abstract book for poster presentations · analysis of the ipsl-cm6-lr ensemble of historical...

CMIP6ModelAnalysisWorkshop25-28March2019,Barcelona

Abstractbookfor

posterpresentations

Session1-MIPoverviews,Modelgroupoverviewsandinfrastructure

1-P01

TropicalPrecipitationVariabilityIntheCASFGOALS-f3-H

BaoQing(1),HeBian,WangXiaocong,WuXiaofei(2),LiJinxiao,WangLei,LiuYimin,WuGuoxiong

1-LASG,InstituteofAtmosphericPhysics,ChineseAcademyofSciences(China),2-SchoolofAtmosphericSciences,ChengduUniversityofInformationTechnology(China)

CASFGOALS-f3-Hisanext-generationClimateSystemModelfromtheInstituteofAtmosphericPhysics(IAP) StateKey LaboratoryofNumericalModeling forAtmospheric Sciences andGeophysical FluidDynamics(LASG),which isparticipating inCMIP6HighResMIP.FGOALS-f3-H is featuredonthehighresolutionupto0.25¡globallyandascale-awareparameterizationofresolvingcumulusprocesses.ThepresentAMIPandrelevantclimatecontrolrunhasbeencarriedoutonChina'sTianhe-2machinefor65-year simulation. The tropical precipitation characteristics of 0.25¡ FGOALS-f3-H have beeninvestigatedusingthehigh-resolutionTRMM-basedprecipitationproducts,SSTproductsandtropicalcyclonebesttrackdata(IBTrACS).TheevolutionofthemajortropicalsystemsincludesITCZ,MJO/TISO,thefrequencyandintensityoftropicaldailyprecipitation,andtropicalcyclone/typhoon.Theresultsindicatethat0.25¡FGOALS-f3-HmitigatesthedoubleITCZproblem,andwellreproducestheeastwardpropagatingMJOintheborealwinterandnorthwardpropagatingTISOoverequatorialIndianoceanintheborealsummer,whicharetakenasthelong-standingchallengesinthefieldofclimatemodeling.ThefrequencyandintensityoftropicaldailyprecipitationarealsocomparabletoTRMMproducts.Inaddition,thebehaviorofthetropicalcyclone/typhoonisreported.Finally,somemodelsbiasrelatedwithtropicalprecipitationarediscussedandthecurrentandfutureapplicationsareintroduced.

Keywords:FGOALSf3H,HighResMIP,TropicalPrecipitationVariability


1-P02

CMIP6EvaluationwiththeESMValTool

BockLisa(1),EyringVeronika(1)(2),LauerAxel(1),RighiMattia(1),EsmvaltoolDevelopmentTeam

1-DLRInstitutfürPhysikderAtmosphäre(Germany),2-UniversityofBremen,InstituteofEnvironmentalPhysics(IUP)(Germany)

In recent years the scope of model evaluation has expanded dramatically, with new insightfulapproaches inaddition to traditionalmethods that remain importantand thereforeare frequentlyapplied toanynewsimulation.This representsanopportunity tomakewell-establishedaspectsofmodelevaluationmoreroutinesothattheycanbemadeavailablemorequickly,withlesseffort,andbearesourcetomodelanalystsanddevelopers.ToaccomplishthisintimeforCMIP6,theEarthSystemModel Evaluation Tool (ESMValTool, https://www.esmvaltool.org/) has been developed as open-source software bymore than 41 institutions. It has undergone rapid development since the firstreleasein2016andisnowawell-testedtoolthatprovidesend-to-endprovenancetrackingtoensurereproducibility. It consists of a backend that performs common pre-processing operations and adiagnostic part that includes tailored diagnostics and performance metrics for specific scientificapplications.Thediagnosticpart includesa largecollectionofstandardrecipesforreproducingtheanalysis of many variables across atmosphere, ocean, and land domains, with diagnostics andperformance metrics focusing on the mean-state, trends, variability and important processes,phenomena,aswellasemergentconstraintsanddetaileddiagnosticsformonsoon,ElNinoSouthernOscillation (ENSO), theMadden-Julian Oscillation (MJO). A growing list of more targeted analysispackages such as the NCAR Climate Variability Diagnostics Package CVDP are also integrated theESMValTool.ItcanalsoreproducethemajorityofanalysesoftheIPCCAR5climatemodelevaluationchapterwhichallowscomparisonofthenewCMIP6simulationsastheyarepublishedtotheESGF.Inthispresentationwewill introducethenewESMValToolversion2.0andwillshowresultsfromtheCMIP6evaluationasitstandsatthetimeoftheworkshop.

Keywords:modelevaluation,routineevaluation,ESMValTool,CMIP6


1-P03

AnalysisoftheIPSL-CM6-LRensembleofhistoricalexperiments

BoucherOlivier(1),ClimateModellingCentreandIPSL

1-InstitutPierre-SimonLaplace(France)

IPSLCMChasperformedanensembleof31membersofhistoricalsimulationswhichdifferbytheirinitialconditionssampledfroma longpiControlexperiment. Initialanalysis indicatethat i)whileallmemberswarm,thedegreeofwarmingbetween1850-1900andpresent-dayvariesbetween1.0and1.4¡C depending on the members and ii) AMOC keeps some memory of the initial conditionsthroughoutthehistoricalsimulations.Akeyquestionwewilltrytoansweriswhetherthehistoricalmembersthatperformbestintermsofsimulatedwarmingarealsothosethathaveaclimatevariabilityinphasewiththeobservations.Asecondquestiontoansweriswhethermodesofnaturalvariabilityareforcedtosomeextentbysomeoftheforcingagents.

Keywords:historicalsimulations,IPSL,CM6,LR,aerosols


1-P04

CommunityEarthSystemModelversion2(CESM2)

DanabasogluGokhan(1),LamarqueJean-Francois(1)

1-NationalCenterforAtmosphericResearch(UnitedStates)

Anevaluationof theCMIP6DECKand severalMIPs simulationswith theCommunityEarthSystemModelversion2(CESM2)willbepresented.Thesimulationsuseboththelow-topandhigh-topfullchemistry versions of the atmospheric component. In comparison to its previous version, CESM2contains many substantial science and infrastructure improvements and capabilities in all itscomponents.Thesenewadvancementsinclude:anatmosphericmodelcomponentthatincorporatessignificant improvementsto itsturbulenceandconvectionrepresentations,openingthewayforananalysisofhowthesesmall-scaleprocessescanimpacttheclimate;improvedabilitytosimulatemodesoftropicalvariabilitythatcanspanseasonsandaffectglobalweatherpatterns;alandicesheetmodelcomponentforGreenlandthatcansimulatethecomplexwaytheicesheetmovesanddoesabetterjobofsimulatingcalvingoftheiceintotheocean;aglobalcropmodelcomponentthatcansimulatebothhowcroplandaffectsregionalclimate,includingtheimpactsofincreasedirrigation,andhowthechanging climatewill affect crop productivity; awavemodel component that simulates howwindcreateswavesontheocean,animportantmechanismformixingoftheupperocean;anupdatedrivermodelcomponentthatsimulatessurfaceflowsacrosshillsidesandintotributariesbeforeenteringthemainriverchannel;andanewsetofinfrastructureutilitiesthatprovidemanynewcapabilitiesforeasierportability,casegenerationandusercustomization,testingfunctionality,andgreatlyincreasedrobustness and flexibility. Inclusion of new physics and components as well as of improvedrepresentationsofnumerousotherprocessesresultinbettersimulationcharacteristicsinmanyfieldsincomparisonwithavailableobservations.

Keywords:coupledmodels,CESM2


1-P05

ThephysicalperformanceandvariabilityoffirstEC-EarthtransientsimulationensembleunderCMIP6.

DoscherRalf(1),FuentesFrancoRamon(2),WyserKlaus(3),ZimmermannKlaus(1)

1-SwedishMeteorologicalandHydrologicalInstitute(SMHI)(Sweden),2-SMHI(Sweden),3-SwedishMeteorologicalandHydrologicalInstitute(Sweden)

The EC-Earth 3 model for CMIP6 has been developed and tuned in several Earth System Modelconfigurations,withtheGlobalClimateModel(GCM)andaninteractivevegetationmoduleascorephysical configurations, supplemented by configurations with/without atmospheric composition,oceanbiogeochemistryandGreenlandglacier.ThisposterdocumentsandexploresthefirstensembleofCMIP6transientsimulationswithEC-Earth3inGCMmodeincludingdynamicvegetation.Analysisencompasseslargescalebiases,oceanoverturningvariabilityandthemainmodesofhemisphericscalevariabilityandhowwelltheyarerepresentedintheensembleoftransientsimulationsandhowtheychange over proceeding time intervals. The variability analyses includes the performance ofparticularly ENSO, AMO, NAO and PNA. Regarding anomalies (precipitation and air surfacetemperature)associatedtoENSOduringDJF, teleconnections toNorthAmericaandtheNorthandtropicalAtlanticareevaluatedinthetransientrunsasfarasobservationsandreanalysisdataexist.Wealsoexaminethespatialpattern,intensityandpowerspectraoftheSSTandSLPanomaliesrelatedtoAMO,aswellasrelatedtoPNA.Resultswillallowtojudgetherobustnessandvariabilitywithintheensemble,whichisimportantknowledgebeforestartingfuturescenariosimulationsfordifferentSSPsundertheumbrellaoftheCMIP6ScenarioMIP.

Keywords:CMIP6transientsimulation,ensemble,biases,AMOC,ENSO,AMO,NAOandPNA,teleconnections


1-P06

IndianOceanDipoleanditslinkagetoSouthAsianMonsooninIITM-ESM

A.g.Prajeesh(1),P.Swapna(2),KrishnanR(2)

1-IndianInstituteofTropicalMeteorology(India),2-IndianInstituteofTropicalMeteorology(India)

ThefidelityinreproducingtropicalIndianOcean(IO)variability,particularlytheIndianOceanDipole(IOD),anditslinkagestoSouthAsianMonsoonRainfall(SAMR)isinvestigatedinthephase6CoupledModelIntercomparisonProject(CMIP6)historicalsimulationsoftheIITMEarthSystemModel(IITM-ESM)1,2.Themodelsimulatedseasurfacetemperature(SST)showsabasinwidenegativebiasoftheorderof1¡C.Conversely,thermoclinedepthshowspositivebiasreachingupto100minmanyregionssuch as the Arabian Sea and the southeast IO. Nonetheless, model successfully simulates manyfundamentalcharacteristicsofIODowingtobetterSST-thermoclinecouplingandmuchsmallerbiasesintheIODregion.AlthoughthesimulatedIODamplitudesarelargerthantheobserved,modelskillfullyrepresentstheIODspatialpattern,periodicityandphase-locktothemonsoonseasonalcycle.IITM-ESMsimulatesarealisticIOD-SAMRrelationship,contrarytomostoftheCMIP5modelswhichfailtorepresenttheobservedIOD-SAMRcorrelationduetotheoverlystrongcontrolofENSO3.Inagreementwiththis,wefindaweakcorrelation(r=0.35)betweenIODandENSOinthemodel.Further,thestudytriestofindpotentialfactorsresponsibleforthisbetterrepresentationofIOD-SAMRrelationshipinthemodel.

Keywords:ENSO,thermocline,IndianOcean,teleconnection,Monsoon,IOD,IOD


1-P07

WCRPCORDEX:ADiagnosticMIPforCMIP6

GutowskiWilliam(1),ScienceAdvisoryTeamCordex,LakeIrène(2)

1-IowaStateUniversity(UnitedStates),2–SMHI,Sweden

TheCOordinatedRegionalDownscalingEXperiment(CORDEX) isaprojectundertheWorldClimateResearch Programme and a diagnostic model intercomparison project (MIP) in CMIP6. CORDEXcoordinates with ScenarioMIP and is structured to allow cross comparisons with HighResMIP andinteractionwith theCMIP6VIACSAdvisoryBoard.TheRegionalChallengesestablishedbyCORDEXcontribute to this effortbyprovidingabasis forusingCMIP6GCMoutput toproducedownscaledprojected changes in regional climates and assess sources of uncertainties in the projections. TheRegionalChallenges includefocusonspecificregionalphenomena,suchasconvectivesystemsandclimatesofurbanareas,smallislandsandinlandwaters.Theyprovideabasisfordistillingoutputfrommultiplemodelsintoclimatechangeinformationforvulnerability,impactsandadaptationstudiesandforimprovingsimulations.Topromotethiseffort,theCORDEXScienceAdvisoryTeamhasdevelopeda CORDEX Output for Regional Evaluations (CORE) framework that will produce a baseline set ofhomogeneoushigh-resolution,downscaledprojectionsforregionsworldwideusingCMIP6outputasitbeomesavailable.AnimmendiateintentistoprovideoutputthatcansupportIPCCAR6assessments.For this reason, the core framework is intended to be ambitious but sufficiently cost-effective toproducetimelyoutputfortheAR6.COREisnotintendedbyitselftobethenextphaseofCORDEX,butCOREsimulationscanserveafurtherpurposeofprovidingafoundationineachCORDEXregionforfurtherdownscalingactivities.ThecurrentstatusoftheCOREprojectandsupportbyCORDEXoverallforCMIP6objectiveswillbereportedatthemeeting.

Keywords:regionalclimateprocesses,vulnerability,impactsandadaptation,high,resolutiondownscaling,IPCCAR6


1-P08

TheMRIEarthSystemModelver.2.0(MRI-ESM2.0):Basicevaluationofthephysicalcomponent

YukimotoSeiji(1),KawaiHideaki(1),TsuyoshiKoshiro(1),OshimaNaga(1),YoshidaKohei(1),UrakawaShogo(1),TsujinoHiroyuki(1),DeushiMakoto(1),TanakaTaichu(1),HosakaMasahiro(1),

YabuShokichi(2),YoshimuraHiromasa(1),ShindoEiki(1),MizutaRyo(1),ObataAtsushi(1),AdachiYukimasa(2),IshiiMasayoshi(1)

1-MeteorologicalResearchInstitute[Tsukuba](Japan),2-JapanMeteorologicalAgency(Japan)

The Meteorological Research Institute Earth System Model version 2.0 (MRI-ESM2.0) has beendevelopedbasedonpreviousmodels,MRI-CGCM3andMRI-ESM1,whichparticipatedinCMIP5.Thesemodels underwent numerous improvements meant for highly accurate climate reproducibility.Accumulationofvariousimprovementsconcerningclouds,suchasanewstratocumuluscloudscheme,led to remarkable reduction in errors in shortwave, longwave, andnet radiation at the topof theatmosphere,becomingsufficientlysmallcomparedwiththose in theCMIP5models.The improvedradiation distribution brings the accurate meridional heat transport required for the ocean andcontributesareducedsurfaceairtemperature(SAT)bias.MRI-ESM2.0displaysrealisticreproductionof both mean climate and interannual variability. For instance, the stratospheric quasi-biennialoscillationcannowberealisticallyexpressedthroughtheenhancedverticalresolutionandintroductionof non-orographic gravitywave drag parameterization. For the historical experiment,MRI-ESM2.0reasonably reproduces global SAT change for recent decades; however, interdecadal variation isoverestimatedcomparedwithobservations.MRI-ESM2.0hasbeenimprovedinmanyaspectsoverthepreviousmodels,MRI-CGCM3/MRI-ESM1,and isexpectedtodemonstratesuperiorperformance inmanyexperimentsplannedforCMIP6.

Keywords:climatemodeling


1-P09

GFDL'scontributionstoCMIP6-highlightsfromGFDLCM4andESM4

JohnJasmin(1)

1-NOAA/GeophysicalFluidDynamicsLaboratory(UnitedStates)

NOAA/Geophysical FluidDynamics Laboratory (GFDL) has participated in and contributed to everyCoupledModel Intercomparison Project (CMIP) to date. For Phase 6 of CMIP (CMIP6), GFDL hasdevelopedtwonewfully-coupledmodels:ahighresolutionphysicalclimatemodel(CM4),andanearthsystem model focusing on increased comprehensiveness (ESM4). CM4 and ESM4 will be used toparticipateintheDECK,aswellas13specializedModelIntercomparisonProjects(MIPs).WepresentanoverviewofmodeladvancesinCM4andESM4highlightingimprovementsattainedbythemodels(e.g.ENSO,MJO),aswellasongoingchallenges(e.g.dryAmazon).ResultswillalsobecomparedtoAR5modelsacrossasuiteofclimatevariables,andclimatebenchmarkssuchasTCR,ECS,andTCRE.Dependingondataavailability,resultswillalsobecomparedtootherCMIP6models.

Keywords:ESM4,GFDL,CM4,GFDL,CMIP6


1-P10

TheUKEarthsystemmodelcontributiontoCMIP6:Firstresults

JonesColin(1)

1-UniversityofLeeds(UnitedKingdom)

The1stUKEarthsystemmodel(UKESM1)isbuiltonthephysicalclimatemodelHadGEM3-GC3.1withthe addition of an interactive treatment of the global carbon cycle and stratosphere-tropospherechemistrycoupledtoanadvanced,two-momentaerosolscheme.UKESM1constitutestheprimaryUKcontributiontoCMIP6.WepresentresultsfromtheUKESM1CMIP6DECKsimulations,particularlyananalysisofover1000yearsofpiControlsimulation,assessingthelong-termstabilityofthesimulatedphysical and biogeochemical climate, the global carbon cycle and vegetation. We further discussprocesses contributing to bothmulti-decadal and centennial timescale variability in this unforcedsimulation.WethenpresentresultsfromtheUKESM1abrupt-4xCO2simulation,definingtheEffectiveClimateSensitivity(ECS)ofthemodelandthe1pctCO2simulation,fromwhichtheTransientClimateResponse(TCR)iscalculated.Finally,wepresentanoverviewoftheUKESM1ensembleofhistoricalsimulations,emphasizingevaluationoflarge-scale,forcedclimatesignals.Implicationsofthemodel'ssensitivity to CO2, non-CO2 trace gases and aerosol emissions for policy-relevantmetrics, such asallowablecarbonemissionscommensuratewithdifferentglobalwarminglimitswillbediscussed.

Keywords:Forcedclimatesignals,climatesensitivity,transientclimateresponse


1-P11

OverviewofUSDepartmentofEnergy'seffortsonModelDiagnosticsandMetricsforUnderstandingandQuantifyingModelBiases

JosephRenu(1)

1-USDepartmentofEnergy(UnitedStates)

TheU.S.DepartmentofEnergy'sModelAnalysisprogramareaoftheEarthandEnvironmentalSciencesProgram has been supporting research to develop diagnostics andmetrics for understanding andquantifyingmodelbiasesinprocessesspanningacrossthecoupledearthsystem.Theseprocessesarebroadly related to the water cycle, extreme events, large-scale circulation features, modes ofvariability,cloudprocessesandfeedbacks,landsurfaceprocessesandland-atmosphereinteractions,terrestrial and ocean biogeochemistry, high latitude processes, and process interactions andfeedbacks. Feature tracking and computation and visualization of diagnostics and metrics areimplementedinseveralpackagessuchasthePCMDIMetricsPackage(PMP),InternationalLandModelBenchmarking (ILAMB), Toolkit for Extreme Climate Analysis (TECA), FLEXible object TRacKeR(FLEXTRKR), and Community Data Analysis Tools (CDAT). These tools have been used by a largecommunityofresearchers,facilitatingmodelevaluationandintercomparisonandprovidinginsightsfor model development and improvement. The Model Analysis program area is also supportingresearch to develop use-inspired metrics to evaluate model aspects relevant to stakeholders ofregionalclimateinformation.Thispresentationwillprovideahigh-leveloverviewandsummarizekeyactivities supportedby theModelAnalysis effort andprovideexamplesof insights gained throughapplicationoftheanalysistoolsinsingle-andmulti-modeloutput.

Keywords:Modelbiases,diagnosticstools,metrics,understanding


1-P12

TheroleoftheIPCCDataDistributionCentreinsupportingassessmentsofclimatechange

JuckesMartin(1),StockhauseMartina(2),ChenBob,PascoeCharlotte,CallaghanSarah,EdmundsRorie

1-ScienceandTechnologyFacilitiesCouncil(UnitedKingdom),2-DeutschesKlimarechenzentrum[Hamburg](Germany)

TheDataDistributionCenter (DDC) of the Intergovernmental Panel onClimate Change (IPCC)wasestablishedin1997tofacilitatethetimelydistributionofaconsistentsetofup-to-datescenariosofchangesinclimateandrelatedenvironmentalandsocio-economicfactorsforuseinclimateimpactsassessments,particularlytosupporttheperiodicIPCCAssessmentReports.Thisworkcouldbeseenashelping to bridge the gap between intercomparison of models which is conducted within CMIP(CoupledModelIntercomparisonProject)andthebroaderproblemofassessingclimatechange.Thispresentation will explore two key aspects of this work: firstly, supporting the exchange of databetween thephysical climate, climate impactsandsocio-economic sciencedomainsand, secondly,ensuring that data used in assessments,whichmaybe fromdynamic research projects, is reliablycuratedtoensurelongtermaccessandpropertrackingofdatausage.Theseactivitiesplayacriticalroleinenablingrobustassessmentsbyanextendedcommunityeffortwhilemaintainingtransparentdataprovenance.OurworkisunderpinnedbyengagementwiththeWorldDataSystem.Inbothcases,thesegoalsaresupportedbyapplyingwellestablishedmethodologiesofdataarchivalandcuration.Dataexchangeissupportedbyengagingwiththesciencecommunitiestoensurethatkeydatasetsarewelldocumentedandmaintained.TheworkoftheDDChasbeenmadepossiblebytheworkoftheIPCCTaskGrouponDataandScenarioSupport for ImpactandClimateAnalysis (TGICA)whichhaspublishedarangeofguidancedocumentsondatause.Thedatacurationworkhasasignificantroleinensuringthattheinterestsofthedataprovideraresupported.Inparticular,thedataprovidershouldgetcreditfordatawhichhaslongtermvalue.TheDDCisworkingwithDataCitetoprovideDOIdatareferencesforlargeclimatedataprojections.

Keywords:dataprovenance


1-P13

PMIP4-CMIP6simulationsoftheLastGlacialMaximumclimate:firstresults

KageyamaMasa(1),KapschMarie(2),LhardyFanny(3),MikolajewiczUwe(2),RocheDidier(3)(4),ParticipantsPaleoclimateModellingIntercomparisonProject

1-Laboratoiredessciencesduclimatetdel'environnement(France),2-MaxPlanckInstituteforMeteorology(Germany),3-LaboratoiredesSciencesduClimatetde

l’Environnement[Gif-sur-Yvette](France),4-VUUniversityAmsterdam(Netherlands)

TheLastGlacialMaximum(LGM,~21,000yearsago)isoneofthefiveperiodsselectedforthePMIP4-CMIP6exercise.Itisacoldclimate,forcedbylowgreenhousegasconcentrationsandhighicesheetsinthenorthernmid-latitudes,inadditiontotheGreenlandandAntarcticicesheets.Inthiswork,wewillpresentthefirstPMIP4-CMIP6simulationsfortheLGM.Wewillexaminethelarge-scalefeatures:land-seacontrasts,polaramplification,positionofthejet-streams,intensityoftheAtlanticMeridionalOverturning Circulation and compare their changes to available reconstructions and to thosesimulated under 4xCO2 forcing. We will also analyse the impacts of the uncertainty in ice sheetreconstructions on these large-scale features. Indeed, the PMIP4-CMIP6 simulations have beendesignedtoaddressthisquestionbyallowingfortwodifferentreconstructions(Kageyamaetal.,GMD,2018).

Keywords:LGM,land,seacontrast,polaramplification,atmosphericlarge,scalecirculation,AMOC,4xCO2


1-P14

DiagnosisofmodelbiasimprovementofKIOSTEarthSystemModel

KimYoungho(1),YignNoah,Myong-InLee,HoJinLee,DaehyunKim

1-YoungHoKim(SouthKorea)

The Korea Institute ofOcean Science& Technology has developed the KIOST Earth SystemModel(KIOST_ESM) to participate in the CMIP6 and to develop a climate prediction system. Though theKIOST_ESM is based on the GFDL CM2.5 (Delworth et al., 2012), we applied a new atmosphericconvectionscheme(Park,2014),planetaryboundarylayerscheme(BrethertonandPark,2009),oceanmixedlayerscheme(Nohetal.,2016)andsoilrespirationscheme(Kimetal.,2018)totheKIOST_ESM.TheSouthernOceanwarmbias,whichisthetypicalclimatemodelbias,intheKIOST_ESM,hasbeenmostlydisappearedaswellasthemodelperformancefortheMJO(Madden-JulianOscillation),ENSO(El-NinoSouthernOscillation)hasbeengreatly improved.TheKIOST_ESM,however, still show thedoubleITCZ(IntertropicalConvergenceZone)biasandthecold-tonguebiasintheequatorialeasternPacific thoughmitigated somewhat. To diagnose themodel bias improvement of the KIOST ESM,sensitivityexperimentswereconducted.Asaresult,theSouthernOceanwarmbiasisgreatlyaffectedbyboththeatmosphericconvectionandoceanmixedlayerschemeswhilethemodelperformancefortheMJOismostlydominatedbytheatmosphericconvectionscheme.TodevelopaclimatepredictionsystemapplyingtheKIOST_ESM,oceandataassimilationwasapplied.Wehaveproducedanewlong-termclimatereanalysisdata.BycomparingwiththepreviousKIOSTclimatereanalysis,thenewclimatereanalysisshowsthatthesalinitybiasandwindbiasaroundtheequatorialPacificandSouthernOceanaresignificantly improved,whichappearstoberelatedtothe improvedDouble ITCZandSouthernOceanwarmbiasinKIOST_ESM.Currently,KIOSTisrunningtheKIOST_ESMtoparticipateinCMIP6.WewillfinishtheDECKsimulationinFebruary,2019andotherCMIPexperimentsin2019.

Keywords:EarthSystemModel,atmosphericconvection,oceanmixedlayer,southernoceanwarmbias,MaddenJulianOscillation


1-P15

DevelopmentofProcess-orientedDiagnosticsthroughNOAA'sClimateModelDevelopmentTaskForce

KrastingJohn(1),NeelinDavid(2),BarrieDaniel(3),GettelmanAndrew(4),MaloneyEric(5),MingYi(1),WingAllison(6)

1-NOAAGeophysicalFluidDynamicsLaboratory(UnitedStates),2-UCLA-Dept.ofAtmosphericandOceanicSciences(UnitedStates),3-NOAAClimateProgramOffice(UnitedStates),

4-NationalCenterforAtmosphericResearch[Boulder](UnitedStates),5-ColoradoStateUniversity[FortCollins](UnitedStates),6-FloridaStateUniversity(UnitedStates)

Coupled climatemodels and their components are subject to systematic biases. Continuedmodeldevelopment may improve these biases, but it is important to determine whether or not theunderlyingphysicalprocessesarecorrectandconsistentwiththemodelsimulations.Inothercases,model improvements introduce additional degrees of freedom that need to be evaluated andconstrainedbyobservations.Severalcoordinatedanalysiseffortsexistwithinthecommunitythataretargeted toward model evaluation and performance benchmarks. The MDTF Diagnostics Packagediffersfromexistingpackagesasitsprimaryfocusisthedevelopmentandstandardizationofprocess-orienteddiagnostics(PODs).Throughaproposal-drivenapproach,theModelling,Analysis,Predictions,and Projections (MAPP) Division of NOAA's Climate Program Office (CPO) currently supports thedevelopmentofPODsaimedatimprovingmodelsimulationsofdiversesetofquantities.Weanticipatedevelopmentover thenext threeyearsPODsrelatedto:ENSO, tropicalandextratropicalcyclones,Arctic sea ice, lake-effect snowfall, forced shifts in rainfall patterns, warm-cloud microphysicalprocesses,weathertyping,theNorthAmericanmonsoon,andocean-relateddiagnosticsincludingsealevel and the Atlantic Meridional Overturning Circulation. The proposed diagnostics will spantimescales from the fast timescales of parameterized convection and cloud processes throughseasonalpredictionanddecadal/centennialclimateprojections.Aprototypeanalysisframeworkexistsforrunningtheprocess-orienteddiagnosticsonGFDLandNCARmodeloutput,butweseektobroadenthisframeworkandincorporateresultsfromabroadercollectionofCMIPmodelsandexperiments.We also seek engagement with other community diagnostic efforts to improve interoperability,sharing of diagnostics, and to identify synergistic opportunities towards a more complete modelanalysisenterprisethatspansthe“development-to-production”continuum.

Keywords:modeling,diagnostics,analysis,process,oriented,bias


1-P16

AdvancingourunderstandingoftheimpactsofhistoricandprojectedlanduseintheEarthSystem:TheLandUseModelIntercomparisonProject

(LUMIP)

LawrenceDavid(1),HurttGeorge,ShevliakovaElena,CalvinKatherine,PongratzJulia,JonesChris,LawrencePeter,ArnethAlmut,BrovkinVictor,JonesAndrew


Human land-use activities have resulted in large changes to the Earth surface, with resultingimplications for climate. In the future, land-use activities are likely to intensify to meet growingdemandsforfood,fiber,andenergy.TheLandUseModelIntercomparisonProject(LUMIP)aimstofurtheradvanceunderstandingofthebroadquestionofimpactsofland-useandland-coverchange(LULCC)aswellasmoredetailedsciencequestionstogetatprocess-levelattribution,uncertainty,anddatarequirementsinmoredepthandsophisticationthanpossibleinamulti-modelcontexttodate.ThereisafocusonseparationandquantificationoftheeffectsonclimatefromLULCCrelativetoallforcings,separationofbiogeochemicalfrombiogeophysicaleffectsoflanduse,theuniqueimpactsofland-cover change versus land-management change,modulation of land-use impact on climate byland-atmospherecouplingstrength,andtheextentthatimpactsofenhancedCO2concentrationsonplantphotosynthesisaremodulatedbylanduse.LUMIPinvolvesthreemajorsetsofscienceactivities:(1) development of an updated and expanded historical and future land-use dataset, (2) modelexperimentsdesigned toaddress LUMIPquestions,and (3)developmentofmetricsanddiagnosticprotocols that quantify model performance with respect to LULCC. Experiments include idealizedcoupledandland-onlymodelsimulationsdesignedtoadvanceprocess-levelunderstandingofLULCCclimateimpacts,withafocusontheimpactoflandmanagement.Additionalexperimentsquantifythehistoricimpactoflanduseandthepotentialforfuturelandmanagementdecisionstoaidinmitigationofclimatechange.Here,wewillpresentinitialresultsfromtheLUMIPsimulations.

Keywords:land,usechange,landmanagement,mitigation


1-P17

TheEnergyExascaleEarthSystemModel(E3SM)version1:EvaluationandAnalysisofClimateSensitivity

GolazChris(1),CaldwellPeter,VanRoekelLuke,LeungL.Ruby

1-LawrenceLivermoreNationalLaboratory(UnitedStates)

This presentation describes the first version of the U.S. Department of Energy (DOE) new EnergyExascale Earth SystemModel (E3SMv1).We focus on the standard resolution of the fully-coupledphysical model designed to address DOE mission relevant water cycle questions. Its componentsincludeatmosphereandland(1-degreeresolution),oceanandsea-ice(60kminthemid-latitudesand30kmattheequatorandpoles),andrivertransport(0.5-degree)models.Thisbaseconfigurationwillalsoserveasafoundationforadditionalconfigurationsexploringhigherhorizontalresolutionaswellas augmented capabilities in the form of biogeochemistry and cryosphere configurations. TheperformanceofE3SMv1isevaluatedbymeansofastandardsetofCoupledModelIntercomparisonProject Phase 6 Diagnosis, Evaluation, and Characterization of Klima (CMIP6 DECK) simulationsconsisting of a long pre-industrial control, historical simulations (ensembles of fully coupled andprescribedSSTs)aswellas idealizedCO2forcingsimulations.ThemodelperformswelloverallwithbiasestypicalofotherCMIP-classmodels.WhiletheE3SMv1historicalensemblecapturesthebulkoftheobservedwarmingbetweenpre-industrial(1850)andpresent-day,thetrajectoryofthewarmingdivergesfromobservationsinthesecondhalfofthe20thcenturywithaperiodofdelayedwarmingfollowedbyanexcessivewarmingtrend.Usingatwo-layerenergybalancemodel,weattributethisdivergencetothemodel'sstrongaerosol-relatedeffectiveradiativeforcing(ERFari+aci=-1.65Wm-2)andhighequilibriumclimatesensitivity(ECS=5.3K).

Keywords:E3SMmodelfeatures,DECKexperiments,20thcenturywarming,climatesensitivity


1-P18

BESMdevelopmentstowardsCMIP6

PauloNobreandco-authors

1-InstitutoNacionaldePesquisasEspaciais-INPE,RodoviaPresidenteDutra,Km39,CachoeiraPaulista,SP12630-000,Brasil

TheimprovementsoftheBrazilianEarthSystemModel(BESM)towardsparticipatingintotheCMIP6project arepresented. Thepresentation is aimedat responding thequestion:Howdoes theEarthsystem respond to forcing scenarios? Lessons learned from the coupled ocean-ice-atmosphereresponsesofBESMtoglobal radiative forcingsarepresented.Resilientcharacteristicsof theglobalcoupled (e.g. atmospheric radiation and AMOC) and uncoupled (air temperature and ocean ice)oscillationsofBESMrepresentationoftheglobalcoupledclimatesystemarepresentedfromthousandyearslongexperimentsunderCMIP5design;tobereplicatedunderCMIP6.Amongtheimprovementspresentedarenewcloudmicrophysics, aswell as the inclusionof lagrangian capability forpassivetracers in the atmospheric component model. The ocean model is MOM5 under free surfaceassumption and the concurrent integration of ocean biogeochemical TOPAZmodel. Based on theintegrationswithMOM5/TOPAZ,adiscussionabouttheeffectsoftheincreaseofatmosphericCO2ontheoceanbiogeochemicaldynamicsarealsopresented.Aswell,highperformancecomputingaspectsarepresentedintheequation:computationtimexbinaryreproducibilityoncoupledmodellongruns


1-P19

ComparisonofEarthsystemmodelsthrougheffectivedocumentationofmodelsandinsightabouttheimplementationofforcings

PascoeCharlotte(1)(2),LawrenceBryan(3),GuilyardiEric(3)(4),GreensladeMark(4),HassellDavid(3),ChrisBlanton(5)

1-NationalCentreforAtmosphericScience[CEDA](NCAS)(UnitedKingdom),2-STFCRutherfordAppletonLaboratory(UnitedKingdom),3-NationalCentreforAtmosphericScience[Reading](NCAS)(UnitedKingdom),4-CNRSLOCEAN(France),5-NOAAGeophysicalFluidDynamics

Laboratory(GFDL)(UnitedStates)

ForeffectivecomparisonofEarthsystemmodelsinlargemulti-modelensemblesonemustfirsthaveanunderstandingofthemodeldesign,accurateknowledgeoftherequirementsofexperimentsandinsight intohowforcingconstraintsare implemented.Earthsystemmodelling involvesmanyroles,from those involving model construction, to those involving devising, executing, and exploitingnumericalexperiments. Inmostcasestheseroles involvemultiplepeople,andinmanycasesthesepeopleareindifferentinstitutionsandmayonlyinteractviapublishedinformation(whetherjournalpapers,technicalnotes,orwebsites).Thesemultiplepeople,sourcesandmethodologiesleadtoscopeformisinterpretationofintentandmisunderstandingofwhatwasactuallydone.WepresenttheES-DOCdocumentationmethodologyandinformationsharingecosystemwhichallowstheEarthsystemmodellingcommunitytoaccuratelydocumentandshareinformationaboutboththedesignoftheirmodelsandalsoabouthowtheirmodelsareconfiguredtoconformtocommonforcingprotocols.WedescribehowES-DOCconceptshavebeenappliedtothedesignphaseofCMIP6,improvingnotonlythedocumentationofCMIP6,butalsoimprovingtheexperimentsthemselves.Asaconsequencewebelieve itwillbeeasier forboththeMIPdesignersandparticipantstobeconfidentthattheyhaverequested,understoodand/orexecutedwhatwasrequired.

Keywords:CMIP6,ES,DOC,forcings,experiments,models,documentation


1-P20

GaugingsystematicbiasesacrossCMIPgenerations

PeterGleckler(1),DoutriauxCharles,LeeJiwoo,DurackPaul,ZhangYuying

1-ProgramforClimateModelDiagnosisandIntercomparison(UnitedStates)

ThePCMDIMetricsPackage(PMP) isanopen-sourcePython-basedcapabilityusedtoproduceanddocumentobjectivesummariesofEarthSystemModel (ESM)agreementwithobservations.This isaccomplishedviaadiversesuiteof relativelyrobusthigh-levelsummarystatisticsacrossspaceandtime scales. These results can be used to quantify model improvements across generations (e.g.,CMIP1-CMIP6) and to identify the relative strengths and weaknesses of different models. In thispresentationwehighlightnewmetricsthathavebeenincorporatedintothePMP,withaparticularinterest in thepossible reductionof systematicbiases in thenewergenerationof (CMIP6)models.Some of these results are based on recent PCMDI research, but a particular emphasis is oncollaborationswithanassortmentofinternationalexpertteams.Examplesincludemetricsforextra-tropicalmodesofvariability,ENSO,MJO,monsoonsandhighfrequencycharacteristicsofsimulatedprecipitation. PCMDI curates data from all generations of AMIP and CMIP going back to the early1990s.Theearlier(preCFconventions)dataisgraduallyandmeticulouslybeingbroughtuptomeetcommunitystandards.Thisworkpositionsustodocumentmodelimprovementsovernearly30yearsofmodeldevelopmentwithattentionnowturningtoCMIP6asnewsimulationsarebeingpublishedtotheEarthSystemGridFederation(ESGF).Thesemulti-modelandmulti-generationsummariescanbe helpful for the IPPC process and many other applications. Acknowledgements: This work wasperformed under the auspices of the U.S. Department of Energy by Lawrence LivermoreNationalLaboratory under Contract DE-AC52-07NA27344. We also acknowledge support from the U.S.DepartmentofEnergy,OfficeofScience,ClimateandEnvironmentalSciencesDivision,RegionalandGlobalModelingandAnalysisProgram.

Keywords:systematicerrors


1-P21

TheimportanceofdatareferencesinCMIP6datausageandIPCCclimateassessments

StockhauseMartina(1),LautenschlagerMichael

1-DeutschesKlimarechenzentrum[Hamburg](Germany)

Climate research like CMIP (CoupledModel IntercomparisonProject) and IPCC (IntergovernmentalPanelonClimateChange)AssessmentReports,currentlyintheir6thcycle,dependheavilyondata,whichiswell-curatedandremainsavailableonthelong-term(seeabstractbyJuckesetal.,2018).Toimprove the traceability of climate research andassessments, theCMIP6data sourcesneed tobereferenced.Datareferencesandthusgivingcredittodataproviders,theirfundersandcontributorshavebeenpartofprinciplesforGoodScientificPractice(e.g.DFG,2013)forseveralyears,buthavenotyetbecomeanaturalandeffectivepartofresearchworkflows.ForCMIP6forthefirsttimeinthehistory of CMIP, it is possible to cite the evolving CMIP6 data on model/MIP and experimentgranularities(StockhauseandLautenschlager,2017).Agoodexamplehowtocitedatainanarticle'sreferencelistisprovidedinthepaperbyDuracketal.(2018).TheCMIP6datacitationserviceenablesdatauserstofinddatareferences:intheESGF(EarthSystemGridFederation)portalsfordatasets,usingGoogle's Dataset Search, or on the FurtherInfoUrl (global attribute of CMIP6NetCDF files)landingpage.Accompanyinginitiativesandprojectssuchasthe“EnablingFAIRData”project(Stalletal.,2018)aredevelopingbestpractices for the integrationofdataandotherdigital resources intoauthorguidelinesandthegeneralresearchworkflowincollaborationwithacademicpublishersandrepositories.TherebydataFAIRness(Findable,Accessible,Interoperable,Reusabledata)intheEarthsciencesisadvanced.References:DFG(2013):https://doi.org/10.1002/9783527679188.oth1.Durack,P.J.,etal.(2018):https://doi.org/10.1029/2018EO101751.Stall,S.,etal.(2018):https://doi.org/10.1029/2018EO109301.Stockhause,M.,andM.Lautenschlager(2017):https://doi.org/10.5334/dsj-2017-030.

Keywords:datapublication,FAIRdata,CMIP6,IPCCAR6,GoodScientificPractice,CMIP6CitationService


1-P22

DiagnosingclimateresponseandfeedbackinresponsetoidealizedCO2forgininK-ACE

SunMinah(1),SungHyunMin,KimByeonghyeon,KimJisun,LeeJohan,ByunYoung-Hwa

1-NationalInstituteofMeteorologicalSciences(SouthKorea)

NIMS/KMA(NationalInstituteofMeteorologicalSciences/theKoreaMeteo-rologicalAdministration)hasparticipatedintheCMIP3and5usingECHO-GandHadGEM2-AOmodels.ToparticipateintheIPCC6threporttheK-ACE(KMA'sAdvancedClimateEarthsystemmodel),anEarthSystemModel,hasbeendevelopedbyNIMS/KMAandwesimulatedtheDECK(theDiagnostic,EvaluationandCharacterizationofKlima)experiments,whichisanessentialexperimentforCMIP6(Eyringetal.,2016).K-ACEconsistsofaphysicalbasismodelforUKESM,oceanmodelforMOM4.1,andseaiceforCICE.Inthisstudy,weanalyzedtheresponseofthemeanfieldsoftheatmosphereandoceantoCO2forcingandclimatesensitivityusingpre-industrialcontrolsimulationandCO2-forcedexperiments.TheCO2experimentsconsistof1%CO2/yrexperiment,whichincreasecarbondioxideby1%eachyear,andAbrupt4xCO2experiment,whichquadrupledtheconcentrationofcarbondioxideasof1850.EachCO2experimentwas prescribed with an only different carbon dioxide concentration of pi-control simulation andintegrationperiodwas151yearsfrom1850.WeusedtheECS(EquilibriumClimateSensitivity)andTCR(TransientClimateResponse)asawaytoquantifyandcompareresponsestoforcinginthemodelthroughCO2experiments,whichwerecomparedwiththemodelsparticipatinginCMIP5toassesstheperformanceofK-ACE.TheECSandTCRestimatedbytheCO2experimentsoftheK-ACEwas5.1K,and3.6K,respectively,andtheECSoftheCMIP5modelsrangedfrom2.1Kto4.7K,andtheTCRrangedfrom1.1Kto2.6K.TheresponsetotheCO2forcinginK-ACEwassomewhathigherthantheCMIP5range.

Keywords:KACE,ECS,TCR


1-P23

input4MIPs:GettingCMIPforcingdatainbettershape

DurackPaul(1),TaylorKarl(1),AmesSasha(1),LeeJiwoo(1)


input4MIPs(inputdatasetsforModelIntercomparisonProjects)isanactivitytomakeavailableviatheEarthSystemGridFederation(ESGF)theboundaryconditionandforcingdatasetsneededforCMIP6.Theprojectwasinitiatedin2016andhassubsequentlybecometheprimarysourceofinputdataforCMIP6modelingcenters.Variousdatasetsareneededforthepre-industrialcontrol(piControl),AMIP,1 percent compounding CO2 (1pctCO2), abrupt four times CO2 (abrupt-4xCO2) and historicalsimulations, and additional datasets are needed for many of the 23 CMIP6-endorsed modelintercomparisonprojects(MIPs)experimentsthatcomprisetheCMIP6project.ThispresentationwillprovideanoverviewoftheprogramevolutionandthestatusofforcingdatasetsbeingusedinCMIP6simulations.Akeyadvantageoftheproject, is thattheforcingdatasetsused insimulationswillbetraceable,withcomprehensivedocumentationprovidedbyES-DOCs.AstheseforcingdatasetsbeingusedinvarioussimulationsarealsohostedinESGFalongsidetheCMIP6data,theyarereadilyavailabletoanalystsduringsimulationassessment.Suchend-to-enddataprovenanceanddocumentationwillallowforanimprovedassessmentofthemodelsuiteresponsestoforcings.Inaddition,byformalizingandcentralizingtheforcingdatasetsbeingusedbysimulations,dataerrorsandinconsistencieshavebeenbetterandmorequicklyidentified,providinganimprovementtoforcingdatasetconsistencyinCMIP6whencontrastedtoearlierCMIPphases.

Keywords:CMIP6,forcing,volcano,change,variability,ocean,atmosphere,landuse


1-P24

AnoverviewofthefirstresultsfromScenarioMIPexperiments

TebaldiClaudia(1)


We will characterize the main features of the new scenario experiments organized under theScenarioMIP design. In particular, we will focus on analyzing globally/hemispherically/regionallyaveragedquantities(temperature,precipitation,indicesofextremes)andspatialpatternsofchangebytheendof the21stcentury.WewilluseCESM2outputand--asmuchaspossible --additionalavailablemodelsinordertoperformafirstexplorationofstructuralmodeluncertainties.AcomparisontoCMIP5RCPsreachingsimilarradiativeforcinglevelsbytheendofthecentury(RCP2.6and8.5inparticular)willbeperformed,inordertodetectsignificantdifferences,ifany,particularlyinpatternsof warming and precipitation changes by the end of the century, used by the well-establishedemulation techniques.Thequestionof the influenceof regional forcings (like land-useandaerosolemissions)whichhavechangedwithinthenewSSP-drivenemissiontrajectoriesisacentralonewithinthe new scenario framework, and even if we do not expect to be able to address it fully andexhaustivelybythisfirstanalysis,theconsiderationofthenewspatialpatternsandtheircomparisonwitholder,oracrossscenarios,mayprovideafirst-orderassessmentoftheroleofsuchforcings.

Keywords:SSPs,RCPs,Scenarios,ScenarioMIP,climatechangepatterns


1-P25

TheVulnerabilityImpactsAdaptationandClimateServicesAdvisoryBoard;TowardsusingCMIP6outcomeinVIACSapplications

TeichmannClaas(1),RuaneAlexC.(2)

1-ClimateServiceCenterGermany(GERICS)(Germany),2-NASAGoddardInstituteforSpaceStudies(UnitedStates)

TheposterwillpresentachievementsofPhaseIoftheVulnerability,Impacts,Adaptation,andClimateServices(VIACS)AdvisoryBoardforCMIP6,aswellastheenvisagedtransitionand initial ideasandperspectives forPhase II. TheVIACSAdvisoryBoardwas launched in2015as anendorsedCMIP6diagnosticMIP.ThemaingoalistofacilitatecommunicationbetweenVIACScommunities(whouseclimate model outputs) and climate modelers (who create outputs) across CMIP6 leadership andendorsedMIPs.Phase1oftheVIACSAdvisoryBoardcoincidedwiththedesignphaseofCMIP6,andthuswasprimarilyfocusedonprovidingVIACSperspectiveintotheconstructionandprioritizationofmodel experiments andmodel outputs for downstream applications. Themain focus included anevaluationof VIACS interest in all CMIP5output variables andMIP experiments. Thiswas done inrequests for additional tailored output variables; the creation of specific applications-orientedpackagesofoutputdatatobeavailableviatheCMIP6dataarchive;andanindicationofspecificVIACScommunities likely touseeachMIPsimulationexperiment.TheVIACSABalsocalled forenhancedevaluation of key impacts-relevant model diagnostics, perhaps including impacts-orientedobservationalproducts(Obs4Impacts).PhaseIIoftheVIACSABwillfocusonenhancedevaluationofCMIP6outputsthatarenowavailableandkeyonlinediagnosticsrelevanttoVIACSapplications.WewillalsoeasethetransitionofCMIP6modellers'findingstotheVIACScommunitiesinordertoraiseawarenessofknowncaveats,andwillengagedifferentVIACScommunitiestotracktheuseofCMIP6outputsincorporatingclimatevariability,predictability,scenarios,anduncertaintiesintheirresearchandapplications.Ofparticularinterestaretheinteractionsbetweenlong-termclimatetrendsandtheshiftingprofileofhazardsandrisk.

Keywords:vulnerability,impact,adatation,climateservices


1-P26

AIRSObs4MIPsV2DatasetandCMIP6ModelTemperatureandHumidityBiases

TianBaijun(1)

1-JetPropulsionLaboratory(UnitedStates)

The“ObservationsforModelIntercomparisonProjects”(Obs4MIPs)collects,organizesandpublishesvariouswell-establishedsatellitedatasetsforCoupledModelIntercomparisonProject(CMIP)modelevaluation.WehavegeneratedandpublishedtheAtmosphericInfraredSounder(AIRS)Obs4MIPsV2dataset including themonthlymean tropospheric air temperature, specific humidity, and relativehumidityprofiles fromSeptember2002toSeptember2016. In thispaper,wewilldiscuss theAIRSObs4MIPsV2datasetandsomepreliminaryresultsonCMIP6modeltemperatureandhumiditybiasesbasedonthisdataset.

Keywords:AIRS,Obs4MIPs,CMIP6


1-P27

Thenatureof60-yearoscillationsoftheArcticclimateaccordingtothedataoftheINMRASclimatemodel

VolodinEvgeny(1)

1-InstituteofNumericalMathematics[Moscou](Russia)

Usingthedataofpre-industrialexperimentwiththeINM-CM5climatemodelfortheperiodof1200years,westudythemechanismofnaturaloscillationsofArcticclimatewiththeperiodofabout60years.ItisshownthatforaquarteroftheperiodpriortotheArcticwarmingthereisaflowofAtlanticwaterintotheArcticoceanbeingmoreintensethanusual,thesalinityanddensityarelessthanusualnearthecoastandshelfborder.AstheresultofadvectionofAtlanticwaterafterArcticwarming,thewaternearthecoastandshelfborderbecomesmoresaltyandheavy,whichleadstoaweakeningofthe flow of Atlantic water and the change of oscillation phase. The conclusions are confirmed bycalculations of the generation of anomalies of temperature, salinity, and velocity of currents bydifferent terms, aswell as estimationof the contributionof various components to the changeofoscillationphase.

Keywords:mechanism,naturalclimatevariability,Arctic,NorthAtlantic


1-P28

ProjectcompliantclimatemodeloutputanalysiswithCDO's

WachsmannFabian(1),SchupfnerMartin(1),LegutkeStephanie(1)

1-DeutschesKlimarechenzentrum[Hamburg](Germany)

CMIP6definesadatastandardinordertofacilitatedataanalysisacrossresultsfromdifferentmodels.TheGermanFederalMinistryofEducationandResearch(BMBF) is fundingactivitieswhichprovidesupporttoGermanmodelinggroupsinmanyaspectsofstandardcompliantCMIP6dataproduction.OneaspectistoassistthemodelinggroupsatthecustomizationoftheCMIP6datarequest.Anotherfocus is on the extension of the ClimateDataOperator (CDO) toolkit in order to enablemodelinggroupstousetheCDOsforproject-standardcompliantformattingofmodeloutput,inadditiontodataanalysis anddiagnostics. TheCDOs is aworldwidepopular infrastructure softwaredevelopedandmaintained at theMax Planck Institute forMeteorology (MPI-M). It comprises a large number ofcommandlineoperatorsforgriddeddata,e.g.forstatistic,interpolation,orarithmetics.InorderfortheoperatorstobeusableforCMIP6dataprocessing,itsmetadataaswellasdatamodelisadaptedtothatofCMIP6.Inaddition,anewoperator‘cmor’hasbeendevelopedwhichcallsthesubroutinesoftheCMOR(ClimateModelOutputRewriter)librarydevelopedatPCMDI(ProgramforClimateModelDiagnosis&Intercomparison).Itallowstheusertocreatedifferentprojectstandardsbyreadingatablethatdescribes thetargetprojectstandardandwhich isprovidedasan inputargument.The ‘cmor’operator may be piped with other CDO operators, i.e. no temporary files are created. It is fullyintegratedintheCDOsoftwareinfrastructure,thereforetheuserbenefitsfromtheextensivesupportfacilitiesprovidedbytheMPI-MandtheDKRZ.

Keywords:Postprozessing,Tool,Standard


1-P29

TheCloudFeedbackModelIntercomparisonProject(CFMIP):CurrentstatusforCMIP6

WatanabeMasahiro(1),TselioudisGeorge(2),AndrewsTimothy(3),Bodas-SalcedoAlejandro(4),BonySandrine(5),BrethertonChristopher(6),BrientFlorent(7),CeppiPaulo(8),HwangYen-Ting(9),KangSarah(10),KayJennifer(11),KleinStephen(12),MauritsenThorsten(13),WebbMark(3),

WingAllison(14)

1-TheUniversityofTokyo(Japan),2-NASAGoddardInstituteforSpaceStudies(GISS)(UnitedStates),3-MetOfficeHadleyCentre(UnitedKingdom),4-MetOfficeHadleyCentre(UnitedKingdom),5-LMD/IPSL,CNRS(France),6-UniversityofWashington(UnitedStates),7-Centre

NationaldeRecherchesMeteorologiques(France),8-ImperialCollegeLondon(UnitedKingdom),9-NationalTaiwanUniversity(Taiwan),10-UlsanNationalInstituteofScienceandTechnology(SouthKorea),11-UniversityofColoradoatBoulder(UnitedStates),12-LawrenceLivermore

NationalLaboratory(UnitedStates),13-StockholmUniversity(Sweden),14-FloridaStateUniversity(UnitedStates)

TheprimarygoalofCFMIP is to investigate themain climate responses related to clouds, suchaschangesinthermodynamics,circulation,andthewatercycle,inordertoimproveourunderstandingof cloud-radiative feedback mechanisms in a changing climate and use this knowledge to assessclimatesensitivity.Toachievethesegoals,CFMIPbringstogetherclimatemodelling,observational,and process modelling communities and provides community support(https://www.earthsystemcog.org/projects/cfmip/).CFMIPstartedin2003asaCMIP-endorsedmodelintercomparisonprojectandnow,inrelationtoCMIP6,proposedthethirdphaseofCFMIP(CFMIP3)experiments (Webb et al. 2017 GMD). The CFMIP3 Tier 1 experiments consist of a compact andinexpensive set: amip+4K, amip4xCO2, amipFuture (building on the DECK amip experiment), andcorresponding idealizedmodelexperiments inaquaplanet configurations.Theseexperimentsareacontinuum from CMIP5/CFMIP2 and address the question “What are the physical mechanismsunderlying the range of cloud feedbacks and cloud adjustments predicted byGCMs?"Weprovideoutputs fromtheCFMIPObservationalSimulatorPackage(COSP)andotherprocessdiagnostics forquantitative evaluation of modelled clouds with observations. Currently, 15 modeling groups arecarryingouttheCFMIP3experimentsandareexpectedtouploadtheiroutputstoESGFbythemiddleof2019.Inaddition,thereareanumberofTier2experimentstoaddressfurthersciencequestions.Someexperimentsrequire150-yearAOGCMsimulationswhileothersfollowtheAMIPconfigurationexpandingtheTier1set.Finally,CFMIPsupportsinformallyproposedexperimentsinordertoexpandourunderstandingofcloudprocessesandtheircouplingwithcirculation.OurplansforCMIP6willbesummarizedandpositionedinthewidercontextoftheWCRPGrandChallengeonClouds,Circulationand Climate Sensitivity. Also, results from the CMIP6/CFMIP3 experiments will make a significantcontributiontotheIPCCAR6.

Keywords:clouds,precipitation,circulation,climatefeedback,radiativeforcing


1-P30

MainProgressoftheBeijingClimateCenterClimateSystemModel(BCC-CSM)fromCMIP5toCMIP6

WuTongwen(1),LuYixiong(1),YongjieFang(1),XinXiaoge(1),JieZhang(1)

1-BeijingClimateCenter,ChinaMeteorologicalAdministration(China)

Main progresses of Beijing ClimateCenter (BCC) climate systemmodel from thephase five of theCoupledModelIntercomparisonProject(CMIP5)toitsphasesix(CMIP6)arepresented,intermsofphysical parameterizations andmodels performance. BCC-CSM1.1 and BCC-CSM1.1m are the twomodelsinvolvedinCMIP5,andBCC-CSM2-MR,BCC-CSM2-HR,andBCC-ESM1.0arethethreemodelsconfiguredforCMIP6.Historicalsimulationsfrom1851to2014fromBCC-CSM2-MR(CMIP6)andfrom1851to2005fromBCC-CSM1.1m(CMIP5)areusedformodelsassessment.Theevaluationmatricesinclude(a)energybudgetattopoftheatmosphere, (b)surfaceairtemperature,precipitation,andatmosphericcirculationforglobalandEastAsiaregions,(c)seaiceextentandthicknessandAtlanticMeridionalOverturningCirculation(AMOC),and(d)climatevariationsatdifferenttimescalessuchasglobalwarming trend in the 20th century, stratospheric quasi-biennial oscillation (QBO),Madden-JulianOscillation(MJO)anddiurnalcycleofprecipitation.ComparedtoBCCCMIP5models,BCCCMIP6modelsshowsignificantimprovementsinmanyaspectsincluding:troposphericairtemperatureandcirculationatglobalandregionalscaleinEastAsia,climatevariabilityatdifferenttimescalessuchasQBO,MJO,diurnalcycleofprecipitation,andlong-termtrendofsurfaceairtemperature.

Keywords:BCCmodel,fromCMIP5toCMIP6,Historicalrun


1-P31

OverviewoftheGlobalMonsoonsModelIntercomparisonProject(GMMIP)forCMIP6

ZhouTianjun(1),TurnerAndrewG(2),KinterIiiJames(3)

1-InstituteofAtmosphericPhysics,ChineseAcademyofSciences(China),2-3.NationalCentreforAtmosphericScienceandDepartmentofMeteorology,UniversityofReading,Reading,UK;(United

Kingdom),3-4.CentreforOcean-Land-AtmosphereStudiesandDepartmentofAtmospheric,OceanicandEarthSciences,GeorgeMasonUniversity,(UnitedStates)

TheGlobalMonsoonsModel Inter-comparisonProject (GMMIP)hasbeenendorsedbyCMIP6.ThefocusofGMMIPisonmonsoonclimatology,variability,predictionandprojection.MorethantwentyinternationalmodellinggroupsarecommittedtojoiningGMMIP.GMMIPaimstoanswerfourprimaryscientificquestions:1)Whataretherelativecontributionsofinternalprocessesandexternalforcingthataredrivingthehistoricalevolutionofmonsoonofthelate-19ththroughearly-20thcenturies?2)Towhatextentandhowdoesatmosphere-oceaninteractioncontributetotheinterannualvariabilityand reproducibility of monsoons? 3) How can high-resolution and associated improved modeldynamicsandphysicshelptoreliablysimulatemonsoonprecipitationanditsvariabilityandchange?4)What is the effect of the orography of the Himalaya-Tibetan Plateau on the development andmaintenanceoftheAsianmonsoon?Similarly,whatistheimpactoforographyelsewhereonotherregionalmonsoons? Inorderofdecreasingpriority,GMMIPexperimentsaredesigned toexamine:Tier-1)Modelskillatsimulatingtheclimatologyandinterannual-to-multidecadalvariabilityofglobalmonsoons forcedbySSTsduring thehistorical climateperiod ;Tier-2)The roleof the InterdecadalPacificOscillationandAtlanticMultidecadalOscillationindrivingvariationsoftheglobalandregionalmonsoons;Tier-3)Theeffectsoflargeorographicterrainontheestablishmentofthemonsoons(suchasexaminingtherelativeinfluenceoftheHimalayasandTibetanPlateauontheregionalcomponentsoftheAsianmonsoon).TheoutputsoftheCMIP6DECK,historicalsimulationandotherendorsedMIPswillbeusedinthediagnosticanalysisofGMMIPtogiveacomprehensiveunderstandingoftherolesplayed by different external forcings, and potential improvements in the simulation of monsoonrainfallathighresolution.SomeemergingresultsbasedonthesubmittedCMIP6datawillbeshownandthepotentialcontributiontoIPCCWG1AR6willbehighlighted.

Keywords:GlobalMonsoonsModelIntercomparisonProject,GMMIP,internalvariability,externalforcing

Session2-ForcingandFeedbacks

2-P01

ClimatesensitivityandfeedbacksintheIPSL-CM6climatemodel

AlbrightAnnaLea(1),BonySandrine(1),DufresneJean-Louis(1),MusatIonela(1),FairheadLaurent(1),IdelkadiAbderrahmane(1),PoppMax(1),VialJessica(2)

1-LaboratoiredeMeteorologieDynamique(LMD/IPSL)(France),2-MaxPlanckInstituteforMeteorology(Germany)

AnewversionoftheIPSL-CMclimatemodelhasbeendevelopedforCMIP6.ThisversiondiffersfromtheCMIP5versionbyitsenhancedhorizontalandverticalresolutions,andbysignificantchangesinitsphysicalparameterizations, including in the representationsof clouds, convection, turbulence, andradiation. This model version will participate in CMIP6, CFMIP3, and RCE-MIP coordinatedexperiments. Here we present an analysis of the climate sensitivity, feedbacks, and adjustmentsproducedbythisnewmodelinarangeofidealizedexperimentsandconfigurations:coupledocean-atmosphere,atmosphere-only,aqua-planet,andRCE.Weinvestigatethereasonsforthehighclimatesensitivity (4.7 K) produced by this model, diagnose feedbacks and adjustments using the kernelmethod,andassesstheirrobustnessacrossasuiteofexperiments.Particularattentionwillbepaidtothe responseof tropical clouds to increasedCO2andglobalwarming.After analyzing the IPSL-CM6model,weperformamulti-modelanalysisbasedonavailableCMIP6simulations.

Keywords:climatesensitivity,feedbacks,clouds,tropicalclouds,adjustments,IPSL,CM6


2-P02

Anewmulti-variablebenchmarkforLastGlacialMaximumsimulations

CleatorSeanF.(1),HarrisonSandyP.(2),NicholsNancyK.(3),PrenticeIainColin(4),RoulstoneIan(1)

1-UniversityofSurrey(UnitedKingdom),2-UniversityofReading(UnitedKingdom),3-UniversityofReading(UnitedKingdom),4-ImperialCollegeLondon(UnitedKingdom)

We present a new global reconstruction of seasonal climates at the LastGlacialMaximum (LGM,21,000 yr BP)made using 3-D variational techniques,with pollen-based site reconstructions of sixclimatevariablesandtheensembleaverageofthePMIP3/CMIP5simulationsasaprior.WeassumeasetofGaussiancorrelationsintheerroroftheprior,inordertoproduceaclimatereconstructionthatis smoothed both from month to month and from grid cell to grid cell. The pollen-basedreconstructions includemeanannual temperature (MAT),meantemperatureof thecoldestmonth(MTCO),meantemperatureofthewarmestmonth(MTWA),growingseasonwarmthasmeasuredbygrowingdegreedaysaboveabaselineof5¡C(GDD5),meanannualprecipitation(MAP)andamoistureindex (MI), which is the ratio of MAP to equilibrium evapotranspiration. Different variables areavailableatdifferentsites,butourapproachbothpreservesseasonalrelationshipsandallowsthefullsuiteofseasonalclimatevariablestobederivedateachlocation.Weaccountfortheecophysiologicaleffects of low [CO2] on vegetation inmaking reconstructions of precipitation andmoisture index,whichresultsinthereconstructionofwetterclimatesthanwouldotherwiseseemfromthevegetationcomposition. Finally, by comparing the error contribution to the final reconstruction, we provideconfidence intervals on these reconstructions and delimit geographical regions for which thepalaeodataprovidenoinformationtoconstraintheclimatereconstructions.ThenewreconstructionswillbeusedasarobustbenchmarkforevaluationofthePMIP4/CMIP6entry-cardLGMsimulations.

Keywords:Benchmark,Reconstruction,LGM,DataAssimilation,Pollen


2-P03

BiogeochemicalfeedbacksinCMIP6EarthSystemModels

CollinsWilliam(1),LamarqueJean-Francois(2),SchulzMichael(3)

1-UniversityofReading(UnitedKingdom),2-NationalCenterforAtmosphericResearch(UnitedStates),3-MeteorologiskInstitutt(Norway)

InthisposterweshowthefirstresultsanalysingbiogeochemicalfeedbacksfromthefirstfewEarthSystemmodeltohavesubmittedresultstoCMIP6.Weexplainhowananalysisframeworkhasbeendevisedtodirectlycomparethesebiogeochemicalfeedbackswiththephysicalclimatefeedbacksandalsotoframethemintermsofcarbonbudgets.TheCMIP6EarthSystemmodelsincorporateavarietyoffeedbacksthatmayaffecttheoverallsensitivityoftheirequilibriumsurfacetemperatureresponseto a climate forcing. As well as the carbon cycle feedback, a multitude of other biogeochemicalfeedbackshavealsobeenidentifiedcouplingthefatesofshort-livedandlong-livedspecies.Climatechangecanaffectthechemistry,transportandremovalofreactivegases(suchasmethaneandozone)andaerosols,andimpactthesourcestrengthofnaturalemissionsofmethane,ozoneprecursorsandaerosols. Theevaluationof their importance for future climate changehasbeenverypatchy,withquantificationforafewofthefeedbacksassessedinAR5(Ciaisetal.2013)fromaveryfewmodels,andafewsinglemodelstudiesofspecificfeedbackssincethen(e.g.Nowacketal.2015).Thissciencewill take a large step forward with the CMIP6 project. All Earth System models are required toparticipate in the DECK 4xCO2 and 1% CO2 so for the first timewewill analyse themulti-modelresponsesofchemistry,naturalemissionsandremovalofreactivegasesandaerosolstoaspecifiedchangingclimate.Tocompletethefeedbackloop,theclimateeffects(effectiveradiativeforcing-ERF)of thesespeciesarequantified inAerChemMIP.SpecificAerChemMIPexperimentsaredesignedtoquantifytheclimateeffectsofnaturalemissionsof:dust,seasalt,DMS,biomassburning, lightningNOx,andbiogenicVOCs.

Keywords:Feedbacks,Aerosols,Chemistry


2-P04

Energyconservingandphysicallyconsistentmethodforisolatingtheimpactsofsea-icechangesinamulti-modelframework

CvijanovicIvana(1),LevineXavier(1),OrtegaPablo(1),LucasDonald(2)

1-BarcelonaSupercomputingCenter-CentroNacionaldeSupercomputacion(Spain),2-LawrenceLivermoreNationalLaboratory(UnitedStates)

Modelingapproachesaimedatisolatingtheimpactsofsea-icelossmoreoftenthannottendtoimposenon-physicalenergyfluxperturbationsinordertoachievethechangesinsea-icecover.Thismakesitdifficulttodiscerniftheobservedclimaticresponsereallyoriginatesfromthesea-icecoverchangesorifitisalteredbytheimposedenergyfluxperturbations.Whileother,easytoimplementapproaches,such as ‘painting the ice black' or ‘turning the ocean white' have the advantage of not imposingspurious heat flux perturbations, they are unphysical and fail to capture the range of processesresulting from the actual disappearance or recovery of sea-ice cover. Here we illustrate a newapproachthatisbothenergyconservingandphysicallyrealistic,anddiscussitsapplicabilityinamulti-model framework.Ourmethodemploys smallperturbations to selected sea-icephysicsparametervaluesthatresultinsea-icecoverchanges.Thesea-icephysicsparameterperturbationsareimposedonly within their respective expert defined ranges, allowing for a physically realistic experimentaldesign(withinthelimitsofthemodeltheyareappliedin).Importantly,thisapproachdoesnotrequireanykindofartificialalterationofhighlatitudeenergyfluxestoachievethesea-icecoverchangesthusalso ensuring energy budget conservation. The impacts of the sea-ice physics perturbations arediscussedbasedonsimulationsperformedwithCCSM4andEC-Earthmodels,withtheaimofsettingabenchmarkforothermulti-modelapplications.Preliminaryanalysissuggeststhatnotonlydothesea-icephysicsparameterperturbationspresentausefultoolforisolatingtheimpactsofsea-icelossandsea-iceforcingonclimatebutthattheyalsoprovideanewpathforimprovementsinmodelingofsea-icechanges.

Keywords:energybudgetconservation,seaicephysicsparameterperturbations,seaiceloss


2-P05

AssessingthelinearityandadditivityofwatercyclechangessimulatedbyCNRM-CM6-1

DouvilleHervé(1),RoehrigRomain(1),Saint-MartinDavid(1),VoldoireAurore(1),AliasAntoinette(1),ColinJeanne(1),DecharmeBertrand(1),SéférianRoland(1),

SénésiStéphane(1),RibesAurélien(1),GeoffroyOlivier(1),ChadwickRobin(2),WebbMark(2)

1-CNRM,UniversitédeToulouse,Météo-France,CNRS(France),2-MetOfficeHadleyCentre(UnitedKingdom)

The confidence in global and regional projections of freshwater resources is rooted in theunderstanding of how the different components of the global water cycle respond to differentradiativeforcings.TheCFMIPandDAMIPapproachesareincreasinglyusefulforthispurpose.WithinCFMIP,bothtier1andtear2simulationshavebeendesignedtoassesstheadditivityand/orlinearityofthewatercycleresponsetoincreasingCO2concentrations.ThiswillbeillustratedwithpreliminaryresultsoftheCNRM-CM6-1globalclimatemodel.Beyondglobalmeanprecipitationandtheso-calledhydrologicalsensitivity,othermorepolicy-relevantfeaturesoftheglobalwatercyclewillbeexplored.ThefastadjustmenttoCO2increasewillbedistinguishedfromtheslowerresponsetotheglobaloceanwarming.Moreover, itwill be shown that the tier 2AGCMexperiments enable thebreakdownofsimulatedwatercyclechanges intomultiplecontributions, includingtheradiativeandphysiologicalCO2 effects, the role of a uniformwarming of sea surface temperature (SST), and the role of theanomalous SST pattern or of the sea ice retreat. Within DAMIP, the hydrological responses toanthropogenic and natural radiative forcings can be assessed, aswell as their additivity and theirpotentialcontributionstorecentlyobservedwatercyclevariations.ThesmallensemblesachievedwithCNRM-CM6-1arenotnecessarilysufficientfortheattributionofregionalchangesbutalreadysuggestpromisingattributionstudiesattheglobalscale.

Keywords:watercycle,hydrologicalsensitivity,linearity,drivers,attribution


2-P06

NorthAtlanticresponsetoexternalforcingandroleoftheanthropogenic-aerosols

GastineauGuillaume(1),MenegozMartin,RobsonJon,BellucciAlessio,CassouChristophe

1-LOCEAN(France)

Results from the IPSL-CM6 large-ensemble of historical simulations and associated detection-attribution model intercomparison project (DAMIP) simulations indicate that the anthropogenic-aerosolshavecausedanintensificationoftheAtlanticMeridionalOverturningcirculation(AMOC)attheendofthetwentiethcentury.Theanthropogenic-aerosolshavealsoimportantimpactsontotheglobalatmosphericcirculation.TheseeffectshaveovercometheinfluenceofgreenhousegasesthathavecontributedtoreducetheAMOC.Both, theradiativeeffectofanthropogenic-aerosolandtheassociateddynamicalchangeswillbeexploredwithintheavailableCMIP6database.TheimpactfortheEuropeanclimatewillbealsoexplored.WewillfinallyquantifyiftheaerosolsforcingexplaintheobservedhistoricalAtlanticMultidecadalvariability,andinvestigatetheassociateduncertainty.

Keywords:detection,externalforcings,Climatevariability,attribution,AMOC


2-P07

ChangesinGrowthRateandSeasonalCycleAmplitudeofColumnCO2inCMIP5modelsandSatelliteData

GierBettinaK.(1)(2),BuchwitzMichael(2),CoxPeterM.(3),EyringVeronika(2)(1),FriedlingsteinPierre(3),ReuterMaximilian(2)

1-DLRInstitutfürPhysikderAtmosphere(Germany),2-UniversityofBremen,InstituteofEnvironmentalPhysics(IUP)(Germany),3-CollegeofEngineering,MathematicsandPhysical

Sciences,UniversityofExeter(UnitedKingdom)

EarthSystemModels(ESMs)displayalargespreadinmodeledatmosphericCO2,resultinginuncertainprojectionsforthefuture.UtilizingtheEarthSystemModelEvaluationTool(ESMValTool,Eyringetal.,2016),weevaluateemissiondrivenhistoricalsimulationsfromCMIP5and,astheybecomeavailable,CMIP6,withsatellitedata.ThecomparisonfocusesonthegrowthrateandseasonalcycleamplitudeofXCO2,whicharegoodcandidatestobeusedinthedevelopmentofemergentconstraints.EmergentconstraintsarerelationshipsacrossanensembleofmodelsbetweenanunobservableEarthsystemsensitivity and an observable trend or variation in the contemporary climate. Both models andobservationsshowcharacteristicseasonalcyclesinCO2,withlowervaluesinthesummerwhenstrongphotosynthesis causes plants to absorb carbon dioxide, and higher values in the winter whenphotosynthesisstops.Theseasonalcycleamplitude,definedasthepeak-to-troughamplitudeoftheseasonalcycle,thereforedependsonthestrengthofthesummerphotosynthesisandthedurationofthegrowingseasonandislargerintheNorthernthanintheSouthernHemisphere.Theobservationsandmodelswillbeanalyzedfurthertodeveloprobustemergentconstraintsformulti-modelclimatemodelprojections,similartoWenzeletal.2014and2016.InthosestudiessensitivitiesofatmosphericCO2 were used to constrain the long-term sensitivity of tropical land carbon storage to climatewarmingandthemagnitudeoftheCO2fertilizationeffect.References:Eyring,V.,etal.,Geosci.ModelDev.,9,1747-1802(2016).Wenzel,S.,P.M.Cox,V.Eyring,andP.Friedlingstein,JGRBiogeoscience119(5),794-807,(2014).Wenzel,S.,Cox,P.M.,Eyring,V.,andFriedlingstein,P.,Nature538,499?501,(2016).

Keywords:SeasonalCycle,GrowthRate,CMIP5,XCO2,CO2,satelliteobservations


2-P08

ImprovingdustforcinginGFDLESM4bycouplingdustemissionfromthedynamiclandmodel(LM4.1)anddepositiontotheocean

biogeochemistrymodel(COBALT).

GinouxPaul(1),DunneJohn(1),HorowitzLarry(1),KrastingJohn(1),MalyshevSergey(1),NaikVaishali(1),ShevliakovaElena(1),StockCharles(1)

1-NOAAGeophysicalFluidDynamicsLaboratory(UnitedStates)

MineraldustparticlesinteractwiththeEarth'sClimateSystembyabsorbingandscatteringsolarandterrestrial radiations. Deposition of dust on open oceans provides nutrients for phytoplanktonblooming. These interactions may feedback on dust emissions mostly through modulation of thehydrologicalcycle.SeveralstudieshavedocumentedthefailureofCMIP5modelstoreproducelong-termvariabilityofdust concentration,whichmaybe related to vegetation changes in response toperturbationsinthehydrologicalcycle.Becausevegetationanditscharacteristicsareaffectedbybothdirecthuman influences (e.g.deforestation)andclimatechange (e.g.diebackduetodrought), it isimportant to include vegetation dynamics as one of the predictors of dust emission to accuratelysimulatepast,present,andfuturedustforcing.ThenewGFDLEarth'sSystemModel(ESM)version4containsfullyinteractiveandconsistentdustlifecycle:fromitsemissionscalculatedbythedynamicland model (LM4.1) to its long-range transport in the atmospheric model (AM4) followed by itsdepositiontotheoceanwhere it feedsthetracersof thebiogeochemistrymodel (COBALT). In thisposter,wepresentananalysisoftheamplificationofdustforcingbyvegetationandland-usechanges,anditsconsequencesonradiativeforcingandoceanproductivity,basedonGFDLESM4simulationsinsupportofCMIP6.

Keywords:Aerosolforcing,MineralDust,EarthSystemModel,DynamicVegetationandlanduse,Oceanbiogeochemistry


2-P09

Theimpactoffixedozonein4xCO2simulations

HardimanSteven,AndrewsTimothy(1),BushellAndrew(2),DunstoneNick,DysonHarold,JonesGareth,KnightJeff,NeiningerErica,RingerMark,ScaifeAdam,WoodRichard

1-MetOfficeHadleyCentre(UnitedKingdom),2-MetOffice(UnitedKingdom)

CMIP6protocolsuggestsprescribing1850ozoneconcentrationsin4xCO2simulations.Thisleadstoamismatch between the dynamical tropopause, which rises due to climate change, and the ozonetropopause,whichremainsfixed.Theresultisunphysicallyhigh(stratospheric)ozoneconcentrationsin the upper troposphere, leading to biases in high level cloud, cold point temperature, andstratosphericwatervapour.IntheMetOfficeclimatemodelthisimpactstheclimatesensitivityandsurfaceclimate.Inthefuture,thisuseofpre-industrialozoneconcentrationsin4xCO2simulationsislikelytocausesimilarproblemsinotherclimatemodelswithoutinteractiveozoneschemes.

Wedescribeamethodtointeractivelyredistributeozonein4xCO2simulations,whichremovestheinconsistencybetweenthedynamicalandozonetropopauseheightswhilstretainingtheprescribedCMIP6ozonedistributionsascloselyaspossible.Thisremovesanyunphysicalconsequencesofthetropopause mismatch from such simulations, whilst still allowing a fair comparison against otherCMIP6modelsimulations.

Aftereachmodelyear,themonthlymean,zonalmean,dynamicaltropopauseisformedbasedontheprevious two model years. At each latitude, the ozone tropopause is defined at 1km below thedynamicaltropopausebysettingozoneconcentrationsthereto80ppbv,andsmoothingappropriately.The mass of ozone removed from the troposphere is added to the stratosphere, by multiplyingstratosphericozoneconcentrationsbythesamepercentageeverywhere.Thustotalmassofozoneisconserved.Thisredistributionisthenappliedtothe3Dmonthlymeanozoneconcentrations,andthusazonalstructuresintheCMIP6prescribedozonearealsopreserved.Theclimatemodelisthenrunforthefollowingyear,usingthisredistributedozone,andthenthewholeprocessisrepeatedwithintheautomatedsystem.

We present results from pre-industrial and 4xCO2 simulations. This method can also be used for1%/yearCO2andtransientclimatesimulations.

Keywords:ozone,climatesensitivity


2-P10

GlobalmonsoonchangesinCMIP6GMMIPpacemakerexperiments

HuangXin(1),ZhouTianjun(1),LiuBo(1)

1-LASG,InstituteofAtmosphericPhysics,ChineseAcademyofSciences(China)

Changesinmonsoonprecipitationarevitalforannualrainfallandthuswatersupplyforagriculture,foodandenergyoftheworld'smosthighlypopulatedanddevelopingregions.TheGlobalMonsoonsModelInter-comparisonProject(GMMIP)hasbeenendorsedbytheCoupledModelInter-comparisonProject(CMIP)asoneoftheparticipatingmodelinter-comparisonprojects(MIPs)inthesixthphaseofCMIP(CMIP6).Thetier-2experimentsaredesignedtoexaminetherolesoftheInterdecadalPacificOscillation (IPO)andAtlanticMultidecadalOscillation (AMO) indrivingvariationsof theglobalandregionalmonsoons.UsingtwosetsofGMMIPexperimentsbasedonNCARCommunityEarthSystemModel (CESM) version1.2, namely historical experiment (hist) and tropical Pacific SST restoringexperiment(hist-resIPO),wefirstlyinvestigatetherelativecontributionsofinternalmodeandexternalforcingtotheobservationaldecreasingtrendofIndiansummermonsoonprecipitationinthesecondhalfofthe20thcentury.Weshowevidencethattheobserveddecreasingtrenddoesnotarisefromthe response toexternal forcingalone.Rather, thedrying itself appears to result fromsuppressedmoisture convergence, associated with an IPO-induced weaker Walker circulation, resulting indepressedconvectionandadvectionofrelativelydrierairfromtheEurasianlandmass.Wethenextendtoglobalmonsoontoexaminethecoherenthistoricalvariability.Differencesbetweenhistandhist-resIPO experiments indicate that global monsoon index in hist-resIPO is better correlated withobservationthanhistexperimentinthepastcentury.BasedonthecomingCMIP6,weexpectthelargesuite of GMMIP pacemaker experiments carried out by different modeling centers for betterunderstanding the mechanisms of precipitation changes in the global monsoon system anddisentanglingtheinfluenceofSSTvariabilityoninterannual-to-interdecadalmonsoonsvariabilityfromexternalforcing.

Keywords:Globalmonsoon,GMMIP,Pacemakerexperiment,Externalforcingandinternalvariability


2-P11

Responseofpermafrostunderdifferentsolargeoengineeringmethods

JiDuoying(1),ZhangQian(1)

1-BeijingNormalUniversity(China)

Weconductamodel-basedassessmentofchanges inpermafrostextentandcarbonstorageundersolardimmingandstratosphericaerosolinjectionmethodsformingpartoftheGeoengineeringModelIntercomparison Project Phase 6(GeoMIP6). The solar dimming experiment G1 is designed tocompletely offset the global mean radiative forcing due to a CO2-quadrupling experiment(abrupt4_CO2),whileinG4experiment,theradiativeforcingduetotherepresentativeconcentrationpathway4.5(RCP4.5)scenarioispartlyoffsetbyalayerofsulphateaerosolsinthestratosphere.BothG1 and G4 geoengineering simulations lead to lower temperatures at higher latitudes, but withresidual polar amplification effect. It is not clear how the permafrost extent and its carbon stockchangeundersolardimmingandstratosphericaerosolscenarios.Inthisstudy,weuseBNU-ESMwithimprovedpermafrostparameterizationschemetosimulatetheresponseofpermafrosttodifferentsolargeoengineeringmethods.

Keywords:earthsystemmodel,geoengineering,permafrost


2-P12

Projectionuncertaintiesinthenextgenerationofclimatemodelsandensembles

KnuttiReto(1),LorenzRuth(1),BrunnerLukas(1)

1-EigenössischeTechnischeHochschuleZürich(Switzerland)

Asourunderstandingimproves,climatemodelscontinuetoincreaseincomplexitytosynthesizeallthatknowledge.Whenmoreobservationsbecomeavailableandcomputational capacity increases,moreprocessesareconsideredandmodelsarerunathigherresolutions.Thehopeisthatmodelswillconvergetoreality.Butaretheyreally,andhowdoweknow?Whileindividualmodelsimproveinhowthey represent quantitieswe can observe, their spread in future projections has been remarkablystable over decades, both globally and locally. Model agreement on future projections has notincreasedfromCMIP3toCMIP5.However,modelspreadinanensembleofopportunitymaybetoobigifpoormodelsareincluded,ortoosmallifallmodelsaremissingimportantfeedbacks.HerewefirstsystematicallyquantifymodelagreementwithobservationsonkeyquantitiesliketemperatureandprecipitationtoassessmodelperformanceofthenewCMIP6ensemble,comparedtoCMIP3andCMIP5.Second,weestimatemodelspreadinfutureprojections,usingconsistentmethodologiesandtakingintoaccountdifferencesinemissionscenarios,forthethreemostrecentintercomparisons,andcomparethesetoanuncertaintyassessmentusingmodelweightingandemergentconstraintsthattakes into account model performance and model dependence. The separation of differences inindividualmodelresponsestoforcingfromdifferencesinhowanensembleingenerated,differencesinemissionscenarios,andcontributionsfromnaturalvariabilityisanimportantstepinunderstandinghow andwhy the newCMIP6 projections differ fromearlier CMIP intercomparisons. It provides anecessary step towards a quantification of projection uncertainty in future IPCC reports, and anopportunity to reflect on the potential and challenges in reducing these uncertainties in futuregenerationsofmodelsandensembles.

Keywords:projection,uncertainty


2-P13

CMIP5subtropicalmarinelowcloudfeedbackinterpretedthroughaunifiedpredictiveindex

KoshiroTsuyoshi(1),KawaiHideaki(1),YukimotoSeiji(1)

1-MeteorologicalResearchInstitute[Tsukuba](Japan)

Ithasbeen recognized formanyyears that lowcloud feedback isoneof themajor sourcesof theuncertainty in future climate change. The strength of local temperature inversion is a key climateparameterthatcontrolsthe lowcloudamount.Theestimated inversionstrength(EIS),apredictiveindexbasedonthedifferenceinpotentialtemperaturesbetweenthe700-hPalevelandthesurface,isstronglypositivelycorrelatedwithlowcloudamount.Veryrecently,wedevelopedanewpredictiveindexforlowcloudamount,theestimatedcloud-topentrainmentindex(ECTEI).ThisisarefinementofEISthattakesintoaccountacloud-topentrainmentcriterionusingaspecifichumiditygapbetweenthe700-hPalevelandthesurface.Inthisstudy,thechangesinlowcloudamountandthesepredictiveindicesinawarmerSSTclimateareinvestigatedoverthesubtropicaloceansforeightCMIP5modelsusingamipandamip4Ksimulations.MostmodelsshowthatthelowcloudamountdecreaseswhenSSTincreases,despiteEISincreases.Ontheotherhand,ECTEIdecreases.Thesurfacespecifichumidityismainly controlled by the saturation specific humidity at the surface andmuch greater than thespecifichumidityinfreetroposphereinmostcases.Therefore,thehumiditygapincreaseswithSSTandleadstoreductionoflowclouds.TheresultsuggeststhatthiseffectexceedstheincreaseinlowcloudsduetoenhancementoftemperatureinversionstrengthinawarmerSSTclimate.Furthermore,usingamipFutureandamip4xCO2simulations,theinfluenceofSSTpatternandlowcloudfeedbackinincreased CO2 climate will be discussed. This work clearly indicates the usefulness of ECTEI forinterpretationoffactorscontrollinglowcloudamountandlowcloudfeedback.

Keywords:lowcloudfeedback,lowcloudamount,inversionstrength,cloudtopentrainment


2-P14

Inter-modelspreadininstantaneousradiativeforcingacrossmultipleclimatedrivers

KramerRyan(1),SodenBrian(2),SmithChristopher(3),MyrheGunnar(4),ForsterPiers(3)

1-NASAGoddardSpaceFlightCenter(UnitedStates),2-RosenstielSchoolofMarineandAtmosphericScience(UnitedStates),3-UniversityofLeeds(UnitedKingdom),4-Centerfor

InternationalClimateandEnvironmentalResearch[Oslo](Norway)

Theeffectiveradiativeforcingisawidelyusedconceptfordiagnosingtheresponseoftheclimatetodifferent forcing agents and has proven to be important for understanding why climate modelsrespond differently to identical emission scenarios. It consists of two parts - the instantaneousradiativeforcing,whichmeasurestheperturbation inradiativefluxesduesolelytoachange intheforcing agent, and rapid adjustments, which measure the radiative perturbations induced by theatmosphere'sresponsetotheinstantaneousradiativeforcing.Recenteffortstodiagnoseuncertaintyineffectiveradiativeforcinghavelargelyfocusedonrapidadjustments,inpartbecauseinstantaneousradiativeforcingisnotroutinelycalculatedinclimatemodelsimulations.Hereweuseradiativekernelstosystematicallydiagnoseinstantaneousradiativeforcinginmulti-modelensemblesofclimatemodelsimulationsfromCMIP5andthePrecipitationDriverandResponseIntercomparisonProject(PDRMIP)undervariousidealizedforcingscenarios.Weshowthatdifferencesininstantaneousradiativeforcing,not rapid adjustments, are the dominant contributor to inter-model spread in effective radiativeforcing. Because instantaneous radiative forcing is relativelywell constrained by radiative transfertheory,itprovidesatractablesolutiontoreducingtheintermodelspreadineffectiveradiativeforcingand,consequently,futureclimatechangeprojections.

Keywords:instantaneousradiativeforcing,radiativekernels,effectiveradiativeforcing,rapidadjustments


2-P15

Consistencyandrobustnessofemergentconstraintsforequilibriumclimatesensitivity

LauerAxel(1),EyringVeronika(1)(2),SchlundManuel(1)

1-DLRInstitutfürPhysikderAtmosphäre(Germany),2-UniversityofBremen,InstituteofEnvironmentalPhysics(IUP)(Germany)

AnemergentconstraintreferstotheuseofobservationstoconstrainasimulatedfutureEarthsystemfeedback.Ifphysicallyplausiblerelationshipscanbefoundbetween,forexample,changesoccurringonseasonalorinterannualtimescalesandchangesfoundinmodelsimulationsofanthropogenically-forcedclimatechange,anobservationalconstraintofmulti-modelprojectionsmightbepossible.Thisoffers thepossibility to reduceuncertainties in climateprojections. Emergent constraints canhelpguidingmodeldevelopmentontoprocesses crucial to themagnitudeand spreadof future climatechangeandtopointoutfutureobservationalpriorities.In this study, selected emergent constraints for equilibrium climate sensitivity published in recentliterature are tested in termsof their consistency and robustness. For this purpose, the emergentconstraints have been implemented into the Earth System Model Evaluation Tool (ESMValTool,https://www.esmvaltool.org/)toallowforconsistentcalculationsandprocessingofmodeldataandobservationaldatasets.AfocuswillbeontestingconsistencyofemergentconstraintsacrossdifferentphasesofCMIP.ResultsfromemergentconstraintsderivedfromCMIP5simulationswillbecomparedagainsteachotherandcross-checkedagainstresultsobtainedfromtheCMIP3modelensembleandfrom CMIP6 as the simulations become available. Uncertainties introduced by using specificobservational datasets will be assessed by comparing the results obtained with alternativeobservations of the same variables and, where available, using uncertainty estimates for theobservations.Theaimofthisstudyistoassesstheconsistencyandrobustnessofselectedemergentconstraints for equilibrium climate sensitivity and to identify opportunities and challenges of thisapproachwiththehelpofCMIP6simulations.

Keywords:emergentconstraint,ESMValTool,equilibriumclimatesensitivity


2-P16

ComparisonsofSimulatedCloud-Radiation-Circulation-PrecipitationCouplingoverTropicalPacificOceansinGlobalClimateModelsbetweenCMIP5

andCMIP6:PreliminaryResults

LiJuilinLi(1),LeeWe-Liang(2),RichardsonMark(1),JiangJonathon(1),WangYi-Hui(3),YuJia-Yuh(4),FetzerEric(1),StephensGraeme(1)

1-JetPropulsionLaboratory(UnitedStates),2-RCEC,Sinaca(Taiwan),3-CaliforniaPolytechnicStateUniversity,(UnitedStates),4-DepartmentofAtmosphericSciences,NationalCentral

University(Taiwan)

Most CMIP3 and CMIP5 models exclude falling ice from their radiation code. Through controlledsimulations with CESM1-CAM5, which is one of the few models that allows falling ice radiativeinteractions,wefoundthatfallingicereducesbothoutgoinglongwaveandsurfaceincidentshortwavein convective regions such as the ITCZ and SPCZ. This stabilizes the local atmosphere and forcesnonlocalresponsesinsurfacewinds,temperaturesandprecipitation,counteractingseveralcommonCMIP3andCMIP5biases.ToourknowledgemostCMIP6modelsalsoexcludeprecipitationradiativeeffects.HereweexamineavailableCMIP6fullycoupledhistoricalsimulationsusingCloudSat-CALIPSO,CERESandthenewERA5reanalysis.WefocusonthetropicalPacificasitisaheatengineforglobalclimate system.Our preliminary results indicate continued biases associatedwith the exclusion ofprecipitatinghydrometeors,particularlyintheITCZ,SPCZandwarmpool.Inaddition,aweaklowlevelmeanflowfromthewarmpoolassociatedwithmoistureconvergenceisfound.Thiswesterlybiaswithwarmandmoistair transportmightcontributetothemodel'snortheastwardoverextensionof theSPCZandtheconcomitantwarmbiasinseasurfacetemperaturesandexcessiveprecipitation.

Keywords:Cloud,radiation,precipitationcoupling


2-P17

ClimatesensitivityandcloudfeedbacksinCESM2andE3SM

MedeirosBrian(1),BacmeisterJulio(1),HannayCecile(1),KleinStephen(2),ZelinkaMark(3)

1-NationalCenterforAtmosphericResearch(UnitedStates),2-LawrenceLivermoreNationalLaboratory(UnitedStates),3-LawrenceLivermoreNationalLaboratory(UnitedStates)

TheCESM2andE3SMarecloselyrelatedclimatemodels.Theiratmosphericcomponentsarenearlythe same, but E3SM hasmore vertical levels, and themodels have been tuned separately. Othercomponentsdiffersubstantially,notablytheoceansareentirelydifferent.Thisstudyexploresclimatesensitivityandcloudfeedbacksinthesetwomodels.WestartfromtheDECKexperimentswithbothmodelstoestablishthatusingconventionalmeasures,thesemodelsproducesimilarclimatesensitivity(whichisaround5K,onthehighsideoftheCMIP5ensemble).Changesincloudradiativeeffectsuggestthat the response of subtropical stratocumulus and transitional cloud types are an importantcontributiontothehighsensitivity.Adynamicalregimesanalysisisusedtoexploretheconnectiontocloudregimesinmoredetail.DoesthesimilaritybetweenCESM2andE3SMsignalreducedparametricuncertaintyinthisfamilyofmodels?Itappearsnot,andthesimilaritymaybeacoincidence.Thisisdemonstrated using CESM2 with a different set of parameters but a substantially lower climatesensitivity.Additionalexperimentsfurthersuggestthevariableclimatesensitivityisassociatedwiththeresponseoflow-leveltropicalclouds,justashasbeendescribedinmulti-modelensembles.

Keywords:climatesensitivity,cloudfeedback


2-P18

UKESM1:Afirstassessmentofthepre-industrialtopresent-dayanthropogenicforcinganditsattributiontodifferentforcingagents

O'connorFiona(1),FolberthGerd(1),HardacreCatherine(1),JohnsonBen(1),JohnsonColin(1),MulcahyJane(1),RobertsonEddy(1),RumboldSteven(2),SellarAlistair(1),JonesColin(3)

1-MetOfficeHadleyCentre(UnitedKingdom),2-NationalCentreforAtmosphericScience(UnitedKingdom),3-NationalCentreforAtmosphericScience(UnitedKingdom)

Aquantitativeunderstandingoftheroleofdifferentforcingagentsinbothhistoricalandfutureclimatechange remains a key motivation and scientific question for the forthcoming 6th CoupledModelIntercomparisonProject(CMIP6).FundamentaltothisquestionistheimpactontheEarth'sradiativebalanceofphysicalandchemicalperturbationsfromanthropogenicactivities.Inthiswork,effectiveradiativeforcings(ERFs)arequantifiedfordifferentanthropogenicforcingagentswiththeUK'sEarthSystemModel,UKESM1.ByusingasinglemodellingframeworkandadoptingtheprotocolfromtheRadiativeForcingModelIntercomparisonProject(RFMIP),pre-industrialtopresent-dayanthropogenicERFsarecalculatedconsistentlyforallclimateforcers.Theforcingagentsconsideredherearethelong-lived well-mixed greenhouse gases, stratospheric and tropospheric ozone, aerosols, and land usechange. Inparticular,additionalUKESM1simulationsareusedtoattribute themethaneERF,asanexample, to forcingbymethane, troposphericandstratosphericozone,aerosols,andstratosphericwatervapour.Resultsshowthattheozoneforcingattributabletomethane isclosetozero,duetochangesinstratosphericandtroposphericozoneoffsettingeachother.However,thepre-industrialtopresent-day change in methane gives rise to a positive aerosol forcing through aerosol-cloudinteractionsbychangingtherelativecontributionsofthedifferentsulphurdioxideoxidationpathwaysandtheaerosolsizedistribution.

Keywords:effectiveradiativeforcing,Earthsystem,anthropogenic,forcingagents,UKESM1


2-P19

Usingsimpleindicesofglobalclimatechange:thePMIP4andCMIP6simulationsandpaleoclimatedatatoevaluatehowtheEarthsystem

respondstostrongforcings

Otto-BliesnerBette(1),BradyEsther(1),TomasRobert(1),FengRan(2),ZhuJiang(3)

1-NationalCenterforAtmosphericResearch(UnitedStates),2-UniversityofConnecticut(UnitedStates),

3-UniversityofMichigan[AnnArbor](UnitedStates)

Simulationsofpastclimatesprovide'out-of-sample'testsofEarthsystemmodelsbeingusedtopredictfuture climate trajectories.Data is available to confirm the sensitivityof thesemodels to forcings,albeitwith thedatabecomingsparser the fartherback in time.The4thphaseof thePaleoclimateModelingIntercomparisonProject(PMIP4)willprovidethisperspectivetoCMIP6andIPCC,asPMIPhas done since the 1990s. This poster will present a new set of paleo-simulations that are beingcompletedwithCESM2:samemodel,sameresolutionastheDECKandfuturescenariosimulations.These include: theearlyEoceneClimaticOptimum(~53-51millionyrsago,4xCO2,deepmip-stand-4xCO2),mid-Pliocene(3.2millionyrsago,400ppmCO2,midPlioceneEoi400),LastInterglacial(127,000yrsago,lig127k),LastGlacialMaximum(21,000yrsago,190ppmCO2,lgm),andmid-Holocene(6000yrs ago, midHolocene). Simple indices of the surface temperature response to high CO2/lowCO2/seasonality of incoming solar radiation: Northern versus Southern Hemisphere temperaturechange,NorthernHemispheremeridionaltemperaturegradient,land-oceancontrasts,tropicalPacificeast-westzonalgradient,polaramplification,willbecomputedforthesepastclimatesimulationsandcompared to paleoclimate data. In addition, these indiceswill be computed and compared to theCESM24xCO2(abrupt-4xCO2)simulation.Amulti-modelcomparisonwithpreviousPMIPsimulationsforthemid-Holocene,LastGlacialMaximum,andmid-Pliocenewillhighlighttheevolutionofmodelperformancesinsimulatingclimatestatesdifferentfromthemodernone.Astheybecomeavailable,these simple indices will be computed for the suite of PMIP4/CMIP6 simulations expected to besubmittedbyinternationalmodelinggroupsintimefortheIPCC.

Keywords:PMIP4,model,datacomparison,pastclimate


2-P20

InfluenceofCMIP6ForcingonHistoricalandDecadalHindcastSimulationswithMPI-ESM

PohlmannHolger(1),BittnerMatthias(1),WienersKarl-Hermann(1),ModaliKameswarrao(1),KadowChristopher(2),IllingSebastian(2),PankatzKlaus(3),MüllerWolfgang(1),

MarotzkeJochem(1)

1-Max-Planck-InstitutfürMeteorologie(Germany),2-InstitutfürMeteorologie,FreieUniversitätBerlin,Berlin(Germany),3-DeutscherWetterdienst(Germany)

In this study we compare an ensemble of historical simulations using CMIP5 external forcing,performedwithin theGermanMedium-termClimate Predictions (MiKlip ? fona-miklip.de) project,with another ensemble using CMIP6 forcing performed for the contribution to the CMIP6. Bothensemblesareproducedusing thesamehigh-resolutionversionofourMaxPlanck InstituteEarth-SystemModel(MPI-ESM-HR).Theanalysisofthemodelperformanceanduncertaintiesofdifferentmetricsandtrendsisappliedtothecompletehistoricalperiod(1850-2005)aswellassub-periods(e.g.,1960-2005)toaccountfortheincreaseofobservations.Furthermore,wecomparetwoensemblesofdecadalhindcastsimulationswitheachother,alsoproducedwiththesameMPI-ESM-HRmodel,usingeitherCMIP5orCMIP6forcing.Theanalysisincludesdifferentstate-of-the-artpredictionskillmetricsfordecadalclimatehindcastevaluation.Toaccountforthesmallensemblesizethesignificanceoftheresultsisanissue.Weestimatetheimpactoftheinitializationbycomparingthepredictionskillrisingfromtheexternalforcinginthehistoricalsimulationswiththepredictionskillinthedecadalhindcastsimulations.ThecomparisonoftheresultscausedbythedifferentexternalforcingbearsustheuniquepossibilitytosystematicallyevaluatetheimpactofthechangefromCMIP5toCMIP6forcingonMPI-ESMsimulations.

Keywords:DecadalHindcastSimulations,HistoricalSimulations,CMIP6forcing


2-P21

DerivingEarthSystemFeedbacksonmultipletimescales

SandersonBenjamin(1)

1-CentreEuropéendeRechercheetdeFormationAvancéeenCalculScientifique(France)

Foroveradecade,theglobalequilibriumresponseofaclimatemodeltoincreasinggreenhousegaseshas been often been categorized by a single temperature - a "climate sensitivity" calculated byassuminga linear forcing-feedback relationship in a simulationwhereatmosphericCarbonDioxideconcentrationsarequadrupledandthesystemisallowedtoequilibrate.However,recentworkhasshownthatfeedbacksvaryasafunctionoftimeandspatialpattern,callingintoquestiontheutilityand accuracy of this approach.Here,we propose a simplemethod to decompose the equilibriumresponseoftheclimatesystemintofeedbackswhichoccuronarangeoftimescalesusingcurrentlyavailablesimulations.WhenappliedtotheCMIP5ensemble,theapproachsuggeststhatEarthSystemresponse can be categorized by a discrete range of adjustment timescales. Existing 150 yearsimulations constrain response on shorter timescales, but leave uncertainty in feedbacks on thecenturytimescale.However,ouranalysissuggeststhattheequilibriumresponseofsomemodelsinthe CMIP5 archive is underestimated by up to 2K by assuming a single feedback strength. Thedecompositionalsoallows foraveryaccuratepredictionofTransientClimateResponseusingonlyinformationfromaquadrupledCO2simulation.ExistingemergentconstraintsonTCRareshowntobroadly constrain the faster timescales of climate response,with the slowermode only becomingdetectableafteremissionsceaserising.

Keywords:climatesensitivity,feedbacks,transientclimateresponse,carbonbudgets


2-P22

HistoricalaerosolforcingdiagnosisinCMIP6,AerChemMIPandAeroCommodels

SchulzMichael(1),MyhreGunnar(2),CollinsBill(3),LamarqueJean-Francois(4)

1-NorwegianMeteorologicalInstitute(Norway),2-CenterforInternationalClimateandEnvironmentalResearch[Oslo](Norway),3-UniversityofReading(UnitedKingdom),

4-NationalCenterforAtmosphericResearch[Boulder](UnitedStates)

AnewgenerationofaerosolmodelshasbeenimplementedinstateoftheartCMIP6models.Atthesame time CMIP6 and AerChemMIP experiments are for the first time coordinated with detailedaerosolmodelinginAeroCommodels.Specialattentionhasbeenmadetoproposerefineddiagnosticsin CMIP6 anAeroCom,which allow for anunderstandingof bias in aerosol forcing and associatedaerosolparameters.TheaerosolforcinghistoryinavailableCMIP6outputisdocumentedandanalysedas a function of emissions, aerosol mixing ratios, optical properties, cloud properties. AeroComhistoricalfixedSSTsimulationsareinvestigatedtoseeifcoupledAOGCM'saresignificantlydifferentindescribingaerosolforcinghistory.ObservationaldataavailableintheAeroComdatabaseareusedtoevaluate theseparametersandanybiasagainstobservations.Theconsequencesofbias for theuncertaintyinaerosolforcinghistoryarediscussed.

Keywords:aerosolevaluation,AeroCom,AerChemMIP,historicalforcing


2-P23

PolaramplificationandatmosphericmeridionalenergytransportinCMIP6DECKsimulations

SemmlerTido(1)

1-AlfredWegenerInstituteforPolarandMarineResearch-AWI(GERMANY)(Germany)

HowdoesArcticandAntarctic sea icedevelop in thenewsetofCMIP6simulations,howdoes theatmospheric meridional energy transport change and how does the vertical structure of theatmospherechange?Toanswerthisquestion,asmanyavailableCMIP6historical,1%CO2peryear,and 4*CO2 simulations as possible will be evaluated regarding these parameters. It has beenacknowledgedthattheenergytransportfromthemid-latitudesintothepolarregionsisoneimportantcontributortopolaramplification.Whiletheatmospherictemperaturesignatureofseaicedeclineisabottom-heavywarming,i.e.mostpronouncedintheboundarylayer,themeridionalenergytransportcanincreasethetemperaturefurtheraloft.Thereforeitisofinteresttoassessthemeridionalenergytransportand itsrole inpolaramplificationfromtheCMIP6simulations.Theanalysis inchanges inpolaramplificationandmeridionalenergytransportandchangingtemperaturegradientsindifferentatmospheric layers will also help to stimulate discussion about possible impacts on mid-latitudeweathercharacteristicssuchassynopticactivityandextremeevents.

Keywords:Polaramplification,energytransport


2-P24

Globalsoilmoisture-carbonfeedbacks:PlannedjointanalysesfromLS3MIPandC4MIP

SoniaI.Seneviratne(1),VictorBrovkin(2),PierreFriedlingstein(3),ChrisJones(4),VivekArora(5),HyungjunKim(6),andGerhardKrinner(7)

1)ETHZurich,Zurich,Switzerland;2)MPI-Meteorology,Hamburg,Germany;3)U.Exeter,Exeter,UK;4)MetOffice,Exeter,UK;5)CanadianCentreforClimateModellingandAnalysis,Victoria,Canada;

6)U.Tokyo,Tokyo,Japan;7)CNRS-IGE,Grenoble,France

Theterrestrialbiosphereisamajorcarbonsinkasitabsorbsabout25percentofthehuman-inducedCO2 emissions. However, the rate of land carbon uptake remains highly uncertain, and showssubstantialspreadinEarthSystemModel(ESM)projections.Recent studies have revealed that terrestrial water storage substantially affects changes inatmospheric CO2 concentrations both on interannual time scale in observations (Humphrey et al.2018),andinlong-termprojectionsinESMs(Greenetal.2019).However,DynamicGlobalVegetationModels(DGVMs)havebeenfoundtounderestimatethisrelationship(Humphreyetal.2018).ThispresentationwillprovideanoverviewoftheseresultsandofplannedanalyseswiththeLS3MIPandC4MIPexperimentsthatwillassesssoilmoisture-carbonrelationshipsintheCMIP6ensemble.References:Green,J.K.,S.I.Seneviratne,A.M.Berg,K.L.Findell,S.Hagemann,D.M.Lawrence,P.Gentine,2019:Largeinfluenceofsoilmoistureonlong-termterrestrialcarbonuptake.Nature,565,476–479Humphrey,V.,J.Zscheischler,P.Ciais,L.Gudmundsson,S.Sitch,S.I.Seneviratne,2018:SensitivityofatmosphericCO2growthratetoobservedchangesinterrestrialwaterstorage.Nature,560,628-631.

Keywords:


2-P25

ToolsforcomputingradiativeforcingandradiativefeedbacksfromCMIP6output

SodenBrian(1),KramerRyan(1)

1-UniversityofMiami(UnitedStates)

Radiativeforcingisafundamentalquantityforunderstandingbothanthropogenicandnaturalchangesin climate. However, its explicit calculation can be both time consuming and computationallyexpensive.Asaresult,mostmodelingcentersneglecttocomputetheinstantaneousradiativeforcing(IRF),leavingusersofthemodeloutputunawareofdifferencesinIRFbetweenmodelsimulationsevenwhenintegratedunderidenticalemissionscenarios.Recentstudieshavehighlightedtheimportanceofsuchdifferencesincontributingtoinconsistenciesinmodelprojectionsofbothhistoricalandfutureclimate change. Radiative feedbacks are also pivotal in regulating the response of the climate toexternal forcings, however differences in the methodology used to compute feedbacks betweenvariousmodelinggroupscanbiasthereportedfeedbacks.AspartoftheNOAAModelDiagnosticsTaskForce, a set of Python-based tools have been developed to allow users to compute both IRF andradiativefeedbacksusingstandardmodeloutputfromCMIP6DECKandrelatedMIPs.Thesetoolsuseradiativekernelstodecomposechangesinbothtop-of-atmosphereandsurfaceradiativefluxesintocontributionsfromindividualfeedbacktermsandtheIRF.Byprovidingaconsistentmethodologyforcomputingbothfeedbacksandforcing,thesetoolscanhelptofacilitateabetterunderstandingofthecauseofintermodeldifferencesintheresponsetoforcings.Theposterwilldescribehowtoaccessthesoftware, reviewthecalculationof thesediagnosticsandhighlight theirapplicationforavarietyoftopicsrelatedtoclimatesensitivity,hydrologicsensitivity,andinteractionsbetweenclouds,aerosols,radiationandtheatmosphericcirculation.

Keywords:Radiativeforcing,RadiativeFeedbacks


2-P26

RepresentationandtrendsofOrganicAerosolsinCMIP6AerChemMIPSimulationsusingtheWholeAtmosphereCommunityClimateModel

(WACCM6)

TilmesSimone(1),LamarqueJean-Francois(1),EmmonsLouisa(1),GettelmanAndrew(1),HodzicAlma(1),MillsMike(1),KinnisonDouglas(1),ParkMijeong(1)


The Community Earth System Model Version 2 (CESM2) Whole Atmosphere Community ClimateVersion6(WACCM6),isparticipatinginCMIP6DECKandotherMIPexperiments.WACCM6includescomprehensivechemistryforthetroposphere,stratosphere,mesosphere,andlowerthermosphereandsimulates interactiveaerosols inbothtroposphereandstratosphere.Anewsecondaryorganicaerosol (SOA) scheme has been implemented inWACCM6 to allow the coupling between aerosolprecursoremissionsandSOAformation,usingthevolatilitybinset(VBS)modelframework.Thenewapproachallows interactivecouplingbetweenbiogenicemissionsandorganicaerosols (OA),whichalso interact with chemistry, radiation and therefore climate. Here, we focus on Tier 1 modelsimulations of the Aerosol ChemistryModel Intercomparison Project (AerChemMIP).We contrasthistoricalsimulationswithandwithoutpre-industrialneartermclimateforcersforbothAMIP-typeand fully coupled simulations between 1850 and 2014, in order to explore the importance ofanthropogenicemissionsontrendsofOAgloballyandfordifferentregionsandtheirradiativeimpact.Besides standard AerChemMIP simulations, we performed additional simulations withoutanthropogenic OA precursor emissions with WACCM6, between 1960 and 2014, and discuss thecontributions of differentOAprecursor sources to the evolution ofOAover different regions.WefurtherpresentanevaluationofsimulatedOAforpresentdayusingATomaircraftobservationsandMODISAODsatelliteobservationsandplantoincludeotheravailablemodelresultsintheanalysis.

Keywords:AerChemMIP,OrganicAerosols,WACCM


2-P27

HistoricalandFutureChangesinTroposphericOzoneusingaparameterisedApproachwiththeCMIP6emissionsdataset

TurnockSteven(1),WildOliver(2),O'connorFiona(1)

1-UnitedKingdomMetOffice[Exeter](UnitedKingdom),2-LancasterEnvironmentCentre(UnitedKingdom)

A simple parameterisation of source-receptor relationships for tropospheric ozone (O3) has beendevelopedbasedonsimulationsfromarangeofmodelsparticipatingintheTaskForceonHemisphericTransportofAirPollutants(TF-HTAP)experiments.Perturbationstoprecursoremissions(NOX,COandVOCs)andmethane(CH4)abundanceareusedtoquantifychangestotroposphericO3globallyandacross16regions.Thecontributionfromlocal,remoteandCH4sourcestosurfaceO3canbeprovidedforeachreceptorregionalongwithanestimateofuncertaintybasedonthespreadoftheresultsfromthemodelsusedasinput.HereweusechangesinO3precursoremissionsandCH4abundancesfromthe CMIP6 emissions dataset to predict both the historical and future impact on surface andtroposphericO3.NinedifferentSharedSocio-economicPathways(SSPs)fromthescenariodatasetareusedtopredictthefutureimpactontroposphericO3from2014to2050.Scenariosthatincludeweakairpollutant and climatemitigationmeasuresare inadequate in limiting the futuredegradationofsurfaceO3airqualityandenhancementofnear-termclimatewarmingoverallregions.Middle-of-the-road and strong mitigation scenarios are able to reduce both surface O3 concentrations and O3radiative forcing, providing benefits to future air quality and near-term climate forcing. TheparameterisationisasimpletoolandisusedheretostudytheimpactsandassociateduncertaintiesofthelocalandhemisphericemissioncontrolstrategieswithintheCMIP6scenariosonbothsurfaceO3airqualityandthenear-termclimateforcingbytroposphericO3.

Keywords:ozone,airquality,climate,radiativeforcing


2-P28

AnalysisofCMIP6atmosphericmoisturefluxesandtheimplicationsforprojectionsoffuturechangeinregionalrainfall

WattersonIan(1),RashidHarun(1),KeaneRichard(2)

1-CSIROClimateScienceCentre(Australia),2-MetOffice(UnitedKingdom)

TheprojectionsforfuturechangeinregionalrainfallpresentedbytheIPCCFifthAssessmentReportarelargelybasedonclimatemodelsimulationscontributedtoCMIP5.Formanyregions,inparticularAustralia,thewiderangeofpossiblechangesinrainfallfromthemodels,bothincreasesanddecreases,posesamajorresearchchallenge.Partofthevariationinsimulatedrainfallchangeisassociatedwiththehorizontaltransportoftropicalmoisture.ThewatercyclewasdiscussedinAR5,butamulti-modelanalysisofmoisturefluxeswasnotpresented.Australianstudies,atleast,werehamperedbythelackofsuitabledatafromCMIP5.ThenewCMIP6archiveincludesvertically-integratedmoisturefluxdata,so itshouldbefeasibletodoamulti-modelassessmentofthis flux, forthefirst time.Thevariablenamesareintuawandintvaw,andtheyareincludedintheEmontable.GiventhatthefluxhasbeenrequestedonlybyPMIP,theassessmentmayneedtofocusontheDECKexperiments.Westudydatafrom1pctCO2,fromavailablemodelsincludingtheU.K.MetOffice'sGC2andAustralia'sACCESS-CM2.Feasibly,thepresentclimatecanberepresentedbyearlydecadesofthis150-yearexperiment.Thestandardizedchangeor‘trendperdegree',underrisingCO2concentrations,canbedeterminedfromtheregressionofyearlyvalueswiththeglobalmeantemperature.Weassesstheflux,theatmosphericmoisturebudget,andrainfall.Itisfoundthatthefuturechangeintheconvergenceofthefluxishighlyspatially correlated with the precipitation change. From such multi-model results, prototypeprobabilisticprojectionsofthechangeinthesevariableswillbedeveloped.Thereisthepotentialforsomeconstraintonprojectionsoffuturerainfallbasedonanimprovedunderstandingofthemoisturebudget.Thiswouldhaveaconsiderablebenefittousersofsuchprojections.Keywords:Moisturetransport,Regionalrainfall,Projections,Multimodelanalysis


2-P29

AerosolforcingofextremesummerdroughtoverNorthChina

ZhangLixia(1),WuPeili,ZhouTianjun

1-InstituteofAtmosphericPhysics,ChineseAcademyofSciences(China)

ThefrequencyofextremesummerdroughthasbeenincreasinginNorthChinaduringthepastsixtyyears,whichhascausedseriouswatershortages.Itremainsunclearwhetheranthropogenicforcinghas contributed to the increasingextremedroughts.Using theNationalCenters forEnvironmentalPrediction and the National Center for Atmospheric Research (NCEP/NCAR) re-analysis data andCoupledModelIntercomparisonProjectPhase5(CMIP5)modelsimulationswithvariouscombinationsofhistoricalforcings,theauthorsinvestigatedthedrivingmechanismbehindtheobservedchanges.Metrological drought is usuallymeasured by precipitation anomalies,which show lower fidelity incurrentclimatemodelscomparedtolargescalecirculationpatterns.BasedonNCEP/NCARre-analysis,a linear relationship is firstlyestablishedbetween theweakest regionalaverage850hPa southerlywindsandextremesummerdrought.Thismeridionalwindsindex(MWI)isthenusedasaproxyforattribution of extreme North China drought using CMIP5 outputs. Examination of the CMIP5simulationsrevealsthattheprobabilityoftheextremesummerdroughtswiththefirstpercentileofMWIfor1850-2004underanthropogenicforcinghasincreasedby100%,onaverage,relativetoapre-industrialcontrolrun.ThemorefrequentoccurrenceofextremelyweakMWIsordroughtoverNorthChinaisascribedfromweakenedclimateandEastAsiansummermonsoon(EASM)circulationduetothedirectcoolingeffectfromincreasedaerosol.

Keywords:drought,aerosolforcing,extremes,Asianmonsoon

Session3-Uncertainty,biasesandconstraints

3-P01

Consistentborealwinterforecastskillincurrent(non-CMIP6)climatepredictionsystemsonseasonalscales

AcostaNavarroJuanCamilo(1),OrtegaPablo,BatteLauriane,MassonnetFrancois,TorralbaVeronica,SmithDoug,Doblas-ReyesFrancisco

1-BarcelonaSupercomputingCenter-CentroNacionaldeSupercomputacion(Spain)

Climatepredictionsystems(i.e.initializedclimatemodels)showhightomoderatepredictivecapacityon seasonal scales on important large scale climatic phenomena (e.g. ElNino SouthernOscillation(ENSO),ortheNorthAtlanticOscillation).Skillfulspatialpredictionofclimaticvariablesinthenorthernhemisphereistypically lowerandhighlyvariabledependent.Forexamplesurfacetemperature,sealevelpressureandprecipitationshowadecreaseinskillinthatdescendingvariableorder.AnanalysisofCNRM-CM6,GloSea5andEC-Earth3.2onseasonalforecastmoderesultsinconsistentlargescaleandspatialdeterministicandstatisticalforecastskillacrossmodelsandvariables,despiteofhavingvery different systematic biases.Weexplored the sources of predictability in the systems and thepossible mechanisms leading to consistent results across models, despite of the systematic biasdisagreement.

Keywords:temperature,Circulation,Precipitation,DCPP,Predictability,Predictions,Seasonal


3-P02

AframeworkforunderstandingthequalityofSouthernoceancirculationincoupledclimateandEarthSystemModelsimulations.

BeadlingRebecca(1),RussellJoellen(1),StoufferRon(1),GoodmanPaul(1),MazloffMatthew(2)

1-UniversityofArizona(UnitedStates),2-ScrippsInstitutionofOceanography(UnitedStates)

Thelarge-scalecirculationintheSouthernOcean(SO)isacriticalcomponentofEarth'sclimatesystemdue to its dominant role in exchanging carbon and heat between the ocean and atmosphere andtransforming global oceanwatermasses. Evaluating the ability of fully coupled climatemodels toaccurately simulate SO circulation and properties relative to observations is crucial for buildingconfidence in climate model projections and advancing model fidelity. However, identifying andanalyzingmodel biases relative to observationally-basedmetrics is just the first step. By analyzingmultiplebiasescollectivelyacrosslargeensemblesoffullycoupledclimateandEarthSystemModels,mechanismsgoverning thediversemean-state SO circulation foundacross coupledmodels canbeidentified.Theanalysispresentedhere (1)assesses theabilityofa largeensembleof fullycoupledclimate models contributed to both CMIP5 and CMIP6 to simulate observationally-based metricsassociatedwithanaccuraterepresentationofpresentdaySOcirculation,and(2)presentsaframeworkbywhichthequalityof thesimulationcanbecategorizedandmechanismsgoverningtheresultingcirculation canbededuced. Theanalysis highlightshowdifferent combinationsofbiases in criticalmetrics including zonal mean westerly surface wind stress strength and position, properties andtransportofNorthAtlanticDeepWater (NADW)entering theSO inbothdensityanddepth space,propertiesofintermediateandmodewaters,andsurfacebuoyancyfluxes,resultsindistinctmean-state Antarctic Circumpolar Current transports. Furthermore, how these distinct mean-state SOcirculationsacrossmodelsimpactthecarbonandheatbudgetoftheSOisquantified.

Keywords:CMIP5,EarthSystemModels,SouthernOceanCirculation,CMIP6,AntarcticCircumpolarCurrent


3-P03

EvaluationofCMIP6climatemodelsinpredictingmonsoonrainfallbasedonbiascorrectedclusteringapproach

BhomiaSwati(1),KishtawalChandraMohan(2)

1-SpaceApplicationsCentre,ISRO(India),2-SpaceApplicationsCentre,ISRO(India)

Inthepresentstudybiascorrectedcluster(BCC)analysishasbeenperformedonthestate-of-the-artglobal climate models (GCMs) simulations, participating in the Coupled Model Inter-comparisonProjectPhase6(CMIP6).GCMsaredevelopedindependentlyatdifferentinstitutes,eachofwhichvaryin their resolution, treatment of physics and dynamics, representation of orography, physicalprocesses etc.Due to these inter-model differences, it is essential to consider the rangeofmodelsimulationsratherthandependingonsingleclimatesimulation.Inordertogetconfidencefromthemultiplemodels,biascorrectedclusteringapproachhasbeenexperimentedinthepresentwork.Indoingso,biasofeachmodelwascorrectedinthefirststepfollowedbyclustering,whichisaflexibleandunsupervisednumericaltechniquethatinvolvesthesegregationofdataintostatisticallysimilargroupsorclusters.Afterthesegregationofmodeloutputintoclusters,eachclusterwasassociatedwith the confidence index, a procedure that assesses significance of the clusters. The historicalsimulationsweredividedinthetrainingaswellastestingperiods,whereinduringthetrainingperiodtherobustclusterateachlocationwereidentifiedandtheskilloftheidentifiedclusterswascheckedduringthetestingphaseinpredictingthesummermonsoonrainfall.Themulti-modelmeancalculatedusing members of the most-populous cluster viz. the cluster with highest confidence index wasidentifiedateachlocationandwasthencomparedwiththesimpleall-modelmean.

Keywords:ClimateModels,CMIP6,Multimodels,ConfidenceIndex,Clustering,Biascorrection,Monsoonrainfall


3-P04

Reducinguncertaintyinnear-termEuropeanclimateprojectionsusingamodelweightingapproach

BrunnerLukas(1),LorenzRuth(1),KnuttiReto(1)

1-EigenössischeTechnischeHochschuleZürich(Switzerland)

Climatechangeadaptationandmitigationrequiresreliableprojectionsaboutthefuturestateoftheclimate system. The upcoming CoupledModel Intercomparison Project phase six (CMIP6) aims toprovideanupdateon suchprojectionsbasedonover100models frommore than40 institutions.Combiningtheseprojectionsintoactionableinformationaboutnear-termchangesonaregionalscaleis, however, challenging givenhigh internal variability andmodel spread. In the EuropeanClimatePredictionsystem(EUCP)projectwe focusonchanges in theEuropeanSREXregions from2041to2060fortemperatureandprecipitation.Weapplyamodelweightingscheme(Lorenzetal.,2018)thataccountsforpastmodelperformanceaswellasmodelinter-dependence.Inthat,itisbasedontheassumptionthatsomemodelsperformbetterthanothersforcertainsettingsandthatmanymodelsare not fully independent from each other. Using the CMIP5 ensemble, we show that regionallyweightingmodelsbyperformanceand independence reduces future spreadandcanshift thebestestimatesignificantlycomparedtoasimpleunweightedmean.Wealsodiscusstheadvantagesandlimitationsofapplyingourapproachtomodelensembleswithmultiple initial-conditionmember inanticipationoftheupcomingCMIP6usingalargeensembleofCESM.

Keywords:modelweighting,cmip5,cmip6,Europe,EUCP,near,term


3-P05

AnanatomyoftheforecasterrorsintheseasonalpredictionsystemwithEC-Earth

Cruz-GarciaRuben(1),OrtegaPablo(1),Acosta-NavarroJuan(1),MassonnetFrancois(2),Doblas-ReyesFranciscoJ.(1)(3)

1-BarcelonaSupercomputingCenter-CentroNacionaldeSupercomputacion(Spain),2–UniversitécatholiquedeLouvain(Belgium),3-Instituci—CatalanadeRecercaiEstudisAvanats(Spain)

Initializationisanessentialstepwhenperformingclimatepredictions,andtheuseofthelatesthigh-quality observations and their assimilation in the model realm is of paramount importance. Lessattention has been paid to other essential aspects of initialization that are equally important. Forexample,incompatibilitiesbetweentheinitialconditions(ICs)productsusedforthedifferentmodelcomponentscancause important initializationshocks,hindering thepredictioncapacityduring thefirstweeksoftheforecast.Inthisstudyweinvestigatethedifferentcontributionstoforecasterrorsina seasonal prediction systemwith EC-Earthwhere sea ice is initialized via Ensemble Kalman filterassimilation of ESA-derived sea ice concentrations, that was produced within the context of theClimateModelUserGroup (CMUG)ClimateChange Initiative (CCI) fromESA.The largest sourceofforecasterrorinArcticseaiceappearsinregionsofhighobservationaluncertainty,whichhinderstheefficiencyoftheassimilationprotocol.WealsoinvestigatedthedevelopmentofthesystematicerroranditscompetingeffectwiththeshockthatemergesfromtheincompatibilitybetweenoceanandseaiceICs.After26(21)daysthesystematicmodelerrorbecomesthelargestcontributortotheforecasterrorfortheMay(November)initializedforecasts,withtheinitial-incompatibilitydominatinginthepreviousdays.Moreover,thedevelopmentofbotherrorsissensitivetothemonthofinitialization:theshockismorepronouncedinNovemberthaninMay.Themajordifferencesbetweenbothmonthsrelatetothesystematicerror,whichismuchhigherinNovember,andalsotothedirectionoftheshockwithrespecttotheseasonaltrend.Inbothcasestheshockleadstoseaicemelting,but,unlikeinMay,inNovemberithappensinacontextofseaiceexpansion.ThisopposingeffectduringNovembermightbeaccentuatingthegenerationofthedrift.

Keywords:seasonalprediction,initializationshock,systematicerror,Arcticseaice


3-P06

Aframeworktodeterminethelimitsofachievableskillforinterannualtodecadalclimatepredictions

DonatMarkus(1),LiuYiling,TaschettoAndrea,Doblas-ReyesFrancisco,AlexanderLisa,EnglandMatthew

1-BarcelonaSupercomputingCenter-CentroNacionaldeSupercomputacion(Spain)

Currentdecadalpredictionsystemsshowregionallysignificantskillinpredictingnear-surfaceclimateconditions,butmuchoftheskillondecadaltimescalesoriginatesfromwarmingtrendsduetoexternalforcing and the benefit from initialisation remains uncertain. Furthermore, initialisation of thepredictions is imperfect due to incomplete observational coverage and inconsistencies betweenobserved and model-simulated climate characteristics, and it is unknown what skill would beachievablegivenimprovedinitialisation.Hereweproposeanewframeworktoinvestigateachievableskillofdecadalpredictionsbycomparingperfect-modelpredictionexperimentswithpredictionsoftherealworld,toidentifymarginsforpossibleimprovementstopredictionsystems.Inaddition,weassesstheaddedvaluefrominitialisationoverchangesduetoclimateforcingindecadalpredictionsfocusingon annual average near-surface temperature. We find that ideal initialization may substantiallyimprovethepredictionsduringthefirsttwoforecastyearsparticularlyinpartsoftheSouthernOcean,IndianOcean,thetropicalPacificandNorthAtlantic,andsomesurroundinglandareas.Onlongertimescales,thepredictionsrelymoreonmodelperformanceinsimulatinglow-frequencyvariabilityandlong-termchangesduetoexternalforcing.ThisframeworkidentifiesthelimitsofpredictabilityusingtheNCARCommunityEarthSystemModel(CESM)version1.0.5,andclarifiesthemarginsofachievableimprovementsfromenhancingdifferentcomponentsofthepredictionsystemsuchasinitialization,response to external forcing and internal variability. We encourage similar experiments to beperformedusingotherclimatemodels,tobetterunderstandthedependenceofpredictabilityonthemodelused.Inparticular,itwouldbeusefultocomplementinitialiseddecadalpredictionscontributingtotheDecadalClimatePredictionProject(DCPP)componentAaspartofCMIP6byasimilarsetofperfect-modelexperiments,tomanagetheexpectationsregardingtheachievableskillofinterannualtodecadalpredictionsbasedoncurrentclimatemodels.

Keywords:predictability,decadalprediction,initialization,perfectmodel,climatevariability


3-P07

UnderstandingCMIPSimulationBiaseswithNCAR'sClimateModelAssessmentTool

FasulloJohn(1)


Evaluatingclimatemodelswithphysicallyrelevantglobal-scaleobservationsoftheenergyandwatercycles is important to both identify biases and establish confidence in their projections. However,severalissuesmustbedealtwithinordertodosoaccuratelyandmeaningfully.Here,theNCARClimateModelAssessmentTool(CMAT)isdescribedanditsapplicationtoCMIPsimulationsacrossgenerations(3,5,6)ispresentedwiththegoalofidentifyingongoingmodelimprovementsandpersistingbiases.AkeymotivationforCMATistheneedtoexaminemodelfidelityinboththeclimatologicalmeanandacrosstimescalesofvariability,whichinCMATincludebothseasonalandinterannualvariations(i.e.ENSO).Stateoftheartobservationsareused,includingERA5andCERES-EBAF4,andcontextformodel-observation differences is provided for by fully accounting for various uncertainties, including theinfluence of internal variability (estimated from the CESM Large Ensemble). Summary graphics foraddressing whether a given model's biases are statistically meaningful, or whether significantdifferences exist between models, are provided to assess both absolute and relative modelperformance. An assessment of transient historical simulation characteristics is also produced,includingcomparisonstokeyobservations(e.g.globalsurfacetemperature,oceanheatcontent)andadetailedassessmentofsimulationdrift.LastlyanobjectivemethodforscoringCMIPmodelsisusedtoprovideperspectiveonthelargenumbersofsimulationsincludedinitsarchivesandtheassociatedprogressionofskillandreductioninbiasacrossgenerationsrelatedtoatmosphericdynamicsandtheenergyandwatercycles.ComparisonofAMIPandfullycoupledrunsprovidesadditionalinsightintotheoriginofpersistentbiases.Figure1:Biases inCESM2shortwavecloudforcing(SWCF)basedonCERESEBAF4retrievals.Significantbiasesareindicated(>internalvariability=stippled;+observationaluncertainty=hatched).ThemeridionalstructureofSWCFinCESM2improvesuponalong-standingbiasinCMIPmodels.

Keywords:climatemodelevaluation,observedclimatechange


3-P08

ApplicationofaBigDataapproachtoconstrainprojection-basedestimatesofthefutureNorthAtlanticCarbonUptake

GorisNadine(1),TjiputraJerry(2),JohannsenKlaus(3)

1-NORCEClimateandBjerknesCentreforClimateResearch(Norway),2-NORCEClimate&BjerknesCentreforClimateResearch(Norway),3-NORCE(Norway)

Withincreasingmodelcomplexityandresolution,thesizeofCMIP6-outputisestimatedtogrowbyafactorof10to20whencomparedtoCMIP5-output.Thislargeamountofdatamakesthetargetofunderstanding and constraining the model-spread highly challenging and calls for the usage ofmathematicaloptimizationalgorithms.Weproposetocombineprocess-basedmodelevaluationswithaBigDataapproachandshowafirstapplication.Asaproofofconcept,westudiedanensembleof11CMIP5-models for a high atmospheric CO2 scenario. The simulated future carbon uptake of theconsideredensemblehashighuncertaintiesintheNorthAtlantic.Ourprocess-basedmodelevaluationshows that the large futuremodel spread is caused by different efficiencies in carbon drawdown,tightlylinkedtowintermixingandbiology.Wefindthatboththecontemporary(i)summeroceanicpCO2-anomalyand(ii)fractionoftheanthropogeniccarbon-inventorybelow1000-mdeptharegoodindicatorsforthefuturecarbon-uptakeofamodel(withmulti-modelcorrelationcoefficientsabove0.7).Subsequently,weuseageneticalgorithmforbigdataanalysisandspatiallyoptimizetheresultsforbothindicators.Afterseveraltechnicalpitfallsrelatedtotheposingoftheproblemandassociatedmodifications, the algorithm took only a few minutes to successfully identify the subpolar gyre(indicatori)andtheNorthAtlanticDeepWater(indicatorii)asoptimal.Withtheuseofobservation-based estimates of both indicators in their optimal regions, a constrainedmodel-ensemblewith asignificantlyreduceduncertaintywascreated.WenotethatBigDatawillbehighlysuitedforCMIP6model data as it allows for an automatic and fast exploration of variables across the ensemble ofmodels. However, we experienced that the close interaction between climate modelers andcomputationalscientistswasandwillbecriticaltoachieveameaningfulandsuccessfuloutcome.

Keywords:Oceaniccarbonuptake,reducinguncertainty,observationalconstraints,processbasedevaluation,modelspread,BigData


3-P09

BenchmarkingCMIPTerrestrialCarbonCycleandBiogeochemistryModelswiththeILAMBPackage

HoffmanForrestM.(1)(2),CollierNathan(2),MuMingquan(3),XuMin(2),YangCheng-En(2)(1),Keppel-AleksGretchen(4),LawrenceDavidM.(5),KovenCharlesD.(6),RileyWilliamJ.(6),

RandersonJamesT.(3)

1-UniversityofTennesseeKnoxville(UnitedStates),2-OakRidgeNationalLaboratory(UnitedStates),3-UniversityofCaliforniaIrvine(UnitedStates),4-UniversityofMichiganAnnArbor

(UnitedStates),5-NationalCenterforAtmosphericResearch(UnitedStates),6-LawrenceBerkeleyNationalLaboratory(UnitedStates)

The International LandModelBenchmarking (ILAMB)project is amodel-data intercomparison andintegrationactivitydesignedtoinformimprovementoflandmodelsand,inparallel,tosuggestdesignsofnewmeasurementcampaignstoreduceuncertaintiesassociatedwithkeylandsurfaceprocesses.As Earth system models (ESMs) become increasingly complex, there is a growing need forcomprehensiveandmulti-facetedevaluationofmodelpredictions.Toadvanceourunderstandingofbiogeochemical processes and their interactions with climate under conditions of increasingatmosphericcarbondioxide(CO2),theReducingUncertaintiesinBiogeochemicalInteractionsthroughSynthesisandComputation(RUBICO)ScientificFocusAreaisdevelopingmodelevaluationmetricstoconstrainmodelresultsandinformmodeldevelopment.Betterrepresentationofbiogeochemistry-climatefeedbacksandecosystemprocessesinESMsisessentialforreducinguncertaintiesassociatedwith projections of climate change during the remainder of the 21st century and beyond. In thispresentation,we demonstrate the use of the ILAMB software systemby evaluating the fidelity ofterrestrial carbon cycle models coupled within Earth system models used to conduct historicalsimulations for theCoupledModel IntercomparisonProject (CMIP).WehighlightsystematicmodelbiasescommontomanylandmodelsanddescribehowILAMBmaybeusedtovalidatemodelsfrommodel?data intercomparison experiments like C4MIP, LUMIP, and LS3MIP and to serve as a guideduringmodeldevelopment.ILAMBisbeingextendedtocomputestandardfeedbackparametersandmultivariatefunctionalmetricsthatyieldinsightsintomodelprocessrepresentationsandemergentecosystempropertiesthatmayconstrainfuturepredictability.Thesenewcapabilitieswillbeillustratedwith CMIP model output in preparation for publication of results in support of the IPCC SixthAssessmentReport.

Keywords:ILAMB,benchmarking,landmodel,feedbacks,carboncycle,C4MIP,LUMIP,LS3MIP


3-P10

Anemergentconstraintonoceanacidificationinthesubsurfacelayersbasedonmulti-modelanalysis

KawamiyaMichio(1),WatanabeMichio

1-JapanAgencyforMarine-EarthScienceandTechnology(Japan)

IntheFifthAssessmentReportofIntergovernmentalPanelonClimateChange,oceanacidificationinthe surface ocean projected by CMIP5 models has been analyzed intensively, but not much wasmentioned for the subsurface layers,where thereare concernsover lossofdeep-water coral reefhabitat.Hereweevaluate,andproposeanemergentconstrainon,theprojectedoceanacidificationratesofsubsurfacelayersof?thewesternNorthPacificwherethestrongestsinkofatmosphericCO2isfoundinthemid-latitudes.Weusetheoutputsofprojectionsunderthehighestemissionscenarioof the representative concentrationpathwaysperformedbyEarth systemmodels (ESMs). The lowpotentialvorticitywatermasscalledtheNorthPacificSubtropicalModeWater(STMW)showslargedissolvedinorganiccarbon(DIC)concentrationincrease,andisadvectedsouthwestward,sothat,intheseatothesouthofJapan,DICconcentrationincreasesandoceanacidificationoccursfasterthaninadjacentregions.IntheSTMWoftheIzu-Ogasawararegion,theoceanacidificationoccurswithapHdecreaseof~0.004/year,amuchhigherratethanthepreviouslyestimatedglobalaverage(0.0023/year),sothatthepHdecreasesby0.3-0.4duringthetwenty-firstcenturyandthesaturationstateofcalcite(?Ca)decreasesfrom~4.8downto~2.4.WefindthattheESMswithadeepermixedlayerintheKuroshioExtensionregionshowalargerincreaseinDICconcentrationwithintheIzu-OgasawararegionandwithintheRyukyuIslandsregion. It issuggestedthatmixed layerdepth intheKuroshioExtensionmayserveasanemergentconstraintforfutureprojectionofsubsurfaceoceanacidification.AsimilaranalysisschememayapplytotheNorthAtlanticwheremodewatersareformedalongtheGulfStream.

Keywords:oceanacidification,emergentconstraint,mixedlayerdepth,carbonuptake,subsurfaceocean


3-P11

Long-termBalancesandVariabilitiesofSurfaceEnergyandWaterCycles:PreliminaryResultsfromLS3MIPandGSWP3

KimHyungjun(1),KrinnerGerhard(2),SeneviratneSonia(3),VanDenHurkBart(4),DerksenChris(5),OkiTaikan(1),WorkingGroupLs3mip

1-InstituteofIndustrialScience(Japan),2-InstitutdesGeosciencesdel'Environnement(France),3-InstituteforAtmosphericandClimateScience[Zürich](Switzerland),4-RoyalNetherlandsMeteorologicalInstitute(Netherlands),5-EnvironmentandClimateChangeCanada(Canada)

LandSurface,Snow,Soil-moistureModel IntercomparisonProject (LS3MIP)consistsofvarioussub-componentsofexperiments.Amongthem,‘land-hist'isthefirst‘landonly'contributiontotheCoupledModel Intercomparison Project (CMIP)with roles such as 1) to evaluate how state-of-the-art landsurface schemes coupled with climate models represent land processes appropriately and 2) toestimatelong-termbalancesofvariabilitiesoftheinteractionsamongtheprocesses.Inthisstudy,wereportthepreliminaryresultsof‘land-hist'experimentofLS3MIPbasedontheGlobalSoilWetnessProject3(GSWP3)climatedriversetforahistoricalperiod(1901-2014).Italsoprovidesthemeanstoquantifysystematicbiasesanduncertainties,whichisofparticularinterestforclimatechangeimpactassessments.

Keywords:LS3MIP,GSWP3,landhist


3-P12

Anewdiagnostictoolfortheenergybudgetsandtransportsinclimatemodels

LemboValerio(1),LucariniValerio,LunkeitFrank

1-CenterforEarthSystemResearchandSustainability(Germany)

We present here a novel diagnostic tool for studying some fundamental aspects related to thethermodynamicsoftheclimatesystem.Thetoolhasbeendevelopedaspartofthelargecommunityeffort"ESMValTool",aimedatprovidingastandardizedsetofdiagnosticsfortheevaluationofCMIP6modelsimulations.Thetoolincludesanumberofindependentmodules.Oneallowstoassessthelocalandglobalatmospherichydrological cycle, in termsofwatermassand latentenergy.AnotheroneprovidesestimatesofradiativeandheatbudgetatTop-of-Atmosphere,withintheatmosphereandattheEarth'ssurface.Italsocomputesthemeridionalenthalpytransportsandtheiryearlymeanpeaklocationsandmagnitudes.AnothermodulecomputestheatmosphericLorenzEnergyCycle,isolatingstationary and transient eddy components. The last module accounts for the material entropyproduction,i.e.theentropyproductionassociatedwithirreversibleprocesses.Thisisaccomplishedintwoways,eitherthroughconvergenceofradiativefluxes,orthroughdirectcomputationoftheenergyexchangesineachprocess.Inthelattercase,thekineticenergydissipationtermisaddressedbymeansofthestrengthoftheLorenzEnergyCycle.Inputfieldsaretheradiation,precipitationandpressurefieldsatthemonthlytimeresolution,velocityandtemperaturefieldsatthesurfaceandinpressurelevels at the daily resolution. If a land-sea mask is provided, the tool is also able to performcomputationsandprintoutputsseparatelyonlandandoceans.Meetingtherequirementofprovidingsyntheticglobalmetricsformodelintercomparison,resultsaregivenasasetofparametersandareautomaticallycomparedwhenamulti-modelensembleisgivenasinput.ThediagnosticswillbesoonmadepubliclyavailableintheESMValToolPythonpackage,oncetheportingtoESMValToolversion2willbefinalized.

Keywords:modeldiagnostics,ESMValTool,energybudget,energytransports,watermassbudget,entropy


3-P13

Canwebeatclimatemodeldemocracyinmulti-modelensembleprojections?

LorenzRuth(1),BrunnerLukas(1),KnuttiReto(1)

1-InstituteforAtmosphericandClimateScience,ETHZurich(Switzerland)

Toeffectivelyplanadaptationtoclimatechangeweneedclimateprojectionswhichcoverallpotentialoutcomesbutareasnarrowaspossible.Uncertainties inclimateprojectionsarearesultofnaturalvariability, scenario uncertainty andmodel uncertainty.Model uncertainty could be decreased bygivingmoreweight to thosemodels that aremore skillful for a specific process or application. Inaddition,models in an ensemble of opportunity are not necessarily independent. This fact needsincreasedattentionintheupcomingCMIP6ensemble,whichwillincludemorethan100models,manyof them similar to each other. We compare multiple approaches on how multi-model ensembleaveragescanbecalculated;e.g.,aweightedmulti-modelmeantakingintoaccountperformanceandindependence and themean of the ten ‘bestmodels’ (based on RMSE over the historical period)compared to thearithmeticmulti-modelmeanas in IPCCAR5.We investigatehowthesedifferentapproachesinfluencethelong-termprojectionoftemperatureoverCentralEuropeandtesttheskilloftheprojectionsinaperfectmodeltest.Theweightedmulti-modelensembleimprovestheskilloftheprojectiononaverage.However,spreadislargedependingonwhichdiagnosticsweusetoinformthemethod,howmanyofthesediagnosticsweuse,orwhichmodelisusedastruth.Thelatterimpliesthattheresultsaresensitivetowhichobservationaldatasetisusedasreference.Hence,dependingonobservationaluncertainty,itmightnotbestraightforwardtodeterminewhatagoodmodelis.Inaddition, there is no clear relationship between how many diagnostics we use and skill of theprojection. Therefore,we conclude that close attention has to be paid on choosing diagnostics toinformtheweightingwhicharerelevantforthetargetprojection.

Keywords:weighting,uncertainty,long,termprojections


3-P14

Biaspatternsof6dailylandsurfacevariablesinCMIP5modelsandconsequencesofbiasadjustmentintermsofchangesandassociated

uncertaintyattheendofthecenturyunderRCP8.5

LoukosHarilaos(1),NoelThomas(1),VautardRobert(2),VracMathieu(2),DenvilSébastien(3)

1-theclimatedatafactory(France),2-LaboratoiredesSciencesduClimatetdelƒnvironnement[Gif-sur-Yvette](France),3-InstitutPierreSimonLaplace(France)

WepresentananalysisofbiasesofCMIP5modelsprojectionsunderemissionscenarioRCP8.5for6surfacevariables(surfaceminimum,maximumandmeantemperature,precipitation,surfacewindandsolar radiation) after bias adjustment. To our knowledge, this is the first bias adjustment exercicecoveringsomanyvariablesatdailyscalefortheCMIP5RCP8.5projections.Wefirstshowhowbiasadjustmentimprovescomparisonwithreanalysesoverthehistoricalperiodandanalysedifferencesbymodelandvariablebothattheglobalandregionalscales.Wethenshowhowtheclimatologicalchangesattheendofthecentury(differencebetweenthe30-yearaveragesof2071-2100comparedto1971-2000)andtheensembleuncertaintyareaffected.ToproducethebiascorrectedCMIP5datasetweusedtheCumulativeDistributionFunctiontransform(CDF-t)method(Michelangelietal.,2009,Vracetal.,2016,Famienetal.,2017).CDF-tisavariantofthequantilemapping(QM)methodwhichconsists in comparing the cumulative distribution function (CDF) of a climate variable (e.g.,temperature)atlargescale(e.g.,fromGCM)totheCDFofthesamevariablefromobservations.TheCDF-tmethodhasbeenextensivelyusedintheliteratureandformanyvariables(e.g.,Kallacheetal.,2011;Vracetal.,2012;Lavaysseetal.,2012;Vautardetal.,2013;VracandFriederichs,2015).CDF-testimatesthe‘futureobserved’CDFthroughamathematicaltransformationbeforeperformingaQMand thusdoesnotdependon theassumptionofa samecorrection forpastand futureclimate,astraditionalQMmethodshaveto.Asobservation-basedreferencedatasetweusedtheWATCH-Forcing-Data-ERA-Interimdataset(WFDEI;Weedonetal.,2014). It includeseightmeteorologicalvariables.Weuseddailyaveragesfrom1979to2012,forthegloballandsurfaceat0.5¡x0.5¡resolution.

Keywords:CMIP5,bias,adjustment,uncertainty,quantilemapping,WFDEI


3-P15

Simulationsandevaluationsoftheversion1.0oftheE3SMLandModel(ELM)fortheLS3MIP

MaoJiafu(1),RicciutoDaniel,ShiXiaoying,HoffmanForrest,ThorntonPeter

1-EnvironmentalSciencesDivisionandClimateChangeScienceInstitute,OakRidgeNationalLaboratory(UnitedStates)

The Land Surface, Snow and Soil Moisture Model Intercomparison Project (LS3MIP) diagnosesinteractionsbetween landandatmosphereandassesses the landcomponentsof theEarthsystemmodels for theCMIP6.Oneof thekeycomponentsof theLS3MIP is toconductoffline landmodelexperimentsdrivenbycommonobservationaldrivers,attributingthecausesbehindmodeldifferencesto the driver or structural deficiencies.With the International LandModel Benchmarking (iLAMB)package,wewillinvestigateandpresentcomprehensivebenchmarkingresultsoftheELMagainstbestavailableobservations,intermsofthemeansstatesandmultiyearvariationsoflandsurfaceenergy,water,andbiogeochemicalbudgets.Sincethreemeteorologicalforcings(e.g.,theGSWP3,Princeton,andCRU-NCEP)willbeutilizedtodrivetheELM,wewillexaminethesensitivityofsimulatedfluxesandstockstotheseforcingsandpartitionuncertaintyfrombothinternalandexternalmodelsources.Also,wewilldemonstratepossibledrivingmechanismsunderlyinglong-termchangesofmajorlandsurfacevariablesusingthefactorialELMsimulations (e.g.,climatechange, landuseand landcoverchange,nitrogendeposition,andaerosoldeposition).

Keywords:LandSurfaceModel,ELM,E3SM,LS3MIP,Benchmarking


3-P16

ClimateresponsetothePinatuboandTamboraeruptionsinEC-Earth3.2

MartinEneko(1),BilbaoRoberto(1),MenegozMartin(1)(2),OrtegaPablo(1)

1-BarcelonaSupercomputingCenter-CentroNacionaldeSupercomputacion(Spain),2-InstitutdesGéosciencesdel'Environnement(France)

Explosivevolcaniceruptionscauseepisodesofstrongnegativeradiativeforcingintheclimatesystemby injecting large amounts of sulphur dioxide into the stratosphere, producing an aerosol whichreflectspartof the incomingsolar radiation.Thisnegative radiative forcingcausesaglobal surfacecooling, although there is substantial uncertainty in the magnitude and in the regional climateresponse in climatemodels and observations are limited. The aim of this work is to analyse andcomparetheclimateresponsetothePinatubo(1991)andTambora(1815)eruptions,thelatterbeinga significantlygreatereruption than the former, inanensembleofexperimentsdonewith theEC-Earth3.2 atmosphere-ocean general circulation model (AOGCM). Both the Pinatubo and TamboravolcanicforcingshavebeenestimatedwiththeEasyVolcanicAerosol(EVA,Tooheyetal.,2016)tool,andthePinatuboforcinghasbeenvalidatedagainsttheCMIP6estimate.Previousstudieshaveshownthatthebackgroundclimateconditionsmaymasktheclimateresponsetovolcaniceruptionsofsimilarmagnitude to Pinatubo,making it difficult to detect dynamical signals (e.g.Ménégoz et al., 2017).Preliminaryresultsshowthatatmospherictemperatureanomalies(troposphericandstratospheric)inresponsetotheTamboraeruptionaresignificantlygreaterinmagnitudeandpersistforlonger,thusmakingthevolcanicimprintsdetectable.Weanalysethevolcanicallyforcedatmosphericandoceaniccirculationchanges,andhowtheyprojectontothemainmodesofclimatevariabilitysuchasElNi–oSouthernOscillation (ENSO)andtheNorthAtlanticOscillation (NAO).Thisanalysis iscarriedout inpreparationfortheCMIP6VOLMIP(Zanchettinetal.,2016)experiments.

Keywords:volcano:volcanicforcing:volmip:climatechange


3-P17

EstimatingtheUncertaintyinClimateProjections

MilinskiSebastian(1),FyfeJohnC.(2),MarotzkeJochem(1)

1-Max-Planck-InstitutfürMeteorologie(Germany),2-CanadianCentreforClimateModellingandAnalysis(Canada)

TheIPCCAR5usedapragmaticapproachtoquantifyingtheuncertaintyinclimateprojections.Onerealisationpermodelperscenariowaspickedanddefinedtheensemble.The5?95%ensemblespreadwas used to characterise the uncertainty, which was inflated to account for other sources ofuncertaintybyinterpretingthe5?95%spreadasthe16?83%(likely)range.Whilestraightforwardandclearlycommunicated,thisapproachhadseveraldrawbacks,andtherehasbeensubstantialprogresssince the AR5 on how to assess the uncertainty of projections. Large ensembles of single climatemodelshaveopenedupanewperspectiveoninternalvariabilityandmodeldifferencesandallowacleanseparationofthesecontributorstouncertainty.Initialisedpredictionexperimentsanddetectionand attribution approaches exploit observations to provide information on the future. Modelevaluationanddiagnosisofmodelindependencehavebeenappliedtoensembleweightingschemes.Herewe sketch how this recent progressmight be synthesised into an estimate of uncertainty inclimate projections.We first apply our approach to CMIP5 simulations; later application to CMIP6simulationsshouldbestraightforward.

Keywords:separatinguncertainties,internalvariability


3-P18

CMIP5/CMIP6model-analogseasonalforecastskill:ametricformodelevaluationofENSOdynamics

NewmanMatthew(1),DingHui(2),AlexanderMichael(3),WittenbergAndrew(4)

1-UniversityofColorado/CIRESandNOAA/ESRL/PSD(UnitedStates),2-UniversityofColorado/CIRESandNOAA/ESRL/PSD(UnitedStates),3-NOAA/ESRL/PSD(UnitedStates),


Recently, Ding et al showed that tropical Indo-Pacific SST forecast skill (both deterministic andprobabilistic)fromtheoperationalmodelsoftheNorthAmericaMulti-ModelEnsemble(NMME)canbematchedorevenexceededwithasimple"model-analog”method,appliedtothecorrespondinglongcontrolrunsfromeachNMMEmodel.Inthismethod,forecastinitializationismadedirectlyonamodel'sownattractor,usingananalogapproachwheremodelstatesclosetotheobservedinitialstateare drawn from a “library'' obtained from prior uninitialized CGCM simulations. The subsequentevolutionofthosemodel-analogsyieldsaforecastensemble,withoutadditionalmodel integration.This suggests seasonal forecasts can be made using almost all CMIP models. We have maderetrospective tropical Indo-Pacific forecasts for the 1961-2015 period from28 CMIP5models (andavailable CMIP6 models), using model-analogs. All show skill for sea surface temperature (SST)forecasts,butthereisaconsiderablerangeamongthemodels,withevengreaterskilldifferencesforprecipitation. Model-analog forecasts from the ten “best” CMIP5 models have skill for SST andprecipitationcomparabletothatofboththeNMMEmodel-analogforecastensembleand(since1982)traditional assimilation-initializedbias-correctedNMMEhindcasts. ENSO forecast skill hasno trendoverthe55-yrperiod.Itsdecadalvariationsappeartoberandom,althoughskilldoesimproveduringepochsofincreasedENSOactivity.IncludingtheCMIP5-projectedeffectsofexternalradiativeforcingsimprovesthetropicalSSTskillofthemodel-analogforecasts,butnotwithintheENSOregion.Finally,wedemonstratethatthe55-yrtropicalPacificoceanhindcastskill,plottedonaTaylordiagramforleadsof1-12months,isalsoarobustmetricofhoweachCMIP5/6model'sattractorcorrespondstonature'sattractor,sinceitdirectlymeasuresthefidelityofeachmodel'sENSOdynamics(ratherthanstaticmeasuresofENSO)totheobservedevolutionofallENSOevents.

Keywords:forecasts,ENSO,metrics,predictability


3-P19

UncertaintyinEarthSystemModels:BenchmarksforOceanModelPerformanceandValidation

OgunroOluwaseun(1),ElliottScott(2),CollierNathan(1),WingenterOliver(3),DealClara(4),WeiweiFu(5),HoffmanForrest(1)

1-ClimateChangeScienceInstitute,OakRidge(UnitedStates),2-LosAlamosNationalLaboratory(UnitedStates),3-NewMexicoInstituteofMiningandTechnology[NewMexicoTech](United

States),4-UniversityofAlaskaFairbanks(UnitedStates),5-UniversityofCalifornia,(UnitedStates)

ThemeanoceanCO2sinkisamajorcomponentoftheglobalcarbonbudget,withoceanreservoirsholdingaboutfiftytimesmorecarbonthantheatmosphere.MarinephytoplanktonplaysasignificantroleinthenetcarbonsinkasonequarterofanthropogenicCO2emissionsendupintheocean.LifeintheoceanincreasestheefficiencyofmarineenvironmentstotakeupmoreCO2andultimatelyreducestheimpactsofpersistentriseinatmosphericconcentrations.However,challengeswithappropriaterepresentationof physical andbiological processes in Earth SystemModels (ESM)undermines theefforttoquantifyseasonaltomultidecadalvariabilityinoceanuptakeofatmosphericCO2.Inabidtoimproveanalysesofmarinecontributionstocarboncyclefeedbacks,wehavedevelopednewanalysismethodsandmetricsaspartoftheInternationalOceanModelBenchmarking(IOMB)efforttomeetthegrowingdiagnosticandbenchmarkingneedsofoceanbiogeochemicalmodels.Thepackagewasemployedtovalidateoceanmodelresultswithobservationdatasets.Internationaloceanmodelsthatcontributed results to the fifth phase of Coupled Model Intercomparison Project (CMIP5) wereanalyzed. Our analyses suggest that biogeochemical processes that determine CO2 uptake by theglobaloceanarenotwellrepresentedinmostESMs.Polarregionscontinuetoshownotablebiasesinbiogeochemicaltracers.Someofthesedisparitiescouldhavefirstorderimpactsontheconversionofatmospheric CO2 to organic carbon. Combined effects of twoormore of these forcings on oceanbiogeochemical cycles are challenging to predict as additive and antagonistic effectsmayoccur. Abenchmarking tool for accurate assessment and validation of marine biogeochemical outputs isindispensableaswecontinuetoimproveESMdevelopmentsandunderstandcarboncyclefeedbacksfromtheocean.AnalysisofmarinebiogeochemicalfeedbacksinsuiteofCMIP6experimentscouldalsobeimproved.

Keywords:Marinebiogeochemicalcycle,modelevaluation


3-P20

Investigatingdriversofmidlatitudecirculationbiasesinclimatehindcastensembles

SobolowskiStefan(1),LiCamille(2),ChenLilan(3),OgawaFumiaki(2)

1-NORCEandtheBjerknesCentreforClimateResearch(Norway),2-UniversityofBergenandtheBjerknesCentreforClimateResearch(Norway),3-NanjingUniversity(China)

Theprojectedresponseoftheatmosphericcirculationtoincreasinggreenhousegasconcentrationsishighlyuncertain.Oneoftheprimaryreasonsforthisisthatthestate-of-the-artmodelsweemploytoinvestigatetheseresponsesstruggletorepresentbasicfeaturesofthemidlatitudecirculation.Biasesalsohavedetrimentaleffectsonpredictiveskillatclimatepredictiontimescalesofseasonstodecades.Despitethis,physicalunderstandingbothofthecontrolsonthesefeaturesandthedriversoftheirbiases is still limited. The present study investigates ensembles of climate hindcast simulationsperformedby theNorwegian Earth SystemModel. This is themodel used in the CMIP6 endorsedDecadalClimatePredictionProject.TheseensemblesarecomparedtobothfreerunsandAMIP-stylesimulationswithERA-Interimservingasgroundtruth.WeexaminetheNorthPacificandNorthAtlanticjets inbothwinterand summer.Wealso identifywhere theobservations liewithin thepredictivedistributionoftheensemble.ResultsshowthatthewintertimeNorthAtlanticjetistoozonal,extendstoofarintoEuropeandisshiftednorthwards.VirtuallytheentireNorthAtlanticsectorliesoutsidethepredictivedistributionoftheensembleandperformanceactuallydegradesinsimulationswithtighterconstraints on the assimilation. By contrast the wintertime North Pacific jet is rather betterrepresentedbothwithrespecttopatternaswellasmagnitudeofthebiases.Thisislikelyduetothebetter-representedteleconnectionsbetweenthetropicalandextratropicalPacific.However,thereisanasymmetrytothebiaseswiththeNorthPacificshowingthelargestbiasesinsummerandtheNorthAtlanticshowingthelargestbiasesinwinter.Furtherinvestigationsuggeststhebiasesresideintheatmospheric component of the model and too weak ocean-atmosphere interactions. Physicalmechanismsare investigated throughdecompositionof themomentumbudgetaswell asdiabaticprocesses.

Keywords:climateprediction,climatevariability,climatedynamics


3-P21

PotentialPredictabilityHorizonofatmosphericCO2concentrationsinCMIP6simulations

SpringAaron(1)(2),IlyinaTatiana(2)

1-InternationalMaxPlanckResearchSchoolonEarthSystemModelling(Germany),2-MaxPlanckInstituteforMeteorology(Germany)

DecadalpredictionswithEarth-System-Model(ESM)simulationscanbeapowerfultoolinpredictingthenear-termevolutionof theglobalcarbonsink. Inparticular,ESMscanbeusedtodiagnosetheinternal-variabilityofcarbonsinksandtoconstrainpredictionsofatmosphericCO2concentrations.Previous carbonpredictability studies findpredictive skill in air-seaCO2 fluxes (NorthAtlantic: 4-7years[Lietal.,2016];globalocean:4-7years[Seferianetal.,2018;Lovenduskietal.,2018]).Yettheywerebasedonconcentration-drivensimulations, inwhichair-seaandair-landcarbonfluxesdonotalteratmosphericCO2concentrations.Hence, asofnow,nopredictability studyexistswitha fullycoupledcarboncyclebasedonemission-drivensimulations.Weperformensemblesimulationsinaperfect-modelframeworkbasedontheCMIP6versionofMPI-ESM-LRunderpre-industrialforcingwithfreelyvaryingatmosphericCO2(esmControlwithinC4MIP).Ourresultsplaceatrend-and-bias-freeupper limit on potential predictability of atmospheric CO2 concentrations.We find an eight yearpredictabilityhorizonoftheglobalatmosphericCO2signal.ThepredictabilityhorizonofCO2fluxoverlandismuchshorterthanovertheoceans,whileCO2fluxvariabilityonlandismuchhigherthanintheocean.Comparingtheglobally-accumulatedmagnitudes(fractions)ofreducedinternalvariabilityduetoinitializationforthefirstleadyearresultsina0.1PgC(75%)reductionofair-seaCO2fluxeswhileair-landCO2 fluxes are reducedby 0.6PgC (65%),which indicates a 13%-87%attributionof 1-yearatmospheric CO2 predictability to the ocean-land carbon sinks.We find evidence for establishingpredictive skill for atmospheric CO2 in emission-driven simulations. Furthermore, the ESM-basedpredictabilityhorizonsurpassesthestatisticpredictionbyBettsetal.[2016]anddampedpersistence.As simulations under the CMIP6 protocol become available, further multi-model potentialpredictabilitycomparisonswillshedlightontheoriginsofpotentialpredictabilityinthecarboncycle.

Keywords:predictabilityhorizon,predictability,potential,landcarbonsink,oceancarbonsink,carbonsink,carbon,co2flux,co2,C4MIP,ensemble,emission,driven


3-P22

DevelopmentofanewclimatemodelemulatorbasedonCMIP6multi-modelensemble

TsutsuiJunichi(1)

1-CentralResearchInstituteofElectricPowerIndustry(Japan)

Thepresentstudyaimstoproduceanupdatedlistofforcing-responseparametersofAOGCMswithadditionalquantitiesaboutdifferencesbetweenthenewandpreviousmulti-modelensembles,CMIP6andCMIP5,andtobuildaclimatemodelemulatorreflectingthesemodelensemblestobeusedforprobabilisticassessmentofthetemperatureresponsetoclimatemitigationscenarios.ThemethodisbasedonthepreviousstudydescribedinTsutsui(2017,ClimaticChange)withsomeimprovements.The forcing-responseparametersarediagnosed throughsimultaneouscurve fitting toglobalmeantimeseriesofthesurfacetemperatureandthetopofatmosphereenergyimbalancefromtwobasicCO2-forcedexperiments, an abruptquadrupling and1%-per-year increase in the atmosphericCO2concentration. This scheme can serves as an improved alternative to the widely-used regressionmethod.ThevariationofanensembleofAOGCMsisquantifiedthroughastatisticalapproachincludingprincipalcomponentanalysisoftheimpulseresponsefunctionthatisusedinthefittingscheme.Theprincipal components also form a basis of the probabilistic assessment. One application of theemulator isanexaminationofthequantitythatrelatesthetemperaturetoCO2emissions, termedtransientclimateresponsetocumulativeCO2emissions(TCRE).Assumingmedianquantitiesforthecarboncyclemoduleintheemulatorandusingthesurfaceskintemperatureasatemperaturevariable,we have found that the range of TCRE from the 14th to 87th percentile of the CMIP5 variationcorrespondstotheassessedlikelyrangeintheIPCCfifthassessmentreport,andthemedianisalmostidenticalbetweenthem.TheissueofhowthenewensembleofAOGCMsaswellasthetemperaturedefinitionwouldaffectthisrangewillbeaddressedattheworkshopusingavailableCMIP6output.

Keywords:effectiveradiativeforcing,climatefeedbackparameter,transientclimateresponse,climatemodelemulator,TCRE


3-P23

BenchmarkingthesimulatedglobalcarboncycleofCMIP6ESMsusingatmosphericCO2flaskmeasurements

RicaudPhilippe(1),SéférianRoland(2),QuesadaSamuel(3),Saint-MartinDavid(2),DecharmeBertrand(2),AroraVivek(4),BrovkinVictor(5),IlyinaTatiana(5),SchwingerJorg(6),

TjiputraJerry(7),RaddatzThomas(5)

1-Grouped'étudedel'atmosphèremétéorologique(France),2-Grouped'étudedel'atmosphèremétéorologique(France),3-EuropeanCentreforMedium-RangeWeatherForecasts(United

Kingdom),4-CanadianCentreforClimateModellingandAnalysis(Canada),5-MaxPlanckInstituteforMeteorology(Germany),6-DepartmentofEarthScience[Bergen](Norway),

7-UniversityofBergen(Norway)

Atmospheric CO2 is the ultimate product of global carbon cycle and climate. Its spatio-temporalevolutionisnotonlydrivenbythelarge-scaleclimatedynamicsbutalsobythenetfluxesofCO2seenbytheatmosphere,resultingitselffromanthropogenicCO2emissions(suchasfossilfuelemissionsandaircraftemissions)andlandandoceannaturalsinksofCO2.Theirrepresentationinstate-of-the-art emission-driven Earth system models consists thus in a global scale metrics to provide acomprehensiveverificationofsimulatedglobalcarboncycleandEarthclimateallatonce.Here,weproposetobenchmarktwogenerationsofEarthsystemmodels(CMIP5andCMIP6)usingsurfaceflaskCO2 measurements available through the cooperative international air sampling network(https://www.esrl.noaa.gov/gmd/ccgg/). To this end, we use available emission-driven historicalsimulationsascoordinatedbytheCMIPandC4MIPframeworks.Skill-assessmentmetricssuchaslineartrends,seasonalcycleandannualgrowthrateareusedtotrackhowfartheESMskillhaschangedbetweenCMIP5andCMIP6.EmergentpropertiessuchastheresponseoftheannualgrowthrateofCO2toElNi–oSouthernOscillationorthechangeinseasonalamplitudetorisingannualmeanCO2areusedtocharacterizethereliabilityoftheglobalcarboncycleassimulatedbythetwogenerationsofEarthsystemmodels.

Keywords:emission,drivensimulation,atmosphericCO2,skill,assessment,model,dataevaluation

Session4-Highresolution

4-P01

Mesoscaleair-seainteractionsinKuroshioExtensionregionduringwinterseasonsimulatedbyaHigh-resolutionCoupledGCM

AnBo(1)(2),YuYongqiang(1)(2),LinPengfei(1)(2)

1-UniversityofChineseAcademyofSciences,Beijing,China(China),2-StateKeyLaboratoryofNumericalModelingforAtmosphericSciencesandGeophysicalFluidDynamics,Instituteof

AtmosphericPhysics,ChineseAcademyofSciences(China)

Mesoscaleoceaniceddiesareveryactivealongthewesternboundarycurrentsandtheirextensions,wheretherearehugeenergyandwaterexchangesbetweentheatmosphereandoceanbymesoscaleair-sea interactions. In this study, the ability of a high-resolution coupled GCM to simulate themesoscale air-sea interactions is assessed. The model results used here are from high resolutioncoupled GCM, in which the oceanic component model is LASG/IAP Climate system OceanModel(LICOM),withhorizontalresolution0.1¡x0.1¡,andtheatmosphericmodelisCommunityAtmosphericModel(CAM)with0.23¡x0.31¡resolution.ThisstudyfocusesonthemesoscaleinteractionsinKuroshioExtension region (30¡N-50¡N 140¡E-180¡E) during winter season(ONDJFM). A seven-day temporalrunning mean and LOESS (locally weighted regression) filter are applied to obtain the mesoscaleinformation in theatmosphereandocean.Bycalculatingtheregressioncoefficientsbetweenhigh-pass-filteredseasurfacetemperature(SST)andturbulentheat flux (THF),aswellasestimatingthesecondary circulation of atmospheric response tomesoscale eddies,mesoscale air-sea interactionprocessesareexplored,andsomeprincipalresultscouldbeconcludedthat:1)Thishigh-resolutioncoupledGCMisabletosimulatethebasicmesoscaleair-seainteraction.2)TheatmosphericsecondarycirculationsaboveboththewarmeddyandcoldeddycasesintheKuroshioExtensionregioninwintercouldalsobesimulatedbythehigh-resolutionGCM.3)Themesoscaleair-seainteractionsinthetwocasesselectedabovebelongtothe“verticalmomentummixingmechanism”,withsignificantverticalmotionsattheedgeandverticalsecondarycirculationsaloft.

Keywords:mesoscaleeddy,airseainteraction,highresolutioncoupledmodel


4-P02

RunningtheEC-Earthmodelatultra-highresolution:challengesandbenefits

ArsouzeThomas(1),BrodeauLaurent,CastrilloMiguel(2),AcostaMario(2),RenaultLionel(3),Doblas-ReyesFrancisco(2)

1-BarcelonaSupercomputingCenter(Spain),2-BarcelonaSupercomputingCenter-CentroNacionaldeSupercomputacion(Spain),3-UCLA(UnitedStates)

WithintheframeworkofthePRIMAVERAH2020projectandtheHighResMIPcoordinatedexercise,BSC has developped a coupled version of the EC-Earth 3.2 climate model at a groundbreakingresolution of about 15 km for all the climate system components (ocean, sea ice, land andatmosphere).Weherefocusonthetechnicalchallengesassociatedwithsuchacomputationalcostlyconfiguration,bothintermsofdevelopmentandproduction,inparticularinregardtootherexistingconfigurations.Also,westressout the importanceofmechanical feedbacks fromtheocean to theatmosphereontotheoceandynamicsbothintermsoflargescaleoceaniccirculationandmesoscaleactivity.Todoso,wecompareasetoftwocompanionsimulationswithandwithouttheexchangeofoceanic surface currents from the oceanic component to the atmospheric component. Finally,wecomparethissimulationtoitslowresolutioncounterparts,toallowidentifyingtherobustbenefitsofincreasedmodelresolution,bothintermsofprocessesresolvedandclimatepredictability.

Keywords:Highresolution,HighResMIP,Currentfeedbadck,PRIMAVERA


4-P03

NorthAtlanticpost-tropicalcyclonesinreanalysisdatasets

BakerAlexander(1),HodgesKevin(1),HaarsmaRein(2),SchiemannReinhard(1),VidalePierLuigi(1)

1-NationalCentreforAtmosphericScience(UnitedKingdom),2-RoyalNetherlandsMeteorologicalInstitute(Netherlands)

Mid-latitudecyclonesofatropicaloriginarean importantnaturalhazardaffectingwesternEuropeandthenortheastUnitedStates.Hereintermed'post-tropicalcyclones',thesehigh-impactweatherevents expose these regions to hurricane-force wind speeds and extreme precipitation. Theirfrequency is projected to increase with anthropogenic climate change because warming-inducedpoleward and eastward expansion of tropical cyclone genesis areas will allow more systems topropagatetothemid-latitudebaroclinicregion.Aquantitativeunderstandingofnaturalpost-tropicalcyclonevariability?againstwhichglobalclimatemodelperformanceandprojectionsunderclimatechangemaybeevaluated?isthereforerequired.Here,weuseanobjectivefeature-trackingalgorithmto identify post-tropical cyclones in reanalysis datasets. We then analyse post-tropical cycloneseasonality,interannualvariability,structuralcharacterisitics,andlifecycles,aswellastheassociatedprecipitationovertheNorthAtlanticbasin,focussingonthosesystemswhichimpactwesternEuropeandthenortheastUnitedStates.WeevaluatetheperformanceofthereanalysesagainstInternationalBest Track Archive for Climate Stewardship (IBTrACS) data. Finally, we compare the frequency oftropical-to-extratropical lifecycles between reanalyses and, to relate these results to hydrologicalimpacts, quantify the contribution of post-tropical cyclone-associated precipitation to totalprecipitation.Wehopetostimulatefurtherdiscussionofthesesystems,notleasttheirtracks,tropical-to-extratropicaltransition,andrepresentationinglobalclimatemodels.

Keywords:Atlantic,cyclones,reanalysis,precipitation


4-P04

Air-SeainteractionsovertheGulfStreaminanensembleofHighResMIPpresentclimatesimulations

BellucciAlessio(1),AthanasiadisPanos(1),ScoccimarroEnrico(1),RuggieriPaolo(1),FedeleGiusy(2),Haarsma(3),Garcia-SerranoJavier(4),DeVriesIris(5),GualdiSilvio(1)

1-CentroEuroMediterraneosuiCambiamentiClimatici(Italy),2-UniversityofCa'Foscari[Venice,Italy](Italy),3-RoyalNetherlandsMeteorologicalInstitute(Netherlands),4-Barcelona

SupercomputingCenter(BSC-CNS)(Spain),5-DepartmentofMeteorology[Stockholm](Sweden)

A dominant paradigm for mid-latitude air-sea interactions identifies the synoptic-scale weather“noise” as themain driver for the observed ocean surface variability. The underlyingmechanism,outlinedintheseminalpaperbyHasselmann(1976;hereafterH76),hasbeenchallengedinanumberofstudies.BasedontheoreticalargumentsandobservationalevidenceBishopetal.(2017;hereafterB17)pointtoamoreactiverolefortheoceanweather(mesoscaleoceaniceddies)inmodulatingtheair-sea interactions, as opposed to the passive ocean H76 view. B17 use the lead-lag covariancepatternsofSST(andSSTtendency)andsurfaceturbulentheatfluxestoidentifyareaswhereair-seafluxesareeitheratmosphere-orocean-driven.Basedonthisapproach,substantialdeviationsfromH76arefoundoverregionscharacterizedbybaroclinicallyunstablewesternboundarycurrentsandtheAntarcticCircumpolarCurrent.InthisanalysiswefollowtheapproachoutlinedinB17toassesstheroleofmodelresolutioninrepresentingthenatureofocean-atmosphereinteractionovertheGulfStream in the HighResMIP (Haarsma et al., 2016) multi-model ensemble of present-climatesimulations,performedundertheframeworkoftheH2020PRIMAVERAProject.Specialemphasiswillbedevotedtodisentanglethephysicalprocessesatplayondifferentspatio-temporalscales.

Keywords:air,seainteractions,GulfStream,modelresolution,Highresmip


4-P05

Influenceofchangesinlarge-scalecirculationonsurfacewindprojectionsforwindpoweroverEurope

BrayshawDavid(1),GonzalezPaula,ZappaGiuseppe

1-Departementofmeteorology,UniversityofReading(UnitedKingdom)

Europeanpowersystemshaveasignificant-andgrowing-exposuretochangesinnear-surfacewindspeeds. Despite this, there is little process-based research on the impact of climate change onEuropean wind, with GCM simulations showing inconsistent responses.This paper quantifies thecontributionofchanges in large-scalecirculationonnear-surfacewindspeedsusingreanalysis, theCMIP5 ensemble and new PRIMAVERA high-resolution simulations. In the present-day, reanalysisproducts reveal a strong connection between near-surface wind variability and the large-scalecirculationondailytomonthlytimescales,particularlyinthecoldseason.Thisrelationshipcanbewellcapturedbythefirsttwoempiricalorthogonalfunctions(EOFs)of850hPaNorthAtlanticzonalwind.Incomparisonwith reanalysis,historicGCMsimulations reproduce thecharacteristicsof the large-scale circulation to surface-wind relationship well but the strength of the connection is typicallysomewhatweaker.Usingregression-basedmethods,therawprojectedclimatechangesinwindspeedoverWesternEuropearethendecomposedintotwocomponents:oneassociatedwiththelarge-scalecirculationchanges,anda ‘residual'.The largescalecirculationcomponent isshowntoexplainthewindinessprojections inthefreetroposphereoverWesternEurope,butcannotcompletelyexplainthecorrespondingnear-surfacewindspeedprojections(anegativeresidualisfoundinthemajorityofGCMs).ThissuggeststhatprojectedC21nearsurfacewindspeed-changesoverWesternEuropearetheresultoftwodistinctprocesses:thefirstassociatedwithchangesinthelarge-scaleatmosphericcirculation,andasecondlikelytobeassociatedwithmorelocalchangesinnear-surfaceboundary-layerstructure.ImprovedunderstandingofbothprocessesandtheirrepresentationinGCMsisneededforconfidentwindprojectionsonmulti-decadaltimescales.

Keywords:climatevariability,windpower,surfacewind,windspeed,NorthAtlanticJet,climatechange


4-P06

ImpactofmodelresolutiononArcticseaiceandNorthAtlanticOceanheattransport

DocquierDavid(1),GristJeremyP.(2),RobertsMalcolmJ.(3),RobertsChrisD.(4)(3),SemmlerTido(5),PonsoniLeandro(1),MassonnetFrancois(1),SidorenkoDmitry(5),

SeinDmitry(5)(6),IovinoDoroteaciro(7),FichefetThierry(1)

1-UniversitécatholiquedeLouvain(Belgium),2-NationalOceanographyCentre(UnitedKingdom),3-MetOfficeHadleyCentre(UnitedKingdom),4-EuropeanCentreforMedium-RangeWeatherForecasts(UnitedKingdom),5-AlfredWegenerInstitute,HelmholtzCentreforPolarandMarine

Research(Germany),6-ShirshovInstituteofOceanology(Russia),7-FondazioneCentroEuro-MediterraneosuiCambiamentiClimatici(Italy)

Arcticsea-iceareaandvolumehavesubstantiallydecreasedsincethebeginningofthesatelliteera.Concurrently,thepolewardheattransportfromtheNorthAtlanticOceanintotheArctichasincreased,partly contributing to the loss of sea ice. General circulation models (GCMs) with finer and finerresolution help understanding the complex interplay of processes at high latitudes. Here, weinvestigatetheimpactofmodelresolutiononArcticseaiceandAtlanticOceanheattransport(OHT)byusingfourdifferentstate-of-the-artcoupledGCMs(withtwoconfigurationseach)representingtheatmosphere,oceanandseaice.ThemodelsparticipateintheHighResolutionModelIntercomparisonProject(HighResMIP)ofthesixthphaseoftheCoupledModelIntercomparisonProject(CMIP6).Modelresultsovertheperiod1950-2014arecomparedtodifferentobservationaldatasets.Wefindthatafineroceanresolutiondrives lowerArcticsea-iceareaandvolumeandgenerallyenhancesAtlanticOHT,bringingthemodelsincloseragreementwithobservations.Therepresentationofoceansurfacecharacteristics,suchassea-surfacetemperature(SST)andvelocity,isgreatlyimprovedbyusingafineroceanresolution.ThisstudyhighlightsaclearanticorrelationatinterannualtimescalesbetweenArcticseaice(areaandvolume)andAtlanticOHTinallmodels.However,thestrengthofthisrelationshipisnotsystematicallyimpactedbymodelresolution.ThehigherthelatitudetocomputeOHT,thestrongerthe relationship between sea-ice area/volume and OHT. Sea ice in Barents, Kara and Greenland-Iceland-NorwegianSeasismorestronglyconnectedtoAtlanticOHTthanotherArcticseas.

Keywords:Modelresolution,Arcticseaice,Oceanheattransport


4-P07

ImpactofstochasticphysicsonclimatesimulationswithEC-Earth:lookingattheatmosphere

FabianoFederico(1),MecciaVirna,CortiSusanna

1-IstitutodiScienzedellAtmosferaedelClima[Bologna](Italy)

The inclusion of stochastic physics (SP) schemes in climate models has been found to produceimprovementsinmodelperformanceswithrespecttotheobservations,partiallyfillingthegapintherepresentation of sub-grid scale processes. We analyze two sets of EC-Earth coupled climatesimulationsperformedundertheSPHINXproject(Davinietal.,2017):onesetofsimulationsincludestheStochasticallyPerturbedParametrizationTendencies(SPPT)schemeappliedtotheatmosphericmodel;theothersetisrunwiththestandarddeterministicmodel,withoutSP.ForbothsetstheRCP8.5scenarioforcingisapplied.TheexperimentswithSPshowasignificantlylowerclimatesensitivitywithrespecttothestandardexperiments.Wefocushereonthedifferencesintheradiativefeedbacksbetweenthetwoensembles.Itisfoundthatthelargestdifferencesarepresentinthetropicsandweinvestigatetheimplicationsofthisresultsontheenergybalanceandmeridionaltransportinthewholeatmosphere.Besides,weanalyze theperformanceofbothsetsofexperiments in representing thewintertime(DJF)weatherregimesintheNorth-AtlanticandNorth-Pacificsectors.Thecomparisonisdone in terms of the regime patterns, the associated frequencies and residence times and thetransitionprobabilitiesbetweendifferentregimes.Wethenstudythelinkofeachregimewithregionaltemperature/precipitationextremesinbothensembles.

Keywords:stochasticphysics,radiativefeedbacks,weatherregimes


4-P08

Aerosolmidlatitudecycloneindirecteffectsinobservationsandhigh-resolutionsimulations

FieldPaul(1)(2),MccoyDaniel(1),SchmidtAnja(3),GrosvenorDan(1),BenderFrida(4),ShipwayBen(2),HillAdrian(2),WilkinsonJonathan(2),ElsaesserGregory(5)

1-UniversityofLeeds(UnitedKingdom),2-MetOffice(UnitedKingdom),3-UniversityofCambridge[UK](UnitedKingdom),4-DepartmentofMeteorology[Stockholm](Sweden),

5-ColumbiaUniversity(UnitedStates)

The response of midlatitude cyclone cloud properties to a change in cloud droplet numberconcentration(CDNC) ispresented.AspartofthePRIMAVERAproject, idealizedexperimentsset inconvection-permitting global aquaplanet simulations with different CDNC values are compared tothirteen years of remote-sensing observations. Simulations at 7 km horizontal resolution arecontrasted with traditional GCM-representative 200 km horizontal resolution simulations. Thiscomparisonshowsqualitativeagreement,butdemonstratesmorerealisticrepresentationofaerosol-cloud interactions at higher resolution. Observations and simulations agree that increased warmconveyorbelt(WCB)moistureflux intocyclones isconsistentwithhighercycloneliquidwaterpath(CLWP).Forcyclonesthathavethesamemeancyclone-widerainrate,higherCLWPisexhibitedforhigherCDNC.TheCLWPadjuststoallowtherainratetobeequaltothemoisturefluxintothecyclonealong thewarm conveyor belt. This results in an increasedCLWP for higher CDNCat a fixedWCBmoisturefluxinbothobservationsandsimulations.IfobservedcyclonesinthetopandbottomtercileofCDNCarecontrasteditisfoundthattheynotonlyhavehigherCLWP,butalsolargercloudcover,andalbedo.

Keywords:convectionpermittingglobalmodel,lifetimeeffect,PRIMAVERA


4-P09

Impactofchangesinatmosphericandoceanmodelresolutiononmodesofvariabilityinhistoricalcoupledmodelsimulations

FuentesFrancoRamon(1),KoenigkTorbenDoescherRalf,ZimmermannKlaus

1-SMHI(Sweden)

Weusehistoricalsimulationsfromcoupledgeneralcirculationmodels(GCMs)participatingintheEUH2020 project PRIMAVERA, to assess the impact of changes in ocean and atmospheric modelresolutionondifferentcoupledmodesofvariability. Inan inter-modelcomparisonacrossdifferentatmosphericandoceanmodel resolutions,weanalysedifferenceson the representationof spatialpatterns,intensitiesandfrequenciesofElNinoSouthernOscillation(ENSO),theAtlanticMultidecadalOscillation(AMO),thePacificDecadalOscillation(PDO),theNorthAtlanticOscillation(NAO)andthePacificNorthAmericapattern(PNA)comparedwithobservationsandreanalysis.RegardingtheENSOspatialpattern,wefoundthatwithhigheroceanoratmosphericmodelresolutiontheGCMstendtoshowwarmerSeaSurfaceAnomalies(SST)intheeasternequatorialPacific.However,wedidnotfindaclearimprovementontheENSOcycle,andontheENSOpowerspectrawithincreasedresolution.SST anomalies related to PDO are in general underestimated by all GCMs and there is no clearimprovementwith increasedmodel resolution. In contrast, with increased resolutionmost of theGCMsshowanimprovementoftheAMOrelatedSSTpattern,withsomebiasesoverthenorthAtlantic.The atmospheric response to ENSO tends to be weaker in all GCMs compared to the reanalysis,showing a systematic weaker Aleutian low response to ENSO with increased model resolution.HowevermostoftheGCMsshowanimprovementintheprecipitationanomaliesrelatedtoENSOwithincreasedresolution,particularlyoverNorthAmerica.WeakersealevelpressureanomaliesassociatedtoNAOandPNAarefoundsystematicallyacrossmodelswithincreasedresolution,onlyNAO(asENSO)showsanimprovedrepresentationofitsassociatedprecipitationwithincreasedresolution.

Keywords:impactofresolution,highresolution,Climatevariability


4-P10

TowardsanenergeticallyconsistentverticaloceanmixingschemeinMPI-ESM

GutjahrOliver(1),JungclausJohann(1),BrueggemannNils(2),HaakHelmuth(1)

1-MaxPlanckInstituteforMeteorology(MPI-M)(Germany),2-InstitutfürMeereskunde,UniversityofHamburg(Germany)

Unresolvedsmall-scale turbulentkineticenergy (TKE)needs tobeparameterized inoceanmodels.Below themixed layer, a constant background diffusivity is commonly assumed. This backgrounddiffusivity is supposed to parameterize the breaking of internal gravity waves, but it lacks anyspecificationofitsenergysource.Althoughsmall,thisdiffusivityisyetimportantforforcingthelarge-scaleoceancirculationbutisinconsistentinthatanartificialenergysourceiscreatedinoceanmodels.Motivatedbythisfact,wereplacedtheartificialbackgrounddiffusivityintheMaxPlanckInstitute?Earth System Model (MPI-ESM) by a prognostic equation for the internal wave energy and itsdissipation,usingtheInternalWaveDissipation,EnergyandMixing(IDEMIX)scheme(OlbersandEden2013,doi: 10.1175/JPO-D-12-0207.1).By following theHighResMIPprotocol,weperformeda twinexperiment with 150 yearlong globally coupled control simulations with theMPI-ESM1.2-HR. Thereference simulation uses the TKE schemewith a constant background diffusivity, whereas in thesensitivitysimulationthebackgrounddiffusivityisreplacedbyaprognosticequationforinternalwavebreakingfromIDEMIX.Wefindasignificanteffectonthelarge-scalecirculation,whenprognosticallyparameterizing thebreakingof internalwaves. Inparticularaffected is thecirculation in theNorthAtlanticandintheArcticOcean,aswellasthewatermassesinthesebasins.

Keywords:PRIMAVERA,oceanverticalmixing,internalgravitywaves,turbulence,parameterizations,MPI,ESM


4-P11

Extra-tropicaltransitionofAtlantichurricanesinPRIMAVERAHighResMIPTier1simulations

HaarsmaRein(1),KapetanakisDimitris(1),BakerAlex(2)

1-RoyalNetherlandsMeteorologicalInstitute(Netherlands),2-UniversityofReading(UnitedKingdom)

Wehaveanalyzed theHighResMIPTier1 simulationsperformedby thePRIMAVERApartnerswithrespecttotheextra-tropicaltransitionoftropicalcyclones.WehavefocusedontheNorthAtlanticandextra-tropicalcyclonesthatreachEurope.Forthosestormswehaveanalyzedtheircharacteristicsandcompared themwith observations and reanalyses products. Specifically we addressed the role ofwarm-seclusionmechanism in the extra-tropical transition using Hart diagrams.Most PRIMAVERAmodelsareabletorepresentreasonablywelltheobservedcharacteristicsofextra-tropicaltransition.InagreementwiththestudyofDekkeretal.(2018)about50%ofthestormsthatreachEuropearewarm-seclusionstorms.

Keywords:Hurricanes,extra,tropicaltransition,Europe


4-P12

CriticalSouthernOceanclimatemodelbiasestracedtoatmosphericmodelclouderrors

HyderPat(1),EdwardsJohn(1),HewittHelene(2),AllanRichard,GregoryJonathan,WoodRichard(1)

1-MetOffice(UnitedKingdom),2-MetOffice(UnitedKingdom)

TheSouthernOceanisapivotalcomponentoftheglobalclimatesystemyetitispoorlyrepresentedinclimatemodels,withsignificantbiasesinupper-oceantemperatures,cloudsandwinds.CombiningAtmospheric and Coupled Model Inter-comparison Project (AMIP5/CMIP5) simulations, withobservationsandequilibriumheatbudgettheory,weshowthatacrosstheCMIP5ensemblevariationsin sea surface temperature biases in the 40 60¡S Southern Ocean are primarily caused by AMIP5atmospheric model net surface flux bias variations, linked to cloud-related short-wave errors.Equilibrationof thebiases involves local coupled sea surface temperaturebias feedbacksonto thesurfaceheat flux components. In combinationwithwind feedbacks, thesebiasesadverselymodifyupper-ocean thermal structure. Most AMIP5 atmospheric models that exhibit small net heat fluxbiasesappeartoachievethisthroughcompensatingerrors.WedemonstrateusingtheHadleyCentreCMIP6modelthattargeteddevelopmentstocloud-relatedparameterisationsprovidearoutetobetterrepresenttheSouthernOceaninclimatemodelsandprojections.

Keywords:SSTbias,surfacefluxes,southernocean


4-P13

EffectsofOrographyontheMeanSouthAsianMonsoonCirculationandRainfallinCMIP6Models

KinterJim(1),GaalRachel(1)

1-GeorgeMasonUniversity[Fairfax](UnitedStates)

What is the effect of the orography of the Himalaya/Tibetan Plateau on the development andmaintenanceoftheAsianmonsoon?EmbeddedwithinthisquestionarethehypothesesthatthesouthAsianmonsoonisprimarilytheresultoftheelevatedheatsourceontheTibetanPlateau(YeandWu1998 and references therein) and the Himalayas acting as a physical barrier to atmospheric flow,creating a thermal insulator for monsoon development (Boos and Kuang 2010). There is somedisagreement about the relative roles of these processes in determining themeanmonsoon. TheGlobalMonsoonsModelInter-comparisonProject(GMMIP),anendorsedCMIP6project,focusesonmonsoonclimatology,variability,predictionandprojection.TheGMMIPprotocol(Zhouetal.2016)includestwomodelexperiments(amip-TIPandamip-TIP-nosh)designedtoexaminetheeffectsofthelargeregionalorographyontheestablishmentofthesouthAsianmonsoon.Theamip-TIPexperimentfollowsWuet al. (2007) andBoos andKuang (2010) bymodifying the topographyof the Tibetan-Iranian Plateau (TIP) in the model, leveling off the TIP to 500m, with other surface propertiesunchanged.Intheamip-TIP-noshexperiment,thesurfacesensibleheatfluxatelevationsabove500movertheTIP isnotallowedtoheattheatmosphere(Wuetal.,2012).Theatmosphericcomponentexperienceszerosurfaceupwardsensibleheatflux,whereasthelandcomponentisasusual.Inthispaper,weanalyzetheresultsoftheamip-TIPandamip-TIP-noshexperimentsofGMMIP.Theanalysiswillfocusonthemeanmonsooncirculationandrainfall,andthesurfaceenergybudget,includingbothensemblemeanandintra-modeldifferences.

Keywords:monsoon


4-P14

DeepwaterformationintheNorthAtlanticOceaninhighresolutionglobalcoupledclimatemodels

KoenigkTorben(1),FuentesFrancoRamon(1)

1-SMHI(Sweden)

Duringthepastdecades,theideaofaweakeningandevencollapsingAtlanticMeridionalOverturningCirculation(AMOC)asaresponsetoglobalwarming,anditspossibleimpactontheclimateofEurope,has been a recurrent andheateddebatewithin the climate community. Themain argument for apossible decline of theAMOC is the reduction of deepwintertime convectivemixing in theNorthAtlantic.However,manyglobalclimatemodelsshowlargebiasesanduncertaintiesintheoceanicdeepconvection.Here,weanalyzetheimpactofresolutionontherepresentationoftheNorthAtlanticdeepconvectionincoupledclimatemodels.

Coupledhistoricalandcontrolsimulationsindifferentresolutions,performedaspartoftheH2020-EU-PRIMAVERA-project(followingtheHighResMIP-protocol),areusedfortheanalysis.

IncreasedresolutionleadstoincreaseddeepconvectionintheLabradorSeaanddecreasedconvectionin the Greenland Sea. Compared to observations from ARGO-floats, the high resolution modelsstronglyoverestimatetheconvectionintheLabradorSea.ThebiasissmallerintheGreenlandSeaandsomewhatreducedcomparedtothestandardresolutionmodelsimulations.

TheconvectionintheLabradorSeaislargelygovernedbytheoceanheatreleasetotheatmosphere.Northwesterlyatmosphericflows,oftenconnectedtoapositivestateoftheNorthAtlanticOscillation,increasetheoceanheatreleaseandthusthedensityoftheoceansurfaceintheLabradorSea.Thehigh-resolutionmodelsshowstrongersurfaceheatfluxesthanthestandardresolutionmodelsintheconvectionareas,whichagreeswiththestrongerconvectionintheLabradorSea.AlsointheGreenlandSea,highresolutionleadstoanincreasedoceanheatreleasetotheatmosphere.However,here,therelationbetweensurfaceheatfluxesandconvectionismodeldependent.

Keywords:highresolutionmodelling,oceanicconvection,AtlanticMeridionalOverturningCirculation,NorthAtlantic


4-P15

AnalysisofMesoscaleConvectiveSystemsinMPAS-CAM5HighResolutionandConvectionPermittingSimulations

LeungRuby(1),Feng,SakaguchiKoichi,SongFengfei,HouzeRobert

1-PacificNorthwestNationalLaboratory(UnitedStates)

TheModelforPredictionAcrossScales(MPAS)isanonhydrostaticglobalvariableresolutionmodelingframework.UsingMPAScoupledwiththeCommunityAtmosphereModel(CAM5)physicspackage,asuiteofsimulationsisperformedincludingAMIPsimulationsforCMIP6HighResMIPat30kmand120kmandsimulationsatgridspacingrangingfrom4kmto50kmusingregionalrefinementoverNorthAmerica.Thesesimulationsprovideanopportunitytosystematicallyevaluatetheimpactsofmodelresolutiononsimulationofextremeevents.Wewillpresentanalysisofmesoscaleconvectivesystems(MCSs),whichareresponsibleformostoftheextremeprecipitationeventsduringthewarmseasoninthecentralandeasternU.S.Withadistinctnocturnalmaximum,failuretosimulateMCSsintheU.S.hasledtoatypicalbiasindiurnalcycleofprecipitationinCMIP3andCMIP5models.ToevaluatetheimpactsofmodelresolutiononsimulationsofMCSs,wecomparetheobservedandsimulatedMCSsandtheirlarge-scaleenvironmentsacrossarangeofresolutions.ApplyinganMCStrackingalgorithm(FLEXTRKR)toobservedandsimulatedoutgoing longwaveradiationandprecipitation,wecompareobserved and simulated spatial distribution ofMCS precipitation and its contribution to the totalprecipitation and MCS statistics such as MCS rain rate, size, and propagation speed. Using selforganizingmaps(SOM),weidentifythelarge-scaleenvironmentassociatedwithMCSsinobservationsandcomparethefrequencyofoccurrenceofMCSfavorableenvironmentinobservationsandmodelsimulationstounderstandmodelbiasesinsimulatingMCSsintheU.S.Ouranalysisdemonstrateslargedifferences in springvs. summerMCS large-scaleenvironmentandMCScharacteristics, suggestingmorechallengesformodelstosimulatesummerthanspringMCSs,eventhoughMCSsinbothseasonscontributealmostequallytototalprecipitationinthecentralandeasternU.S.

Keywords:mesoscaleconvectivesystems,impactsofmodelresolution,MCSlarge,scaleenvironment


4-P16

Climatemodelingwithamulti-gridapproach

MaurerVera(1),BrauchJennifer(1),FruhBarbara,StegerChristian(1)

1-DeutscherWetterdienst(DWD)(Germany)

TheIcosahedralnonhydrostatic(ICON)modelisajointdevelopmentofDeutscherWetterdienst(DWD)and of Max-Planck-Institute for Meteorology. The numerical-weather-prediction version of ICON,ICON-NWP,hasbeenrunningoperationallyatDWDformorethanthreeyearsnowusingamulti-gridapproach,i.e.withahorizontallyhigherresolveddomainoverEuropeconnectedtoaglobaldomainvia2-waynesting.

Inthefirstpartofourproject,wetransferredthemulti-gridapproachofICON-NWPtotheatmosphericpart of ICON-ESM with the aim of producing high-resolution climate projections for Europe. Thetargetedresolutionis10-20km.Inthesecondpart,weinvestigatetheinfluenceofhigherhorizontalresolutionon(1)extremeeventsoverEuropesuchasheatwavesandheavyprecipitationusingtheatmosphericpartofICON-ESMandon(2)modelbiasessuchasthenorthern-hemispherecontinentalsummerwarmbiaswhichhasalreadybeenpresentinMPI-ESMandseemstopersistinICON-ESM.

Keywords:highresolution,climateprojection,ICON,ESM


4-P17

Cloudfeedbacksinextratropicalcyclonesandanti-cyclones:insightfromlong-termsatellitedataandhigh-resolutionglobalsimulations

MccoyDaniel(1),FieldPaul(2),RobertsMalcolm(3),VidalePierLuigi(4),VanniereBenoit(4),KodamaChihiro(5),ElsaesserGregory(6),KahnBrian(7),ZelinkaMark(8),MauritsenThorsten(9),VoldoireAurore(10),HaarsmaRein(11),HillAdrian(2),ShipwayBen(12),WilkinsonJonathan(2),

Bodas-SalcedoAlejandro(13)

1-UniversityofLeeds(UnitedKingdom),2-MetOffice(UnitedKingdom),3-MetOffice(UnitedKingdom),4-UniversityofReading(UnitedKingdom),5-JapanAgencyforMarine-EarthScienceandTechnology(Japan),6-ColumbiaUniversity[NewYork](UnitedStates),7-JetPropulsionLaboratory

(UnitedStates),8-LawrenceLivermoreNationalLaboratory(UnitedStates),9-DepartmentofMeteorology[Stockholm](Sweden),10-Grouped'étudedel'atmosphèremétéorologique(France),11-RoyalNetherlandsMeteorologicalInstitute(Netherlands),12-MetOffice(UnitedKingdom),

13-MetOfficeHadleyCentre(UnitedKingdom)

Extratropical cloud feedbacks represent one of the leading uncertainties in calculating climatesensitivity.Cloudbehaviorintheextratropicsisacomplexmixoflarge-scalesynopticactivityandcloudmicrophysicsencompassingiceandliquidphases.Ithasbeensuggestedthatice-to-liquidtransitionsdrivethenegativecloudfeedbackacrosstheextratropics.CMIP5modelsdisagreedbyupto35Kastothetemperaturewhereiceandliquidwereequallylikelyinclouds.Thismodellinguncertaintyanditspotential impacton climate sensitivitymake it important tounderstand.Weuse a simple regime-compositingapproachtoseparatecyclonesandanti-cyclonesandexaminethecloudfeedbackwithineachregime.Wecontrastglobalclimatemodels(GCMs)withhorizontalresolutionsfrom7kmuptohundreds of kilometers with Multi-Sensor Advanced Climatology Liquid Water Path (MAC-LWP)microwaveobservationsofcyclonepropertiesfromtheperiod1992-2015.Inter-cyclonevariabilityinbothobservationsandmodelsisstronglydrivenbymoisturefluxalongthecyclone'swarmconveyorbelt (WCB). Stronger WCB moisture flux enhances liquid water path (LWP) within cyclones. Wepropose two cloud feedbacks acting within extratropical cyclones: a negative feedback driven byClausius-Clapeyronincreasingwatervaporpath(WVP),whichenhancestheamountofwatervaporavailable tobe fluxed intothecyclone;anda feedbackmoderatedbychanges in the lifecycleandvorticityofcyclonesunderwarming,whichchangestherateatwhichexistingmoistureis importedintothecyclone.WeshowthatchangesinmoisturefluxcanpredictthemajorityoftheobservedtrendinSouthernOceancycloneLWPoverthelasttwodecadesandtheresponsetoforcinginCMIP5GCMs.AsetofperturbedphysicssimulationsintheUMisundertakentoseparateeffectsrelatedtoice-to-liquidtransitionsfrommeteorologicalvariability.

Keywords:cloudfeedbacks,mixedphasecloudfeedback,cloudphysics,climatesensitivity,PRIMAVERA,NICAM,ICON,CASIM,high,resolution,midlatitudes,extratropics,cyclones,observations


4-P18

ImpactofstochasticphysicsonclimatesimulationswithEC-Earth:lookingattheocean.

MecciaVirna(1),FabianoFederico(1),CortiSusanna(1)

1-IstitutodiScienzedell’atmosferaedelClima[Bologna](Italy)

Stochasticphysicsschemesprovideamorerealisticrepresentationoftheunresolvedscalesinglobalcirculationmodelsby improvingbothmeanclimateandclimatevariability.Westudythe impactofincludinga stochasticphysics scheme in theatmospheric componentofEC-Earthon thesimulatedocean state. The experiments consist of coupled climate simulations in which three ensemblemembersconstitutethecontrolrunsandthreeensemblemembersincludestochasticphysics.Forthelatter,theStochasticallyPerturbedParametrisationTendencies(SPPT)schemeisincorporatedintheatmosphericcomponentofEC-Earth.Theperiodofsimulationspansfrom1850to2100andthefuturescenario corresponds to RCP8.5. The two ensembles present a different climate sensitivity. Wecomparebothoceanstatestoinvestigatetheoceanmodelresponsetoaperturbedatmosphere.Thesurfacenetdownwardheat flux in theocean results higher for the control runs. Thedifference isapproximately4x10^23Jintheentireperiodof250years.Themeridionalheattransport(MHT)intheoceansresultsalsohigherinthecaseofthecontrolruns.ThePacificandAtlanticoceanscontributeroughlywith 57% and 43%of the total difference, respectively. The larger amount ofMHT in theAtlanticinthecontrolrunsrespecttothestochasticruns,causesdifferencesintheAtlanticMeridionalOverturningCirculationaswell.AnabruptlossofwinterseaiceintheArcticoccursattheendofthesimulatedperiod for the control runsonly. In this case, the global annual surface air temperaturereachesthethresholdvalueforanabruptcollapseofwinterseaiceintheArctic.Thisfeatureisnotobserved when the stochastic physics is included, at least during the analyzed period. Includingstochastic physics in EC-Earth might yield differences in the simulated climate comparable todifferencesbetweenmodels.

Keywords:Stochasticphysics,EC_Earth,Ocean.


4-P19

BombCyclonesinPRIMAVERASimulations

MinobeShoshiro(1),GaoJiaxiang,RobertsMalcolm(2),HaarsmaRein(3),PutrasahanDian(4),RobertsChristopher(5),ScoccimarroEnrico(6),

TerrayLaurent(7),VidalePierLuigi(8)

1-HokkaidoUniversity(Japan),2-MetOffice(UnitedKingdom),3-RoyalNetherlandsMeteorologicalInstitute(Netherlands),4-CREFO,OISE,UniversitédeToronto(Canada),

5-EuropeanCentreforMedium-RangeWeatherForecasts(UnitedKingdom),6-CentroEuroMediterraneosuiCambiamentiClimatici(Italy),7-EuropeanCenterforResearchandAdvancedTraininginScientificComputation(France),8-UniversityofReading(UnitedKingdom)

Bombcyclonesorexplosivelydevelopingextratropicalcyclonescancauseserioushazardsforsociety.Thus,itisimportanttounderstandhowtheglobalwarminginfluencebombcyclones.Inthisstudy,wetry to address two questions: 1) Howwill bomb cycloneswill change due to the globalwarming?2)Howarebombcyclonesinfluencedbymodelresolutions?Toaddressthesequestions,weanalyzeoutputs of PRIMAVERA project under collaboration between PRIMAVERA and CLIVAR. A novelexperiment design of PRIMAVERA is that the same experiment is repeated for lower and higherresolutionsofeachmodel,andthisallowsustoaddressthesecondquestion.Atthetimeofabstractsubmission,wehaveinvestigatedthesecondquestiononlybyanalyzinghistoricalsimulationoutputsoffiveAGCMs,butwewillextendouranalysisforfuturesimulationsandwillanswerthefirstquestiontoo. Bomb cyclones are identified by a detection algorithm over the Northern Hemisphere. Thelocationsofbombdevelopmentsareclusteredover thewesternNorthAtlanticandwesternNorthPacific,roughlycollocatedwiththeKuroshioandtheGulfStreamandtheirextensions,consistentwithobserved features. It is found that thenumberofbombcyclones increase in thehigher resolutionversionofmodelsthaninthelowerresolutionversionformostofmodelsandacrossdifferentmodels.Ifweestimatethebombcyclonenumberasafunctionofmodelresolution,theexpectednumberof25kmresolution ismorethan30% larger thatof100km.Althoughthis isacrudeestimation, it isstrongly suggested that typical standard CMIP5/CMIP6 model resolution can substantiallyunderestimatebombcyclones.

Keywords:modelresolutions,extratropicalcyclones,bombcyclones,airseainteraction,CLIVAR


4-P20

Temporalandspatialintermittencyofsub-dailyprecipitationinAustralianmonsoonandmaritimecontinentlinkedtoGCMprecipitationbiases

MoiseAurel(1),MartinGill,KlingamanNicholas

1-BureauofMeteorology[Melbourne](Australia)

DespitesomeimprovementincoupledGCMdevelopmentseenoverrecentyears,largeprecipitationbiases across tropical regions remain, especially across the Maritime Continent and AustralianMonsoondomain.Oftenclimatemodelbiasesonmulti-yearandglobalscalesbothdevelopwithinafewdays of the start of the simulation and are closely related to deficiencies in the simulationofprocessesonmuchshorterandsmallerscales.Differentmodelparameterizationscanproduceeitherveryintermittentorverypersistentrainfallatthelevelofthemodel'stime-stepandgridscale,andalsoproduceapoorrepresentationoftheprocessesandtimingassociatedwiththediurnalcycleofconvectionoverland.Suchdeficienciescanhaveasignificantimpactontheregional-scalecirculationand water cycle and thereby undermine confidence in projections of the spatial and temporalcharacteristics of heavy rainfall in a future climate. In our study,we examined the distribution ofrainfallintensity,anditscoherenceacrosstemporalandspatialscales,overtheMaritimeContinentdomain and the tropical Australia monsoon region in a subset of recent-generation coupledatmosphere-oceanglobalclimatemodels.Weusearecently-publishedsetofdiagnostictools(ASoP)which allows sub-daily rainfall data from the models to be compared with those from satelliteobservations(here:CMORPHandTRMM)andprovidesinsightintohowtherainfallvariabilitychangesasthetemporalandspatialscalesincrease.Resultswillbepresented.

Keywords:precipitationbias,monsoon,diurnalcycle


4-P21

MoisturetransportassociatedtoTropicalCyclones.

PeanoDaniele(1),ScoccimarroEnrico(2),BellucciAlessio(2),CherchiAnnalisa(3)(1),D'ancaAlessandro(4),AntonioFabrizio(4),FioreSandro(4),GualdiSilvio(2)(3)

1-CentroEuro-MediterraneosuiCambiamentiClimatici(Italy),2-CentroEuroMediterraneosuiCambiamentiClimatici(Italy),3-IstitutoNazionalediGeofisicaeVulcanologia-SezionediBologna

(Italy),4-Euro-MediterraneanCenteronClimateChange(Italy)

Tropical cyclones (TCs) transport energy and moisture along their pathways interacting with theclimate system. The impact on June-to-September precipitation due to zonal moisture transportassociatedwithTCsintheMaritimeContinentandthemeridionalmoisturetransportcarriedbyTCsintheNorthAtlanticareinvestigatedinasetofAMIP(atmosphere-onlysimulationsforcedwithobservedseasurfacetemperatures)andcoupledsimulationsperformedatdifferentspatialresolutions.Theroleof TCs in removing humid air from the Maritime Continent during the boreal summer season ishighlighted comparing precipitation and vertically integrated zonalwater transport between yearswithintenseTCactivityandlowTCactivityyearsalongtheperiod1980-2015.This ismadeevidentbasedonJRA-55atmosphericreanalysisandconfirmedbymodellingexperimentswhereresultsfromaTCsresolvingmodelarecomparedtoalowerresolutionmodel.Theradialaveragedistributionofspecific humidity along the TC tracks is also inspected and used in evaluating the TCsmeridionalmoisture transport in the North Atlantic. When compared to observation (IBTRACS and JRA-55reanalysis),thesimulatedwatercontentassociatedtoTCsdisplaysreasonablygoodperformanceinhigh-resolution model configuration in Tropical areas. In extra-tropical region, instead, modelsoverestimatetheTCswatercontentcomparedtoobservation.ThiseffortispartofHighresMIPanditisdevelopedintheframeworkoftheEU-fundedPRIMAVERAproject.

Keywords:Tropicalcyclones,moisturetransport,MaritimeContinent,NorthAtlantic


4-P22

Quantifyingtropicalcyclonerainfallandsizeinhighresolutionclimatesimulations

ReedKevin(1),StansfieldAlyssa(1),ZarzyckiColin(2),UllrichPaul(3)

1-SchoolofMarineandAtmosphericSciences(UnitedStates),2-NationalCenterforAtmosphericResearch[Boulder](UnitedStates),3-UniversityofCalifornia[Davis,USA](UnitedStates)

Recently,generalcirculationmodelsrunathighhorizontalresolutionshavebeenusedtostudythelarge-scale controlsof global and regional tropical cyclone (TC) activity inpast, present and futureclimate.ThemainfocusofthisworkistobetterunderstandtheextremeprecipitationassociatedwithTCs, with an emphasis on landfalling TCs in the Eastern U.S. In particular, this study uses variousconfigurationsoftheCommunityAtmosphereModelversion5(CAM5),acomprehensiveatmosphericgeneral circulationmodel, forcedwithprescribed sea-surface temperatures (SSTs)andgreenhousegases for present climate. Three different variable-resolution gridswith horizontal grid spacing ofapproximately 25 km over all, or parts, of the North Atlantic are investigated. The impact of thevariable-resolutiongridchoiceonsimulatedTCsandtheirassociatedprecipitationisinvestigatedusingtheupdatedTempestExtremessoftwaretocalculatestormsizeandextractstormrelativerainfall.ThisisalsoanimportantfirststepininvestigatinghowTCvariability,sizeandextremerainfallwillbealteredin a future climate. This work is synergistic with High Resolution Model Intercomparison Project(HighResMIP)activitiesastheTempestExtremesframeworkisdesignedtoallowforefficientmodelintercomparisonofextremeevents.

Keywords:Extremes,HighResolution,TropicalCyclones


4-P23

CoordinatedGlobalHighResolutionClimateModelling?PRIMAVERAandCMIP6HighResMIP

RobertsMalcolm(1),TeamPrimavera

1-MetOffice(UnitedKingdom)

Oneoftheuncertaintiesintherolethatmodelhorizontalresolutionplaysinglobalclimatesimulationhasbeenthelackofacoordinatedexperimentaldesignandalargeensembleofmulti-model,multi-resolutionsimulationstoenabledetailedanalysisoftherelevantprocesses.TheCMIP6HighResMIPprotocol for1950-2050hasprovidedthe former,while theEuropeanH2020projectPRIMAVERA iscontributing 7 different models to the multi-model ensemble (in addition to other internationalgroups),atresolutionsrangingfrom250km-1to25km-1/12.

WithafocusontheAtlanticandEuropeanclimatevariabilityandchange,ouranalysishasexaminedmean statebiases suchas SST,precipitation,Gulf Streampositionandvariability, aswell as fluxesbetween the Atlantic and Arctic systems. Aspects of variability such as the Atlantic MeridionalOverturningCirculation(AMOC)andAtlanticMultidecadalOscillation(AMO)havebeenexamined,aswellasAtlantictropicalcyclones.

Oneofthemainchallengesinanalysingdatasetssuchasthesearethedatavolumes?weexpecttostoreatleast3PBofdata.Ourcoordinatedanalysishasusedacommonplatform(CEDA-JASMIN),onwhichourdata,algorithmsandprocessingresourceareavailable.Wealsohaveongoinganalysiswithavarietyofotherpartners (e.g.CLIVAR, tropical cyclonecommunity, IPCCchapters), forwhich theapplicationofstandardisedmethodsandalgorithmswillenableustoshareresultsmoreeasily-thisapproachemphasisesmovingthealgorithmtothedataratherthanviceversa.Thedatasets,bothrawmodeloutputandsomederivedproductssuchasstormtracks,willbeuploadedtotheCMIP6ESGFandCEDAarchives.

In addition, we have been investigating the number of ensemblemembers necessary to producerobustresultsinboththeimpactofhorizontalresolution,aswellasseparatingclimatechangesignalsandtrendsfrominternalvariabilityandweathernoise.

Keywords:HighResMIP,resolution,multi,model


4-P24

SensitivityofAtlanticOceanbiasestohorizontalresolutioninprototypeCMIP6simulationswithAWI-CM

SeinDmitry(1),RackowThomas(1),SemmlerTido(1),DanilovSergey(1),KoldunovNikolay(1),SidorenkoDmitry(1),WangQiang(1),JungThomas(1)

1-AlfredWegenerInstituteforPolarandMarineResearch-AWI(GERMANY)(Germany)

Mostoftheclimatemodelsshowsubstantialbiasesinthedeepoceanthatarelargerthanthelevelofnaturalvariabilityandtheresponsetoenhancedgreenhousegasconcentrations.Hereweanalyzetheinfluence of horizontal resolution in a hierarchy of fivemulti-resolution simulationswith the AWIClimate Model (AWI-CM), which employs a sea ice-ocean model component formulated onunstructuredmeshes.Theoceangridsizesconsideredrangefromanominalresolutionof1¡(CMIP5-type)uptolocallyeddy-resolving.Weshowthatincreasingoceanresolutionlocallytoresolveoceaneddiesleadstoamajorreductionindeepoceanbiases.Adetaileddiagnosisofthesimulationsallowstoidentifytheoriginsofthebiases.Wefindthattwomajorsourcesatthesurfaceareresponsibleforthedeepbias in theAtlanticOcean. Furthermore, theSouthernOceandensity structure isequallyimprovedwithlocallyexplicitlyresolvededdiescomparedtoparameterizededdies.Partofthebiasreductioncanbetracedbacktowardsimprovedsurfacebiasesoveroutcroppingregions,whichareincontact with deeper ocean layers along isopycnal surfaces. Our prototype simulations provideguidancefortheoptimalchoiceofoceangridsforAWI-CMtobeusedinthefinalrunsforphase6ofthe'CoupledModelIntercomparisonProject'(CMIP6)andfortherelatedflagshipsimulationsinthe'HighResolutionModelIntercomparisonProject'(HighResMIP).Quiteremarkably,retainingresolutiononlyinareasofhigheddyactivityalongwithexcellentscalabilitycharacteristicsoftheunstructured-meshseaice-oceanmodelenablesustoperformthemulti-centennialclimatesimulationsneededinaCMIPcontextat(locally)eddy-resolvingresolution.

Keywords:CMIP,AtlanticOcean,AWICM


4-P25

AttributionofrecentchangesinextremeweatheroverEurope

TerrayLaurent(1)

1-Cerfacs(France)

Theissueofrecentchangesinmidlatitudeextremeweatherandtheircauseshasbeenanimportanttopicinrecentyears.Inparticular,alotofattentionhasbeengiventothepossiblelinksbetweenArcticamplification (AA),changes inatmospheric flowwaviness (amplifiedRossbyplanetaryandsynopticwaves)andextremeweather.Werevisit thistopicbasedonthreescientificquestions:What is theobservationalevidencesupportingtheclaimformoreextremeweather inrecentdecades?Canwedetectandattributechangesinextremeweather(andflowwaviness)tochangesinexternalforcingsand/orinternalvariability?Howrobustistheobservational/modelevidencelinkingextremeweathertoamplifiedRossbywaves?Herewefocusonthe1950-2017(andfuture)periodsandtemperatureextreme events at large spatial scale by considering regions such as Europe. Often contrasted(anomalouslywarm/cold)extremeweathereventsattheselargespatialscalesoccursimultaneouslyowing to an amplified ridge/trough pattern across a continent.We then define a simple extremeweather index (EWI) based on area-average absolute anomalies using several observed dailytemperaturedatasets.WethenshowthatEWImonthlywintertrendsarenotstatisticallydifferentfrom zero while EWI positive summer (July-August) trends are depicted over Europe. We theninvestigatetheattributionquestionusingdifferentsingleandmulti-modelensembles,bothSST-forced(HiResMIP/PRIMAVERA) and coupled (HiResMIP/PRIMAVERA, CESM-LENS, all and single-forcingsimulations)andexaminethesensitivityoftheresultstoatmospherichorizontalresolutionbasedonthePRIMAVERAmulti-modelensemble.Finally,weuseawavinessindex(RWI)basedonRossbywavepacketenvelope toshowthatextremeEWIsummerdaysarecommonly linked to (andcausedby)anomalouslyhighRWIvalues.WethenshowthattherecurrentlyisnostrongobservationalevidencesupportingtheclaimedlinkbetweenAAandchangesinwavinessandextremeweatherinsummer.

Keywords:extremeweather,atmosphericcirculation,attribution,horizontalresolution,Europe


4-P26

ProjectionofTropicalCycloneActivityintheWesternNorthPacificUsingaSingleColumnOceanCoupledModel

TuChiaying(1),HsuHuang-Hsiung(2),LinShian-Jiann(3)

1-ResearchCenterforEnvironmentalChanges,AcademiaSinica(Taiwan),2-ResearchCenterforEnvironmentalChanges,AcademiaSinica(Taiwan),


GFDL high-resolution (25-km) AGCMHiRAM and HiRAM coupled with SIT (Snow/Ice/Thermocline)singlecolumnoceanmodelwereusedforclimateexperimentsfollowingHighResMIPprotocal.IntheTier1experimentHiRAMwellsimulatesmeanclimatology,AsianMonsoonseasonalevolution,frontalactivity, and tropical cyclone-intraseasonal oscillation relationship. Strength of simulated extremeprecipitation iscompatiblewithTRMMprecipitation. In theTier2historicalexperimentHiRAM-SITproduces similar climatology as HiRAM, but better diurnal cycle of precipitation, near surfacetemperatureandMJOeastwardpropagation.DuetothelackofoceancirculationintheHiRAM-SITcoupledmodel,simulated0.25-degSSTwerenudgedtoCMIP51-degensembleSSTintheTier2futureexperimenttoavoidclimatedrift.ItisnoticedthattropicalcycloneactivityinthewesternNorthPacificis projected to be significantly weakened in the projected future climate. This is due to theequatorwardcontractionofconvectionandthecorrespondinganomaloussubsidencepolewardoftheequatorial convection belt. Strongest response occurs in the western North Pacific and results insignificantlyweakenedconvectionandwestwardextensionof thesubtropicalanticyclone.Vorticitybudgetanalysisfindsthatthebackgroundstatedecidestheamplitudeofresponse.Theanticyclonicresponse is the largest in thismonsoon trough (cyclonic) region. This is true even if the low leveldivergence anomaly is zonally symmetric. Dynamical Interaction between vorticity and divergenceanomalies, the associated thermodynamic components, and the dynamical thermodynamicinteractionlikelyfurtherenhancetheresponseinthemonsoontroughandcyclonicregions.

Keywords:HiRAM,TCactivity


4-P27

TheroleofStochasticPhysicsandmodelresolutionforthesimulationofTropicalCyclonesinAGCMs

VidalePierLuigi(1)(2),HodgesKevin(3),RobertsMalcolm(4),DaviniPaolo(5),StrommenKristian(6),WeisheimerAntje(7),CortiSusanna(5)

1-NCAS(UnitedKingdom),2-UniversityofReading(UnitedKingdom),3-NationalCentreforAtmosphericScience(UnitedKingdom),4-MetOfficeHadleyCentre(UnitedKingdom),5-Istitutodi

Scienzedell’AtmosferaeDelClima[Torino](Italy),6-UniversityofOxford[Oxford](UnitedKingdom),7-EuropeanCentreforMedium-RangeWeatherForecasts(UnitedKingdom)

Overthepastdecades,climatemodelshavestartedtomakeuseofresolutionsthatenablegeophysicalvortices with characteristics that increasingly resemble realistic tropical cyclones. For instance, anumberof studies fromtheUSCLIVARHurricaneWorkingGrouphavedemonstratedhowtropicalstormtracks,includingtheirgeographicaldistribution,variability,trends,arecrediblyrepresentedbyAGCMsemploying20kmmeshsize.ThenumberofTropicalCyclonesineachhemisphere(andevenfor key basins) starts to be credibly simulated at 20kmandbeyond.However, themajority of themodels analysed can still only simulate storms up to Category 3,while Category 4 and above aresignificantlyorentirelyunderestimated.Thepush for increasing resolutioncontinues,withAGCMsstartingtooperateunder10km,withtheaimofimprovingrealism,butatthesametimeithasbeensuggestedthattheuseofStochasticPhysicsmayactasasurrogateforhighresolution,byprovidingamoredynamicenvironmentforinstabilitiestogrow.ResultsfromtheanalysisofalargeensembleofAMIP-typesimulationswiththeECMWFIFS(projectSPHYNX)andtheMOHadleyCentreHadGEM3familyofmodels(PRIMAVERAproject)indicatethattheuseofStochasticPhysicsatmoderatetohighresolutionprovidesa30-50%increaseinthenumberofsimulatedTropicalCyclones,whichisroughlyequivalent to a doubling of resolution. An investigation of the mechanisms behind this modelbehaviour, using a hierarchy of sensitivity studies, points to the prominent importance of theStochasticKineticEnergyBackscatter schemeforTC formation,but this seemstobe largelymodeldependent,asmanyofthesettingswerepreviouslytunedformaximisingNWPperformanceateachcentre.Theseresultsjustifyapushforamoresystematicinvestigation,coordinatedviatheEU-Horizon2020PRIMAVERAproject.

Keywords:TropicalCyclones,HighResolution,StochasticPhysics


4-P28

Roleofoceanmesoscaleeddiesfortheresponseofclimatesystemtostronggreenhousegasforcing

VonStorchJin-Song(1),PutrasahanDian,A.,LohmannKatja,JungclausJohann,HaakHelmuth,GutjahrOliver

1-MaxPlanckInstituteforMeteorology(Germany)

Theefficiencyoftheoceaninuptakingheatandcarboncannotablyalterthetransientandequilibriumclimateresponse.Recentstudiesshowthattheseuptakescanbestronglyaffectedbyoceanicmeso-scaleeddies.Currentgenerationclimatemodelsgenerallydonotresolveoceanicmeso-scaleeddiesandhavetorelyoneddyparametrizations.Itisfoundthattheeffectofparameterizededdiesdiffersfromthatofresolvededdies.Thus,itispossiblethattheclimateresponseprojectedbyaclimatemodelwithaneddy-resolvingoceandiffersfromthatprojectedbyaclimatemodelwithanon-eddy-resolvingocean.Thusfar,thispossibilityhasnotbeensystematicallystudied.Here,wequantifywhetherandtowhatextentclimateresponseisaffectedbyoceanicmeso-scaleeddiesusingasetofabrupt4xCO2-runs carried out with the Max Planck Institute - Earth Systems Model (MPI-ESM) with T127atmosphericresolution(~1degree)andvaryingoceanresolutions(1,0.4and0.1degrees).Theanalysisoftheserunsaimsatidentifyingtheeffectofmeso-scaleeddies,notonlyonthefastandslowclimateresponses measured by air temperature, but also on the responses of oceanic circulations, heatpenetration into the deep ocean, air-sea interactions, and regional responses for regions directlyinfluencedbystronglyeddyingboundarycurrentssuchastheGulfStream.

Keywords:climateresponse,oceanmesoscaleeddies,oceanheatuptake


4-P29

EvaluationofextremeprecipitationandtemperaturesassimulatedbytheavailableCMIP6HighResMIPmodels

WehnerMichael(1),GlecklerPeter(2)

1-LawrenceBerkeleyNationalLaboratory(UnitedStates),2-LawrenceLivermoreNationalLaboratory(UnitedStates)

UsingrecentlydevelopedtoolsoftheCoordinatedModelEvaluationCapabilities(CMEC)program,weevaluatetheabilityofavailableHighResMIPmodelstosimulateextremesofpentadalprecipitationand 3 day temperaturemaxima andminima. Using a non-stationary extreme value approach, wecompare long period return values of these quantities as simulated by the models to griddedobservational products over North America, Europe and parts of Asia. Evaluation of these resultsagainstpreviousgenerationlowerresolutionmodelswillbepresentedaswellasadiscussionofthepossiblesourcesoferrorsaswellasimprovements,ifany.

Keywords:extremeprecipitation,extremetemperature,climatemodelperformance


4-P30

Improvedmeltpondsinclimatemodels

WyserKlaus(1),WangShiyu(1),KoenigkTorben(1)(2),DocquierDavid(3),LecomteOlivier(4),SterlinJean(3)

1-SwedishMeteorologicalandHydrologicalInstitute(Sweden),2-DepartmentofMeteorology[Stockholm](Sweden),3-UniversitécatholiquedeLouvain,EarthandLifeInstitute,Georges

Lema”treCentreforEarthandClimateResearch(Belgium),4-UniversitécatholiquedeLouvain,EarthandLifeInstitute,GeorgesLema”treCentreforEarthandClimateResearch(Belgium)

ThereductionofArcticseaiceisanalarmingsignoftheon-goingclimatechangewithconsequencesnot just for the Arctic but for the entire climate system through the albedo feedback. A realisticrepresentationoftheseaicealbedoisthereforekeytobetterunderstandpresentandfuturechangesintheArctic.Animportantfeatureofseaiceduringthesummerseasonistheoccurrenceofmeltpondsthatmaketheiceappeardarkerandincreasetheabsorptionofsolarradiation.Manyclimatemodelsaccountforthiseffecteitherimplicitlybyreducingtheicealbedoduringsummerorwithanexplicitmeltpond scheme. Within the framework of the PRIMAVERA project, UCL has developed a newtopographic meltpond scheme for the LIM3 sea ice model in which the meltpond evolution iscomputedfromtheicetopography.Theschemeassumesthatthesurfacemeltwatersarecollectedinthedepressionoftheicetopography,andthattheicethicknessdistributioncanbeusedtodeterminethereservoircapacityofthedepressions.Theamountofwater intheponds iscalculatedfromthedetailedbudgetofinputmeltwaterandlossesduetoiceprocesses.ThisnewmeltpondschemehasbeenintegratedintheEC-Earthmodelwithstandardandwithhighresolution.ThefirstpreliminaryanalysisshowareductionoftheArcticsea icevolumewiththenewmeltpondschemeindicatingalargersensitivitytoanincreasedclimateforcing.WealsofindabetterrepresentationoftheseasonalcycleoftheArcticseaicewheninteractivemeltpondsareused.WefurtherinvestigatetheresponseoftheArcticclimatetothechangesintheseaiceinducedbymeltponds,andassesstheimpactthereofontheEuropeanclimate.

Keywords:Arcticclimate,seaicemodel,PRIMAVERA


4-P31

UsingESMValTooltoAssesstheImpactofResolutionandForcingsonOceanandSeaIcePropertiesintheSouthernOcean

ZimmermannKlaus(1),FuentesFrancoRamon(2),KoenigkTorben,DoescherRalf,WyserKlaus(3)

1-SwedishMeteorologicalandHydrologicalInstitute(SMHI)(Sweden),2-SMHI(Sweden),3-SwedishMeteorologicalandHydrologicalInstitute(Sweden)

ESMValToolhasbeenestablishedasausefulstandardtoolformodelevaluation.Here,weusethenewversion2ofESMValTool,inparticularourextensionofitscapabilitieswithregriddingforirregulargridsascommonlyusedinoceanmodelsinthepreprocessorstageofthetool.

ThisnewabilityisemployedtoperformanalysisofoceanandseaicepropertiesinAntarcticaandtheSouthernOcean,trackingtheprogressandchangesofthenewestmodelgenerationusedinCMIP6ascomparedtoCMIP5.Furthermore,theinfluenceofchangesinmodelresolutiononthesecrucialpartsof the earth system is explored using the high resolution studies from PRIMAVERA project, thusimprovingtheunderstandingoftheroletheyplayinclimatemodels.

TheflexiblearchitectureofESMValTooltogetherwithitsextensivesupportforobservationaldatasetsandtheeasyintegrationofmodeldatathatcomplieswiththeCFconventions,meansthattheanalysiscaneasilybestreamlinedandappliedtonewobservationaldatasetsastheybecomeavailable,andreadilybeperformedacrossothermodeldatasets,e.g.fromthewidegamutofCMIP6experiments.

Inthisway,wegaininsightintothemodelresponsetodifferentforcings,herebycomparingthepre-industrialcontrolsimulationswiththehistoricalrunsfromCMIP6.

Keywords:southernocean,resolution,forcings,esmvaltool,antarctic,seaice

Session5-Variabilityandextremes

5-P01

Combingdecadalpredictionsandnear-termprojectionstoobtainreliableinformationfortheupcoming30-40years

BefortDanielJ.(1),O'reillyChristopherH.(1),WeisheimerAntje(1)

1-Atmospheric,OceanicandPlanetaryPhysics,DepartmentofPhysics,UniversityofOxford(UnitedKingdom)

Thiswork isembeddedintheEUHorizon2020projectcalledEUCP,aimingtodevelopaEuropeanpredictionsystemofthenext40years.Themaingoalofthisworkistoinformthecommunityabouttheskilland limitationsofcurrentdecadalpredictionsandtopresent first resultsshowingtowhatextentthesepredictionsarebeneficial forconstrainingnear-termprojections. Weexploretheskilland reliability of current decadal predictions from the CMIP5 archive with a strong emphasis onEurope.Severalmethodsaretestedtomergethepredictionswiththerespectivemodel'snear-termprojections.Thisincludesmergingdecadalpredictionsandnear-termprojectionsbyweightingmodelsaccordingtotheirabilitytosimulatepastclimates.Allresultsareverifiedintermsofprobabilisticskillscores, e.g. ranked probability score and reliability scores. First results applying simple calibrationmethodstotheCESMLargeEnsembledemonstratepotentialimprovementsofreliabilityandskillofnorthernEuropeantemperatureprojections.However,eventhoughreliabilityisimprovedusingthesemethods, the resulting projection remains overconfident. From themedium-range up to seasonaltime-scales, the use of stochastic physics to represent model uncertainties has been shown tosubstantially improve reliability of the prediction. However, stochastic physics parameterisation iscurrentlynotincludedindecadalpredictionsystems.Hereweanalysepotentialbenefitsofincludingstochasticphysicsbeyondseasonaltime-scalesbycomparingdecadalpredictionswithcurrentlong-termseasonalforecastsfortheupcoming13months.

Keywords:decadalprediction,neartermprojections,skill,reliability


5-P02

ExtremeSummerTemperaturesintheNorthernHemisphereandtheirLinktotheAtlanticMultidecadalVariabilityinDecadalHindcasts

BorchertLeonard(1),PohlmannHolger(1),Suarez-GutierrezLaura(1)(2),NeddermannNele-Charlotte(2)(3),MüllerWolfgang(1)

1-MaxPlanckInstituteforMeteorology(Germany),2-InternationalMaxPlanckResearchSchoolonEarthSystemModelling(Germany),3-InstituteforOceanography,UniversitätHamburg(Germany)

Itremainsanunsolvedchallengetopredictsummertemperatureextremesontheseasonal-to-decadaltimescale.Onepossiblewaytounderstandtheemergenceofsummertemperatureextremes,andthereforetheirprediction,istoidentifyconnectionsbetweenlarge-scalefeaturesofclimatevariabilityandsuchtemperatureextremes.Here,weshowevidencethatthephaseoftheAtlanticMultidecadalVariability (AMV) influences the occurrence and decadal prediction skill of summer temperatureextremesontheNorthernHemisphere.Weusea10-memberensembleofyearlyinitializeddecadalhindcastswiththeCMIP6versionoftheMPI-ESM-HRmodelcoveringtheperiod1960-2017.WeshowfordifferentregionsintheNorthernHemispheremid-tohigh-latitudes(CentralEurope,Scandinavia,North-East Asia, and the Central Midwest USA) that the likelyhood of summer (JJA) temperatureextremespredictedbythehindcastsincreasesfrom15%innegativeAMVphasesto22%inpositiveAMV phases. The amount of summer temperature extremes predicted by the model is highlycorrelatedbetweenthefourregions,becausetheAMVandtheoccurrenceofsuchextremesbetweentheseregionsarestronglyconnectedviaacircumglobalatmosphericRossbyWave,thecircumglobalwavetrain.Predictionsofsummertemperaturevariabilityareskillfulforupto8yearsaheadintheregions in Scandinavia, North-East Asia and the USA. Moreover, prediction skill for summertemperaturesintheseregionsstronglydependsonthephaseoftheAMV:summertemperaturesinScandinavia and theUSA show high predictive skill in positive AMV phases,whileNorth-East Asiashows high predictive skill in negative AMV phases. Conversely, due to a large number of othermechanisms influencing summer temperature extremes over Central Europe, we find no decadalpredictionskillforEuropeansummertemperatureextremesinthismodelsimulation.

Keywords:DecadalPrediction,SurfaceTemperatureExtremes,AMV,Wavetrain


5-P03

DisentanglingtheCO2seasonalcycleformitsterrestrial,oceanicandanthropogenicsources.

CadulePatricia(1),PeylinPhilippe(2)

1-IPSL(France),2-IPSL-LSCE(France)

TheterrestrialbiosphereandtheoceanplayakeyroleintheEarth'sclimate.Together,theyabsorbapproximately half of the anthropogenic carbon dioxide emissions released into the atmosphere.Understanding and simulating the carbon cycle is a crucial issue, which inherently leads tobenchmarkingmodelcarbonfluxesthroughtheuseofatmosphericCO2data,whichprovidesastrongconstraintonthefluxesatcontinentalscales.However,therearepotentiallimitations.Indeed,modelsmay reproduce the atmospheric CO2 seasonal cycle at many stations, but perform poorly whencomparedtocarbonfluxestimatesatcontinentalscale(atmosphericinversions).Wehavedevelopedamethodology,which combines these two approaches and builds on the concept ofMonths andRegionsofInfluence.Carbonfluxesofeachmonthofthe1980-2009periodandofeachofthe22so-calledTRANSCOMregionsweretransportedseparately.ThisresultedinafinegraindecompositionofthesimulatedCO2at31worldwidemonitoringstations.Thisstudyusestheoutputsof23CMIP5ESMsandtheresultsof3atmosphericCO2inversions,andwillbeextendedtoCMIP6models.Wefocusonlandcarbon fluxesandanalyse the contributionof themost influential regionsandmonths to theamplitude(andtrendthereof)oftheCO2seasonalcycle.Thisprovidesfurtherinsightsonthereasonsforwhichamodelcorrectlyreproduces,ornot,theatmosphericsignalandallowstocomparewithsimilarcontributionsobtainedfromatmosphericinversionfluxes.Ultimately,thisenablestopinpointtowardthe likelyphysicalandbiogeochemicalprocessestobe improved.AtBarrow,weshowthatmodels share roughly the same region-months of influence than the MPI_JENA inversion (80%similarity)andoverestimatetheseasonalamplitudeby3.4ppm(-6.2to27.5)withmost importantbiasesinborealregionsduringmidandlategrowingseasonperiod.

Keywords:Benchmarking,EarthSystemModels,carbonfluxes,terrestrialecosystems,climatechange


5-P04

TheBrewer-DobsoncirculationinCMIP6models

CalvoNatalia(1),AbalosMarta(1)

1-Dpto.FisicadelaTierrayAstrofisica,UniversidadComplutensedeMadrid(Spain)

The Brewer Dobson circulation (BDC) is themeanmeridional circulation in the stratosphere. It isresponsibleformasstransportinthestratosphereandthusitisimportantforthedistributionoftracegasesinthestratosphereanditsthermalstructureandhasimplicationsfortroposphere-stratosphereinteractions.Inthelastdecades,theBDChasreceivedlargeattentionasmostclimatemodelssimulateastrengtheningoftheBDCinresponsetoincreasinggreenhousegasemissionsinfuturescenarios.Thesestudieshavefocusedmainlyontheshallowbranchoftheresidualmeancirculationwhilethedeepbranchhasbeenlessexplored.Inaddition,thespecificmechanismsresponsiblefortheenhancedwavedrivingthatacceleratethestratosphericcirculationarestillnotwellunderstoodandpreviousmodelintercomparisonworksshowawidespread.TheCoupledModelIntercomparisonProjectphase6 provides a new opportunity to assess the robust features and quantify uncertainties in theclimatologyandtrendsoftheBDCinthelatestgenerationofcoupledatmosphere-oceanmodelswithawell resolved stratosphere. Our studywill focus on both the shallow and deep branches of theresidualcirculation,andinvestigatethewaveforcingoftheresidualcirculationanditsfuturechangesandthedynamicalmechanisms.Inadditiontotheresidualcirculationdiagnostics,themeanageofairwillbeexaminedtoinformonthenetmasstransport,includingtwo-waymixing.

Keywords:meanmeridionalcirculation,Dobsoncirculation,Brewer,wavedriving


5-P05

ProcesseslinkingtheintensityoftheAtlanticMultidecadalVariabilitytotheclimateimpactsoverEuropeasassessedfrom

CMIP6/DCPP-Cpacemakerexperiments

CassouChristophe(1),QasmiSaid(1),BoéJulien(1)

1-CNRS-Cerfacs(France)

Targeted sensitivity experiments have been proposed within CMIP6/DCPP to isolate the climateresponse to the AtlanticMultidecadal Variability (AMV) through amodeling framework based onpacemaker-type simulations. Those consist in restoring the modeled North Atlantic Sea SurfaceTemperature(SST)toananomalouspattern(providedthroughinput4MIPsforcoordination)that isrepresentativeof theobservedAMV,whiletherestof thesystemremains fullycoupled.FollowingDCPP-C guidelines, so-called ensembleAMV-experiments (hereafter 1xAMV) have been conductedwiththeCNRM-CM5model.AdditionalensembleshavebeenproducedbymultiplyingtheanomalousAMV-relatedanomaliesby2and3(hereafter2xAMVand3xAMV)toevaluatethesensitivityofthemodel response to the intensity of the AMV and to help isolate the mechanisms at work. Weconcentrateonthewinterseasonandshowthatresponsesin1xAMVareoverallweakintemperatureandprecipitation,whereasmorerobustandsignificantsignalsemergein2xAMVand3xAMV.PositiveAMV then leads to large-scale continental warming (maximum loading along the Atlantic flank ofEurope) and enhanced rainfall (except over Scandinavia). To deepenour understanding of the fullAMV-forced response,wehave adapted a so-called flowanalogsmethod to isolate the respectivecontribution of the dynamical versus thermodynamical processes.We show that the AMV-forcedatmospheric anomalous circulation tends to cool down the European continent, whereas thethermodynamical response leads to warming through the advection of warmer and more humidoceanicairpenetratinginlandandthroughthemodificationofthesurfaceradiativefluxeslinkedtoaltered cloudiness and snow-cover reduction. In 1xAMV, dynamical and thermodynamicalcontributionscanceleachover,whereasthelatterclearlyovercomesthedynamicalcoolingin2xAMVand3xAMVandexplainstheoverallwarming.Ourstudyclearlyemphasizesthenon-linearityoftheAMV-forcedresponsewithrespecttotheintensityoftheAMV-forcing,andinparticularitsdynamicalcomponent.

Keywords:decadalvariability,AMV,DCPP


5-P06

TopPrecipitationExtremesfromEventProspectives:Observation,Simulation,andAttribution

ChenCheng-Ta(1),LoShih-How

1-NationalTaiwanNormalUniversity,DepartmentofEarthSciences(Taiwan)

Eventsofextremeprecipitationhaveahugeinfluenceonsociety.Extremerainfall leadstofloodingandlandslideincludingrisktohumanlife,damagetobuildingsandinfrastructure,andlossofcropsand livestock. From past IPCC reports, it is commonly expected that frequency and intensityprecipitationextremeswillbothincreaseastheclimatewarms.Althoughthereareplentyofstudiesworkingonthetopic,theuniquenessofourworkisonbothusinghigh-resolutionglobalandregionalclimatemodelsandthesearchofeventsbyspatialandtemporalconnectionofextremeheavyrainfalloccurrences.High resolutionmodels in general better capture the the intensity andoccurrenceofheavyrainfall.Theadvantageofeventidentificationisthepossibilitytolinkthevolumetotherankingandpotentialdamagesoftheextremeeventsinadditionaltousualassessmentofmeanchangesinfrequencyandintensity.Withthegrowinghistoricalrecordsandobservationalanalysis,onecanmakecomparison and objectively rank the historical events according to their severeness. The sameapproachcanbeappliedtoCMIPclimatemodelsimulationtoobtainspatialandtemporalevolutionof extreme precipitation events. The event associated environmental thermal and dynamicalconditionsindifferentclimateregionscanbeconstructedtofurtherdelineatetheunderlyingphysicalprocesses andhow they change through time in thepast and into the future.Additional effort toquantifytheattributablerisksofextremeeventsduetohumanimpactorotherindividualforcingcanbestudiedusingdesignednumericalexperimentswithlargeensemblemembersinCMIP.

Keywords:extremerainfall


5-P07

Decadalvariabilityinweatherregimesandteleconnectionsinreanalysisdatasetsandclimatesimulations.

CortiSusanna(1),FabianoFederico(1),MecciaVirna(1)

1-IstitutodiScienzedell’atmosferaedelClima[Bologna](Italy)

Atmospheric intraseasonal variability, especially in the extratropics during the cold season, ischaracterized by preferred large-scale quasi-stationary flow patterns that have non-Gaussiancharacteristics. They are known as weather regimes and can be associated with significanttemperatureandprecipitationanomalies.Severalobservation-basedandmodelstudieshaveshownthatElNino-SouthernOscillation(ENSO)andMaddenandJulianOscillation(MJO)forcingaffectstherelativefrequencyofoccurrenceofcirculationregimes.TheENSOforcingisparticularlyimportantforthePacificNorth-American(PNA)region,whoseintraseasonalvariabilityisdirectlylinkedtotropicalPacificSSTanomalies.TheMJOforcingaffectstheoccurrence(andsub-seasonalpredictability)oftheNorthAtlanticOscillation(NAO)and,toalesserextent,theEuro-AtlanticBlocking.Ontheotherhand,severalrecentstudieshaveshownthatwintertimelargescaleextratropicalcirculationoverboththePacificNorthAmericanandtheEuro-Atlanticsectorsexhibitsanon-negligibleinterdecadalvariability,whichseemstoberelatedtoadecadalmodulationofthetropical-extratropicalteleconnections. Inthisstudyweusecentury reanalysisdatasetsanda rangeofclimatesimulations to investigate theinterdecadalvariabilityofextratropicalweatherregimesandtheirassociatedtropicalandextratropicalteleconnections.Theimplicationsoftheresultsofthisstudyforassessingfutureclimatechangearediscussed

Keywords:NaturalClimatevariability,Decadalvariability,Tropical,extratropicalteleconnections,Weatherregimes


5-P08

Forcedresponse,warmingpausesandsurgeeventsintemperatureandheavyprecipitationextremes

FischerErich(1)

1-ETHZurich(Switzerland)

Regionaltrendsinextremescanbesubstantiallyamplifiedormaskedbyinternalvariability.Thus,itisparticularly challenging to isolate the forced response in extremes both from observations andindividualmodel simulations. Reliable estimates of the forced response of extremes are howeverparticularly important for detection and attribution as well as for uncertainty quantification.Ultimately,itisforcedresponsethatdetermineslong-termchangesinreturnlevelsofextremes.BasedonthefewCMIP5modelsthatprovideddailyoutputforseveral initialconditionmembers,wehaddemonstrated that thedisagreementbetween individual simulationsprimarily arises from internalvariability. CMIP6 provides multiple initial condition members for substantially more models andallowstobetterseparateunforcedinternalvariabilityfromtheforcedresponseinextremes.Iwillbeusingthis foramorecomprehensivequantificationofuncertaintyandrobustnessofprojectionsofextremes.Duetointernalvariability,regionaltrendsinextremesmaydeviatestronglyfromtheforcedresponse in individual simulations and in the real world. Thismay lead tomulti-decadal warmingpausesorstagnationperiodsinextremes,similartothehiatusorwarmingholesofhotextremes.Suchstagnationperiods,duringwhichtheforcedresponse ispotentiallyhiddenby largevariability,maylead to a serious underestimation of the changes in the probability of extremes. Likewise, such astagnationperiodsareoftenfollowedbyasurgeeventduringwhichextremesbreakpreviousrecordsbyalargemargin.CMIP6incombinationwithexistinglargeinitialconditionlargeensembleswillbeusedtoquantifythetimescalesandprobabilityofsuchpausesandsurgeperiodsinextremestooccur.The findingsof this studydirectlyaddress caveats identified in the frameworkof theWCRPGrandChallengesofExtremesandintheearlydraftingofIPCCAR6chapters.

Keywords:Extremes,heatwaves,uncertainties,internalvariability,forcedresponse


5-P09

Naturaldecadalsea-levelvariabilityintheIndianOcean:LessonsfromCMIPmodels

GangadharanNidheesh(1),LengaigneMatthieu(2),VialardJerome(2),IzumoTakeshi(3),UnnikrishnanAS(4),KrishnanR(1)

1-IndianInstituteofTropicalMeteorology(India),2-Laboratoired’océanographieetduClimat:ExpérimentationsetApprochesNumériques(France),3-Laboratoired’océanographieetduClimat:ExpérimentationsetApprochesNumériques(France),4-NationalInstituteofOceanography(India)

Understandingnaturalsea-levelvariabilityiscrucialtoidentifyanysecularchangeinsealevel,sincethe observed sea-level trends are a combined response to both anthropogenic as well as naturalclimatechanges.However,oneofthemaincaveats,particularlyfortheIO,isthelackoflongsea-levelobservation and inconsistencies in the representation of decadal sea-level variability amongobservation-basedsea-levelproducts.Thoseobservationalissuesareastrongincentivetoinvestigatethe leading modes of IO decadal sea-level variability in climate models from the Coupled ModelIntercomparison Project (CMIP). Two consistent modes of IO decadal sea-level variability, whichcollectively explain about 50% of the total decadal variance, are identified in 26 CMIP controlsimulations.Withopposite sea-level signals in the southwesternandeastern tropical IO,mode1 islargelyrelatedtothedecadalmodulationoftheIndianOceanDipole(IOD).Co-variabilityofdecadalIODwith the El Ni–o-Southern Oscillation (ENSO), yields sea-level signals along the west coast ofAustralia(WAC),transmittedfromthewesternPacificviaIndonesianThroughflow.Mode2consistsofabroadsea-levelpatterneastofMadagascar,mainlydrivenbythewind-stresscurlassociatedwiththeMascareneHighdecadalmodulations.Inaboutone-thirdofthemodels,mode2islag-correlatedtomode1 and in suchmodels, sea-level variations along theWAC propagatewestward as Rossbywaves and contribute to sea-level mode2. These sea-level modes identified in CMIP are broadlyconsistent with those derived from the relatively short altimeter dataset. In a poor observationalsamplingcontext,thissuggeststhattheCMIPdatabasecanprovideguidanceforidentifyingrobustmodesofIOdecadalsea-levelvariability.

Keywords:ElNi–oSouthernOscillation(ENSO),IOD,IndianOcean,sealevel,


5-P10

AMOChysteresisinapre-CMIP6GCMandaproposalforcomparingAMOCfeedbacks.

JacksonLaura(1),WoodRichard(2)

1-MetOfficeHadleyCentre(UnitedKingdom),2-MetOffice(UnitedKingdom)

Therehaslongbeenevidencefrompaleoobservations,theoryandsimplifiedmodelsthattheAtlanticMeridionalOverturningCirculation(AMOC)hasthepotentialtocollapseandforhysteresis(whereitdoesnotrecoverafterforcingisremoved).AlthoughhysteresishasbeenseeninafewlowresolutionandfluxadjustedGCMs,ithasproveddifficulttodemonstrateincomplex,coupledGCMsleadingtosuggestionsthatthepresenceofotherfeedbacksincomplexGCMsmayhinderhysteresis.

Wepresentasuiteof idealisedexperimentsusingHadGEM3-GC2 (which isaprecursor to theMetOffice'sCMIP6eddy-permittingGCM) to test for thepresenceof thresholds in theAMOCusinganensemblewithdifferentratesandlengthsoffreshwaterforcing.WeshowthattheAMOCdoesnotalwaysrecoverafterfreshwaterforcingisstopped,andcanremaininaweakstateforatleastacoupleofhundredyears.TherecoveryornotdependsonhowmuchtheAMOCisweakenedbytheforcingandweshowthatthisoccursthroughtheinterplayofadvectivefeedbacks.

Theseresults,however,areforoneGCMandwewishtounderstandhowthesefeedbacksvaryacrossGCMsandwhetherwecanlearnmoreaboutthefeedbacks,andhencepotentialforbistability,intherealworld.Hencewesetoutaproposalforamulti-modelstudytoexaminethefreshwaterfeedbacks.AlthoughwephysicallyunderstandmanyofthefeedbacksontheAMOC,bothadvectiveandthroughsurfacefluxes,wedonothaveagoodunderstandingoftherelativestrengthsandwhichdominate.CMIP6 provides an opportunity to study these feedbacks in state-of-the-art GCMs and to takeadvantageofimprovementsinresolutionandrepresentationsofprocesses.

Keywords:AMOC,threshold,tippingpoint


5-P11

Globalmonsoonresponsetoseasurfacetemperatureduringthe20thcentury:EvidencesfromAGCMsimulations

JiangJie(1)

1-LASG,InstituteofAtmosphericPhysics,ChineseAcademyofSciences(China)

Multi-decadal variations of global land monsoon are observed during the 20th century, with anincreasingtrendinthefirsthalfofthe20thcentury,followedbyadecreasingtrendupto1980sandshowedanupwardtrendlater.ByexaminingtheoutputsoftwoAGCMs(viz.GISS-E2-RandHadAM3)forcedbyhistoricalseasurfacetemperature(SST),werevealthemechanismsoftheglobalmonsoonprecipitationchangesduring1901-2000andsuggestthatthesignificantchangescanbeattributetoatmosphere's response to SST anomalies. The climatological characteristics and the interdecadalvariationsofobservedgloballandmonsoonprecipitationarereasonablyreproducedbytwoAGCMs.ThechangesofgloballandprecipitationaremainlycontributedbytheNorthernHemispheremonsoondomain accompanied with the variations of NHSM circulation and Walker circulation. At theinterdecadaltimescale,thegloballandmonsoonprecipitationanomaliesandNHSMcirculationarehighlycorrelatedwithSSTAoverNorthAtlanticandtropicalPacific.BothNorthAtlanticandLaNi–apattern-likePacificcanleadmorewatervaportransporttomonsoonregionsbystrengtheningPacificsubtropicalhighsandtradewinds, thenfurther leadtothe increaseofmonsoonprecipitation.TheoutputsofGMMIPexperimentswillbeusedforourfurtherresearch.Theresultsoftier-1experimentswillbeusedtorepeataboveworktoconfirmtheresultsarerobustandnotrelyonindividualmodel.Theresultsoftier-2experiments,whichinclude"hist-resIPO"and"hist-resAMO"runs,willbeusedtorevealthecontributionofSSTanomaliesoverthetropicalPacificandtheNorthAtlantictomonsoonprecipitation changes. We aim to answer: How does atmosphere-ocean interaction affect theinterdecadal variations of globalmonsoon? Canwe predictmonsoon precipitation changes in thecomingdecadesgivenSSTvariability?

Keywords:Globalmonsoon,Seasurfacetemperature,Climatevariability,Monsooncirculation


5-P12

SealevelvariabilityinmarginalseasfromCMIPsimulations.Strengths,weaknessesandwaystosolvethem.

JordaGabriel(1)

1-InstitutoEspanoldeOceanografia(IEO),CentreOceanograficdeBalears(Spain)

Sealevelisakeyvariablefortheclimatesystemthatneedstobemodeledasaccuratelyaspossible.Itisanimportantindicatorofclimatechangeasitintegrateschangesintheoceanheatcontentandlandicemelting,amongothers.Incoastalregions,changesinthemeansealevelandintheextremeeventscan have a critical impact on infrastructures and ecosystems. An adequatemanagement of thosepotential impacts require robust projections from climate models. Up to now, may studies haveanalysed the projections of regional sea level based on CMIP simulations. This has proven to beadequateformostoceanicregionsbutinsemienclosedmarginalseas(e.g.Mediterranean,Caribbean,RedSea),theperformanceofglobalclimatemodelsisusuallypoorduetoitscoarseresolutionandlimitedskillsreproducinglocaldynamics.Understandingandquantifyingthelimitsofapplicabilityofthosemodelsintermsofsealevelvariabilityinhighlypopulatedmarginalseasisthusofparamountimportance.ThisevenmoretrueinatimeinwhichmoreandmorestakeholdersarelookingintoCMIPsimulationstodefineadaptationplansfortheforthcomingsealevelrise.InthispresentationwewillanalysethestrengthsandweaknessesofCMIPmodeloutputsintermsofsealevelvariabilityintheMediterranean Sea as a paradigm of vulnerable highly populatedmarginal seas.Wewill focus ondifferentspatialscales(frombasintocoastal)comparingtheoutputsamongmodelsaswellaswithobservations(fromaltimetryandcoastaltidegauges).Finally,wewillpresentsomeexplanationsforthemodelweaknessesbasedontheirabilitytoreproducesomekeyprocessesintheregion.

Keywords:SeaLevel,MediterraneanSea,MarginalSeas,Modelskills.


5-P13

QuantifyingtheAgreementBetweenObservedandSimulatedExtratropicalModesofInterannualVariability

LeeJiwoo(1),SperberKenneth(1),GlecklerPeter(1),BonfilsCéline(1),TaylorKarl(1)


UsingHistoricalsimulationsofCMIP5andCMIP6,modelsandmultipleobservationally-baseddatasets,weemployskillmetricstoanalyzethefidelityofthesimulatedNorthernAnnularMode(NAM),theNorthAtlanticOscillation(NAO),thePacificNorthAmericapattern(PNA),theSouthernAnnularMode(SAM),thePacificDecadalOscillation(PDO),theNorthPacificOscillation(NPO),andtheNorthPacificGyreOscillation (NPGO).Weassess thebenefitsofaunifiedapproach toevaluate thesemodesofvariability, which we call the common basis function (CBF) approach, based on projecting modelanomalies onto the observed empirical orthogonal function (EOF). The CBF approach circumventsissueswithconventionalEOFanalysis,includingtheneedtocorrectforarbitrarysignsofEOF's,andtheneedtotestifhigher-ordermodelmodesbettercomparewiththeobservedmodes.ComparedtoconventionalEOFanalysisofmodels,theCBFapproachindicatesthatmodelscomparesignificantlybetterwithobservations intermsofpatterncorrelationandroot-mean-squared-error (RMSE)thanheretoforesuggested.TheskillmetricsproposedinthisstudyprovideausefulsummaryoftheabilityofmodelstoreproducetheobservedEOFpatternsandamplitudes.Additionally,theskillmetricscanbeusedasatooltoobjectivelyhighlightwherepotentialmodelimprovementsmightbemade.Weadvocatemoresystematicandobjectivetestingofsimulatedextratropicalvariability,especiallyduringthenon-dominantseasonsofeachmode,whenmanymodelsareperformingrelativelypoorly.Time-permitting, analysis of power spectra of the principal component time series and diagnosis of themodesofvariabilitywillbediscussed.

Keywords:Commonbasisfunction,Metrics,Modesofvariability,EOF,CMIP5modelevaluation


5-P14

InvestigatingtheENSOteleconnectionresponsetoglobalwarmingusingamulti-modellarge-ensembleexperiment

LiCamille(1)(2),MichelClio,BethkeIngo,KingMartin,SimpsonIsla,SobolowskiStefan

1-UniversityofBergen(Norway),2-BjerknesCentreforClimateResearch(Norway)

The El Ni–o-Southern Oscillation dominates interannual variability of the climate system, withwidespread global teleconnections that influence regional precipitation and temperature. ENSOteleconnection patterns are generally projected to shift eastward over the Pacific-North Americasectorunderglobalwarming(mediumconfidenceinIPCCAR5report;approximately25%ofmodelsdisagree), due to either an eastward shift of tropical convection associated with the warm poolexpansion or changes in the midlatitude circulation. Better constraining future changes in ENSOteleconnectionsremainsachallengebecauseofmodeluncertaintyinhow(a)tropicalconvectionand(b)themidlatitudecirculationwillrespondtoglobalwarming,aswellas(c)largeinternalvariabilityoftheteleconnectionpattern,whichmakesitchallengingtoisolatetheforcedsignal.Here,wetackle(b)and(c)usinglarge-ensemblesimulations(atleast100membersperexperiment)from5CMIPmodels.Prescribed-SSTexperimentsaredesigned to testhow teleconnections fromthe sameENSOeventsevolve under different background climates (present and future). The present day teleconnectionpatternisrelativelywellsimulated,withthemodelsplacingtheNorthPacificcentre-of-actionwithin400kmofitspositioninERA-Interim.Aslightnortheastwardshift(approximately250km)oftheNorthPacificcentreisfoundin3outof5models,despitelargeinternalvariability(ensemblespread)inthepositionofthiscentre.Inthesemodels,tropicalconvectionalsoshiftseastward,despitenochangeintheENSOSSTforcing.SomeasymmetryexistsbetweenthetwoENSOphases,withLaNi–aexhibitingsmaller spreadand lessnortheastwardshift thanElNi–o in4outof5models.Associatedshifts intemperatureimpactsareweak,whileprecipitationimpactsinregionsalongtheNorthAmericanwestcoastareaccentuatedandexpandinland.

Keywords:ENSOteleconnections,internalvariability,globalwarming,modelintercomparison,largeensemblesimulation


5-P15

Towardspredictingthevariableoceancarbonsink

LiHongmei(1),IlyinaTatiana(1),MüllerWolfgangA.(1)(2),LandschützerPeter(1)

1-MaxPlanckInstituteforMeteorology(Germany),2-DeutscherWetterdienst(Germany)

Theoceancomprisesoneoftheglobe'smostimportantsinksforanthropogeniccarbonbutthissinkhasbeenshowntovarysubstantiallyoninterannualtodecadaltime-scales.Ourinvestigations,usinga grand ensemble of 100-member simulations based on the Max Planck Institute Earth Systemmodel(MPI-ESM),suggestthatthevariationsoftheoceancarbonsinkcanbeattributedtothelargeinternalclimatevariability.ByassimilatingobservationsofatmosphereandoceanphysicalfieldsintoMPI-ESMbaseddecadalpredictionsystem,wecanfurtherproducevariationsoftheoceancarbonsinkconsistentwithobservation-basedestimates.Ourresultsshowthatthevariationsoftheoceancarbonsinkarepredictableupto2yearglobally,whilethepredictiveskillishigherupto5yearsregionally.Wefurtherseparatethepredictabilityintothermal(i.e.,affectedbyCO2solubility)andnon-thermal(i.e.,affectedbyoceancirculationandbiology)components.Wefindthattheshorter-term(<3years)predictiveskilloftheoceancarbonsink ismaintainedbythethermaldrivers,however,the longer-term (>3years)predictive skill isdeterminedby thenon-thermaldrivers.To furtherassessing thepredictabilityoftheoceancarbonsinkinthecontextofglobalcarboncycleinteractively,wemoveonestep further to run emission-driven simulations of decadal predictions. The emission-drivenassimilation captures the variations of the global carbon cycle in comparable to the global carbonbudget. The decadal predictions based on ESMs are an important tool for monitoring the futureevolutionoftheglobalcarboncycleandtodetermineifweareontherighttrackmeetinglongertermclimatetargets.AnextensionofourpredictionsinthecontextofCMIP6DCPPmulti-modelframeworkisplannedandthiswillfurtherfacilitateimprovementofpredictiveskillintheglobalcarboncycle.

Keywords:predictability,oceancarboncycle,internalvariability,globalcarboncycle


5-P16

TropicalNorthAtlanticasanon-stationarymodulatorofENSO-Europeanrainfallteleconnection

Lopez-ParagesJorge(1),Martin-ReyMarta(2),TerrayLaurent(2),Rodriguez-FonsecaBelen(3),PoloIrene(3),LosadaTeresa(3)

1-UMR5318CECI/CERFACS(France),2-UMR5318CECI/CERFACS(France),3-UniversidadComplutensedeMadrid,DptodeFisicadelaTierrayAstrofisica(Spain)

ElNi–o-SouthernOscillation(ENSO)impactontheNorthAtlanticEuropeansector(NAE)stillrisesmanyunansweredquestions.Nowadays,thereisagrowingevidenceadvocatingforanon-stationaryfeatureaffectingboth,thetroposphericandthestratosphericpathways,oftheENSO-NAEteleconnection.Inparticular, a changing link between ENSO and the spring Euro-Mediterranean rainfall has beendocumented in response to different phases of the Atlantic Multidecadal Oscillation (AMO).Nevertheless,theunderlyingphysicalexplanationsarefartobecompletelyunderstood.Inthisstudyapurely troposphericmechanism, inwhich theTropicalNorthAtlantic (TNA)playsamajor role, ispresented.OurresultsrelyonthedistinctcapacityofENSOtogenerateazonalSSTgradientovertheTNAunderdifferentAMOphases.Consequently,aatmosphericresponse,alsorelatedtoENSO,canbetriggeredfromtheTNAtotheEuropeansector.TheoccurrenceofthisENSO-NAEteleconnectionvia TNA could be subject to inter-decadal changes of ENSO properties and Atlantic backgroundconditions. In this study this issue is analyzed in 1) observations, and 2) low-resolution and high-resolution atmospheric and coupled GCM simulations. It is expected that the novel proposedmechanismgivesastepforwardintheunderstandingoftheroleoftheAtlanticbasinasmodulatoroftheENSO-NAEteleconnection.

Keywords:ENSO,Europeanrainfall,climateteleconnection,AGCM,CGCM


5-P17

VariabilityinthenorthernNorthAtlanticandArcticoceansinthepastmillennium:AreviewofCMIP5/PMIP3efforts

Moreno-ChamarroEduardo(1),OrtegaPablo(1),SwingedouwDidier(2),AmrheinDan(3),JungclausJohann(4),YeagerStephen(5)

1-BarcelonaSupercomputingCenter(Spain),2-CNRS(France),3-UniversityofWashington(Seattle)(UnitedStates),4-MaxPlanckInstituteforMeteorology(Germany),5-NationalCenterfor

AtmosphericResearch(UnitedStates)

Thepastmillenniumclimateshowedlargedecadaltomulti-centennialvariations,liketheLittleIceAge(LIA)coldperiod.Causesbehindsuchvariationsarestilldebated,althoughNorthAtlanticcirculationchanges have often been proposed to be key. Combined proxy model analyses complement andstrengthen the interpretation of past ocean changes and processes, including the role of externalforcings and internal processes in climate variability. We here discuss current knowledge of pastmillenniumNorthAtlanticvariabilityinclimatesimulationswiththeIPSL-CM5A-LR,CESM,andMPI-ESM models performed within the context of the CMIP5/PMIP3 initiatives. No evident modelconsensusisfoundinvariationsintheAMOC,subpolargyre,andDenmarkStraitoverflowstrengths,three key elements of theNorthAtlantic large-scale dynamics. These showdifferent responses toexternal forcingsacrossmodels,evenafteraccounting for samplingeffectsdue todifferentmodelensemble sizes. Differences in the external forcings amongmodel simulations may partly explainmodel differences, highlighting the need for increased standardization of model forcings andexperimental protocols. Furthermore, the potentially large role of internal variability calls forensembleapproachestoisolateexternallyforcedsignalsproperly.Lag-zerolinearregressionsbetweenSSTsandlarge-scaleoceancirculationshowvaryingdegreesofagreementamongmodelsandseemtobe period dependent. This questions the robustness and stationarity of SST fingerprints of oceancirculation. Disagreement also arises in Medieval-Climate-Anomaly-vs-LIA and preindustrial-vs-industrial SST changes between models and a state-of-the-art compilation of high-resolutionpaleoceanographic records. The PMIP4/CMIP6's past1000 experiment proposes for the first timecommonforcingsandrelativelylargerensemblesizes.Yetlongtransientsimulationsarestilllimitedbymodelresolution,whichhamperstherepresentationofkeyoceanicprocessesintheNorthAtlantic.FurtheradvancesinourcurrentunderstandingofNorthAtlanticlastmillenniumvariabilitywillrelyonthesuccessofsuchnewcoordinatedexperiments.

Keywords:Pastmillennium,NorthAtlanticandArcticclimatevariability,CMIP5/PMIP3climatemodels


5-P18

Amulti-modelcomparisonoftheoceancontributionstomultidecadalvariabilityintheNorthAtlantic

OrtegaPablo(1),RobsonJon(2),SuttonRowan(2),MenaryMatthew(3),GermeAgathe(4),BlakerAdam(4),SinhaBablu(4),HirschiJoel(4),HermansonLeon(5),YeagerStephen(6)

1-BarcelonaSupercomputingCenter-CentroNacionaldeSupercomputacion(Spain),2-NationalCentreforAtmosphericScience(Reading)(UnitedKingdom),3-Laboratoired'OcéanographieetduClimat:ExpérimentationsetApprochesNumériques(France),4-NationalOceanographicCentre[Southampton](UnitedKingdom),5-UnitedKingdomMetOffice[Exeter](UnitedKingdom),


In this study,weanalyse the inter-relationshipsbetweentheLabradorSeadensities, theboundarycurrents, the Atlantic Meridional Overturning Circulation (AMOC) and, more generally, the widerclimateoftheNorthAtlanticacrossanensembleoflongCMIP5preindustrialsimulations.Thestudyalso relies on the analysis of two 300-year long high-resolution coupled control simulations (withHadGEM3-GC2andHiGEM, respectively). Theultimategoal is to shed lighton the reasons for themodel spread, and to explore the potential use of different climatological properties as emergingconstraints to identify which models are more compatible with the current (and limited) AMOCobservations,andthusaremorelikelytorepresentrealisticallydecadalvariabilityintheNorthAtlantic.OurresultssuggestthatallmodelsconsistentlyshowastronglinkbetweentheLabradorSeaDensitiesandtheAMOCatsubpolarlatitudes(45N),buthavelittlecoherenceregardingtheirrelationshipwiththesubtropicalAMOC(26¡N,i.e.thelatitudeoftheRAPIDarray).ThislinkwiththesubtropicstendstobehigherinmodelswithastrongeranddeeperAMOCcell,andalsoinmodelswithmoreweaklystratifiedLabradorSeaDensities.Inbothcases,modelsclosesttotheobservationalvaluessupportaweaklinkofLabradorSeaDensitieswiththesubtropics.Interestingly,regardlessofthepreviousinter-model differences, allmodels show a coherent delayed link of Labrador SeaDensitieswith oceantemperatures in theEasternSubpolarGyre,supportingrecentworks that relate therecentrecord-breakingcoldanomaliesintheEasternNorthAtlantictoaslowdownintheAMOC.

Keywords:AtlanticMeridionalOverturningCirculation,DecadalVariability,EmergingConstraints


5-P19

ENSOandPDOmodulationofsuddenstratosphericwarmings:amulti-modelstudy

PalmeiroFroilaM.(1),Garcia-SerranoJavier(2)(1)

1-BSC(Spain),2-METEO-UB(Spain)

InthisworkweevaluatehowElNi–o-SouthernOscillation(ENSO)andthePacificDecadalimpactthestratospheric polar vortex variability, and particularly the occurrence of sudden stratosphericwarmings(SSWs).ThetroposphericvariabilityintheNorthAtlantic-EuropeansectorisassociatedwiththeNorthAtlanticOscillation(NAO),whichdominatesEuropeanclimateoninterannualtime-scalesandisstronglyinfluencedbySSWs,asitwasshownduringlate-winter/springof2018.ENSOisthoughtto have an impact on the NAO strength and duration via a stratospheric pathway, however thisrelationshipmaynotbelinearanddependonotherfactors.ThedirecteffectofthePDOontheNAOand/orthepolarstratospherehasbeenmuchlessanalyzedandremainsunclear.ENSOandthePDOcanmodulatetheoccurrenceofSSWssincebothareassociatedwithanomalouswavepropagationintheNorthPacificthatcaninterferewiththeclimatologicalwavepatternovertheAleutianLowregionandaffect thewave injection into the stratosphere.Toachieveourgoal,weusea setof idealizedexperiments that allows to investigate separately the ENSO and PDO teleconnections to the polarvortex;acontrolexperimentwithclimatologicalSSTs,asensitivityexperimentprescribingacanonicalENSOeventontopoftheseasonalcycle,andtwosensitivityexperimentsprescribingrespectivelythepositiveandnegativephaseofthePDOwithanannualpattern.Thesesimulationswereperformedusing three different general circulation models with a well-solved stratosphere: EC-EARTH/IFS,CNRM/ARPEGEandCMCC/CAM,whichcontributetotheERA4CS-fundedMEDSCOPEproject.

Keywords:teleconnections,ENSO,PDO,SSW,stratosphere


5-P20

ENSOevaluationinCMIPmodels

PlantonYann(1),GuilyardiEric(1)(2),LeeJiwoo(3),GlecklerPeter(3),WittenbergAndrew(4),PowerScott(5),McgregorShayne(6)

1-Laboratoired’OcéanographieetduClimat:ExpérimentationsetApprochesNumériques(France),2-NationalCentreforAtmosphericScience[Leeds](UnitedKingdom),3-LawrenceLivermoreNationalLaboratory(UnitedStates),4-NOAAGeophysicalFluidDynamicsLaboratory(United

States),5-AustralianBureauofMeteorology(Australia),6-Monash(Australia)

Since theestablishmentof thebasicphysicalmechanisms30yearsago,majorprogress inElNi–o-SouthernOscillation(ENSO)researchhasbeenmade.ThisresultedinabetterunderstandingofENSOmechanisms,dynamicsandforecastingandmuchimprovedsimulationsofENSOstatisticsincoupledocean atmosphere general circulation models (CGCMs). Beyond the traditional scientist-drivenexploratory model evaluation, and with the advent of climate services, there is now a need todocumentENSOpropertiesforagrowingnumberofstakeholdersoutsidethemodellingcommunity.This requires to co-construct the related science questions between ENSO experts and modelinformationendusersismanydisciplines.WeherereportontheprogressmadebytheCLIVARENSOResearch Focus group to propose consensus metric collections on specific ENSO related sciencequestions:ENSOperformanceandteleconnectionsinhistoricalsimulationsandENSOprocesses.Firstapplication toCMIP6vs.CMIP5 simulations ispresented.Thepilot technical implementationofanENSOmetricspackageindependentofthesoftwareinfrastructureusedalsopavesthewayforfuturestandardsinmodelevaluation.

Keywords:ENSOmetrics,CMIP5vsCMIP6,ENSOperformance,ENSOteleconnections


5-P21

AtlanticMultidecadaVariabilityandNorthAtlanticStormTrack

RuggieriPaolo(1),BellucciAlessio(1),DarioNicoli(1),AthanasiadisPanos(1),DaviniPaolo(2),GastineauGuillaume(3),GriegerJens(4),HarveyBen(5),HodsonDan(5),O'reillyChristopher(6),Rodriguez-FonsecaBelen(7),Ruprich-RobertYohan(8),SanchezGomezEmilia(9),SmithDoug(10),

SuttonRowan(5),WildSimon(8)

1-CentroEuro-MediterraneoperiCambiamentiClimatici[Bologna](Italy),2-CNRInstituteofAtmosphericSciencesandClimate(Italy),3-Laboratoired’OcéanographieetduClimat:

ExpérimentationsetApprochesNumériques(France),4-InstitutfürMeteorologieWissenschaftlicherMitarbeiter(Germany),5-NationalCentreforAtmosphericScience(Reading)(UnitedKingdom),6-UniversityofOxford[Oxford](UnitedKingdom),7-UniversidadComplutensedeMadrid,DptodeFisicadelaTierrayAstrofisica(Spain),8-BarcelonaSupercomputingCenter-

CentroNacionaldeSupercomputacion(Spain),9-CERFACS[Toulouse](France),10-MetOfficeHadleyCentre(UnitedKingdom)

The influenceof theAtlanticMultidecadalVariability (AMV)on theNorthAtlantic storm trackandrelated impacts on the European climate are assessed via a coordinated analysis of idealisedsimulations with state-of-the-art coupled models. The rationale for the approach is that themodulationofseasurfacetemperaturebytheAMVphasesmodifiestemperaturegradientandland-seathermalcontrastinthecoreoftheAtlanticextratropicalstormtrackregion.Consequentchangesof thegrowthofeddiesand theirpropagationcan lead tomodificationsof themeanpositionandspeedofthelow-leveljet.Datausedareobtainedfromamulti-modelensembleofAMV±experimentsconductedunder the frameworkof theDecadalClimatePredictionProject componentC (DCPP-C).TheseexperimentsareperformednudgingthesurfaceoftheAtlanticocean(temperatureandsalinity)tostatesdefinedbyimposingobserved(ERSSTv4)AMV±anomaliesontothemodelclimatology.Thefocusoftheanalysis isontheAMV-drivenmodificationstohighfrequencyatmosphericeddiesandtheir feedback on the mean flow. The signature of these dynamical changes on anomaloustemperatureandprecipitationoverEuropeisalsoinspected.Theanalysiswillshedlightonapotentialcombinationofdynamical (jetdisplacement)andthermodynamic (advectionof thermalanomalies)AMV impacts on Europe. The potential outcome of this analysis is informative of the processesgeneratinguncertaintyofimpactsinthemulti-modelDCPP-Cframework.

Keywords:AtlanticMultidecadalVariability,Jetstreamvariability,EuropeanClimate


5-P22

AMonteCarloAssessmentofChangesinSummertimePrecipitationCharacteristicsUnderRCP8.5-SensitivitytoAnnualCycleFidelity,

Overconfidence,andGaussianity

SperberKenneth(1),AnnamalaiH.(2),PallottaGiuliana(1)

1-LawrenceLivermoreNationalLaboratory(UnitedStates),2-InternationalPacificResearchCenter(UnitedStates)

Typically, climate change modelling studies assess changes and uncertainty using multi-decadalaveragesofseasonalmeanquantities(e.g.,Tebaldietal.2005;Lopezetal.2006,Collinsetal.2013,Kitoh et al. 2013,Wang et al. 2014, Sperber et al. 2016, Kitoh et al. 2017). Using 5-day averagedprecipitationfromallinitialconditionrealizationsof33CMIP5models(Tayloretal.2012),theadded-valueofthisworkistheassessmentofsummerprecipitationintermsofamount,onsetandwithdrawaldate,andlengthofseason(duration)throughtheanalysisofprobabilitydistributionsofinterannualanomalies,andthespecificationofconfidenceintervalsforthechangesinthemean.Also,wecalculatechanges in theprobabilityofbelow-normal,normal, andabove-normalprecipitation in theRCP8.5climatechangesimulationscalculatedusingHistoricalsimulationthresholdsforthreeequallyprobablecategories(Leith1973).Weexplorethesensitivityoftheanthropogenicclimatechangeto(1)usingallmodels,(2)using1modelpermodellinggrouptocircumventoverconfidenceduetothelackofmodelindependence(Knuttietal.2010),(3)theimpactofonlyusingmodelsthataremostrealisticinthesimulationoftheannualcycle intermsthetimingandvariabilityofonsetanddurationofsummerrainfall,and(4)wecompareresultsbasedonGaussianandnon-parametricstatistics,thelatterbasedonMonteCarlosamplingofthedifferentinitialconditionrealizations.

Keywords:AnthropogenicClimateChange,Precipitation,AnnualCycleSkill,MonteCarloSampling


5-P23

AtlanticMultidecadalVariabilityinCMIP6HistoricalSimulations

HodsonDan(1),RobsonJon(1),BoothBen(2),SuttonRowan(1)

1-NationalCentreforAtmosphericScience(Reading)(UnitedKingdom),2-MetOfficeHadleyCentre(UnitedKingdom)

AtlanticMultidecadalVariability(AMV)isamajordriverofimportantclimateimpactsinmanypartsoftheworld.Asyet,however,thereisnoconsensusonthedominantcausesofAMV.Thereisstrongevidence that AMV can arise from processes that are purely internal to the climate system. Inparticular,decadalvariability intheAtlanticMeridionalOverturningCirculationisbelievedtobeanimportantdriverofAMV.However,thereisalsoevidencethatexternalforcings-bothnatural(solar,volcanic)andanthropogenic(e.g.greenhousegases,aerosols,landuse)-canalsoinfluenceAMV,andmayhavedonesoinsignificantwaysduringthetwentiethcentury.Therehasbeenparticulardebate,arising from CMIP5 results, about the role of anthropogenic aerosols. CMIP6 presents a newopportunity to address these important issues. This paper will present results from analysis ofhistoricalsimulationswiththeUKGC3.1model.Iftimepermits,comparisonswillbemadewithresultsfromUKESMandotherCMIP6models. Thiswork is beingdoneaspart of theUKAtlanticClimateSystemIntegratedStudy(ACSIS)programme(www.acsis.ac.uk).

Keywords:AtlanticMultidecadalVariability


5-P24

Tropicalair-seaCO2fluxvariationsintwoESMswithanoceandataassimilationsystem

TatebeHiroaki(1),WatanabeMichio,KoyamaHiroshi,HajimaTomohiro,WatanabeMasahiro,KawamiyaMichio

1-JapanAgencyforMarine-EarthScienceandTechnology(Japan)

Interannual-to-decadalfluctuationsoftheglobal-meanatmosphericCO2concentrationareobservedassociatedwithinternalclimatevariations,andthesefluctuationssometimescanceloutandatothertimes support the centennial increasing trend of the global-mean CO2 concentration due toaccumulationof anthropogenic CO2emissions.Onedirect causeof the fluctuations is theoceanicabsorptionorreleaseofCO2throughtheglobalair-seaCO2fluxvariationswhicharedominatedbythetropicalPacific.Therefore,deeperunderstandingofthetropicaloceanicprocessesandproperlyinitializingoceanicstatesincludingmarineecosystemareofimportanceinfuturepredictionsoftheglobalCO2concentrationtogetherwithquantificationofhuman-influences.Inthepresentstudy,weevaluatedandexaminedthesimulatedair-seaCO2fluxvariationsinthetropicalPacificintwoESMs(MIROC-ES2L and MIROC-ESM) with the same ocean data assimilation system. Although theassimilation procedures are the same, observed anti-correlated relationship between interannualvariationsoftheupwardair-seaCO2fluxandtheseasurfacetemperature(SST)inthetropicalPacificarewellcaptured inMIROC-ES2L,buttherelationship isreversed inMIROC-ESM.Tropicalclimatic-meanstateofMIROC-ESMshowssignificantbiasesofweakertradewinds,morediffuseequatorialthermocline thanobservations, and simulated amplitudeof interannual variations ofNINO3-SST isabouthalfaslargeasobservations.Whenobservationsareassimilatedintothemodel,thesebiasesleadtonon-negligiblecorrectiontermsonthegoverningequationsofoceantemperatureandsalinity,which induces an anomalous spurious equatorial upwelling during El-Ni–o events. The spuriousupwellingbringsdissolvedinorganiccarbonrichwater inthesubsurface layertothesurfacemixedlayer.Consequently,ananomalousupwardair-seaCO2fluxisoccurredduringEl-Ni–o,asopposedtoobservationsandMIROC-ES2L.Bettermodelingphysicalprocesses in the tropicalclimatesystem issuggestedtobeessentialforbettermarineecosystemmodelingandreanalysis.

Keywords:air,seaCO2flux,tropics,ESM,initialization


5-P25

DecadalpredictabilityintheCMIP6models

TengHaiyan(1)

1-NCAR(UnitedStates)

Inordertoinvestigatethepotentialforinitializationtoimprovedecadalpredictions,wewillquantifytheinitialvaluepredictabilityofupper300mtemperatureinthetwonorthernoceanbasinsfortheCMIP6models,andwewillcontrastitwiththeforcedpredictabilityintheDECKexperiments.Wehavedevelopedthemethodologytoestimateinitialvaluepredictabilityfromlongcontrolruns.WhenweappliedthemethodtotheCMIP5models,wefoundthat1)initializationhasthepotentialtoimproveskillinthefirst5yearsintheNorthPacificandthefirst9yearsintheNorthAtlantic,and2)theimpactfrom initializationbecomessecondarycomparedto the impactofRCP4.5 forcingafter61/2and8yearsinthetwobasins,respectively.WeplantoupdatetheknowledgewiththeCMIP6experiments.Iftherearenotmanymulti-modeloutputsavailablebythetimeoftheworkshop,wewillfocusontheCESM2simulationsanddiscusschangeintheinitialvaluepredictabilityfromCCSM3,CCSM4,CESM1toCESM2.

Keywords:decadalpredictability,decadalprediction


5-P26

SimulationofhistoricalandprojectedclimatechangeinaridandsemiaridareasbyCMIP5models

ZhaoTianbao(1)

1-TianbaoZhao(China)

BasedonClimaticResearchUnitTimeSeries3.1 temperatureandGlobalPrecipitationClimatologyCenterfulldatareanalysisversion6precipitationdata,theabilitiesofclimatemodelsfromthefifthphase of the Coupled Model Intercomparison Project to simulate climate changes over arid andsemiaridareaswereassessed.Simulationsoffutureclimatechangesunderdifferentrepresentativeconcentrationpathways(RCPs)werealsoexamined.Thekeyfindingswerethatmostofthemodelsareabletocapturethedominantfeaturesofthespatiotemporalchangesintemperature,especiallythegeographicdistribution,duringthepast60years,bothgloballyaswellasoveraridandsemiaridareas. In addition, themodels can reproduce the observedwarming trends, butwithmagnitudesgenerallylessthantheobservationsofaround0.10.3C/50a.Comparedtotemperature,themodelsperformworseinsimulatingtheannualevolutionofobservedprecipitation,underestimatingboththevariabilityandtendency,andthere isahugespreadamongthemodels intermsoftheirsimulatedprecipitationresults.Themultimodelensemblemean isoverall superior toany individualmodel inreproducingtheobservedclimatechanges. Intermsof futureclimatechange,anongoingwarmingprojected by the multi-model ensemble over arid and semiarid areas can clearly be seen underdifferent RCPs, especially under the high emissions scenario (RCP8.5), which is twice that of themoderate scenario (RCP4.5). Unlike the increasing temperature, precipitation changes vary acrossareasandaremoresignificantunderhigh-emissionRCPs,withmoreprecipitationoverwetareasbutlessprecipitationoverdryareas. Inparticular,northernChina isprojected tobeoneof the typicalareasexperiencingsignificantlyincreasedtemperatureandprecipitationinthefuture.

Keywords:Climateprojection,Climatesimulation,Climatechange,CMIP5,Aridandsemiaridareas


5-P27

Evaluatingclimatemodelsimulatedextremes

ToretiAndrea(1)

1-EuropeanCommission,JointResearchCentre(JRC)(Italy)

Climate extremes heavily affect all socio-economic sectors causing losses, damages and fatalities.Understandingtheirdynamics,representingtheminclimatemodelsimulations,andthenprojectingtheir evolution under future scenarios is of upmost importance. Therefore, tailored statisticalproceduresmustbedevelopedandappliedtoevaluatemodelsandestimatechangesinthecomingdecades.Recentlyproposedapproaches,abletoidentifyandquantifycomplexnonlinearchangesinclimate extremes, are here reviewed and discussed. Applications to CMIP5 and EURO-CORDEXsimulationsareshown.Then,concurrentextremesareintroducedtogetherwithasuiteofinnovativemethodstodetectnonlinearspatio-temporalchanges.Finally,theimportanceofconcurrentextremesinglobalriskassessmentsisdiscussedbyusinganapplicationtoHELIXprojections.

Keywords:climateextremes


5-P28

Impactofinitialisationonthereliabilityofdecadalpredictions

VerfaillieDeborah(1),Doblas-ReyesFranciscoJ.(1)(2),SolarajuMuraliBalakrishnan(1),DonatMarkus(1),WildSimon(1)

1-BarcelonaSupercomputingCenter-CentroNacionaldeSupercomputacion(Spain),2-Instituci—CatalanadeRecercaiEstudisAvançats(Spain)

Bringing together initialised decadal climate predictions and non-initialised climate projections inordertoprovideseamlessclimateinformationforusersoverthenextdecadesisanewchallengingareaofresearchinthefieldofclimateprediction.Thiscanbeachievedbycomparingthepredictionsbased on global initialised and non-initialised simulations for commonprediction time horizons. Anumberofstudieshavefocussedonassessingtheadded-valueof initialiseddecadalpredictionsvs.non-initialisedsimulations in termsof forecastquality.However, the impactof initialisationonthereliabilityofdecadalforecastshasnotyetbeeninvestigatedsufficiently.Forecastreliabilityquantifiestheagreementbetweenthepredictedprobabilitiesandobservedrelativefrequenciesofagivenevent.Usersareparticularlysensitivetothelackofreliabilityoftheprobabilisticclimateinformationbecauseit implies that the uncertainty formulation is not trustworthy. In this communication,multi-modelinitialised decadal hindcasts from CMIP5 (and if available, CMIP6) will be compared to thecorrespondingnon-initialisedhistoricalsimulations intermsofreliabilityovertheircommonperiod1960-2005. Instruments like rank histograms, reliability diagrams and maps of the reliabilitycomponentoftheBrierscorewillbeusedtoassesstheadded-valueoftheinitialisationintermsofreliability for different variables (surface temperature, precipitation, sea-level pressure, ...) anddifferentforecasttimesatboththeglobalandEuropeanscales.

Keywords:decadalpredictions,forecastquality,reliability,initialisation


5-P29

DecadalClimatePredictionwithEC-Earth

WildSimon(1),AcostaNavarroJuanC(1),AmaralRamosArthurE.(1),BilbaoRoberto(1),BretonnierePierre-Antoine(1),CaronLouis-Philippe(1),CastrilloMiguel(1),Cruz-Garc’aRuben(1),Doblas-ReyesFranciscoJ.(2)(1),DonatMarkusG.(1),ExarchouEleftheria(1),EchevarriaPablo(1),Moreno-ChamarroEduardo(1),OrtegaPablo(1),Ruprich-RobertYohan(1),SicardiValentina(1),

TourignyEtienne(1),VerfaillieDeborah(1)

1-BarcelonaSupercomputingCenter-CentroNacionaldeSupercomputacion(Spain),2-Instituci—CatalanadeRecercaiEstudisAvan�ats(Spain)

ThisstudyprovidesanoverviewofthecontributionoftheBarcelonaSupercomputingCentertotheDecadal Climate Prediction Project (DCPP) using the EC-Earth coupled global climate model(https://www.ec-earth.org). Particular focus will be given on the production and results of theinitialiseddecadalhindcastsfrom1960to2017,ietheDCPPComponentA.Ourensemblehindcastscomprise 10members initialised yearly on 1st November, that is 580 different initialised decadalsimulations.ThedecadalpredictionsareperformedwitharesolutionofT255L91intheatmosphereand1¡and75verticallevelsintheocean.AtmosphericinitialconditionsaregeneratedusingERA-40and ERA-Interim. Ocean and sea-ice initial conditions have been produced using a NEMO-onlysimulationforcedbyDFSatmosphericfieldsandnudgedtowardsORAS4.Wewillshowresultsofthequalityofdeterministicandprobabilistichindcasts,suchasanomalycorrelationcoefficientsorBrierskillscoresofnearsurfacetemperatures,verifiedagainstobservationsandreanalysisproducts.Wewillalsoassesstheimpactofinitialisationonforecastreliabilityquantifyingthestatisticalrelationshipbetweenthepredictedprobabilitiesandtheobservedrelativefrequencyofanevent.SpecialemphasiswillbegiventotherepresentationoftheAtlanticMultidecadalVariabilityasamajorsourceofdecadalpredictabilityandothermodesofvariabilitysuchastheInderdecadalPacificVariabilityortheArcticOscillation.Initialisedhindcastswillfurtherbecomparedtoanensembleofnon-initialisedhistoricalsimulationsassessingthepotentialaddedvalueofinitialisingthemodeltowardstheobservedclimatestateinnear-termclimatepredictions.

Keywords:Earth,EC,AMV,decadalprediction,A,DCPP,DCPP


5-P30

TherecentabruptcoolingoverNorthAtlantic:Aforcedsignalornaturalvariability?

YangShuting(1),DrijfhoutSybren(2)(3),MeckingJenny(3)

1-DanishMeteorologicalInstitute(Denmark),2-RoyalNetherlandsMeteorologicalInstitute(Netherlands),3-UniversityofSouthhampton(UnitedKingdom)

ThesubpolarNorthAtlantic(SPG)isaregionknownasnolong-termwarmingincontrasttotheglobalwarminginresponsetotheanthropogenicforcing.Observationsduringthe20thcenturyindicatesthatSPGexperiencedperiodsofbothabruptcooling(e.g.,around1970,andmorerecentlyround2010)aswellasrapidwarming(e.g.,inmid-1990s),implyingthereexistslargedecadalvariabilityintheregion.Model studies suggested thatabruptSPGcoolingmay result fromeitherdisruptionof theAtlanticMeridionalCirculation (AMOC)oracollapseofSPGconvection,botharepossibleresponsestotheglobal warming trend. In this study, we investigate the relative role of natural variability to theanthropogenic forcing in driving the abrupt changes in SPG. Preliminary analyses ofmulti-centurysimulationswithconstantforcings(ie.,piControlexperiments)usingtwoversionsofEC-Earth(ie.,theCMIP5andapost-CMIP5version)indicatethatsignificantwarmingandcoolingtrendsof15-20yearsoccurfrequentlyandalternatively.Inparticularlythestrengthofthecoolingtrendsarecomparablewiththatobservedinmid-2010s.TheanalysisextendedtotheEC-EarthCMIP5historicalandRCP8.5experiments show similar results, suggesting the recent observed SPG cooling may well bephenomenon of natural climate variability. We currently extend these analyses to the CMIP6experiments of EC-Earth and also other CMIP6models. To understand themechanism leading todifferent trends, the upper ocean heat content, the stratification in SPG, as well as the AMOCassociatedwiththewarmingandcoolingwillbeassessedandcomparedtopresent-dayobservations.TheirconnectionwiththeatmosphericNorthAtlanticOscillationwillalsobeinvestigated.

Keywords:Coolingtrend,SubpolrNorthAtlantic

Session6-FutureProjections

6-P01

OntheCausesofPolewardShiftoftheIndianSummerMonsoonLowLevelJetstream

AchutaraoKrishna(1),SandeepS(1),DileepkumarR(1),ArulalanT(1)(2)

1-IndianInstituteofTechnologyDelhi(India),2-NCMRWF(India)

TheIndiansummermonsoonrainfallhasseenadecliningtrendoverpartsoftheIndiansubcontinentinrecentdecades.Thesedecreasesareoffsetbyincreasesinotherparts.Inclimatemodelsimulations,greenhousegasforcingsincreasethemonsoonprecipitationwhileaerosolforcingsleadtoadecrease.Models forced with best estimates of historical anthropogenic aerosols reproduce the observeddeclineinSouthAsianmonsoonprecipitation(Bollasinaetal.;Science,2011).Polsonetal.,(GRL,2014)implicatedanthropogenicaerosolforcingsinthedeclineinglobalmonsoonprecipitation.Guoetal.,(Atmos.Chem.Phys.,2015)furthershowedthatmodelsthatincludetheaerosolindirecteffectshowa largerdecrease inSouthAsianmonsoonseasonrainfall. The large-scalemonsoonchanges (bothgloballyandregionallyoverSouthAsia)includesubtleregionaldetails.OnesuchisadryingtrendoverthesouthernpartofthewestcoastofIndiaandanincreasefurthernorthinrecentdecades.SandeepandAjayamohan(Clim.Dyn.,2015)showedthatCMIP5andCMIP3modelsimulations(historicalaswellasfuturescenarios)indicatethatthispatternofdeclineinrainfallappearstobeaconsequenceofapolewardshiftinthemonsoonlowleveljetstream.Thegenesislocationofmonsoonlowpressuresystems(LPS)thatdistributemorethanhalftheprecipitationovercentralIndiaarealsofoundtobeshiftedpolewardfromtheBayofBengal to Indiansubcontinenttowardstheendof21stcentury inRCP8.5simulations,consistentwiththeshiftinlowlevelflow(Sandeepetal.,PNAS,2018)indicatingalinkbetweenchangesinprecipitationanddynamicsinadditiontoaerosolmicrophysicaleffects.Inthisstudyexaminethedriversofthesedynamicalchangesusingindividualforcingsimulationsfromthe CMIP5 (and CMIP6 as available) to understand their relative roles in producing the observedpatternsofprecipitationchange.

Keywords:monsoonlowleveljetstreamdynamicalresponsesingleforcing


6-P02

Contrastingmethodsofdetectingandattributingtheimpactofexternalforcings

ArblasterJulie(1)(2),TebaldiClaudia(3),MeehlGerald,BatesSusan,StrandGary

1-MonashUniversity[Melbourne](Australia),2-NationalCenterforAtmosphericResearch[Boulder](UnitedStates),3-NationalCenterforAtmosphericResearch(UnitedStates)

Detectionandattributionisbasedontheassumptionthattheimpactofnaturalandanthropogenicforcingsontheclimatesystemcanbewellapproximatedbylinearlyaddingtheeffectsofindividualforcings. This assumption has been validated for global averages and spatial patterns of (mostly)temperaturerelatedvariables.ModellinggroupsinCMIP5undertookthesedetectionandattributionexperiments in variousways, someaddingeachhistorical forcing toapreindustrial control,otherseliminatingcertainforcingsfromahistoricalall-forcingsimulation.Manystudieshaveassumedthesemethodsproducesimilarresults,howevertheavailableexperimentshavenotallowedthisassumptionto be tested until now. Here for the first timewe compare these twomethods of detection andattributioninensemblesimulationswiththeCESM1-CAM5coupledclimatemodel.Forexample,aretheresultsofasimulation that includesallexternal forcingsbut forozonedepletionsimilar to theresultsofaddingtheoutputofallindividualsingle-forcingexperimentsbutforthosedrivenbyozonechanges?Resultssuggestthatsimilarevolutionsofglobaltemperaturesandprecipitationarefoundbetween the two methods, with some notable differences in the response to aerosols. Ourexperimentsenableustogobeyondtheglobalaveragetemperatureresponseandexploredifferentvariablesatdifferentspatialandtemporalscalessotocharacterizeinasystematicfashionwherethelinearityholdsandwhereitbreaksdown.

Keywords:detection,attribution,internalvariability,ensembles,aerosols


6-P03

AttributionofOceanTemperatureChangetoAnthropogenicandNaturalForcingsusingtheTemporal,VerticalandGeographicalStructure

BilbaoRoberto(1),GregoryJonathan(2)(3),BouttesNathaelle(4),PalmerMatthew(3),StottPeter(3)

1-BarcelonaSupercomputingCenter-CentroNacionaldeSupercomputacion(Spain),2-UniversityofReading(UnitedKingdom),3-UnitedKingdomMetOffice[Exeter](UnitedKingdom),4-LaboratoiredesSciencesduClimatetdelƒnvironnement[Gif-sur-Yvette](France)

Weexaminewhethersignificantchangesinoceantemperaturescanbedetectedinrecentdecadesand if sowhether they can be attributed to anthropogenic or natural factors.We compare oceantemperature changes for1960-2005 in fourobservationaldatasets and inhistorical simulationsbyatmosphere-oceangeneralcirculationmodels (AOGCMs) fromtheCoupledModel IntercomparisonProjectphase5(CMIP5).ObservationsandCMIP5modelsshowthattheupper2000mhaswarmedwithasignalthathasawell-definedgeographicalpatternthatgraduallypropagatestodeeperlayersovertime.Greenhousegasforcinghascontributedmosttoincreasingthetemperatureoftheocean,a warming which has been offset by other anthropogenic forcing (mainly aerosols), and volcaniceruptionswhichcauseepisodiccooling.Bycharacterizingtheoceantemperaturechangeresponsetotheseforcingsweconstructmulti-modelmeanfingerprintsoftime-depthchangesintemperatureandcarry out two detection and attribution analysis. We consider first a two-signal separation intoanthropogenicandnaturalforcings.Then,forthefirsttime,weconsiderathreesignalseparationintogreenhouse gas, anthropogenic aerosols and natural forcings. We show that all three signals aresimultaneously detectable. Using multiple depth levels decreases the uncertainty of the results.Limitingtheobservationsandmodel fieldsto locationswherethereareobservations increasesthedetectabilityofthesignal.

Keywords:ocean:climatechange:detectionandattribution


6-P04

ImplicationofMidHoloceneandLastInterglacialchangesininsolationseasonalityonhighandmidlatitudeclimate

BraconnotPascale(1),Otto-BliesnerBette(2),PmipParticipants

1-LSCE/IPSL(France),2-NationalCenterforAtmosphericResearch(UnitedStates)

Themid-Holocene (MH), 6000 years BP, and the Last Interglacial (LIG), 127000 years BP, are keyreference periods for the 4th phase of the PaleoclimateModeling intercomparison Project (PMIP,Kageyama et al., GMD, 2018; Otto-Bliesner et al., GMD, 2017), and will contribute to the overallunderstandingoftheresponseoftheclimatesystemtoexternalforcingaspartofCMIP6.Thesetwointerglacial periods are characterized by enhanced (reduced) seasonality of the incoming solarradiationat the topof theatmosphere in thenorthernhemisphere (southernhemisphere). In theannualmean, largerobliquityalsoincreasessolarradiationinhighlatitudesanddecreases it inthetropics. Trace gas atmospheric concentrations are similar to pre-industrial values. The poster willpresentthenewsetsofavailablesimulationswiththeaimtocharacterizethedifferencesinsnowandice feedbacks in high latitudes resulting from differences in the magnitude of the changes inseasonality between the two periods. The impact of this feedback on polar amplification will beinvestigated.Alsobothperiodsarecharacterizedbyenhancedmonsoonsinthetropicalregions.Thesefirstanalyseswill investigatehowmonsoonseasonality is relatedtothecharacteristicsof thesolarforcingandtofeedbacksfromoceancirculationandlandsurfaceproperties.Forbothperiods,modelresults canbeanalyzed in the lightofavailablepaleoclimate reconstructions. Inaddition, themid-HoloceneisconsideredasanentrycardforthepaleoclimatesimulationsinCMIP6,becauseithasbeenareferenceperiodforPMIPsincethefirstPMIPphase(JoussaumeandTaylor,1995).AcomparisonwithpreviousMHPMIPsimulationswillhighlighttheevolutionofmodelperformancesinsimulatingaclimatestatedifferentfromthemodernone.

Keywords:monsoon,hydrologicalcycle,climatefeedback,Paleoclimate,polaramplification


6-P05

TheresponseofclimatevariabilityinPMIP4/CMIP6

BrierleyChris(1),HargreavesJulia(2),RehfeldKira(3)

1-UniversityCollegeofLondon[London](UnitedKingdom),2-BlueSkiesResearch(UnitedKingdom),3-HeidelbergUniversity(Germany)

Therehasbeensubstantialeffortdevotedtodetermininghowvariabilitywillchangeasglobalclimatechanges.Thishasusedarangeofinformationrangingfrompastobservations,futuresimulationsandclimate proxy reconstructions. One such example is the recognition that the El Ni–o-SouthernOscillation(ENSO)wasweakerthantodaythroughoutmuchofthepast~10,000years(theHolocene).Analysis of the multi-model simulations performed for the Palaeoclimate Model IntercomparisonProjecthelpedunderstandthemechanismsforthis,buthasbeenchallengingtoreplicateforothermodes(suchastheNorthAtlanticOscillationandtheSouthernAnnularMode).Thenewfeaturesoffourth phase of the Palaeoclimate Model Intercomparison Project (PMIP4) combined with theinfrastructureadvancesassociatedwithCMIP6provideagreatopportunitysystematicallyinvestigatethe forced response of variability. PMIP4 has expanded to encompassmore past climates: Tier 1comprisesoffourequilibriumsimulationsandforcedtransientsimulationofthelastmillennium.Thisprovidesthemanyyearsrequiredtobettercharacterisechangesinlow-frequencymodesofvariability.Additionally, the Climate Variability Diagnostics Package (part of the ESMValTool) has made thecomputationofmanyclimatemodesroutine,aswellasprovidingdiagnostics.Combined,thismakessimultaneousanalysesacrossmultipleexperiments,multipleclimatemodelsandmultiplefeaturesofclimatevariability feasible.Hereweexplore theresponseof theamplitudeofclimatevariability tovariousEarthSystemforcings-suchasorbitalchanges,landconfigurationandgreenhousegases.Wecharacterisehowthemainmodeschangeintheseclimatestates,butalsohowthechangesinregionalinterannual variability respond to the changes in those modes. Additionaly PMIP enables theconstrainingfutureclimatechangesthroughthecombineduseofemergentmulti-modelrelationshipsand palaeoclimate reconstructions. We will highlight instances that show potential to use proxyobservationstoconstrainfutureprojectionsofclimatevariability.

Keywords:pastclimate,variability


6-P06

IncreasedvariabilityofEasternPacificElNi–osurfacetemperatureundergreenhousewarming

CaiWenju(1)(2),WangGuojian,WuLixin,DewitteBoris

1-CentreforSouthernHemisphereOceansResearch(Australia),2-KeyLaboratoryofPhysicalOceanography/InstituteforAdvancedOceanStudies,OceanUniversityofChinaandQingdao

NationalLaboratoryforMarineScienceandTechnology(China)

The El Nino Southern Oscillation (ENSO), the dominant year-to-year climate phenomenon on theplanet,severelydisruptsglobalweatherpatterns,withimpactsonthelivesandlivelihoodsofmillionsofpeoplearoundtheworld.ObservationsshowthattherearetwodistinctENSOregimes,onewithaseasurfacetemperature(SST)anomalycentreintheequatorialcentralPacific,andtheotherwithananomalycentreintheequatorialeasternPacific,referredtoasCPandEPENSO,respectively.Undergreenhousewarming,CPSSTvarianceisprojectedtoincreasewithamodestinter-modelconsensuscontributedbyincreasedfrequencyofstrongLaNi–aevents.However,theissueofhowEPENSOSSTmaychangecontinuestobeplaguedbyapersistentlackofapparentinter-modelagreementintheresponseofeasternPacificSSTtoanthropogenicgreenhousegasforcing.Hereweshowforthefirsttimethatifwefocusontheuniqueanomalypatternforeachmodel,ratherthanonanindexfromafixedgeographicalregion(e.g.Ni–o3),thereisinfactarobustincreaseinEPENSOSSTvariancewithastronginter-modelconsensus.

Keywords:GlobalWarming,ElNino


6-P07

ClimateScenariosfortheFifthUnitedStatesNationalClimateAssessment

EasterlingDavidR.(1),KunkelKennethE.(2),BallingerAndrew(2)

1-NOAANationalCentersforEnvironmentalInformation(UnitedStates),2-CooperativeInstituteforClimateandSatellites,NCStateUniversity(UnitedStates)

TheThirdU.S.NationalClimateAssessment(NCA3)mainlyusedclimatescenariosgeneratedusingtheCMIP3suiteofclimatemodelsimulationswithsomegeneratedusingCMIP5.TheFourthU.S.NationalClimateAssessmentused theCMIP5 suiteof simulations, andalsoused the LocalizedConstructedAnalog(LOCA)statisticaldownscalingdatasettogeneratescenariosofextremessuchaschangesinheavyprecipitationevents.WiththeCMIP6suiteofsimulationsbecomingavailabletheintentoftheU.S.GlobalChangeResearchProgramistoutilizethesesimulationsasmuchaspossiblefortheFifthU.S.NationalClimateAssessment.OnemajorquestionraisedduringtheNCAprocessishowextremes,suchashydrologicalextremes,havechangedandwillchangeinthefuture.Hereweexamineheavyprecipitationeventsbyfindingthelargestmulti-dayeventsforvarioussizedareas(e.g.50,000km2)intheobservedrecordfortheeasternUnitedStates,thenexamineboththesimulationsdirectlyfromCMIP5 and two statistical downscaling methods driven by CMIP5 simulations for their ability toproducesimilarprecipitationevents.Secondly,weexaminetheabilityofbothmodelsanddownscalingmethodstoreproducetheobservedspatialcoherenceofthepointprecipitationamountsacrossthesimulatedprecipitationevents.Lastly,weexpecttohavepreliminaryresultsfromCMIP6simulationstocomplementtheseanalyses.

Keywords:climateassessments,downscaling,ClimateScenarios


6-P08

TransientClimateResponsetoCumulativeEmissionsinCMIP6models.PreliminaryresultsfromtheC4MIPexperiments

FriedlingsteinPierre(1),JonesChris(2),AroraVivek(3)

1-UniversityofExeter(UnitedKingdom),2-MetOfficeHadleyCentre(UnitedKingdom),3-EnvironmentandClimateChangeCanada(Canada)

TheTransientClimateResponsetoCumulativeEmissions(TCRE)isanearlinearrelationshipbetweenanthropogenicemissionsofCO2cumulatedovertimeandtheglobalsurfacetemperaturewarming.Thisquantity is strongly relevant for climatepolicies as it allowsquantifying the remaining carbonbudgetcompatiblewithanyclimatetargetsuchas1.5¡Cor2¡C.

Usingcoupledclimate-carboncyclesimulationsperformedbyEarthSystemModelsofdifferentlevelofcomplexity,IPCCAR5assessedthatTCREislikelyintherangeof0.8¡Cto2.5¡Cper1000GtC.Hereweusepreliminaryresultsfromthe1%CO2increaseperyearsimulationsfromtheCMIP6decktoassessTCRE.WefurtherdecomposeTCREinitstwomaincontributions:climatesensitivity(dT/dCO2)andairbornefraction(dCO2/dEmi),thislatterbeingfurtherdecomposedintheconcentration-carbon(beta)andclimate-carbon(gamma)feedbackscomponents.

InitialresultsfromCMIP6simulationscurrentlyavailablesuggestthatTCREmightbeatthehighendoftheCMIP5range,implyingthattheremainingallowablecarbonemissionsfor1.5and2¡CcouldbelowerthanthemostrecentestimatesfromtheIPCCSpecialReportonGlobalWarmingof1.5¡C.

Keywords:TCRE,C4MIP,Carbonfeedbacks


6-P09

Assessingtherobustnessofmarineheatwaveprojections

FroelicherThomas(1),SandraStriegel(1),GriffiesStephen(2)

1-UniversityofBern(Switzerland),2-NOAAGeophysicalFluidDynamicsLaboratory(UnitedStates)

Someoftherecentlyobservedmarineheatwaves(MHWs)revealedthehighvulnerabilityofmarineecosystemsandfisheriestosuchextremeclimateevents.SatelliteobservationsandasuiteofCMIP5Earthsystemmodel(ESM)simulationsshowthatMHWshavealreadybecomelonger-lastingandmorefrequent, extensiveand intense in thepast fewdecades, and that this trendwill accelerateunderfurtherglobalwarming(Frolicheretal.2018).Between1982and2016,wedetectadoublinginthenumberofMHWdays.Iftemperatureweretoriseby3.5¡Cbytheendofthetwenty-firstcentury,theaverageincreaseintheprobabilityofMHWswouldbe41timeshigherthaninpreindustrialtimes.Atthis level of global warming, the spatial extent of the heatwaves would be 21 times larger, theirduration would increase to 112 days and their maximum intensity would rise to 2¡C. The largestchangesareprojectedforthewesterntropicalPacificandtheArcticOcean.AnimportantassumptionunderlyingtheanalysisisthattheemployedESMssimulateMHWsinasufficientlyrealisticmanner.WhereasthisisthecaseforthefrequencyandthemaximumintensityofMHWs,itisnotthecaseforthedurationandspatialextent,whichmightbecausedbytherelativelycoarseresolutionoftheESMs.HereusingdailyseasurfacetemperatureoutputfromCMIP6ESMsimulations(piControl,historical,ScenarioMIP) as well as a new global high-resolution ESM (GFDL-CM2.6, ocean model horizontalresolutionof0.1¡),wetesttherobustnessoftheseearlierCMIP5ESMresults.Aparticularemphasiswillbegiven to the relative roleof changes inmeanocean temperatures, changes in temperaturevariabilityoracombinationofthetwo.

Keywords:Marineheatwave,Oceanwarming,Climatemodeling,Climatechange,Extremeevents


6-P10

Responsesofterrestrialariditytoclimatechangeandglobaldrylandexpansions

FuQiang(1)

1-UniversityofWashington(UnitedStates)

Thedrynessof terrestrialclimatecanbemeasuredbythearidity index,AI, i.e., theratioofannualprecipitation(P)topotentialevapotranspiration(PET).ThePETrepresentstheevaporativedemandoftheatmosphere,whichdependsonthesurfaceairtemperature,relativehumidity,windspeed,andavailableenergy.Drylandsareareaof landswith lowamountofwater insoilwithAI<0.65,whichcompriseofmorethan40%oftheglobal terrestrialarea,andarehometomorethanonethirdofglobalpopulations.Knowledgeofhowanthropogenicclimatechangewillaffectaridityandextentofdrylandsisessentialforwaterresourceandlandusemanagementintheseregions.FengandFu(2013)forthefirsttimeexaminedthechangeofterrestrialaridityinthepastandinthefutureanditsimpactonthedrylandareabyanalyzingbothobservationsandCMIP5simulations.Theyshowedthatglobaldrylandshaveexpandedinthepastandwillcontinuetoexpandinthe21stcentury.Itwassuggestedthat the relative increase in evaporation over the ocean, which controls the relative change inprecipitation over land, is slower than the increase in PET over land, leading to a drier terrestrialclimate(SherwoodandFu2014;FuandFeng2014;Fuetal.2016).Thispresentationwillexaminehowthe terrestrial aridity responds to globalwarming in termsof P/PET andhow this responsewouldimpactthedrylandareaandtheitscomponentsincludinghyper-arid,arid,semi-arid,andsub-humidregionsbyanalyzingtheCMIP6output.WewillalsocomparetheresultswiththosefromCMIP5.

Keywords:Terrestrialaridity,drylands,climatechange


6-P11

EvaluationofthePMIP4/CMIP6palaeosimulations

HarrisonSandy(1),ComasBruLaia(1),CleatorSean(2),LiGuangqi(1),WeiDongyany(1)

1-UniversityofReading(UnitedKingdom),2-UniversityofSurrey(UnitedKingdom)

Past climate states provide an opportunity to evaluate how well models simulate large climatechanges. Themid-Holocene (MH, 6000 yr BP) and the Last GlacialMaximum (LGM, 21,000 yr BP)providestronglycontrastingclimatestates,havebeenusedastargetsformodelevaluationinpreviousroundsofPMIP/CMIP,andare“entrycards”inthesuiteofpalaeoclimatesimulationsthatarebeingrunbyPMIPmembersinCMIP6.Inthispresentation,wewilldescribethepalaeoclimateobservationsthatareavailableforbenchmarkingandevaluationofthese(andother)palaeoclimatesimulations,withemphasisonnewglobalhydroclimatedatasetsincluding(a)isotopicrecordsfromspeleothems,(b)lakestatusrecordsofhydroclimate,and(c)quantitativeclimatereconstructionsthataccountforthe ecophysiological effects of changing [CO2]. The interpretation of these palaeo-records is notalways straightforward, and requires innovative techniques for data-model comparison, includinguncertaintypropagation, forwardmodellingandprocessdiagnosis, andwewilldiscussdata-modelcomparisonapproachesthatarecurrentlyunderdevelopment

Keywords:processdiagnosis,forwardmodelling,observations,palaeo,benchmarking,modelevaluation,palaeoclimate


6-P12

MonsoonprecipitationresponsestoglobalwarmingandtheirregionaldifferencessimulatedbyCMIPmodels

EndoHirokazu(1),KitohAkio(2)(1)

1-MeteorologicalResearchInstitute(Japan),2-JapanMeteorologicalBusinessSupportCenter(Japan)

Global climatemodels used in a global warming projection have been extensively developed andimproved for thepast severaldecades.However,aprojectionof regionalprecipitationchangestillfaces challenges because of its large spatial variation and large uncertainty. Using climate modelexperiments in the CoupledModel Intercomparison Project (CMIP) phase 5,we have investigatedprecipitation responses to global warming and their uncertainty in various monsoon regionsworldwide, andhave tried tounderstand regionaldifferences in the responses. TheCMIPphase5modelsprojectthatsummermonsoonprecipitationwillincreaseinmostregionsinawarmerclimate,butwithlargeregionalvariationsintermsofthemagnitudesincludingalargerincreaseinAsiaandadecreaseinNorthAmerica.Themodelsprojectthatheavyprecipitationwillincreasemuchmorethanthemeanprecipitationoverallmonsoonregions,witha larger increase inAsia.Amoisturebudgetanalysis reveals that an increase in atmospheric moisture (thermodynamic change) contributespositivelytotheprecipitationchanges,butageneralweakeningofthemonsooncirculation(dynamicchange) acts to oppose the positive thermodynamic change. In most monsoon regions, thethermodynamiccomponentoverwhelmsthedynamiccomponent.Interestingly,intheAsianmonsoonregionsthenegativedynamicchangeismuchlessthaninothermonsoons,resultinginlargerincreasesinprecipitation.Furtheranalysisusingidealizedmulti-modelexperimentsindicatesthatCO2-inducedenhancement of the land-sea thermal contrast and resultant circulation changes are the mostinfluentialintheSouthAsianmonsoon,leadingtothelargestincreaseinprecipitation,whichsuggestsan important roleof the landwarmingon theAsianmonsoon response.Weareexpecting thatananalysis with updated climatemodels participating in the ongoing CMIP phase 6 will support ourfindingsandprovidemoreconfidentresults.

Keywords:Asianmonsoon,monsoonprecipitation,CMIPmodels,globalwarming


6-P13

Howfaristhecarbonsinkpredictableinamulti-modelframework?

IlyinaTatiana(1),LiHongmei,SpringAaron,SéférianRoland,FriedlingsteinPierre,LovenduskiNikki,BoppLaurent,ParkJong-Yeon,MegumiChikamoto,DunneJohn,

YeagerStephen,BernardelloRaffaele


SeveralpredictionsystembasedonEarthsystemmodels includethecarboncyclecomponentsandtherebyenablepredictionsoftheoceanand landcarbonsinks.SuchpredictionsystemsareforcedwithprescribedhistoricalatmosphericCO2concentrationsandassimilateonly thephysicalclimatedata,albeitusingdifferentassimilationandinitializationdesigns.Thelandandoceancarboncy-clemodelsrunaspassivecomponentsadjustingtothestateandthephaseofthephysicalclimatesystem.Basedon theoutcomesof6oceancarboncyclecomponentsand3 landcarboncyclecomponentsincludedindifferentpredictionsystems,weperformafirstattemptofamulti-modelassessmentoftheoceanandlandcarbonsinkspredictiveskills.WeuseavailableobservationalsynthesisproductsandtheestimatesoftheGlobalCarbonBudgettoestablishpredictiveskill.Additionally,thepotentialpredictiveskill isevaluatedagainstthemodels'ownreconstructionorassimilationsimulations.Wefindapotentialpredictiveskillofupto3yearsandofabout1yearfortheglobaloceanandthegloballandcarbonsinks,respectively.Moremodelingcenterswillintegratetheircar-boncyclecomponentsaspartofCMIP6DCPPsimulations,facilitatingamorecomprehensiveassessmentofthecarbonsinkpredictability.Thisnewknowl-edgeispivotalforpredictingthefateofanthropogenicCO2emissionsand for facilitating verification of near-term emission trends in support of the UNFCCC globalstocktakes.

Keywords:predictabilityoftheoceanandlandcarbonsinks,DCPP,C4MIP


6-P14

TheGlobalCarbonCycleemissionsdrivensimulationsintheNASA-GISSclimatemodel

ItoGen(1),RomanouAnastasia(1),KiangNancy(1),AleinovIgor(1),FaluvegiGregory,KelleyMaxwell(1)

1-NASAGoddardInstituteforSpaceStudies(UnitedStates)

Wepresentsimulationsoftheglobal(land-ocean-atmosphere)carboncyclefromtheNASA-GoddardInstitute for Space Studies (NASA-GISS) coupled climate model. This incarnation of CMIP6-levelsimulations aredrivenby emissionsof CO2 fromdifferent sources, including land, ocean, biomassburning,land-usechangeandfossilfuelemissions.Bothlandandoceancarboncomponentshavebeenspun up to equilibrium using prescribed, pre-industrial era forcings and continue with the CMIP6greenhouse gas forcing scenarios. CO2 is fully interactive with atmospheric radiation and affectslongwave absorption and increase of surface air temperature. Results are compared with theprescribed-concentrationdrivenrunsinthesamemodel.

Keywords:CO2,emissions,carboncycle,land,ocean,atmosphere


6-P15

TransientsimulationsovertheCommonErausingcomprehensiveEarthSystemModels:ThePMIP4/CMIP6past2kexperiment

JungclausJohann(1),JahnAlexandra(2),TooheyMatt(3),WagnerSebastian(4),LorenzStephan(1)

1-MaxPlanckInstituteforMeteorology(Germany),2-DepartmentofAtmosphericandOceanicSciencesandINSTAARUniversityofColoradoBoulder(UnitedStates),3-GEOMAR-HelmholtzCentreforOceanResearch[Kiel](Germany),4-Helmholtz-ZentrumGeesthacht(Germany)

TheCommonEra(CE,i.e.thetwomillenniabeforetheindustrialization)isamongtheperiodsselectedbythePaleoModelIntercomparisonProject(PMIP)fortransientexperimentscontributingtoCMIP6and the fourthphaseofPMIP (PMIP4). Inanongoingeffort, thePastGlobalChange2k (PAGES2K)initiativeisprovidinganextensiveproxydatabaseoncontinentalandregionalscaleclimatechangeover the last 2000 years. For PMIP4, novel estimates and updates of external forcing have beencompiled (Jungclaus et al., 2017). The new volcanic and solar forcing data sets also allow for anextensionofthetemporalrange.Therefore,inadditiontotheTier-1categorysimulationpast1000fortheperiod850CEto1849CE,theTier-3past2kexperimentcoverstheentireCE.HerewepresentfirstresultsfromtwoEarthSystemModels,CESM1.1(Zhongetal.,2018)andMPI-ESM1.2,followingthePMIP4past1000andpast2kprotocols.ThesesimulationsextendthepoolofcurrentESMsimulationsintothe1stmillenniumCEandrepresentanimportantbasistoassessthemodels'responsetoexternalforcing. We analyze regional trends and variations over the last 2000 years in comparison withPAGES2kreconstructions.Regardingthefirstmillennium,wefocusontherepresentationoftheeffectsofaclusterofstrongvolcaniceruptionsinthefirsthalfofthe6thcenturyCE.Jungclaus,J.H.etal.,2017: The PMIP4 contribution to CMIP6 Part 3: The last millennium, scientific objective, andexperimental design for the PMIP4 past1000 simulations. Geosci. Model Dev., 10, 4005-4033,doi:10.5194/gmd-10-4005-2017.Zhong,Y.,etal.,2018:AsymmetriccoolingoftheAtlanticandPacificArctic during the last twomillennia: a dual observation-modelling study. Geophys. Res. Lett., 45,doi:10.1029/2018GL079447.

Keywords:pastclimate,responsetoexternalforcing,modeldataintercomparison


6-P16

RegionalandverticalstructureofoceanheatuptakeintheUKESM1CMIP6simulationsofthehistoricalclimate

KuhlbrodtTill(1),JonesColin(2),PalmieriJulien(3),YoolAndrew(3)

1-NationalCentreforAtmosphericScience(NCAS),UniversityofReading(UnitedKingdom),2-NationalCentreforAtmosphericScience[Leeds](UnitedKingdom),

3-NationalOceanographyCentre(UnitedKingdom)

OftheadditionalheatintheEarthsystemabout93%havebeentakenupbytheworldoceanoverthepast couple of decades. Hence this process is crucial for moderating global surface temperaturechange.For thecurrent,CMIP6generationofEarthSystemmodels it isequallycrucial toestimateoceanheatuptakeprocessesaccuratelyinordertodeliverrobustandreliableprojectionsofclimatechangeinthecomingdecades.Previousgenerationsofclimatemodelsdisplayedsystematicalbiasesinoceanheatuptake,e.g.a tooweaksimulatedvertical temperaturegradient,suggestingthattheaddedheatisdistributedtoofardeepinthemodels.Inaddition,theinter-modelspreadinoceanheatuptakeefficiencycontributessignificantlytotheuncertaintyinthetransientclimateresponse.HereweanalysetheUKESM1historicalsimulationsforCMIP6intheseregards.Weshowananalysisofthevertical structure of ocean heat uptake compared to observations. Furthermore, we analyse thehorizontalstructureofoceanheatuptake,puttingafocusontherelativerolesoftheNorthAtlanticandtheSouthernOcean.ItisenvisagedtoentrainmoreCMIP6EarthSystemmodelsinthisstudy,andtoputtheresultsinthecontextofEarth'senergyimbalance.

Keywords:oceanheatuptake,historicalsimulations,UKESM1,Earth'senergyimbalance


6-P17

Amulti-modelanalysisofthehistoricalcarbonfluxesandcompatiblefossilfuelemissionsinCMIP6Models

SpencerLiddicoat(1),JonesChrisD,WiltshireAndy


ManyEarthSystemmodelshaverecentlyperformedaseriesofsimulationsforPhase6oftheCoupledModelIntercomparisonProject(CMIP6).Thisposterpresentsamulti-modelanalysisoftheresultsofmanyCMIP6models'historicalsimulations,coveringtheperiodfrom1850-2015,withafocusonthecarbonfluxesbetweentheatmosphere,landandocean.Wepresenttherangesofhistoricallandandoceancarbonuptakesimulatedbythemodels.Knowingthelandandoceanuptake,wecaninfertheamountoffossilfuelcarbonthatcanbeemittedtomaintaintheCO2concentrationpathwaydrivingthemodels;wealsoshow,therefore,therangeofcompatiblefossilfuelemissionsfromtheCMIP6modelsoverthehistoricalperiod.

Keywords:carbonfluxes,cmip6,earthsystemmodelling,compatibleemissions


6-P18

Stratosphere-TroposphereCirculationChanges

ManziniElisa(1),KarpechkoAlexeyYu.(2)

1-MaxPlanckInstituteforMeteorology(MPI-M)(Germany),2-FinnishMeteorologicalInstitute(Finland)

Thestratosphericvortexisoneofthedynamicalfactorsdrivingnearsurfacelongtermweatherandclimate.ThroughtheanalysisoftheCMIP5multi-modelensemble,clearevidenceoftheimpactsofborealwinterstratosphericdynamicalchangesonsurfaceclimateemerged(e.g.,Manzinietal2014;Simpsonetal2018).However,thesestudieshavealsoillustratedaconsiderabledegreeofuncertaintyintheborealwinterstratosphericvortexresponsetotheanthropogenicincreaseingreenhousegases.Herewebuilduponthesepreviousresultstoassesstheborealwinterstratosphericvortexresponseanditsrelationshiptothesurfaceclimateresponse,bymeansofmulti-modelanalysisofhistoricalandscenarioexperimentsperformedwithinCMIP6andreportconsistencyordifferenceswithrespectstothepreviousphasesofCMIP.Inaddition,makinguseofthenewlyrequesteddiagnosticsofresolvedandparametrizedstratosphericwavedriving(DynVarMIP),weattemptamomentumbudgetoftheatmosphericchangesandsearchforacross-modelrelationshipsbetweenstratosphericchangesandmodelbiasesinthestratosphericvortexbasicstate.Manzini, E., et al. (2014) Northern winter climate change: Assessment of uncertainty in CMIP5projections related to stratosphere-troposphere coupling, J. Geophys. Res. Atmos., DOI:10.1002/2013JD021403.Simpson, I.R., et al. (2018) The Downward Influence of Uncertainty in the Northern HemisphereStratosphericPolarVortexResponsetoClimateChange,J.Climate,DOI:10.1175/JCLI-D-18-0041.1

Keywords:circulation,stratosphere,dynamics


6-P19

PresentandfutureseasonallandsnowcoversimulatedbyCMIPcoupledclimatemodels

MenegozMartin(1)(2),KrinnerGerhard(3),Brutel-VuilmetClaire(4),SantolariaMaria(4),DerksenChris(5),MudrykLawrence(5)

1-InstitutdesGéosciencesdel'Environnement(France),2-BarcelonaSupercomputingCenter-CentroNacionaldeSupercomputacion(Spain),3-InstitutdesGéosciencesdel'Environnement(France),4-InstitutdesGéosciencesdel'Environnement(France),5-EnvironmentandClimate

ChangeCanada(Canada)

Snow is a crucial component of the climate system. It perennially covers the large continental icesheets and the sea ice almost entirely, as well as approximately 45 million km2 of the northernhemisphereduringthecoldseason.Thesnowcoverplaysmultiplerolesintheclimatesystem:throughthesnowalbedofeedback,itamplifiesglobalwarming.Italsoinsulatesthesoil,influencingthecarbonstorageinpermafrost,andaffectstheatmosphericcirculation.Observationsshowageneraldecreaseinfallandspringsnowcoverextentandduration,withregionalandseasonalvariabilitydrivenbythecompetinginfluencesoftemperatureandprecipitation.Earthsystemmodelsincludesnowmodulesfromone-layer tomultiple-layerschemeswitha largerange incomplexity in therepresentationofprocessinvolvingsnow-vegetationinteraction,snowmetamorphism,andthesurfaceenergybudget.Previous studieshighlighted thatCMIP5models simulatewell the relationshipbetween theborealtemperatureand the snowcoverextent in theNorthernhemisphere,butunderestimate thepolaramplification,inducinganunderestimationofthesnowcoverdecreaseoverthelastdecades.Becauseof a coarse resolution, snow cover is also poorly represented over mountainous areas in CMIP5models. TheCMIP5ensemble experiments suggest that themagnitude andeven the signof 30-yrregional trendsof snowcover canbe stronglymodulatedby internal variability. Initial snowcoveranalysisbasedonCMIP6modeloutputswillbepresentedduringtheworkshop,aspartoftheESM-SnowMIPandLS3MIPinitiatives.NewobservationaldatasetsprovidedbytheESACCI-snowprojectwillbeusedassoonastheybecomeavailable.

Keywords:satelliteobservations,CMIPmodels,climatevariability,snowcover


6-P20

Attributionoftheobservedintensificationofextremeprecipitationoverdryandwetregions

PaikSeungmok,MinSeung-Ki(1)

1-PohangUniversityofScienceandTechnology(SouthKorea)

Human-induced global warming is expected to accompany changes in hydrological cycle throughlower-troposphericwatervaporincreases,whichwouldresultinchangesinthespatialdistributionofprecipitation, potentially amplifying differences between dry and wet regions. Recently, extremeprecipitationincreasewasidentified,providingimportantimplicationsforriskoffloodingwithglobalwarming. Inthisstudy,weexaminecausesoftheobserved intensificationofextremeprecipitationoverthedryandwetlandareasforthepast60years(1951-2010)byapplyinganoptimalfingerprintingtechniquetoCMIP5andCMIP6multi-modelsimulations.Theobservedandmodelsimulatedannualmaximum of daily precipitation (Rx1day) is converted into a standardized probability-based indexrangingfrom0to1beforeanalysis,followingpreviousstudies.ResultsbasedonCMIP5multi-modelsshowthatanthropogenicsignalisdetectedintheobservedincreaseinextremeprecipitationoverbothdryandwetregions.Wefurtherdemonstratethatthedetectedanthropogenicinfluenceisseparablefromnatural forcing and that the counteracting drying influenceof anthropogenic aerosols is alsodetectedoverthewetregion.AfewavailableCMIP6modelsconfirmtheobservedintensificationofextreme precipitation over dry and wet regions for the extended period up to 2014. By utilizingindividual forcing simulations provided by the Detection and Attribution Model IntercomparisonProject(DAMIP),wewillconductanupdatedattributionanalysis.Also,resultswillbecomparedwithCMIP5-basedones,includingmodelskillsforextremeprecipitationamplitude,variability,andwet-dryregiondistributions.

Keywords:extremeprecipitation,detectionandattribution,dryandwetregions,anthropogenicforcing


6-P21

FutureevolutionoftheGreenlandicesheetinacoupledclimateandicesheetmodel(CESM-CISM)

MuntjewerfLaura(1),FykeJeremy(2)(3),LipscombWilliam(4),SacksBill(4),LöfverstromMarcus(5),SellevoldRaymond(1),ErnaniDaSilvaCarolina(1),BradleySarah(1),

PetriniMichele(1),VizcainoMiren(1)

1-DelftUniversityofTechnology(Netherlands),2-LosAlamosNationalLaboratory(UnitedStates),3-AssociatedEngineeringGroup,Ltd.(Canada),4-NationalCenterforAtmosphericResearch

[Boulder](UnitedStates),5-UniversityofArizona(UnitedStates)

Therearelargeuncertaintiesintheprojectionsof21stcenturyglobalmeansea-levelrise.Partoftheuncertaintystemsfromthelackofabilitytorobustlyresolveinteractionsandfeedbacksbetweenicesheetsandtheclimatesystem.Couplingan icesheetmodel toaclimatemodelwillhelpus in thisregard. TheIceSheetModelIntercomparisonProjectforCMIP6(ISMIP6)isaCMIP6-endorsedMIPthatfocussesontheGreenlandandAntarcticIceSheets.OneISMIP6goalisasetofexperimentswithclimatemodelscoupledtoicesheetmodels(ISM).Thiswaywecanassesstheclimaticimpactoftheicesheetdynamicresponseontotheclimate.TheAOGCM-ISMexperimentsareidenticalinset-uptothe CMIP6 AOGCM experiments. The Community Earth SystemModel version 2 (CESM2) has thecapabilitytocouplebi-directionallytotheCommunityIceSheetModel(CISM)overtheGreenlandIceSheet.Wecanassessicesheetprocessesandfeedbackmechanisms,suchastheice-albedofeedbackandmultipleelevationfeedbacks.Further,routingoficesheetmeltwatertotheoceanandadaptiveicesheettopographyenablesstudyingtheimpactofachangingicesheetonatmosphericandoceancirculationpatterns.Herewepresent preliminary results of the ISMIP6 effortswithCESM2with adynamicallyevolvingGreenlandIceSheetoftheDECK-runs:pre-industrialcontrol(piControl)steadystatesimulationandthe1%yr-1CO2concentrationincrease(1pctCO2)simulation.Focuswillbeonkey features of the global and polar climate simulation: the time evolution of the North AtlanticMeridionalOverturningCirculation (NAMOC)strength, theatmosphericandoceanmeridionalheattransport,Greenlandalbedo, surfacemelt and surfacemassbalance (SMB), and thedynamical icesheetchangeandcontributiontoeustaticsealevelriseoftheGreenlandIceSheet.

Keywords:ISMIP6,sealevelrise,Greenlandicesheet,AOGCM,ISM


6-P22

Seasonalamplification,phaseshift,anduncertaintiesforoceanacidityduringthe21stcentury

OrrJames(1),KwiatkowskiLester(2)

1-LSCE/IPSL(France),2-LMD/IPSL(France)

ComparisonofnineCMIP5EarthSystemModelsindicatesthattheamplitudeoftheannualcycleofsurface-oceanfreeacidity[H+]isprojectedtoincreaseby81±16%duringthe21stcenturyundertheRCP8.5 scenario despiteminor changes in the seasonality of sea surface temperature and salinity(KwiatkowskiandOrr,2018).Althoughtheannualcyclesofotheroceanacidificationvariablesarealsoaltered,theyexhibitsmallerchangesinamplitude.Forinstance,theannualcycleofsurface-oceanpHisattenuatedby16±7%,acounterintuitiveresultwhencomparedwiththechangeinamplitudeinfree acidity, which is larger and of opposite sign. The projected enhancement of the seasonalamplitudeofoceanacidityislikelytoexacerbatefutureimpactsonmarineorganisms,relativetothecaseofconsideringnochangeinseasonality.Atlowerlatitudes,marineorganismswouldbeexposedtorelativelyhigheracidityinthesummerandloweracidityinwinter;inthehigherlatitudes,itwouldbe the opposite. This CMIP5 analysis is nowbeing extended to the CMIP6models as their outputbecomes available. It is also being complemented by an assessment of shifts in seasonal phasing,showing in some regions complete phase reversals. The extended analysis also includes adecompositionofuncertaintiesfrominternalvariability,differencesbetweenmodels,anddifferencesbetweenscenarios,aclassicapproachbutonethatfocusesonchangesintheannualcycle,nottheannualmean.

Keywords:CMIP5,CMIP6,ocean,acidity,pH,CO2,uncertaintyanalysis,annualcycle,amplitude,phase


6-P23

Regionalanalysisofpresentandfuturemarineproductivity

PalmieriJulien(1),YoolAndrew(2),PopovaEkaterina(3),DeMoraLee(4),SellarAlistair(5),JonesColin(6),CoreGroupUkesm

1-NOC(UnitedKingdom),2-NationalOceanographyCentre[Southampton](UnitedKingdom),3-NationalOceanographyCentre(UnitedKingdom),4-PlymouthMarineLaboratory

(UnitedKingdom),5-MetOffice(UnitedKingdom),6-NationalCentreforAtmosphericScience(UnitedKingdom)

Thefutureevolutionofmarinebiogeochemistryiskeytounderstandinghowtheocean'sroleincarbonuptakeanditsprovisionofresourcesmaychange.Forinstance,anthropogenicdriversareexpectedtoinduce ocean warming, altering vertical stratification and upward nutrient supply, while oceanacidification and deoxygenation are expected to impact marine ecosystems. Underpinning thecredibilityoftheseforecastsistherealismofthemodel-simulatedpresent-day,inparticularthespatialandtemporalrepresentationofmarineproductivity.Hereweaimtoidentifypatternsinthedriversofregional productivity at the present-day. For instance, how well do models agree with observedregionaldistributionsofproductivity?Anddomodelsagreeontherelativeimportanceandrolesofunderlyingdrivingfactorssuchasoceancirculation,nutrientsandmixedlayerdepthforsettingthesepatterns?Extendingtothefuture,weaimfirsttoquantifyregionaltrendsacrossmodels,andthentoevaluatewhethertherearecommonunderlyingmechanismsforthesechangesare.Essentially,aretherecommonpatternsinchange,andarethesedrivenbythesamemechanisms,orareCMIP-classmodelsdivergentgiventhesamefutureforecast?Analysiswill initiallymakeuseofanensembleofCMIP6HistoricalandScenarioMIPsimulationsusingtheUKESM1model,withtheaimtoextendacrosstheCMIP6ensemble.

Keywords:oceanbiogeochemistry,regional


6-P24

AttributingtheIndo-Pacificwarmpoolexpansion:seasonalchangesanditsimpactsonprecipitation

ParkIn-Hong(1),MinSeung-Ki(1)

1-PohangUniversityofScienceandTechnology(SouthKorea)

The Indo-Pacific warm pool (IPWP) plays a critical role for global atmospheric circulation andhydrologicalcycle,andattributionoftheobservedchangesinIPWPintensityandareaisimportantforunderstandingglobalandregionalchangesinprecipitationpatternsandcirculation.Inthisstudy,weconduct an attribution analysis of the observed changes in seasonal IPWP during 1953-2012 bycomparingobservationswithCMIP5andCMIP6multi-modelsimulationsperformedunderdifferentexternal forcings including anthropogenic and natural factors. In order to consider model biases,modelsarefirstevaluatedintermsoftheseasonalclimatologyandaboutahalfofmodelsareselectedandusedfortheattributionanalysis.ObservationsshowthattheexpansionoftheIndianOceanwarmpoolhaslargelycontributedtothetotalIPWPexpansion,morestronglyinborealwinterandautumnseasons.Incontrast,theobservedchangesintheIPWPintensitywassimilaracrossseasonsandbasins.CMIP5modelsincludinganthropogenicforcingsarefoundtocapturetheobservedIPWPchanges,withsomeoverestimationinthePacificandunderestimationintheIndianOcean.AttributionresultsshowthattheanthropogenicsignalsaredetectedrobustlyintheIPWPareaandintensityincreasesforallseasons,indicatingthathuman-inducedgreenhousegaseswasthemajorcauseoftheseasonalIPWPgrowth.Further,theregionalimpactsofIPWPexpansionareanalyzedontheassociatedchangesinprecipitation. It is found thatmodelswith largerexpansion in the IndianOcean than in thePacificOcean, similar to theobserved, tend tohave rainfall increasesover thewestern IndianOceanandMaritimeContinentswhilemodelswithstrongerexpansioninthePacificOceanexhibitprecipitationdecreases over East Asia. CMIP6modelswill be further analyzed for the IPWP bias and trends incomparisonwithCMIP5models.

Keywords:detectionandattribution,anthropogenicforcing,precipitation,IndoPacificwarmpool


6-P25

DetectingchangesinNorthAtlanticvariabilityunderglobalwarming

PutrasahanDian(1),VonStorchJin-Song(1),JungclausJohann(1),GhoshRohit(1)


Under global warming, mean climate state in the North Atlantic and surrounding continents isexpected tochangeconsiderably.However,much less isknownaboutchanges to spatial-temporalcharacteristics of its internally-generated variability, modifications in their teleconnections, andsignificanceofthosechanges.Here,weaddressthesegapsinourunderstandingusingtheMaxPlanckInstitute Grand Ensemble (MPI-GE) that consists of 100-member ensembles for several scenarioexperiments.UsingMPI-GE,wecanquantifyexchangesininternalvariabilityovertimebyperformingrobuststatisticalanalysisalongtheensembledomain,suchasensemblevariance,lead-lagensemblecorrelation,EOFsinensemblespace,etc.Inaddition,alargeensembleenablesmorerobustestimatesof spectral characteristics and boosts confidence in spectral peaks. We present an evaluation ofvariability changes in frequency space for different timescales and provide analyses of evolvingrelationshipsbetweenthevariabilityofclimateindices(e.g.theNorthAtlanticOscillation,NAO)andregionalclimatevariability.Weusethehistorical,RCP8.5and1%CO2-increase-per-yearscenario,andfind that the North Atlantic undergoes substantial changes in the variability of ocean and air-seaexchange quantities, spatial patterns of dominant SST variability and teleconnections betweendifferentclimatequantities.Forexample,AMOCvariabilitydecreaseswithtime,aswithits4-yearlagcorrelationwithair-seaheatfluxovertheLabradorSea.DominantspatialvariabilitypatternssuchastheAtlanticMultidecadalVariabilityloosestheirpresent-daycharacteristics.Wealsofindthatwhilevariability of the NAOmay not change under global warming, its relationship to regional surfacetemperaturevariationsoverEurope,Africa,andNorthAmericaissignificantlyaffected,especiallyondecadaltobi-decadaltimescales.Thesefindingshaveimportantconsequencesfortheunderstandingofregionalclimatechange.Themodificationofspatialrelationshipsunderchangingexternalforcingaskforcautionwhenusingcurrent,well-establishedclimateindicesforassessmentofpastchangesandnear-termprediction.

Keywords:internalvariability,teleconnection,NAO,EOF,spectralcharacteristics


6-P26

Assessingco-behaviourofclimateprocessesoversouthernAfricausingCMIP5Models

QuagraineKwesi(1),HewitsonBruce(1),JackChristopher(1),PintoIzidine(1),LennardChris(1),WolskiPiotr

1-UniversityofCapeTown(SouthAfrica)

Using CMIP5models,we assess how co-behaviour of climate processes is captured inmodels. Asestablishedinourearlierresearch,co-behaviourdoesplayapivotalrole in influencingtheregionalclimateofsouthernAfrica.WeuseSelf-OrganizingMap(SOM)technique,whichisatypeofartificialneuralnetworkandhasalreadybeenintroducedtoclimatologytoclassifycirculationpatterns.ThenweapplyvarimaxrotatedPrincipalComponentAnalysis(PCA),amultivariatestatisticaltechniqueusedfrequentlyinstatisticsandclimatologytoidentifystronglyassociatedpatternsacrossthedatawhilebootstrapping their statistical significance. Through this, 8 possible modes of co-behaviour areidentified.We study co-behaviourbydeveloping climate indices from theCMIP5models for threeimportantprocessesofthesouthernAfricanclimate;ElNi–oSouthernOscillation(ENSO),AntarcticOscillation(AAO)andInter-TropicalConvergenceZone(ITCZ).Thecombinatoryimpactofthemodesof theseprocesses are thenmapped toprecipitation and surface temperaturepatterns across thedistinctregionsoversouthernAfricatobetterunderstandtheirinfluenceonsurfaceexpressions.

Keywords:co,behaviour,cmip5,ENSO,AAO,self,organisingmap


6-P27

DetectionandattributionofanthropogenicdynamicalandthermodynamicalcontributionsinextremeeventsoverEastAsiabasedonCMIP6DAMIP

RenLiwen(1),ZhouTianjun(1)

1-StateKeyLaboratoryofNumericalModelingforAtmosphericSciencesandGeophysicalFluidDynamics(LASG),InstituteofAtmosphericPhysics,ChineseAcademyofSciences(China)

Anthropogenicwarminghasbeenshowntodriverecentsummerextremesindifferentregionsoftheworld.Inthemidsummerof2018,anunprecedentedlylongandintenseheatwavehitnortheastAsia,setting a record for regional temperature. This heat wave was accompanied with an anomalousanticycloneovernortheastAsia,inducedbytheunprecedentednorthwardshiftofthewesternPacificsubtropicalhigh.Usingtheanalogue-basedattributionmethodandreanalysisdata,wefoundthattheatmosphericcirculationwasthemaintriggeringfactorofthisheatevent,whichcouldberesponsibleforthreefourthsofthetemperatureanomaly.Nevertheless,thermodynamicalchangesoftherecentdecadesmadeanirreplaceablecontributiontotheriskratioofheatevents,especiallyforthosemoreextreme events.With the above results,wewillmainly focus on the attribution of anthropogenicdynamical and thermodynamical contributions in extreme events over East Asia based on CMIP6DAMIP. We will firstly apply the analogue-based methods to multi-model historical and histNATsimulations in DAMIP Tier1 Experiment, to disentangle anthropogenic dynamical andthermodynamicalcontributionsofchangesinthelikelihoodofoccurrenceofthisextremeheatevents.We thenquantify the relative roleof greenhouse-gasandaerosol inanthropogenicdynamicalandthermodynamicalcontributionstoextremeheateventsbasedonhistGHGandhistAERsimulations.TheaboveattributionprocessesmayalsobeconductedforextremeprecipitationeventsinEastAsia.

Keywords:attribution,anthropogenic,dynamicalandthermodynamicalcontributions,extremeevents,EastAsia


6-P28

Trackingtheimpactofclimatemodelcomplexityinfutureclimateprojections

SeferianRoland(1),NabatPierre(1),MichouMartine(1),Saint-MartinDavid(1),VoldoireAurore(1),ColinJeanne(1),DecharmeBertrand(1),DelireChristine(1),BerthetSarah(1),

ChevallierMatthieu(1),SénésiStéphane(1),FranchisteguyLaurent(1),AliasAntoinette(1),CassouChristophe(2),CattiauxJulien(1),ChauvinFabrice(1),DouvilleHerve(1),

GeoffroyOlivier(1),GueremyJean-Fran�ois(1),MoineMarie-Pierre(2),MsadekRym(2),PlazzottaMaxime(1),VialJessica(1),RibesAurélien(1),RoehrigRomain(1),Salas-Y-MeliaDavid(1),

SanchezEmilia(2),TerrayLaurent(2),ValckeSophie(2)

1-Centrenationalderecherchesmétéorologiques(France),2-CERFACS[Toulouse](France)

Constraining future climate projections requires detailed information on the model response toexternalforcingwhicharisesfromthecombinationofmodelresolution(lowversushighresolution),tuning/calibration(parametersselection)andcomplexity(EarthsystemmodelsvsGeneralCirculationmodels). Here, we use a large variety of CMIP6 experiments (i.e., DECK, historical, AerChemMIP,RFMIP,C4MIP,ScenarioMIP)toinvestigatehowfarthemodelcomplexityimpactthemodelresponseto external forcing and hence future climate projections. To address this question, we use theAtmosphere-OceanGeneralcirculationmodelandEarthsystemmodeldevelopedatCNRM(namely,CNRM-CM6-1andCNRM-ESM2-1)thatofferafully-traceablenumericalframeworkbasedonthesamesuiteofphysicalparameterizationandmodelresolution.Weshowthatincreasingmodelcomplexitywith the inclusion of Earth system components (e.g., aerosols, vegetation or ocean chlorophylldistributions)impactmoreprominentlytheresponsetoexternalforcingthanthemodeledpresent-daymean state.Differences in present-day climatebetweenCNRM-CM6-1 andCNRM-ESM2-1 aremainlyexplainedbytheland-cover-aerosolsinteractionswheredifferencesinland-coverdistributionsimpact dust aerosols' loads leading in turn to biases in surface radiation and climate. Interactiveaerosols play a secondary role in explaining the difference betweenmodels' response to externalforcing; this latter isdrivenbyEarthsystemfeedbackssuchasCO2-stomata-water feedbackwhichinduceanegativeclimategainof-0.11,dampingglobalwarmingbyabout0.5¡C.

Keywords:Earthsystemmodels,traceability,feedbacks,carboncycle,aerosols


6-P29

Thelifetimeoffossil-fuelderivedcarbon

SierraCarlos(1),MetzlerHolger,MarkusMueller

1-MaxPlanckInstituteforBiogeochemistry(Germany)

Carbonreleasedtotheatmospherebythecombustionoffossilfuelsisexchangedwiththelandandoceancompartments,andeventually leavestheactivepartofthecarboncycle intimeframesthatspanbetweendecadestomillennia.Determiningtheamountoftimethatittakestoremovefossil-fuelderivedcarboncangiveimportantinsightsontheoveralltime-scalesofcarboncyclingintheEarthsystem,andcangivevaluableinformationtopolicymakerstobetteraddressclimatechangemitigationandadaptation.However,thequantificationoftimescalesofcarboncyclingfromEarthsystemmodelsisnottrivial,andgenerallyrequiresaspecialtypeofsimulationsinwhichthemodelsarespunuptoequilibriumandanimpulsiveimpulseisinjectedtothesystemtoobserveitsdynamicsoverthelong-term.Thismethodreliesonassumptionsoflinearityandsteady-state,andcannotbeusedfortransientclimatechangesimulations. Inarecentpublication(Metzleretal.2018,PNAS115:1150-1155),wedevelopedanew theoretical framework for theestimationof time-scales fornonlinearandoutofsteady-state conditions.We intend touse this framework to compute the time it takes to removefossil-fuelderivedcarbonfromtheEarthsystemfordifferentmodelsandtransientsimulationsfromtheCMIP6project.Inthiscontribution,wewillintroduceourcomputationalframeworkandthedatarequirementsneededforthesecomputations.

Keywords:Transittime,agedistribution,dynamicalsystems,statetransitionoperator


6-P30

Multi-modelanalysisoftheclimaticeffectsofidealizedglobaldeforestationexperiments

BoysenLena(1),BrovkinVictor(2),VuichardNicolas(3),LawrenceDave(4)

1-MaxPlanckInstituteforMeteorology(Germany),2-MaxPlanckInstituteforMeteorology(Germany),3-LaboratoiredesSciencesduClimatetdelƒnvironnement[Gif-sur-Yvette](France),


Pastmodellingstudiesshoweddiverseresultsregardingthesignsandmagnitudesofbiogeophysicaleffectsoflanduseandlandcoverchange(LULCC)onclimate.TherelativelysmallamplitudeofLULCCsin historical reconstructions and future scenarios and the various modelling strategies of theirimplementation in Earth SystemModels (ESMs) lead to incoherent signals acrossmodels.Wewillpresent first results of idealized deforestation experiments designed within the Land Use ModelIntercomparisonProject(LUMIP)toaddresstheabove-mentionedshortcomingsofpreviousstudies:Increasing the signal-to-noise ratio and guaranteeing a harmonized implementation acrossparticipatingESMs.Therefore,globalforestextentislinearlydecreasedby20millionkm2overaperiodof50yearsstartingfrompre-industrialclimateconditionsfollowedbyat least30yearsofconstantforestcover.Whileoceanandatmosphereconditionsmayadapttothebiogeophysicaleffectsofthislarge-scaledeforestation,other forcingssuchasatmosphericCO2concentrationarekeptconstant.First analyses of theMPI-ESM reveal statistically significant increases in tropical and decreases inborealnear-surface temperatures (Tas)alongwithuniformdecreasesofprecipitation in regionsofdeforestation. Interestingly,thesechangesinTasdevelopfromthecentre(first10-30years)tothemarginsofdeforestation(after40years),e.g.intheAmazonregion,hintingtotheinfluenceofnon-locallandcoverchangeimpactsonclimatethataddontothelocaldeforestationsignal.WeaimtoincludefurtheranalysesfromdifferentESMs(egIPSL,CESM)aswellasadvanceddetectionmethodsbiogeophysicalandbiogeochemicaleffects.

Keywords:landcoverchange,biogeophysicaleffects,biogeochemicaleffects,idealizedscenario


6-P31

What'supwithwhat'sgoingdown?Trendsinprimaryandexportproduction

YoolAndrew(1),PalmieriJulien(1),PopovaKatya(1),DeMoraLee(2)

1-NationalOceanographyCentre[Southampton](UnitedKingdom),2-PlymouthMarineLaboratory(UnitedKingdom)

Acriticalpartoftheocean'sinteriorcarboncycleisplayedbythemarinebiotaintheupperocean.There,carbonisfixedintoorganicmatterbyphytoplanktonand,whilemostlyrecycledtherethroughthe ecosystem, a fraction is exported into the ocean interior via the so-called biological pump,elevating theocean's natural carbon storage.However, under future change, it is anticipated thatmarineproductivitywillgenerallydecline,potentiallyaltering thesizeof thiscarbonreservoir,andwithimplicationsfordeepecosystems.Hereweanalysethedistributionofsurfaceproductivityandexport production, and the relationships that exist between both. This initially focuses on theidentifying present-day relationships and their driving factors, particularly across biogeochemicalprovinces(e.g.upwelling,oligotrophic,shallowwater,sea-iceandseasonally-mixedregimes).Thisisfollowed by tracking how these relationships evolve into the future under anthropogenic-drivenchange. Inparticular,howthebiologicalpumpchanges;whatthemechanismsdrivingthisare;andwhattheimplicationsfortheoceaninterior(includingO2andconsumercommunities).Suchchangeshavepotentialimplicationsforecosystemservicessuchascarbonstorageandfishproduction.AnalysiswillinitiallymakeuseofalimitedsetofCMIP6HistoricalandScenarioMIPsimulations,withtheaimtoextendacrosstheCMIP6ensemble.

Keywords:Marine,biogeochemistry,production,export

Session7-RegionalandImpacts

7-P01

Multi-ModelClimateVulnerability,ImpactsAndAdaptationAssessmentsOfExtremeOceanEventsInGulf-Of-GuineaCoasts

AkandeSamuel(1)

1–CentreforSpaceResearchandApplications,Akure(Nigeria)

Theclimatechangeimpactsoncoastalerosion,flooding,andsealevelrisesrequiresimulationswithvalidatedmodelsandmultipleclimateprojections.Thestudywasconducted in theGulfofGuinearegionof theNorthAtlanticwhichstretches fromtheSenegal in theWest toGabon to theSouth-EasternpartofAfrica.Itishometoalargenumberpeopleandsocio-economicactivitieswithoverfifty(50) million African population. The multi-model projection methodology is based on a statisticaldownscalingapproach.ThestudyusedCMIP6-basedclimateprojectionsapproachoverWestAfrica.Monthly and annual sea surface temperature datawere obtained from the National Oceanic andAtmospheric Administration (NOAA), ECMWF Model, NCEP-CFS2 over the Gulf-of-Guinea regionbetween 1972 to 2017 (45 years) were analysed in relation to three El Nino-Southern Oscillation(ENSO)phases.ToolsweredevelopedusingtheArcGIS10.5andIdrisiTerrsetsoftwaretoenableusersconduct riskanalysesof coastlinesbyclassifyingvulnerableareas,andalso toproduceoutputs forfurtheranalysis.Therecommendationsfromthisresearchthusenhancethecapacityandknowledgeofgovernmentsandotherpolicymakerswiththeexpectationsofclimatechangeandextremeevents.AcoastalspatialplanningdesignedandthentheIntegratedCoastalZoneManagementframeworkforWestAfricacoastalregionwasdeveloped.Theresultsfromthisstudywouldhelptoimproveweatherpredictionthusdevelopingtheeconomyaswellassavinglivesandpropertiesduringclimate-relatedhazards such as erosions, forest fires and floods. The Coupled Model Intercomparison Project 6(CMIP6)wasfoundtobetheidealframeworkwithinwhichtoconductsuchastudy,duetothestronginter-relationwithmodelsbeingdevelopedregionallyandglobally.Inaddition,theCMIP6provideamutualdialogbetweenclimatemodelersandexperts/userswiththeapplicationforawider(orglobal)coverage.

Keywords:Coastal,Model,Projections


7-P02

Inter-comparisonofthemassbudgetofArcticseaiceandsnowinCMIP6models

BlockleyEd(1),KeenAnn(1)


Seaiceisakeycomponentoftheglobalclimatesystemandaveryvisibleindicatorofclimatechange.Arcticsummerseaicecoverhasdeclinedbyover13%perdecadesincesatelliteobservationbegan,andthereismuchinterestinhowthisdeclinewillcontinueinthefuture.GlobalcoupledmodelsarearguablythebesttoolwehaveformakingpredictionsofArcticseaice,butgenerateawidespreadofprojectionsof21stcenturydecline.Comparingintegratedquantitieslikeiceextentandvolumeisnotsufficient to understand the reasons for differences in model projections. Hence it is becomingincreasingly clear that it is necessary to consider, compare and evaluate the underlying processescausing icegrowthanddecline,andhowtheyare likely tochange inawarmingworld. This inter-comparisonwilluseseaicebudgetdiagnosticsincludedintheSIMIPdatarequesttoevaluatethemassbudgetoftheArcticseaiceandsnowoveradefinedregionoftheArcticOceanfortheperiod1960-2100. Area-weighted monthly-mean budget terms representing the dynamic and thermodynamicprocessescausingiceformationandlosswillbecalculatedbyparticipatingmodellingcentres.Wewillcomparetheseasonalcycleofthesebudgetterms,overadefinedreferenceperiod,todeterminehowcloselymodelsagreeontherelativeimportanceoftheprocessescausinggrowthanddeclineofseaice.Wherepossible,budgetcomponentswillbecomparedwithobservations.Thenwewillcomparetheevolutionofthesetermsasthe icedeclinesduringthe21stcentury, toestablishthedominantchangesinthebudget,andtowhatextentthechangesarerobustacrossmodels.Sofar,10modellingcentreshaveexpressedinterestinjoiningthisinter-comparison.Herewepresentearlyresultsfromthosemodellingcentreswhohavecompletedtheappropriatemodelintegrations.

Keywords:Arcticseaiceprocessesintercomparison


7-P03

ImpactsofClimateChangeonAgriculturalSystems

CabreMariaFernanda(1),NunezMario(2)(3)

1-CONICET-UniversidaddeBuenosAires.CentrodeInvestigacionesdelMarylaAtmosfera(CIMA),BuenosAires,Argentina(Argentina),2-CONICET-UniversidaddeBuenosAires.Centrode

InvestigacionesdelMarylaAtm—sfera(CIMA),BuenosAires,Argentina(Argentina),3-UniversidaddeBuenosAires,FacultaddeCienciasExactasyNaturales,DepartamentodeCienciasdela

AtmosferaylosOceanos,BuenosAires,Argentina(Argentina)

Climate change is expected to impact considerably on agricultural production, activity of greatimportanceintheArgentineaneconomy.Inthiscontext,therehasbeenanincreasinginterestonhowclimatechangecouldaffectareasofagriculturalproductionandyieldcrops.Inthiscontext,perennialcropproduction, suchasviticulturalproduction, is consideredhighlyvulnerable toclimatechange.Recently,CabréandNunez(2018)haveassessedtheimpactsofclimatechangeonviticultureusingclimaticprojectionsfromtheIPSL-CM5A-MRmodelfortwotemporalhorizons(2015-2039,2075-2099)undertwoemissionscenarios(RCP4.5,RCP8.5)[submittedtoClimaticChange].Thismodelhasbeenchosenforbeingthebestforthestudyregionamong24climatemodelsfromCMIP5Project(3CN,2014).TheanalysiswascarriedouttakingintoaccountwhichcouldbepossiblegeographicalshiftsofArgentineanwinegrowingregionsResultsshowedthatasignificantsouthwestwarddisplacementofwinegrowing regions is projected to occur, mainly for 2075-2099 under RCP8.5 scenario. Thissouthwestward displacement may face both opportunities and/or challenges on Argentineanviticulture,becausesomewinemakingregionscouldbefavored;whileotherwinemakingregionscouldbedisadvantaged,especiallyby2075-2099underRCP8.5.Theseoutcomeswerethoughtintermsofchangesinvineyardslocation,varietiesselection,qualityandquantityofgrapevinesandcouldbeveryusefulfortheimpactcommunityrelatedtothewinesector.Inthiscontext,similarstudiesofimpactsofclimatechangeonviticultureoronanyagriculturalsystemcouldbeofvaluableimportanceforourcountry. Accordingly, the results of CMIP6models could be used for going onwith futureworks,validatingtheCMIP6ProjectmodelsfordifferentregionsinArgentina,choosingthebestresultsforpresentclimateandusingthemforfutureclimatewiththenewsocioeconomicscenariosinordertostudythepossibleimpactsoffutureclimateconditionsonagriculturalsystems.

Keywords:ClimateChangeImpacts,Argentina,IPSLCM5AMRmodel,emissionscenariosRCP4.5RCP8.5


7-P04

Predictionskillassessmentoflarge-scalevariabilityinfluenceinsummersoutheasternSouthAmericarainfallinmulti-modelCMIPdecadalpredictions

DiazLeandro(1),VeraCarolina(1),SaurralRamiro(1)

1-CentrodeInvestigacionesdelMarylaAtm—sfera(Argentina)

Theleadingco-variabilitymode(SVD1)betweensummersoutheasternSouthAmerica(SESA)rainfallandtropicalseasurfacetemperatures(SST)anomaliesoverthe1962-2013periodexhibitssignificantvariability ranging from the interannual scale to long-term trends. It showsa clear globalwarmingsignal,mainlyrelatedtowarminginthePacificandIndianOceans,inassociationwitharainfallincreasein SESA. After detrending the series, the spatial distribution of both SST and SESA precipitationanomalies associated with the first mode resembles that typically related with El Ni–o-SouthernOscillation(ENSO).Theobjectiveof thiswork is totoassessthepredictionskillofSVD1 indecadalhindcastsimulationsfromtheCoupledModelIntercomparisonProjectPhase5(CMIP5)andCMIP6(ifavailable)oftheWorldClimateResearchProgramme,usingdifferentmethodologiestodealwiththemulti-member/modelensemble.ThreemethodologiestoperformSVD1willbepresentedthatmadeuseofmulti-member/modelensembleinformationforpredictionskill:i)SVD1computedforthemulti-model mean anomalies; ii) SVD1 calculated after concatenating all ensemble members; iii) SVD1computed projecting all ensemblemembers in the spatialmodes obtained using themulti-modelmeananomaliesinmethodologyi).Methodologiesii)andiii)allowtoobtainprobabilisticpredictioninformation,sothatinternalvariabilityuncertaintiescouldbealsoassessed.Thethreemethodologieswereapplied tobothdetrendedandundetrendedanomalies. Itwas found that initializedCMIP5decadalhindcastsareabletorepresentSVD1spatialstructureswithandwithoutconsideringtrendsforthedifferentmethodologies,improvingresultsfromanalogousuninitializedsimulations.Althoughdetrended SVD1 activity shows skill in the first two prediction years, differences betweenmethodologieswillbediscussed.ThesefactsrepresentapromisingresultforthepredictionsofrainfallintheSESAregiononinterannualandlongertimescales.

Keywords:SouthAmerica,Ensemble,Decadalprediction,Covariability


7-P05

ProjectedtrendsofheavyrainfalleventsfromCMIP5modelsoverCentralAfrica

Fotso-NguemoThierryC.(1)(2),ChamaniRoméo(2)(3),YepdoZephirinD.(1)(2),SonkouéDenis(2),MatsaguimCédricN.(4),VondouDerbetiniA.(2),TanessongRoméoS.(2)(5)

1-ClimateChangeResearchLaboratory(CCRL),NationalInstituteofCartography(Cameroon),2-LaboratoryforEnvironmentalModellingandAtmosphericPhysics(LEMAP),Departmentof

Physics,FacultyofScience,UniversityofYaounde1(Cameroon),3-DepartmentofPhysics,FacultyofScience,UniversityofDouala(Cameroon),4-DepartmentofGeography,FacultyofArtsLettersandSocialSciences,UniversityofYaounde1(Cameroon),5-SchoolofWoodWaterandNaturalResources,FacultyofAgronomyandAgriculturalSciences,UniversityofDschang(Cameroon)

In this study, theprojectionsofdaily rainfall fromanensemblemeanof 20global climatemodels(GCMs)areusedtoexamineprojectedtrendsinheavyrainfalldistributionoverCentralAfrica,undertherepresentativeconcentrationpathway8.5.Forthispurpose,twoanalysesperiodsof40-yearshavebeenselected(2006-2045and2056-2095)tocomputetrendsinthe90thand99thpercentilesofthedailyrainfalldistributions.Wefoundthatlargeincreasetrendismostlyfoundinthe99thpercentileofrainfall events, over southernChad, northernCameroon,northern Zambia, and in theGreat LakesArea.ThiscanbeattributedtotheincreaseofmoistureconvergenceintensifiedbythepresenceoftheCongoBasinrainforest.ItisalsoshownthatthelargestnumberofGCMswithatrendofthesamesignastheaveragetrendisobservedovertheaboveregions.Itisthusclearthattheprojectedincreasetrends in heavy rainfall events may further worse floods which are real problems in the CentralAfrican's countries. Therefore, strong subregional policies are needed to help design effectiveadaptationandmitigationmeasuresfortheregion'scountries.

Keywords:CentralAfrica,CMIP5,heavyrainfall,futurechange,trends


7-P06

(willnotbepresented)

ExtremeeventsintheArcticandtheirassociationtolow-frequencyclimatevariabilitiesandseaicecoverchanges

LevineXavier(1),OrtegaPablo(2)

1-BarcelonaSupercomputingCenter-CentroNacionaldeSupercomputacion(Spain),2-BarcelonaSupercomputingCenter-CentroNacionaldeSupercomputacion(Spain)

Arcticclimatehasexperiencedarapidwarminginpastdecades,associatedwithanextensivelossinseaicecover.ThosechangeshavebeenassociatedwithunusualweathereventsovertheArcticandmid-latitudeclimates,whichhavebeenfoundtonegativelyimpactecosystemsandhumanactivitiesinborealregions.Yet linksbetweenthevariability inthoseextremeweatherevents(droughts/coldevents/heatwaves/extreme rain or snowfall thatmay account for anomalies in both duration andintensity) and changes in the seasonal-mean climate remain poorly understood. In particular, it isunclear whether a drastic reduction in sea ice cover--as observed in recent years-- can have asubstantialinfluenceontheoccurrenceofextremeweathereventsatmid-andhigh-latitudes.Here,Idescribe extreme weather events and their spatio-temporal variations in mid- and high-latitudesregions,inalargeensembleoffixedpresent-dayconditionEC-Earthsimulations.First,Idocumentandexplainlow-frequencyvariationsinextremeweathereventsatmidandhigh-latitudes,andthenseekto relate them to interannualordecadalmodesofArctic climate variability.Associationsbetweenextremeeventsandseaicecoveranomalies--occurringconcurrentlyorpriortothoseextremeevent--arequantifiedtoassessseasonalpredictabilityinbothpresent-dayandfutureclimateconditions.Tounderstandhowextremeeventsrespondtodecadalchangesinseaiceconditions,IthenanalyzeArcticclimate variability in long-run Pre-CMIP6 (H2020 PRIMAVERA) simulations in which sea ice coverevolves freely, and compare them to a large set of PAMIP simulations in which sea ice cover isprescribed and held invariant across years. Finally, model biases in extreme weather events areassessedbycomparingPre-CMIP6andPAMIPsimulationstoobservationsofthepastdecades.

Keywords:Arcticclimatevariability,extremeweatherevents,seaicecover


7-P07

FidelityoftheCASFGOALS-f3inrepresentationofsummerrainfallclimatologyandextremeprecipitationoverTibetanPlateau

LeiWang(1),BaoQing,LiuYimin,JinxiaoLi

1-InstituteofAtmosphericPhysics[Prague](China)

TheTibetanPlateau(TP)isknownasthethirdpoleoftheworld.Andthecomplexland-airinteractionsoveritplayessentialrolesintheformationandvariabilityofnearbygeneralcirculationpatterns,thusaffecttheregionalandevenglobalweatherandclimate,especiallytheEastAsianMonsoon.However,lotsof theclimatemodelsshowan insufficientperformanceto reproduce the land-air interactionsovertheTP.Oneofthepossiblereasonsisthecoarsemodelresolutions.Usinga9-yearsimulation,thisstudyevaluatestheperformanceofextremeprecipitationoverTibetanPlateauintheCASclimatesystemmodelFGOALS-f3withtwodifferenthorizontalresolutions(100kmand25km).FGOALS-f3-L(100kmresolution)participatesinDECKexperimentsforCMIP6,whileFGOALS-f3-H(25kmresolution)participatesinHighResMIPforCMIP6.Theevaluationisaboutthepossiblebenefitsoftheincreasedhorizontal resolutiononprecipitationover the TibetanPlateau and surrounding areas, particularlyfocusingonthesummerclimatologyandextremeprecipitation.ThepreliminaryresultsshowthatahigherresolutionishelpfultoreproducethebiasoftheprecipitationgradientoverTP.AsidentifiedinTRMMsatelliteproducts,FGOALS-f3-Hshowsthemoredetaileddistributionoftheprecipitationinthisregion,withasharpnorthwestwarddecreasinggradientofprecipitationthatstartssoutheastslopeofTP,whileFGOALS-f3-Lwith100kmresolutionfailedtoreproducethecharacteristics.BothFGOALS-f3-H and FGOALS-f3-L is able to reproduce the frequency of occurrence for extreme precipitation,although models slightly overestimate the heavy precipitation events. For example, both of theFGOALS-f3-HandFGOALS-f3-Lisabletocaptureextremeprecipitationup-to200mm/day.Finally,therolesofmodelphysicsandair-seainteractionarediscussedtoreducethebiasofprecipitationonthesouthslopeofTP.

Keywords:summerrainfallclimatology,TibetanPlateau,extremeprecipitation,f3,CASFGOALS


7-P08

ContributionoflanduseandlandcoveralterationstochangesinregionalsurfaceenergybalanceinCMIP6EarthSystemmodels.

MalyshevSergey(1),ShevliakovaElena


Landuseand landcoverchange(LULCC)hasbeen longrecognizedasoneof the important factorsaffecting near-surface climate. Biophysical effects of LULCC has been demonstrated to play animportantroleinwaterandenergyexchangesonregionalscales.TheeffectsofLULCCcanplayevenstrongerroleonlocalscales.Forexample,shortervegetationoncroplandsorpastureswill interactwith theatmospheredifferently than inaneighboring forest,even if thenear-surfaceatmosphericconditionsarethesameorsimilar.Local landsurfaceproperties,suchasalbedo,roughnesslength,leafareaindex,andsoilandcanopywateravailabilityareknowntoinducedifferencesinradiativeandturbulent fluxes and in near-surface climate. Previously, the analysis of local land-atmosphereinteractions was hindered by the unavailability of sub-grid scale information in climate and EarthSystemModels' output. Land Use Model Intercomparison Project (LUMIP) provides sub-grid landsurface climatedataby land-use type: cropland,pasture,urban,and secondaryplusnatural lands.Herewepresentmulti-modelanalysisofCMIP6DECK,historical,andLUMIP futureexperiments toexplore local-scale effects of LULCC on the near-surface climate, including local climate contrast[Malyshevetal.2015].Wecompareandcontrastenergybalancesinneighboringland-usetilesandanalyzephysicalmechanismsthatinducethedifferences.WecomparethelocaleffectsofLULCCfrommodelsimulationswiththeavailableobservationaldata.

Keywords:energybalance,landsurface,landuse,climatechange,surfacefluxes


7-P09

EvaluationoftheCNRM-CM6GlobalClimateModelsimulationoverWestAfricawithinCMIP6

MbayeMamadouLamine(1)

1-AssaneSeckUniversity(Senegal)

WeassessasetofhistoricalsimulationsoftheCNRM-CM6GlobalClimateModel(GCM)whichhasbeenusedinCMIP6.ExtremeprecipitationandtemperatureareanalysedinordertoinvestigatetheabilityofthisglobalclimatemodeltofaithfullyrepresentthepresentdayclimateoverSahel,GulfofGuinea, and West Africa. Such assessment is carried out through extreme climate indices incomparison with Global Precipitation Climatology Centre for precipitation (1988-2005) and withClimaticResearchUnitfortemperature(1981-2010).WefoundthattheCNRM-CM6modelreproduceswellthemainpatternsofconsecutivewetdays,consecutivedrydays,veryandextremelywetdays,and rainfall intensity during the summer season over West Africa. However, the model tends tooverestimatethemagnitudeofthesesclimate indices,exceptdrydays,overtheSahel. Inaddition,heavyrainfallandrainfallintensityareslightlyunderestimatearoundcoastalregions.Theannualcycleof over the Sahel, shows that the model captures well the temporal signal of these extremeseventhoughitexhibitsanoverestimationmainlyfromAugusttoNovemberandanundersestimationfromApriltoJuly.IntheGulfofGuinearegions,CNRM-CM6overestimateswetdays,rainfallintensityand the total rainfall during July-Agust-Septembre. As for temperature, the model generallyunderestimatesoverWestAfrica; thecoldbiasesaremorepronounced intheSahel.However, themodelpresentsthehighestvariabilitywithsubstantialwarmbiasinthesahelianregion.Furthermore,theseasonalcycleofmeantemperatureandthediurnaltemperaturerangeoverbothregionsshowsanunderestimation.ThisevaluationshowsthattheCNRM-CM6isabletorepresentthespatialandtemporalvariabilityofthepresentdayclimateoverWestAfricaeventhoughsomebiasesexist.Forclimatechangeimpactstudies,biascorrectionandstatisticaldownscalingcouldbeappliedtobrigethegapbetweentheGCMandimpactmodels.

Keywords:CMIP6,CNRM,CM6GlobalClimateModel,precipitaion,temperature,WestAfrica


7-P10

RainfallinMetUMoverCentralAfrica:Process-basedEvaluation

NdetatsinTaguelaThierry(1),MbaPokamWilfried(2),WhashingtonRichard(3)

1-LaboratoryforEnvironmentalModellingandAtmosphericPhysics,DepartmentofPhysics,UniversityofYaounde1,Yaounde,Cameroon.(Cameroon),2-LaboratoryforEnvironmentalModellingandAtmosphericPhysics,DepartmentofPhysics,UniversityofYaounde1,Yaounde,Cameroon.(Cameroon),3-SchoolofGeographyandtheEnvironment,UniversityofOxford

(UnitedKingdom)

Climatemodelsaretoolusedforinvestigatingtheresponseoftheclimatesystemtovariousforcingsand formaking projections of future climate. Unfortunately, due to little attention that has beendevoted to Central Africa, only fewmechanisms governing climate in that region are known. Thisgenerally lead tomisrepresentationof climateparameters inmodels over the region. Thepresentstudy shows that, latest versions of the Met Office Unified Model (MetUM), a modeling systemdevelopedandusedattheMetOffice,depictwet(dry)biasovertheCongoBasin(alongtheAtlanticcoast south) inSeptember-November season.Both localand remotedriversofverticalmotionareused here to explain those biases. Results show that, dry bias over the coastal area is due to anoverestimationofmoisturedivergenceoverthatarea.Furthermore,convectionmotionislessdeep,thusfavoringanenlargementintothelandofoceansubsidenceleadingtodryconditions.OverCongoBasin, wet bias is due to an underestimation of the AEJ-S intensity, witch is favorable to moreconvection motion, then more precipitation. Moreover, in the MetUM, the Congo Basin Walkercirculation shift a little eastward in SON season and intensify precipitation over the region.PrecipitationoverCongoBasinisalsolinkedtoIndianOceandipolemode.Itappearthat,inMetUM,the Indian Ocean dipolemode is not well represented, leading to a weakening of the downwardbranche of Indian east-west overturning circulation. Wind over the Indian Ocean is then movingwestward,carringwithhimmoisturewhichcontibutetotheoverestimationofrainfallovertheCongobasin.

Keywords:RainfallBias,MetUM,Verticalmotion,CongoBasinWalkercirculation,IndianOceandipole


7-P11

Howdynamicaldownscalingcanadvanceourunderstandingoflarge-andlocal-scaledriversofregionalclimatechange

NikulinGrigory(1),KjellstromErik(1)

1-SwedishMeteorologicalandHydrologicalInstitute(Sweden)

Inrecentyearsmoreandmoreregionalclimatemodel(RCM)simulationshavebeengeneratedwithintheCoordinatedRegionalclimateDownscalingExperiment(CORDEX)framework.ThefirstanalysisoftheCORDEXensemblesshowsthatRCMsimulationsandtheirrespectivedrivingglobalclimatemodel(GCM)simulationsmayprojectcontradictingclimatechangesignalsonregionalscales,forexampleinprecipitation. Moreover, it was found that different RCMs downscaling the same GCM may alsoproducecontradictingfutureprojectionsinprecipitation.Fromoneside,onecanarguethatdynamicaldownscalingincreasestherangeofuncertaintiesinfutureclimateprojections.However,analternativeinterpretation is that such contradictions in future projections provide a useful insight intouncertainties in regional climate change. It stands clear that local-scale processes definedbyRCMparameterisationscanbemuchmoreimportantthanlarge-scaledriversdictatedbythedrivingGCMs.We use a number of ‘contradicting examples’ based on GCM and Euro- and Africa-CORDEX RCMsimulationstoidentifylocal-scaleprocessesresponsibleforthecontradictions.Wealsodiscusshowsuch‘contradictingexamples’canhelpinimprovementofbothglobalandregionalclimatemodels.

Keywords:downscalingclimateuncertaintiesregionalCORDEX


7-P12

ChallengesforBrazilianEarthSystemModel(BESM)

PereiraBruno(1)(2)

1-LaboratoryofGeoecologyofLandscapeandEnvironmentalPlanning-FederalUniversityofCear‡(Brazil),2-OceanographicInstitute,UniversityofSaoPaulo(Brazil)

In the context of global climate change as evidenced by the reports IPCCs numerical models arefundamental instruments to detect and predict climate anomalies. Computational tools based onnumericalmodelingoftimeseriesofenvironmentaldataaredevelopedtoestimatethefrequencyofextreme climate events. The National Institute for space research of Brazil (INPE) developed theBrazilianmodeloftheEarthsystem(BESM)toprovideinformationaboutchangesinglobalclimaticprocessesandparticularphenomenainthesouthernhemisphere.TheBESMmodelisabletopredictthefrequencyandmagnitudeofseasonalclimateeventsonatimescalethatcanreach100years.Theforecasts generated are basis for guidelines for the Territorial Planning and creation of policie formitigation and adaptation strategies to climate change and extreme events. The BESM programinsertedBrazilintheframeoftheprogramCMIP5,basedincomparisonbetweenmeasurementsthemodels. Among the current challenges in the development of themodel BESM's task of creatingsimulationstoverifywhicharetheextremesthatcantransformclassicprocessessuchasanomaliesofseasurfacetemperaturesintheSouthAtlanticandyourpossibleinfluenceonrainfall,thermohalinecirculation,heattransportbetweentheoceansandatmosphereandair-seaCO2flux,oceancarbonflux and sea ice dynamics. Current BESMversionneed incorporate biogeochemicalmodels to findanswersaboutthebehaviourofecosystemstofaceclimatechange.TheBESMprogramcontributetodevelopquestionsaboutwhataretheeffectsofprocessessuchastheincreaseofthetemperatureintheatmosphereandintheoceanandyourrelationshipwithothercomponentsoftheEarthsystem.

Keywords:computacionaltools,globalclimatechange,numericalmodelling


7-P13

Process-basedmodelevaluationandprojectionsoversouthernAfricafromregionalandglobalclimatemodels

PintoIzidine(1),JackChristopher(1),LennardChris(1),HewitsonBruce(1)

1-UniversityofCapeTown(SouthAfrica)

Decision-scalerelevantclimateinformationonclimatechangeisneededtoinformpolicyanddecisionmakingbutofteninvolveshighuncertaintyduetointernalvariability,emissionsscenariosanddifferingmodelrepresentationsofclimateprocessesatdifferentspatialscales.Inordertoincreaseconfidencein future climate changes it is valuable to disaggregate the inter-model differences into regionalcirculationscalesandlocalprocessscales.Thisisparticularlyvaluablewheninterrogatingdynamicallydownscaled ensembles such as Coordinated Regional Climate Downscaling Experiment (CORDEX)where differences between changes projected by the general circulation model (GCM) and theregional climate model (RCM) are frequently observed. This study explores a methodology toinvestigateprojectedchangeasafunctionofchangesinfrequencyofsynopticcirculationwiththeaidofSelf-OrganizingMaps(SOM).TheapproachexamineshowdynamicallydownscaledfutureclimatefromtwoRCMsoftheCORDEX,drivenwithfourGCMs,cangiverisetosurfaceclimatechangesthatdifferfromthoseofthedrivingGCMs.Thestudyfocusesonchangesinprecipitationandthecirculationprocessesdriving theprojected changes from the regional climate simulations. Results shows thatatmosphericcirculationisrelativelywellsimulatedinbothRCMsandGCMs,i.e,theRCMandGCMareinphaseregardingtocirculationpatterns.Andallofthemprojectsimilarchangesinatmosphericstatesforthisregion.Thisimpliesthatmuchofthedifferenceintheprojectedrainfallcouldratherbeduetothe representation of local subgrid-scale parameterized processes, such as convection and/or therepresentation of coastlines and topography. This suggests that over southern Africa the climatechangesignalcouldbealocalresponsedynamicratherthancirculationdynamic.

Keywords:CORDEX,CMIP5,precipitationchanges,SOM,southernAfrica


7-P14

AnalysisoffuturechangesinextremeclimateindicesinIndonesiaregionusingAIMS

PutraIDewaGedeArya(1),RosidSyamsu(2),SopaheluwakanArdhasena(1),ChoJaepil(3)

1-IndonesianAgencyforMeteorologyClimatologyandGeophysics(BMKG)(Indonesia),2-UniversityofIndonesia(UI)(Indonesia),3-APECClimateCenter(APCC)(SouthKorea)

Climatechangehasbroughtgreatenvironmental impactsthathavecausedeconomic lossesduetoextreme climate phenomena such as floods, droughts and changes in rain and temperaturecharacteristics in Indonesia. Information on projections of changes in rainfall characteristics andtemperature trends is challenging and it is very important to make adaptation, mitigation andoperationalplanningforvarious impactedsectors. Inthisstudyweuseobservationaldatafrom50observationweatherstations in Indonesiafor30yearsfrom1981to2010whichhavebeenqualitycontrolled.Rawmodeldatafrom27historicalGCM(GlobalCirculationModel)areanalyzedbasedonspatial pattern similarity and temporal patterns with patterns of station's observation. Spatialreproducibilityisevaluatedusingvariogramanalysis,andwhiletheuncertaintyrangesisobtainedbyincreasing the number of GCMs. Future changes in rain characteristics and temperature wereexamined use Bias-Correction of SQM (Simple Quantile Mapping) of the CMIP5 (Coupled ModelIntercomparisonProjectPhase5) for (RepresentativeConcentrationPathway)RCP4.5andRCP8.5Scenarios.Futurechanges investigatedintermsofextremeclimate indicatorsbasedontheETCCDI(Expert Team on Climate Change Detection and Indices), including annual total precipitation,consecutivedrydays,consecutivewetdays,maximummonthlymaximumvalueofdailytemperatureandmonthlyminimumvalueofdailyminimumtemperature,forallregionsinIndonesiausingtheMME(Multi-ModelEnsemble).Theoutputsof27GCMswerestatisticallyderivedusingAIMS(APECClimateCenter Integrated Modeling Solution), and we analyze the probability values of extreme climateindicators for theassessmentof climate change impacts in theperiods2011-2040,2041-2070and2071-2100.

Keywords:futurechanges,uncertainty,extremeclimateindices,reproducibility,AIMS


7-P15

IntercomparisonofSea-IceObservationalandCMIP6multi-modeldatasets

RanaArun(1),MassonnetFrancois(2)

1-UniversitécatholiquedeLouvain,EarthandLifeInstitute,GeorgesLemaîtreCentreforEarthandClimateResearch(Belgium),2-UniversitécatholiquedeLouvain,EarthandLifeInstitute,Georges

Lema”treCentreforEarthandClimateResearch(Belgium)

Inthisstudyweinvestigatepossibleuncertainties(andtheircause)anddifferencesinvarioussea-iceobservationalandCMIP6historicaldatasets,includingsea-iceconcentration,areaandextent.Weuseseveralobservationaldatasets(namely,NationalSnow&IceDataCentre-NSIDC,OceanandSeaIceSatellite Application Facility - OSISAF, European Space Agency Climate Change Initiative - ESACCI,IntegratedClimateDataCenter-ICDC,NationalOceanicandatmosphericAdministration-NOAAandMetOfficeHadleyCentre-UKMO)andcomparethemtoCMIP6multi-modelhistoricaldatasets(asmanyaspossiblesubjecttoavailability)frommodeloutput.ThefocusofouranalysisisonboththeNorthern and Southern Hemispheres. The products differ by the retrieval methods that wereemployedaswellasbytheirhorizontalresolution(varyingfromabout0.1_to0.25_).Weillustratedifferencesanduncertaintiesoverthehistoricalperiodbyapplyingseveralunivariate(climatologies,trends, standard deviations) and multivariate (per-pair root mean square error - RMSE, principalcomponentanalysis?PCA,Singularvaluedecomposition?SVDanddendrogramsamongstothers)statisticalindices.Thisapproachallowsustoi)evaluatetheabilityofeachofthedatasetstosimulatetheobservedclimatology in regionsof focus, ii) analyze similaritiesandcontradictions inhistoricaldataset between multiple observational datasets and model output and iii) estimate the level ofobservationaluncertaintyandgaugeitsimpactinclimatemodelevaluationstudies.

Keywords:Intercomparison,Icearea,Sea,Iceconcentration,Sea,Observations,CMIP6


7-P16

EvaluationandprojectedchangesindailyrainfallcharacteristicsoverCentralAfricabasedonamulti-modelensemblemeanofCMIP5simulations

SonkouéDenis(1),MonkamDavid(2),Fotso-NguemoThierryC.(1)(3),VondouDerbetiniA.(1),YepdoZéphirinD.(3)(1)

1-LaboratoryforEnvironmentalModelingandofAtmosphericPhysics(LEMAP),DepartmentofPhysics,FacultyScience,UniversityofYaounde1(Cameroon),2-DepartmentofPhysics,Faculty

Science,UniversityofDouala(Cameroon),3-ClimateChangeResearchLaboratory(CCRL),NationalInstituteofCartography(Cameroon)

Thisstudyusesdailyrainfalldatafromglobalclimatemodelssimulations,participatinginthephase5oftheCoupledModelIntercomparisonProject(CMIP5)andeightrainfallindicesdefinedbytheExpertTeam on Climate Change Detection and Indices (ETCCDI), to investigate the changes in extremeweatherconditionsoverCentralAfrica(CA)undertherepresentativeconcentrationpathway8.5.Theperformanceofthemulti-modelensemble(MME)meanwhichinfactreferstothebestperformingmodelsselectedthroughtheTaylordiagramanalysiswasevaluatedbycomparingwithtwogriddedobservation datasets during the historical period (1998-2005). Results show that although someuncertainties may exist between the gridded observation datasets,MME consistently outperformindividualmodelsandreasonablyreproducedtheobservedpatternofdailyrainfall indicesovertheregion.Theassessmentoftheclimatechangesignalinthoseindiceswasdoneforthemidandlatetwenty-firstcentury(2026-2056and2066-2095respectively),relativetothebaselinehistoricaltimeperiod (1976-2005).Wefoundasignificant increase inPRCPTOToversouthern (northern)CA fromDecember to February (from September toNovember). This ismainly due to the increase of highintenserainfalleventsratherthantheirfrequency.TheresultsalsorevealthattheincreaseinPRCPTOTwascoupledwithincreaseinRX5DAY),R95,andinR95PTOT,withmorerobustpatternsofchangeatthelatetwenty-firstcentury.Theincreaseinextremerainfalleventsislikelytoincreasefloodrisks.Also, thedecrease inCWDandPRCPTOTcoupledwith the increase inCDDobservedcouldworsendrought risk and significantly disrupt priority socio-economic sectors for development. For that,decision-makersareinvitedtoseriouslyconsideradaptationandmitigationmeasures,inordertolimittherisksofnaturaldisastersthatCAcountriesmaysufferinthefuture.

Keywords:CentralAfrica,Globalclimatemodel,CMIP5,Extremerainfallevents,Futurechange


7-P17

HowdifferentArcticdoweseefromCMIP6models?

WangMuyin(1),OverlandJames

1-UniversityofWashington(UnitedStates)

Arctic has been changing fast with its temperature increase at twice the rate of the northernhemisphereandglobalmean.Rapiddeclineofsea-icecover(andsea-icethickness)isanothersymbolicchange that has been observed in the past twodecades. Both CMIP3 andCMIP5model generallyunderestimated the rapid decline of Arctic sea-ice extent, yet a dozen of them do simulate theclimatologicalmeansandmagnitudeofseasonalcycleingoodagreementwithobservations.Inthisstudy we will investigate all available CMIP6 models on their performance of sea-ice simulations(mainlyonsea-icecover)intheirhistoricalrunsandpre-industrialcontrolrunstoaddressthefollowingquestions:1)doCMIP6performbetterthanCMIP3andCMIP5whencomparedwithobservationsintheirhistoricalruns?2)Howmuchofthemodeluncertaintiesarecontributedbytheinternalvariabilitywhichwill bedetermined from thepre-industrial control simulations.Wewill then investigate theprojectionsofthefuturesea-icecoverfromthesemodels,andtofindouthowdifferentofcurrentCMIP6modelprojectionscomparingwithpreviousCMIPs.Thisstudywilltrytoidentifytheoriginsandconsequencesofsystematicmodelbiasesintheseaicesimulations.

Keywords:projections,seaice,Arctic,modeuncertainties


7-P18

SubstantialincreaseinthejointoccurrenceandhumanexposureofheatandhazehazardsoverSouthAsiainthemid-21stcentury

XuYangyang(1)

1-TexasA&MUniversity(UnitedStates)

Greenhousegasemissions, ifnotabatedaggressively,will lead to large increases in frequencyandintensityofheatextremes.Anothermajorhealthhazardisdegradedairquality,leadingtomillionsofpremature deaths yearly (Lelieveld et al., 2015;WHO, 2017). An integrated assessment of humanexposuretothejointoccurrenceofheatandhazeextremesandthefuturechangeshasbeenmissing(except for fewstudiesat local scales;Dohertyetal.,2009; Jacksonetal.,2010).Basedonahigh-resolution decadal-long model simulation using a state-of-the-science regional chemistry-climatemodel that is bias-corrected againstmeteorological reanalysis and evaluated thoroughly, hereweshowthat:whendailyaveragewet-bulbtemperatureof25°Cistakenasthe‘deadly’threshold(Moraetal.,2017),heatextremefrequencyaveragedoverSouthAsiaincreasesfrom48±4days/yearduringDecade 2000 (1997-2004) to 76±3 days/year in Decade 2050 (2046-2054) under RepresentativeConcentrationPathway8.5(RCP8.5)scenario,alongwitha83%increaseinthemeandurationofheatevents.ThehumanexposuretohazyweatheroverSouthAsiaisalsoprojectedtoincreasesubstantiallyunderRCP8.5,incontrasttoprojectedairqualityimprovementglobally.WithdailyaveragedPM2.5surface concentrationof60?g/m3definedas the threshold for such “unhealthy”extremes (CPCB,2009), the area-weightedaverageof hazeextremeswouldoccur 132±8days/year inDecade2050under RCP8.5, which is 64-76% higher than Decade 2000. Even more concerning is the modelprojectionofthejointoccurrenceofheatandhazyhazard(HHH).Theserareeventswouldhavelargeincreasesinthefuturewitha160%increaseinfrequencyanda450%increaseinduration.Aholisticviewofhealthimpactofheatandhazehazard(HHH)isthusurgentlyneeded.

Keywords:impact,model,scenarios


7-P19

FidelityoftheObservational/ReanalysisDatasetsandGlobalClimateModelsinRepresentationofExtremePrecipitationinEastChina

YangJing(1)

1-StateKeyLaboratoryofEarthSurfaceProcessandResourceEcology,andAcademyofDisasterReductionandEmergencyManagementMinistryofCivilAffairsandMinistryofEducation,Facultyof

GeographicalScience,BeijingNormalUniversity(China)

Realistic reproductionofhistoricalextremeprecipitationhasbeenchallenging forbothreanalysisandglobalclimatemodel(GCM)simulations.Thisworkassessedthefidelitiesofthecombinedgriddedobservationaldatasets,reanalysisdatasets,andGCMs[CMIP5andtheChineseAcademyofSciencesFlexibleGlobalOcean”AtmosphericLandSystemModel”Finite-VolumeAtmosphericModel,version2(FGOALS-f2)]inrepresentingextremeprecipitationoverEastChina.Theassessmentused552stations'rain gauge data as ground truth and focused on the probability distribution function of dailyprecipitation and spatial structure of extreme precipitation days. The TRMM observation displayssimilar rainfall intensity frequency distributions as the stations. However, three combined griddedobservational datasets, four reanalysis datasets, andmost of the CMIP5models cannot captureextreme precipitation ex- ceeding 150 mm day21, and all underestimate extreme precipitationfrequency.Theobservedspatialdistributionofextremeprecipitationexhibitstwomaximumcenters,located over the lower-middle reach of Yangtze River basin and the deep South China region,respectively.CombinedgriddedobservationsandJRA-55capturethesetwocenters,butERA-Interim,MERRA,andCFSRandalmostallCMIP5modelsfailtocapturethem.Thepercentageofextremerainfallin the total rainfall amount is generally underestimated by 25%75% in all CMIP5models.Higher-resolutionmodelstendtohavebetterperformance,andphysicalparameterizationmaybecrucialforsimulatingcorrectextremeprecipitation.Theperformancesaresignificantly improved inthenewlyreleasedFGOALS-f2asaresultofincreasedresolutionandamorerealisticsimulationofmoistureandheating profiles. This work pinpoints the common biases in the combined gridded observationaldatasetsandreanalysisdatasetsandhelpsto improvemodels'simulationofextremeprecipitation,whichiscriticallyimportantforreliableprojectionoffuturechangesinextremeprecipitation.

Keywords:Extremeprecipitation,Highresolution,CMIP5,CMIP6,FGOALSf3L


7-P20

AreClimatemodelsreliableinprojectingtheimpactsofhalf-degreewarmingincrementonheatextremesoverChina?

ZhaoSiyao(1)(2),ZhouTianjun(1)(2)

1-LASG,InstituteofAtmosphericPhysics,ChineseAcademyofSciences(China),2-UniversityofChineseAcademyofSciences[Beijing](China)

Projectingtheclimateimpactsofhalf-degreewarmingincrementhasbeenonthehighestlevelonthepost-ParisAgreementscienceagenda,butclimatemodelprojectiongenerallyhasuncertaintiesanditsreliability needs to be assessed. As the real world has already witnessed a 0.5¡C global warmingincrement,wehavecomparedtheimpactsofthepasthalf-degreewarmingontheheatextremesinChinaderivedfromobservationalrecordstotheCESMhistoricalsimulationandCESMlowwarmingexperiments futureprojection.We find thatunder theperspectiveof spatiallyaggregate, theheatextremeschangeoverChinaunder0.5¡Cwarmingincrementaredetectableinobservationrecords.The changes are generally reasonably captured in the CESMhistorical simulation butwith slightlyweakeramplitude.Thechangesofintensityindicesinobservationalrecordsarebetteranaloguestothesimulationthanthefrequencyanddurationindices.Thedaytimeextremeschangeinobservationalrecords can serve as conservative estimations for future projection, while the nighttime extremeindicesshowcomparableorslightlyweakerchangesinthefuturethantheobservations.Weplantoapply thismetricon theGMMIPexperimentsbycomparing theheavyprecipitationchangesunderhalf-degree GMST increment over the global monsoon regions in the HadEX2 and GHCNDEXobservationaldatasetstotheGMMIPmodelsimulationsandthenevaluatethefutureprojections.WealsoplantousethismetricontheCORDEXtoevaluatethemodelabilityinsimulatingthehalf-degreeGMST increment impacts on regional climate extremes including heat extremes and heavyprecipitation,andthenassessthereliabilityoffutureprojectiononaspatialaggregateperspective.Webelievethatcloseresemblancebetweentheobservationalrecordsandthehistoricalsimulationcanaddconfidencetofutureprojectionforclimatechangeadaptionandriskmanaging.

Keywords:regionalclimateextremes,halfdegreewarming,CORDEX,GMMIP


7-P21

EvaluationofCMIP6modelsinthecontextofPrecipitationovertheTibetanPlateau

ZhaoYin(1)

1-LASG,InstituteofAtmosphericPhysics,ChineseAcademyofScience[Beijing](China)

TheTibetanPlateau(TPhereinafter)haveshoweditsimportancetotheAsianhydrologiccycle,aswellasglobalatmosphericcirculation.Butasaresultofthesteepandcomplexterrain,theobservationon TP was insufficient and data from reanalysis and model simulations became preference. WecomparedmultiplesatelliteandreanalysisdatainthecontextoftotalprecipitablewateronTP.Andbasedonit,thespacialuneventrendofprecipitationonTP,characterizedbythedryingsoutheasternplateauandthewettingnorthwesternplateau,wasfurtheranalyzed.Applyingmoisturebudgetandverticalvelocitydiagnosticformula,wefoundthatthedryingsoutheasternTPwascausedbyabnormalcoldadvectionandthewettingnortheasternTPwasaresultofdiabaticfeedback.PreviousstudieshaveshownthatmostmodelshaveoverestimatedtheprecipitationonTP.Onthebasisoftheaboveresults,wewillfocusonthesimulationabilityofCMIP6modelsincontextofTPprecipitationusingthehistoricalsimulationexperiments.WewillevaluateandgrouptheCMIP6modelsaccordingtotheirsimulationabilityofTPprecipitation.Thenthereasonfordifferencewillbeanalyzedutilizingclimatedynamicsdiagnosticmethodswhichwasmentionedabove.Asaresult,wewillfirstlyprovidesomeinformationaboutwhethertheabilityofCMIP6modelshavebeenimprovedincontextofprecipitationsimulationoncomplexterrainareas,thensomeoriginsandconsequencesofsystematicmodelbiaseswillalsobefound,inordertohelpmodeldeveloperstoimprovethemodels.

Keywords:modelevaluation,precipitation,Tibetanplateau,topographicprecipitation,modelbias

abstract book for poster presentations · analysis of the ipsl-cm6-lr ensemble of historical...

Documents