摘要
Aconvection-allowingensembleforecastexperimentonasqualllinewasconductedbasedonthebreedinggrowthmode(BGM).Meanwhile,theprobabilitymatchedmean(PMM)andneighborhoodensembleprobability(NEP)methodswereusedtooptimizetheassociatedprecipitationforecast.Theensembleforecastpredictedtheprecipitationtendencyaccurately,whichwasclosertotheobservationthaninthecontrolforecast.Forheavyrainfall,theprecipitationcenterproducedbytheensembleforecastwasalsobetter.TheFractionsSkillScore(FSS)resultsindicatedthattheensemblemeanwasskillfulinlightrainfall,whilethePMMproducedbetterprobabilitydistributionofprecipitationforheavyrainfall.Preliminaryresultsdemonstratedthatconvection-allowingensembleforecastcouldimproveprecipitationforecastskillthroughprovidingvaluableprobabilityforecasts.Itisnecessarytoemploynewmethods,suchasthePMMandNEP,togenerateprecipitationprobabilityforecasts.Nonetheless,thelackofspreadandtheoverpredictionofprecipitationbytheensemblemembersarestillproblemsthatneedtobesolved.
出版日期
2017年05月15日(中国Betway体育网页登陆平台首次上网日期,不代表论文的发表时间)