【摘要】 由于模式对于强降水落区预报有一定的偏差,TS评分不能完美的刻画模式预报强降水的问题,制定了强降水落区偏离程度的检验方法,基于此种方法对多模式(EnWRF、WRF-RUC、T639和EC-thin)山东省2014、2015年5—9月16次强降水过程预报的降水落区形态进行检验。结果表明:除了副高摆动引起的局地强对流天气外,其他过程模式预报均有指示意义,其中预报效果最好的是EnWRF和EC-thin,降水落区的形态与实况的相似度极高,并且表现出一定的互补性。多数情况下,模式预报的强降水中心整体比实况偏小,EC-thin和EnWRF漏报次数最少、准确次数最多,T639次之,WRF-RUC漏报次数最多并且准确次数最少。对于预报有偏离的过程,各模式整体雨区的偏离方向大多偏西或偏北。更多还原

【Abstract】 Because of forecasting deviation of severe precipitation model with observation,TS verification cannot characterize well the capability of severe precipitation forecasting models.The verification method about precipitation area is formulated,and based on it and the models of EnWRF,WRF-RUC,T639,and EC-thin forecast,16 severe precipitation area forecasts from May to September in 2014 and 2015in Shandong are verified with observation.The results show that except local severe convective weather caused by the motion of subtropical high,the other processes all have referential meaning.EnWRF and EC-thin are the best,with the forecast of precipitation area highly similar to observation,and the two model are complementary.In most cases,the model forecast for precipitation is less than the observation.EC-thin and EnWRF have the least number of missing reports and most accurate cases,and T639 model is the next,whereas WRF-RUC model has the most number of missing reports and lowest accuracy.For the forecasts with deviation,the directions of model forecast precipitation area are generaly more west or north than normal comparing with observation.更多还原