摘要
Homogenizationofclimateobservationsremainsachallengetoclimatechangeresearchers,especiallyincaseswheremetadata(e.g.,probabledatesofbreakpoints)arenotalwaysavailable.Toexaminetheinffuenceofmetadataonhomogenizingclimatedata,theauthorsappliedtherecentlydevelopedMultipleAnalysisofSeriesforHomogenization(MASH)methodtotheBeijing(BJ)dailytemperatureseriesfor1960-2006inthreecaseswithdifferentreferences:(1)13M-consideringmetadataatBJand12nearbystations;(2)13NOM-consideringthesame13stationswithoutmetadata;and(3)21NOM-considering20furtherstationsandBJwithoutmetadata.Theestimatedmeanannual,seasonal,andmonthlyinhomogeneitiesaresimilarbetweenthe13Mand13NOMcases,whilethoseinthe21NOMcaseareslightlydifferent.ThedetectedbiasesintheBJseriescorrespondingtothedocumentedrelocationdatesareaslowas-0.71~0C,-0.79~0C,and-0.5~0Cfortheannualmeaninthe3cases,respectively.Otherbiases,includingthoseundocumentedinmetadata,areminor.TheresultssuggestthatanymajorinhomogeneitycouldbedetectedviaMASH,albeitwithminordifferencesinestimatinginhomogeneitiesbasedonthedifferentreferences.Theadjustedannualseriesshowedawarmingtrendof0.337,0.316,and0.365~0C(10yr)~(-1)forthethreecases,respectively,smallerthantheestimateof0.453~0C(10yr)~(-1)intheoriginalseries,mainlyduetotherelocation-inducedbiases.TheimpactoftheMASH-typehomogenizationonestimatesofclimateextremesinthedailytemperatureseriesisalsodiscussed.
出版日期
2010年04月14日（中国期刊网平台首次上网日期，不代表论文的发表时间）
