworth,H.R.,Grant,E.M.,Hirschi,J.J.-M.,Beal,L.M.,Meinen,C.S.,Bryden,H.L.,2007.TemporalvariabilityoftheAtlanticMeridionalOverturningCirculationat26.5N.Science317,935…938.Cunningham,S.A.,2005a.RRSDiscoveryCruises277(26Mar.…16Apr.2004)and278(19Mar.…30Mar.2004):monitoringtheAtlanticMeridionalOverturningCirculationat26.5N.CruiseReportNo.53,103pp.Cunningham,S.A.,2005b.RRSDiscoveryCruise279,04Apr.…10May2004:atransatlantichydrographicsectionat24.5N.CruiseReportNo.54,198pp.Cunningham,S.A.,Marsh,R.,2010.ObservingandmodelingchangesintheAtlanticMOC.WileyInterdisciplinaryReviews:ClimateChange1,180…191.Curry,R.,Dickson,B.,Yashayaev,I.,2003.AchangeinthefreshwaterbalanceoftheAtlanticOceanoverthepastfourdecades.Nature426,826…829.Dansgaard,W.,Johnsen,S.J.,Clausen,H.B.,Dahl-Jensen,D.,Gundestrup,N.S.,Hammer,C.U.,Hvidberg,C.S.,Steffensen,J.P.,Sveinbjornsdottir,A.E.,Jouzel,J.,Bond,G.,1993.Evidenceforgeneralinstabilityofpastclimatefroma250kyrice-corerecord.Nature364,218…220.Delworth,T.L.,Mann,M.E.,2000.ObservedandsimulatedmultidecadalvariabilityintheNorthernHemisphere.ClimateDynamics16,661…676.Dickson,R.,Yashayaev,I.,Meincke,J.,Turrell,B.,Dye,S.,Holfort,J.,2002.RapidfresheningofthedeepNorthAtlanticOceanoverthepastfourdecades.Nature416,832…837.Ganachaud,A.,Wunsch,C.,2002.Large-scaleoceanheatandfreshwatertransportsduringtheWorldOceanCirculationExperiment.JournalofClimate16,696…705.Graf,J.,Sasaki,C.,Winn,C.,Liu,T.,Tsai,W.,Freilich,M.,Long,D.,1998.NASAscatterometerexperiment.ActaAstronautica43,397…407.Gregory,J.M.,Dixon,K.W.,Stouffer,R.J.,Weaver,A.J.,Driesschaert,E.,Eby,M.,Fichefet,T.,Hasumi,H.,Hu,A.,Jungclaus,J.H.,Kamenkovich,I.V.,Levermann,A.,Montoya,M.,Murakami,S.,Nawrath,S.,Oka,A.,Sokolov,A.P.,Thorpe,R.B.,2005.AmodelintercomparisonofchangesintheAtlanticthermohalinecirculationinresponsetoincreasingatmosphericCOconcentration.GeophysicalResearchLetters32,L12703.doi:10.1029/2005GL023209.akkinen,S.,Rhines,P.B.,2004.DeclineofsubpolarNorthAtlanticcirculationduringthe1990s.Science304,555…559.akkinen,S.,Rhines,P.B.,2009.ShiftingSurfaceCurrentsintheNorthernNorthAtlanticOcean.JournalofGeophysicalResearch114,C04005.doi:10.1029/Hansen,B.,Turrell,W.R.,Østerhus,S.,2001.Decreasingout”owfromtheNordicseasintotheAtlanticOceanthroughtheFaroeBankChannelsince1950.Nature411,927…930.n,H.,Sandø,A.B.,Drange,H.,Hansen,B.,Valdimarsson,H.,2005.In”uenceoftheAtlanticSubpolarGyreonthethermohalinecirculation.Science309,Hirschi,J.,Marotzke,J.,2007.Reconstructingthemeridionaloverturningcircula-tionfromboundarydensitiesandthezonalwindstress.JournalofPhysicalOceanography37,743…763.Hirschi,J.,Baehr,J.,Marotzke,J.,Stark,J.,Cunningham,S.A.,Beismann,J.-O.,2003.AmonitoringdesignfortheAtlanticmeridionaloverturningcirculation.GeophysicalResearchLetters30.doi:10.1029/2002GL016776.Hirschi,J.J.-M.,Killworth,P.D.,Blundell,J.R.,2007.Subannual,seasonalandinterannualvariabilityoftheNorthAtlanticMeridionalOverturningCircula-tion.JournalofPhysicalOceanography37,1246…1265.Hirschi,J.J.-M.,Killworth,P.D.,Blundell,J.R.,Cromwell,D.,2009.SeaSurfaceheightsignalsasindicatorsforoceanicmeridionalmasstransports.JournalofPhysicalOceanography39,581…601.doi:10.1175/2008JPO3923.1.Holliday,N.P.,Hughes,S.L.,Bacon,S.,Beszczynska-Moller,A.,Hansen,B.,Lavin,A.,Loeng,H.,Mork,K.A.,Østerhus,S.,Sherwin,T.,Walczowski,W.,2008.Reversalofthe1960sto1990sfresheningtrendinthenortheastNorthAtlanticandNordicSeas.GeophysicalResearchLetters35,L03614.doi:10.1029/2007GL032675.Huber,C.,Leuenberger,M.,Spahni,R.,Fluckiger,J.,Schwander,J.,Stocker,T.F.,Johnson,S.,Landais,A.,Jouzel,J.,2006.IsotopecalibratedGreenlandtempera-turerecordoverMarineIsotopeStage3anditsrelationtoCH.EarthandPlanetaryScienceLetters243,504…519.doi:10.1016/j.epsl.2006.01.002.Hughes,C.,D.MarshallandR.Williams.2002.AmonitoringarrayalongthewesternmarginoftheNorthAtlantic,ResearchProposalsubmittedtoNaturalEnvironmentResearchCouncil.IPCC,2007.Summaryforpolicymakers.In:Solomon,S.,Qin,D.,Manning,M.,Chen,Z.,Marquis,M.,Averyt,K.B.,Tignor,M.,Miller,H.L.(Eds.),ClimateChange2007:ThePhysicalScienceBasis,ContributionofWorkingGroupItotheFourthAssessmentReportoftheIntergovernmentalPanelonClimateChange.bridgeUniversityPress,Cambridge,NewYork.Johns,W.E.,Kanzow,T.,Zantopp,R.,2005.Estimatingoceantransportswithdynamicheightmoorings:anapplicationintheAtlanticDeepWesternBoundaryCurrentat26N.Deep-SeaResearchI52,1542…1567.Johns,W.E.,Beal,L.M.,Baringer,M.O.,Molina,J.,Rayner,D.,Cunningham,S.A.,Kanzow,T.O.,2007.VariabilityofshallowandDeepWesternBoundaryCurrentsofftheBahamasduring2004…2005:resultsfromthe26NRAPID-MOCarray.2008.JournalofPhysicalOceanography38,605…623.Johnson,H.L.,Marshall,D.P.,2002.AtheoryforthesurfaceAtlanticresponsetothermohalinevariability32,1121…1132.Kanzow,T.,Send,U.,Zenk,W.,Chave,A.D.,Rhein,M.,2006.Monitoringtheintegrateddeepmeridional”owinthetropicalNorthAtlantic:long-termperformanceofageostrophicarray.DeepSeaResearchI53,528…546.Kanzow,T.,Cunningham,S.A.,Rayner,D.,Hirschi,J.J.-M.,Johns,W.E.,Baringer,M.O.,Bryden,H.L.,Beal,L.M.,Meinen,C.S.,Marotzke,J.,2007.Observed”owcompensationassociatedwiththeMOCat26.5NintheAtlantic.Science317,Kanzow,T.,Send,U.,McCartney,M.,2008.OnthevariabilityofthedeepmeridionaltransportsinthetropicalNorthAtlantic.Deep-SeaResearchI55,Kanzow,T.,Johnson,H.,Marshall,D.,Cunningham,S.A.,Hirschi,J.J.-M.,Mujahid,A.,Bryden,H.L.,Johns,W.E.,2009.Basin-wideintegratedvolumetransportsinaneddy-“lledocean.JournalofPhysicalOceanography39,3091…3110.Kanzow,T.,Cunningham,S.A.,Johns,W.E.,Hirschi,J.J.-M.,Marotzke,J.,Baringer,M.O.,Meinen,C.S.,Chidichimo,M.P.,Atkinson,C.,Beal,L.M.,Bryden,H.L.,Collins,J.,2010.SeasonalvariabilityoftheAtlanticmeridionaloverturningcirculationatN.JournalofClimate23,5678…5698.Kawase,M.,1987.Establishmentofdeepoceancirculationdrivenbydeep-waterproduction.JournalofPhysicalOceanography17,2294…2317.Knight,J.R.,Allan,R.J.,Folland,C.K.,Vellinga,M.,Mann,M.E.,2005.Asignatureofpersistentnaturalthermohalinecirculationcyclesinobservedclimate.Geo-physicalResearchLetters32,L20708.doi:10.1029/2005GL024233.ohl,A.,2005.AnomaliesofMeridionalOverturning:mechanismsintheNorthAtlantic.JournalofPhysicalOceanography35,1455…1472.ohl,A.,Stammer,D.,2008.VariabilityoftheMeridionalOverturningintheNorthAtlanticfromthe50-yearGECCOStateestimation.JournalofPhysicalOceanography38,1913…1930.Larsen,J.C.,1992.Transportandheat”uxoftheFloridaCurrentat27Nderivedfromcross-streamvoltagesandpro“lingdata:theoryandobservations.PhilosophicalTransactionsoftheRoyalSocietyofLondonA338,169…236.Lherminier,P.,H.Mercier,C.Gourcuff,Treguier,A.M.,2006.MOCobservationsbetweenGreenlandandPortugalinSummers1997,2002and2004.EGUAbstract,presentedinViennainApril.Marotzke,J.,CunninghamS.A.,Bryden,H.L..2002.MonitoringtheAtlanticmeridionaloverturningcirculationat26.5N,ResearchProposalsubmittedtoNaturalEnvironmentResearchCouncil.Olsen,S.M,Hansen,B.,Quadfasel,D.,Østerhus,S.,2008.Observedandmodelledstabilityofover”owacrosstheGreenland-Scotlandridge.Nature455,519…522.doi:10.1038/nature07302.Østerhus,S.,Gammelsrod,T.,1999.TheabyssoftheNordicSeasiswarming.JournalofClimate12,3297…3304.Roussenov,V.M.,Williams,R.G.,Hughes,C.W.,Bingham,R.,2008.BoundarywavecommunicationofbottompressureandoverturningchangesfortheNorthD.Rayneretal./Deep-SeaResearchII58(2011)1744…1753 Atlantic.JournalofGeophysicalResearch113,C08042.doi:10.1029/2007/Smeed,D.A.etal.,2010.RAPIDDeploymentReport,NationalOceanographyCentre,Southampton,CruiseReportNo.44.Smith,D.M.,Cusack,S.,Colman,A.W.,Folland,C.K.,Harris,G.R.,Murphy,J.M.,2007.Improvedsurfacetemperaturepredictionforthecomingdecadefromaglobalclimatemodel.Science317,796…799.Smith,G.C.,Haines,K.,2008.EvaluationoftheS(T)assimilationmethodwiththeargodataset.QuarterlyJournaloftheRoyalMeteorologicalSociety135,Smith,G.C.,Haines,K.,Kanzow,T.,Cunningham,S.A.,2010.Impactofhydro-graphicdataassimilationonthemodelledAtlanticmeridionalover-turningcirculation.OceanScience6,761…774.Vellinga,M.,Wood,R.A.,2002.GlobalclimaticimpactsofacollapseoftheAtlanticthermohalinecirculation.ClimaticChange54,251…267.Visbeck,M.,E.P.Chassignet,R.Curry,T.Delworth,Dickson,B.,Krahmann,G.,2003.TheoceansresponsetoNorthAtlanticOscillationvariability.In:TheNorthAtlanticOscillation,J.W.Hurrell,Y.Kushnir,G.Ottersen,M.Visbeck(Eds.),GeophysicalMonographSeries,vol.134,pp.113…146.Webb,D.J.,1996.Anoceanmodelcodeforarrayprocessorcomputers.Computers&Geosciences22,569…578.WhitworthIII,T.,Peterson,R.G.,1985.VolumetransportoftheAntarcticCircumpolarCurrentfrombottompressuremeasurements.JournalofPhysicalOceanography15,810…816.Wunsch,C.,Stammer,D.,1998.Satellitealtimetry,themarinegeoid,andtheoceanicgeneralcirculation.AnnualReviewofEarthandPlanetarySciences26,Zhang,2008.Coherentsurface-subsurface“ngerprintoftheAtlanticmeridionaloverturningcirculation.GeophysicalResearchLetters35,L20705.doi:10.1029/2008GL035463.D.Rayneretal./Deep-SeaResearchII58(2011)1744…1753