能源分析使用的全球氣候模型評估-Evaluation of Global Climate Models for Use in Energy Analysis

ThisreportisavailableatnocostfromtheNationalRenewableEnergyLaboratory(NREL)at/publications.

ContractNo.DE-AC36-08GO28308

EvaluationofGlobalClimateModelsforUseinEnergyAnalysis

GrantBuster,1SlaterPodgorny,1LauraVimmerstedt,1BrandonBenton,1andNicholasD.Lybarger2

1NationalRenewableEnergyLaboratory

2U.S.NationalScienceFoundationNationalCenterforAtmosphericResearch

NRELisanationallaboratoryoftheU.S.DepartmentofEnergyOfficeofEnergyEfficiency&RenewableEnergy

OperatedbytheAllianceforSustainableEnergy,LLC

TechnicalReport

NREL/TP-6A20-90166August2024

NationalRenewableEnergyLaboratory

15013DenverWestParkwayGolden,CO80401

303-275-3000?

ThisreportisavailableatnocostfromtheNationalRenewableEnergyLaboratory(NREL)at/publications.

ContractNo.DE-AC36-08GO28308

EvaluationofGlobalClimateModelsforUseinEnergyAnalysis

GrantBuster,1SlaterPodgorny,1LauraVimmerstedt,1BrandonBenton,1andNicholasD.Lybarger2

1NationalRenewableEnergyLaboratory

2U.S.NationalScienceFoundationNationalCenterforAtmosphericResearch

SuggestedCitation

Buster,Grant,SlaterPodgorny,LauraVimmerstedt,BrandonBenton,andNicholasD.

Lybarger.2024.EvaluationofGlobalClimateModelsforUseinEnergyAnalysis.Golden,CO:NationalRenewableEnergyLaboratory.NREL/TP-6A20-90166.

/docs/fy24osti/90166.pdf.

NRELisanationallaboratoryoftheU.S.DepartmentofEnergyOfficeofEnergyEfficiency&RenewableEnergy

OperatedbytheAllianceforSustainableEnergy,LLC

TechnicalReport

NREL/TP-6A20-90166August2024

NOTICE

Thisworkwasauthored[inpart]bytheNationalRenewableEnergyLaboratory,operatedbyAllianceforSustainableEnergy,LLC,fortheU.S.DepartmentofEnergy(DOE)underContractNo.DE-AC36-08GO28308.FundingprovidedbytheDOEOfficeofEnergyEfficiencyandRenewableEnergy(EERE),theDOEOfficeofElectricity(OE),theDOEOfficeofFossilEnergyandCarbonManagement(FECM),andtheDOEOfficeofCybersecurity,EnergySecurity,andEmergencyResponse(CESER).TheviewsexpressedhereindonotnecessarilyrepresenttheviewsoftheDOEortheU.S.Government.

ThisreportisavailableatnocostfromtheNationalRenewableEnergyLaboratory(NREL)at

/publications.

U.S.DepartmentofEnergy(DOE)reportsproducedafter1991andagrowingnumberofpre-1991documentsareavailable

freevia

www.OSTI.gov.

CoverPhotosbyDennisSchroeder:(clockwise,lefttoright)NREL51934,NREL45897,NREL42160,NREL45891,NREL48097,NREL46526.

NRELprintsonpaperthatcontainsrecycledcontent.

iii

ThisreportisavailableatnocostfromtheNationalRenewableEnergyLaboratoryat/publications.

TableofContents

TableofContents iii

ListofFigures iii

ListofTables v

1Abstract 1

2Introduction 1

3DataandMethods 2

4ResultsandDiscussion 8

5Conclusion 16

ListofAcronyms 17

CodeandDataAvailability 17

Acknowledgements 18

References 19

ReferencesforGCMs 22

AppendixA.NERCRegion:MidwestReliabilityOrganization(MRO) 27

AppendixB.NERCRegion:NortheastPowerCoordinatingCouncil(NPCC) 33

AppendixC.NERCRegion:ReliabilityFirst(RF) 39

AppendixD.NERCRegion:SoutheasternElectricReliabilityCorporation(SERC) 45

AppendixE.NERCRegion:TexasReliabilityEntity(TexasRE) 51

AppendixF.NERCRegion:WesternElectricityCoordinatingCouncil(WECC) 57

AppendixG.OffshoreWindRegion:Atlantic 63

AppendixH.OffshoreWindRegion:Gulf 66

AppendixI.OffshoreWindRegion:Pacific 69

ListofFigures

Figure1.ComparisonofGCMtrendsinchangestodailyaveragenear-surfaceairtemperaturefor

CONUS 12

Figure2.ComparisonofGCMdailymaximumairtemperatureeventsforCONUS 12

Figure3.ComparisonofGCMdailyminimumairtemperatureeventsforCONUS 13

Figure4.ComparisonofGCMtrendsinchangestodailyaveragenear-surfacerelativehumidityfor

CONUS 13

Figure5.ComparisonofGCMtrendsinchangestodailyaverageprecipitationforCONUS 14

Figure6.ComparisonofGCMminimumannualrainfallsforCONUS 14

Figure7.ComparisonofGCMtrendsinchangestodailyaverage100-meterwindspeedforCONUS 15

Figure8.ComparisonofGCMtrendsinchangestodailyaverageGHIforCONUS 15

Figure9.NERCRegion:MRO(includedstatesshadedingrey) 27

Figure10.ComparisonofGCMtrendsinchangestodailyaveragenear-surfaceairtemperatureforMRO.

29

Figure11.ComparisonofGCMdailymaximumairtemperatureeventsforMRO 29

Figure12.ComparisonofGCMdailyminimumairtemperatureeventsforMRO 30

Figure13.ComparisonofGCMtrendsinchangestodailyaveragenear-surfacerelativehumidityfor

MRO 30

Figure14.ComparisonofGCMtrendsinchangestodailyaverageprecipitationforMRO 31

Figure15.ComparisonofGCMminimumannualrainfallsforMRO 31

Figure16.ComparisonofGCMtrendsinchangestodailyaverage100-meterwindspeedforMRO 32

Figure17.ComparisonofGCMtrendsinchangestodailyaverageGHIforMRO 32

Figure18.NERCRegion:NPCC(includedstatesshadedingrey) 33

Figure19.ComparisonofGCMtrendsinchangestodailyaveragenear-surfaceairtemperaturefor

NPCC 35

iv

ThisreportisavailableatnocostfromtheNationalRenewableEnergyLaboratoryat/publications.

Figure20.ComparisonofGCMdailymaximumairtemperatureeventsforNPCC 35

Figure21.ComparisonofGCMdailyminimumairtemperatureeventsforNPCC 36

Figure22.ComparisonofGCMtrendsinchangestodailyaveragenear-surfacerelativehumidityfor

NPCC 36

Figure23.ComparisonofGCMtrendsinchangestodailyaverageprecipitationforNPCC 37

Figure24.ComparisonofGCMminimumannualrainfallsforNPCC 37

Figure25.ComparisonofGCMtrendsinchangestodailyaverage100-meterwindspeedforNPCC 38

Figure26.ComparisonofGCMtrendsinchangestodailyaverageGHIforNPCC 38

Figure27.NERCRegion:RF(includedstatesshadedingrey) 39

Figure28.ComparisonofGCMtrendsinchangestodailyaveragenear-surfaceairtemperatureforRF.41

Figure29.ComparisonofGCMdailymaximumairtemperatureeventsforRF 41

Figure30.ComparisonofGCMdailyminimumairtemperatureeventsforRF 42

Figure31.ComparisonofGCMtrendsinchangestodailyaveragenear-surfacerelativehumidityforRF.

42

Figure32.ComparisonofGCMtrendsinchangestodailyaverageprecipitationforRF 43

Figure33.ComparisonofGCMminimumannualrainfallsforRF 43

Figure34.ComparisonofGCMtrendsinchangestodailyaverage100-meterwindspeedforRF 44

Figure35.ComparisonofGCMtrendsinchangestodailyaverageGHIforRF 44

Figure36.NERCRegion:SERC(includedstatesshadedingrey) 45

Figure37.ComparisonofGCMtrendsinchangestodailyaveragenear-surfaceairtemperaturefor

SERC 47

Figure38.ComparisonofGCMdailymaximumairtemperatureeventsforSERC 47

Figure39.ComparisonofGCMdailyminimumairtemperatureeventsforSERC 48

Figure40.ComparisonofGCMtrendsinchangestodailyaveragenear-surfacerelativehumidityfor

SERC 48

Figure41.ComparisonofGCMtrendsinchangestodailyaverageprecipitationforSERC 49

Figure42.ComparisonofGCMminimumannualrainfallsforSERC 49

Figure43.ComparisonofGCMtrendsinchangestodailyaverage100-meterwindspeedforSERC 50

Figure44.ComparisonofGCMtrendsinchangestodailyaverageGHIforSERC 50

Figure45.NERCRegion:TexasRE(includedstatesshadedingrey) 51

Figure46.ComparisonofGCMtrendsinchangestodailyaveragenear-surfaceairtemperatureforTexas

RE 53

Figure47.ComparisonofGCMdailymaximumairtemperatureeventsforTexasRE 53

Figure48.ComparisonofGCMdailyminimumairtemperatureeventsforTexasRE 54

Figure49.ComparisonofGCMtrendsinchangestodailyaveragenear-surfacerelativehumidityfor

TexasRE 54

Figure50.ComparisonofGCMtrendsinchangestodailyaverageprecipitationforTexasRE 55

Figure51.ComparisonofGCMminimumannualrainfallsforTexasRE 55

Figure52.ComparisonofGCMtrendsinchangestodailyaverage100-meterwindspeedforTexasRE.56

Figure53.ComparisonofGCMtrendsinchangestodailyaverageGHIforTexasRE 56

Figure54.NERCRegion:WECC(includedstatesshadedingrey) 57

Figure55.ComparisonofGCMtrendsinchangestodailyaveragenear-surfaceairtemperaturefor

WECC 59

Figure56.ComparisonofGCMdailymaximumairtemperatureeventsforWECC 59

Figure57.ComparisonofGCMdailyminimumairtemperatureeventsforWECC 60

Figure58.ComparisonofGCMtrendsinchangestodailyaveragenear-surfacerelativehumidityfor

WECC 60

Figure59.ComparisonofGCMtrendsinchangestodailyaverageprecipitationforWECC 61

Figure60.ComparisonofGCMminimumannualrainfallsforWECC 61

Figure61.ComparisonofGCMtrendsinchangestodailyaverage100-meterwindspeedforWECC 62

Figure62.ComparisonofGCMtrendsinchangestodailyaverageGHIforWECC 62

v

ThisreportisavailableatnocostfromtheNationalRenewableEnergyLaboratoryat/publications.

Figure63.OffshoreWindRegion:Atlantic(includedareashadedingrey) 63

Figure64.ComparisonofGCMtrendsinchangestodailyaverage100-meterwindspeedfortheAtlantic

OffshoreRegion 65

Figure65.OffshoreWindRegion:Gulf(includedareashadedingrey) 66

Figure66.ComparisonofGCMtrendsinchangestodailyaverage100-meterwindspeedfortheGulf

OffshoreRegion 68

Figure67.OffshoreWindRegion:Pacific(includedareashadedingrey) 69

Figure68.ComparisonofGCMtrendsinchangestodailyaverage100-meterwindspeedforthePacific

OffshoreRegion 71

ListofTables

Table1.SummaryofGCMssurveyedandusedinthisreport 3

Table2.Summaryofvariablesanalyzedalongwithhistoricalbaselinedatasets 5

Table3.SummaryofhistoricalGCMskillusingKSstatisticandbiasmetricsforCONUS.Valuesfora

givenmetricineachrowarerankedfrombesttoworsthistoricalskill(darkbluetodark

red) 11

Table4.SummaryofhistoricalGCMskillusingKSstatisticandbiasmetricsforMRO.Valuesfora

givenmetricineachrowarerankedfrombesttoworsthistoricalskill(darkbluetodark

red) 28

Table5.SummaryofhistoricalGCMskillusingKSstatisticandbiasmetricsforNPCC.Valuesfora

givenmetricineachrowarerankedfrombesttoworsthistoricalskill(darkbluetodark

red) 34

Table6.SummaryofhistoricalGCMskillusingKSstatisticandbiasmetricsforRF.Valuesforagiven

metricineachrowarerankedfrombesttoworsthistoricalskill(darkbluetodarkred) 40

Table7.SummaryofhistoricalGCMskillusingKSstatisticandbiasmetricsforSERC.Valuesfora

givenmetricineachrowarerankedfrombesttoworsthistoricalskill(darkbluetodark

red) 46

Table8.SummaryofhistoricalGCMskillusingKSstatisticandbiasmetricsforTexasRE.Valuesfora

givenmetricineachrowarerankedfrombesttoworsthistoricalskill(darkbluetodark

red) 52

Table9.SummaryofhistoricalGCMskillusingKSstatisticandbiasmetricsforWECC.Valuesfora

givenmetricineachrowarerankedfrombesttoworsthistoricalskill(darkbluetodark

red) 58

Table10.SummaryofhistoricalGCMskillusingKSstatisticandbiasmetricsfortheAtlanticOffshore

Region.Valuesforagivenmetricineachrowarerankedfrombesttoworsthistoricalskill

(darkbluetodarkred) 64

Table11.SummaryofhistoricalGCMskillusingKSstatisticandbiasmetricsfortheGulfOffshore

Region.Valuesforagivenmetricineachrowarerankedfrombesttoworsthistoricalskill

(darkbluetodarkred) 67

Table12.SummaryofhistoricalGCMskillusingKSstatisticandbiasmetricsforthePacificOffshore

Region.Valuesforagivenmetricineachrowarerankedfrombesttoworsthistoricalskill

(darkbluetodarkred) 70

1

ThisreportisavailableatnocostfromtheNationalRenewableEnergyLaboratoryat/publications.

1Abstract

Theinterplaybetweenenergy,climate,andweatherisbecomingmorecomplexduetoincreasing

contributionsofrenewableenergygeneration,energystorage,electrifiedenduses,andthe

increasingfrequencyofextremeweatherevents.Energysystemanalysescommonlyrelyon

meteorologicalinputstoestimaterenewableenergygenerationandenergydemand;however,

theseinputsrarelyrepresenttheestimatedimpactsoffutureclimatechange.Climatemodelsandpubliclyavailableclimatechangedatasetscanbeusedforthispurpose,buttheselectionof

inputsfromthemyriadofavailablemodelsanddatasetsisanuancedandsubjectiveprocess.Inthiswork,weassessdatasetsfromvariousglobalclimatemodels(GCMs)fromtheCoupled

ModelIntercomparisonProjectPhase6(CMIP6).Wepresentevaluationsoftheirskillswith

respecttothehistoricalclimateandcomparisonsoftheirfutureprojectionsofclimatechangefortwoclimatechangescenarios.Wepresenttheresultsfordifferentclimaticandenergysystem

regionsandincludeinteractivefiguresintheaccompanyingsoftwarerepository.PreviousworkhaspresentedsimilarGCMevaluations,butnonehavepresentedvariablesandmetrics

specificallyintendedforcomprehensiveenergysystemsanalysisincludingimpactsonenergydemand,thermalcooling,hydropower,wateravailability,solarenergygeneration,andwind

energygeneration.WefocusonGCMoutputmeteorologicalvariablesthatdirectlyaffecttheseenergysystemcomponentsincludingtherepresentationofextremevaluesthatcandrivegrid

resilienceevents.Theobjectiveofthisworkisnottorecommendthebestclimatemodelanddatasetforagivenanalysis,butinsteadtoprovideareferencetofacilitatetheselectionof

climatemodelsandscenariosinsubsequentwork.

2Introduction

Energysystemanalysescommonlyusehistoricalweatherdatasetsasinputtoenergygenerationanddemandmodels(Brinkmanetal.2021;Carvalloetal.2023;Stencliketal.2021;Sharpetal.2023).Recently,moreworkhasstartedtoincorporatetheimpactsofclimatechangeonthese

inputs(Bloomfieldetal.2016;Yalewetal.2020;Craigetal.2018).GCMsandtheirassociatedpublicly-availabledatasetsfromCMIP6areavaluableresourceforestimatingtheimpactsof

climatechange(Eyringetal.,2016).However,thereareamyriadofuniqueGCMsdevelopedby

climateresearchinstitutionsaroundtheworld.EachGCMisuniqueinitsphysicaland

parametricformulations,itsskillinrepresentinghistoricalclimateindifferentgeographies,anditssensitivitytoanthropogenicgreenhousegasemissions(Flatoetal.,2013).Forexample,a

givenGCMmayrepresentavariableinthehistoricalclimatewithgreatprecisionbutmaybe

greatlybiasedinseveralothervariables(furtherdiscussedinSection

4)

.Tofurthercomplicatethetopic,CMIP6includesseveralpossibleclimatechangescenariosthatattempttocharacterizedeeplyuncertainhumanfactorsrelatedtothedevelopmentalprogressofcivilizationandour

continuedemissions.Scenarioshavebeendevelopedthatprojectdecreasesinemissionsbymid-century,andothersthatprojectemissionsdecreasingonlyneartheendofthecentury(Riahietal.,2017).Expertsandquantitativemodelsalikehaveperspectivesonwhichscenariosaremorelikely(Hausfather&Peters2020),butwecannotknowwithcertaintywhichfuturewewill

experience.

Priorworkstudyingclimatechangeinappliedimpactstudieshashandledthesenuancesthroughthefollowingprocess:1)comparedatafromvariousGCMswithhistoricalreferencedatasetstoidentifythosethatbestrepresenthistoricalclimate2)selectoneormoreGCMswithgood

2

ThisreportisavailableatnocostfromtheNationalRenewableEnergyLaboratoryat/publications.

historicalskillandclimatechangescenariosthatencompassarangeofpossibleclimatefutures,3)downscalethelow-resolutionGCMdataforappliedstudies(whenrequiredbyhigh-resolutionapplications),and4)performthesubsequentappliedanalysisusingdomain-specificmodels(Kaoetal.,2022;RalstonFonsecaetal.,2021).ThecomparisonandselectionofinputsfromGCMs

andclimatescenarios(steps#1and#2)isanuancedprocessthatcommonlyincludesthe

quantitativecomparisonofGCMdatasetsusingafewselectedmetricsoverafocusedregionofinterest(Pardingetal.,2020;Ashfaqetal.,2022;Chhinetal.,2018).However,theselectionofGCMs,climatescenarios,andcomparativemetricsareultimatelysubjectivedecisionsand

representvaluejudgementsinachallenginganalyticalprocesswithnoobjectivelybest

methodology.Further,theimpactsofclimatechangearebeingstudiedinanincreasinglywiderangeofapplicationsandthiscomparisonandselectionprocessisoftenveryspecifictoagivenapplication.

ThisreportfocusesonsupportingtheGCMcomparisonandselectionprocessspecificallyforenergyapplicationsintheContiguousUnitedStates(CONUS).PreviousworkhaspresentedsimilarGCMevaluations,butnonehavepresentedvariablesandmetricsspecificallyintendedforcomprehensiveenergysystemsanalysisincludingimpactsonenergydemand,thermal

cooling,hydropower,wateravailability,solarenergygeneration,andwindenergygeneration

(Pardingetal.,2020;Ashfaqetal.,2022;Martinez&Iglesias,2022).Thosethathavefocusedonsomeaspectofenergyimpactshavetypicallyfocusedononevariableoranothersuchasclimateimpactstohydropowerorwindenergy(Martinez&Iglesias,2022),butnonehavepresented

metricsforvariablesthatrepresentthefullenergygenerationanddemandsystem.Thisreportisintendedtofillthatgapandfacilitatemoreinformedselectionsofclimatechangeinputsfor

comprehensiveenergyanalyses.

Thisreportisstructuredasfollows.Section

3

detailsthedatasetsusedinthisreportandthe

methodsusedforGCMevaluation;Section

4

presentsanddiscussestheresultsoftheGCMskillevaluationandthecomparisonoftheirprojectionsfortheContiguousUnitedStates(CONUS);Section

5

concludesthereport;TheappendicespresentsimilarresultstoSection

4

butfor

specificsubregionswithinthelargerCONUSdomain.

3DataandMethods

ThisreportleveragespubliclyavailableclimatechangeprojectionsfromGCMsintheCMIP6

archiveandhistoricaldatafromreferenceandreanalysisdatasets.First,weexploretheavailabledatasetsassociatedwitheachGCMintheCMIP6archiveanddeterminewhichdatasetsare

viableforenergysystemsanalysis.

Forthepurposesofthiswork,welookforGCMdatasetsthatareofcurrentstate-of-the-art

spatiotemporalresolution(e.g.,100kmdaily),thatcontainallvariablesnecessarytomodel

energygenerationanddemand(e.g.,temperature,humidity,precipitation,windspeed,andsolarirradiance),andthathavepublicrecordsintheCMIP6archiveforseveralkeysimulations.Notethatdifferentdownscalingmethodologies(e.g.,dynamicaldownscalingwithregionalclimate

models,RCMs)mayrequirevariablesotherthanthosepresentedhere.However,westillfocusonthissubsetbecausetheyhavedirectimpactsontheenergysystem.

3

ThisreportisavailableatnocostfromtheNationalRenewableEnergyLaboratoryat/publications.

Forthiswork,weselectedtheCMIP6historicalsimulationthatisintendedtorepresentthe

historicalandcurrentclimate,andtwoSharedSocioeconomicPathways(SSPs):SSP24.5,andSSP58.5.NotethattheseSSPshavecorrespondingRelativeConcentrationPathway(RCP)

scenariosfromCMIP5.WeselectedthefirstvariantfromeachGCMexceptforCESM2andCESM2-WACCMwhichhadothervariantswithmorecompletedataavailability.Forthe

comparisonoffutureprojections,weselectdatafromSSP24.5andSSP58.5.Weselectthesetwoscenariosbecauseoftheextensiveuseofthesescenariosinpriorclimateimpactsanalysis(Craigetal.,2020;Kaoetal.,2022;RalstonFonsecaetal.,2021;Martinez&Iglesias2022).

SSP24.5istypicallydescribedasa“middle-of-the-road”emissionsscenariowheretrends

generallyfollowadynamics-as-usualscenario,whileSSP58.5isanaggressivehigh-growthandhighfossilfuelfuturewiththemostoverallemissionsofanyscenario(Riahietal.,2017).

Toinformenergysystemanalysesinwhichdecisionsonenergyinfrastructurearebeingmade

todayandinthecomingseveraldecades,wefocusonprojectionsfromthehistoricalclimate

throughmid-century(e.g.,through2059).DespitethesignificantlydifferentemissiontrajectoriesinSSP24.5andSSP58.5,thetwoscenariosvaryonlyslightlybymid-century(asshownin

Section

4)

withamoredramaticbifurcationoccurringinthelatterhalfofthecentury.

Aftersurveying33GCMswithdatainCMIP6,weselect13GCMsthathavepubliclyavailabledatathatmeettheabovecriteria.Asummaryofthisprocess,theGCMsevaluated,andthe

GCMsselectedispresentedin

Table1

below.

GCMsthatdidnotmeetthecriteriaforthisworkmayhaveadditionalvariablesandscenariosavailablefromdifferentdataarchives.TheseGCMsmaybeusefulforclimateimpactstudies,butbasedontheirdatasetsavailableintheCMIP6archivetheywerenotusedinthiswork.

Table1.SummaryofGCMssurveyedandusedinthisreport.

GCMName

Used

NotesandReference

AWI-CM-1-1-MR

No

Historicalsimulationdoesnotincludeirradiance,precipitation,orhumidity(Semmleretal.,2019).

ACCESS-CM2

No

SSPdataislowspatialresolution(Dixetal.,2019).

BCC-CSM2-MR

No

Doesnotincludehumidity(Xinetal.,2019).

CAMS-CSM1-0

No

Doesnotincludehumidity(Rongetal.,2019).

CanESM5

No

Nodataatdesiredspatiotemporalresolution(Swartetal.,2019)

CESM2

Yes

Usedvariantr4i1p1f1.Othervariants(r1i1p1f1,r2i1p1f1,andr3i1p1f1)donotincludedailymin/maxtemperatures(Danabasoglu,2019a).

CESM2-WACCM

Yes

Usedvariantr3i1p1f1.Othervariants(r1i1p1f1andr2i1p1f1)donotincludedailymin/maxtemperatures(Danabasoglu,

2019b).

CMCC-CM2-SR5

No

Doesnotincludedailymin/maxtemperatures(Lovatoetal.,2020).

CMCC-ESM2

No

Doesnotincludegeopotentialheight(Lovatoetal.,2021)

CNRM-ESM2-1

No

Doesnotincludeanyrelevantvariablesatdesiredspatiotemporalresolution(Voldoire,2019).

4

ThisreportisavailableatnocostfromtheNationalRenewableEnergyLaboratoryat/publications.

GCMName

Used

NotesandReference

E3SM-1-0

No

SSPdataatdesiredspatiotemporalresolutiondoesnot

includeirradiance,windspeeds,andhumidity(Baderetal.,2022a).

E3SM-1-1

No

SSPdataisatamonthlyfrequency(Baderetal.,2020).

E3SM-1-1-ECA

No

SSPdataatdesiredspatiotemporalresolutiondoesnot

includeirradiance,windspeeds,andhumidity(Baderetal.,2022b).

E3SM-2-0

No

NodataforSSP58.5(E3SMProject,DOE,2022)

E3SM-2-0-NAARRM

No

NodataforSSP58.5(Tangetal.,2023)

EC-Earth3

Yes

(EC-EarthConsortium,2019a)

EC-Earth3-CC

Yes

(EC-EarthConsortium,2021b)

EC-Earth3-Veg

Yes

(EC-EarthConsortium,2019b)

EC-Earth3-Veg-LR

No

Nodataatdesiredspatiotemporalresolution(EC-EarthConsortium,2020)

FGOALS-f3-L

No

SSPdataisatmonthlyfrequency(Yu,2019).

GFDL-CM4

Yes

(Guoetal.,2018)

GFDL-ESM4

Yes

(Johnetal.,2018)

HadGEM3-GC31-MM

No

NodataforSSP24.5andincompletetimeserieswithlessthan365daysperyear(Jackson,2020)

INM-CM4-8

Yes

(Volodinetal.,2019a)

INM-CM5-0

Yes

(Volodinetal.,2019b)

IPSL-CM6A-LR

No

Nodataatdesiredspatiotemporalresolution(Boucheretal.,2019).

KACE-1-0-G

No

SSPdataisatlowspatialresolution(Byunetal.,2019).

MIROC6

No

Nodataatdesiredspatiotemporalresolution(Shiogamaetal.,2019).

MPI-ESM1-2-HR

Yes

(Schupfneretal.,2019)

MPI-ESM1-2-LR

No

SSPdataisatlowspatialresolution(Wienersetal.,2019).

MRI-ESM2-0

Yes

(Yukimotoetal.,2019)

NorESM2-MM

Yes

(Bentsenetal.,2019)

TaiESM1

Yes

(Leeetal.,2020)

Foreachvariable,weselectahistoricalreferencedatasetthatcanbeusedtoevaluatethe

historicalskilloftheGCMs.Wechoosedatasetsthatarepubliclyavailable,haveatleasta20-yearhistoricalrecord,andhavebeenusedextensivelyinpreviousenergysystemstudies.WeleveragetheEuropeanCentreforMedium-RangeWeatherForecastsReanalysisv5(ERA5),

Daymet,andtheNationalSolarRadiationDatabase(NSRDB)(CopernicusClimateChangeService,2017;Thorntonetal.,2021;Senguptaetal.,2018).Thevariablesanalyzedandtheircorrespondinghistoricalreferencedatasetsaredetailedin

Table2

below.

Thethreehistoricalreferencedatasetsusedinthisworkareallatfinerspatialandtemporal

resolutionsthantheGCMdatabeingevaluated.Weperformageospatialmappingtoaggregatehigh-resolutionhistoricalpixelstotheirnearestlow-resolutionGCMpixel.Thiscreatesasub-

5

ThisreportisavailableatnocostfromtheNationalRenewableEnergyLaboratoryat/publications.

gridmapping(e.g.,similartoasudokugrid)withoutoverlaporduplicationofthehigh-

resolutionpixels.Asimpleaveragingormin/maxoperationisperformedonthetemporalaxistoaggregatesub-dailydatatotheGCMdailyvalues.

Table2.Summaryofvariablesanalyzedalongwithhistoricalbaselinedatasets.

Variable

Abbreviation

Historical

ReferenceDataset

Resolution

TemporalExtent

Reference

Air

T2M

ERA5

31-kmhourly

1980-2019

Copernicus

Temperature

ClimateChange

(2-meter)

Service,2017

RelativeHumidity(2-meter)

RH2M

ERA5

31-kmhourly

1980-2019

Copernicus

ClimateChangeService,2017

Precipitation

PR

Daymet

4-kmdaily

1980-2019

Thorntonetal.,2021

Global

HorizontalIrradiance

GHI

NSRDB

4-km30-minute

2000-2019

Senguptaetal.,2018

Windspeed(100-meter)

WS100m

ERA5

31-kmhourly

1980-2019

Copernicus

ClimateChangeService,2017

Forthehistoricalskillevaluation,weuse40-yearrecordsfo