Page 132 - 2023年第54卷第1期
P. 132
[22] MUQZ,ZHAOM S,RUNNINGSW.ImprovementstoaMODISglobalterrestrialevapotranspirationalgorithm
[J].RemoteSensingofEnvironment,2011,115(8):1781 - 1800.
[23] BECKHE,VANDAIJM,VINCENZOL,etal.MSWEP:3 - hourly0.25°globalgriddedprecipitation(1979 -
2015)bymerginggauge,satellite,andreanalysisdata[J].HydrologyandEarthSystem Sciences,2017,21
(1):589 - 615.
[24] 胡庆芳.基于多源信息的降水空间估计及其水文应用研究[D].北京:清华大学,2013.
[25] SAVEH,BETTADPURS,TAPLEYBD.HighresolutionCSRGRACERL05mascons[J].JournalofGeophysi
calResearch :SolidEarth,2016,121(10):7547 - 7569.
[26] TAPLEYBD,BETTADPURS,RIESJC,etal.GRACEmeasurementsofmassvariabilityintheearthsystem
[J].Science,2004,305(5683):503 - 505.
[27] SCANLONBR,ZHANGZZ,SAVEH.GlobalevaluationofnewGRACEmasconproductsforhydrologicappli
cations[J].WaterResourcesResearch,2016,52:9412 - 9429.
Accuracyvalidationofremotesensingandreanalysisevapotranspirationdata:
AcomparativestudybasedonGRACEandmonthlywaterbalancemodel
1
1,2
1
1,2
3
1
YUNZhaode,HUQingfang ,WANGYintang ,WUHaiyan,WANGLeizhi,SHANGShouwei
(1.NanjingHydraulicResearchInstitute,StateKeyLaboratoryofHydrology,WaterResources&HydraulicEngineeringScience,
Nanjing 210029,China;
2.YangtzeInstituteforConservationandDevelopment,Nanjing 210098,China;
3.HuzhouMunicipalBureauofWaterResources,Huzhou 313000,China)
Abstract:TheactualevapotranspirationmonthlyseriesET GRACE andET WBM wereconstructedbyusingtheGRACE
terrestrialwaterstoragechangedataandmonthlyWaterBalanceModelwithDoubleParameters (WBM- DP)re
spectively.Taking thisasthe benchmark , the performance oftwo remote sensing evapotranspiration data
(MOD16,SSEbop)andanatmosphericreanalysisdata(GLDAS - Noah)intheupperHanjiangRiverbasinwere
evaluated.TheresultsshowthattheevapotranspirationreferenceseriesobtainedbyGRACEdataandWBM- DPin
theupperHanjiangRiverbasinfrom 2006to2014arecloseattheannualscale,buttherearesignificantdiffer
encesatthemonthlyscale.WBM- DPcansimulatetheseasonalvariationcharacteristicsofterrestrialwaterstorage
betterundertheinfluenceofprecipitationandpotentialevapotranspiration,andWBM - DP - basedevapotranspiration
benchmarksequenceisalsomorereasonable.However,thevariationofwaterstorageinthebasinreflectedbythe
GRACEdataisunreasonable,sothebenchmarksequenceofmonthlyevapotranspirationcalculatedaccordingtothe
waterbalanceequationalsohasanomalies.Onthemonthlyscale ,thedifferenceofthebenchmarkserieswilllead
tonon - uniformchangesintheprecisionindexofevapotranspirationdata ,soitsimpactcannotbeignored.Atthe
sametime ,theresultsofaccuracyevaluationmayalsobedifferent.Ineachmonthofnon - floodperiod,when
ET WBM isusedasthebenchmark,thecomprehensiveaccuracyofMOD16isthebest;whenET GRACE istakenasthe
benchmark ,SSEbopisthebest.ComparedwithET GRACE ,thequantitativeaccuracydifferencesofMOD16,Noah
andSSEboparemoresignificantwhen ET WBM isusedasthebenchmark.Thepaperdeepenstheunderstandingofthe
applicabilityofGRACEdata ,andprovidesareferencefortheaccuracyevaluationanddataoptimizationofsatellite
remotesensingandatmosphericreanalysisevapotranspirationdata.
Keywords:GRACE;WaterBalanceModel;remotesensingevapotranspiration;atmosphericreanalysis
(责任编辑:于福亮)
— 1 2 7 —