北京市“7·21”特大暴雨不同预报产品可预报性对比

Contrasting the Predictability of Several Kinds of Predictions on “7·21” Heavy Rainfall in Beijing

利用ECMWF、NCEP全球预报产品和BJ-RUC区域预报产品,对比了不同模式对北京市"7·21"特大暴雨暖区降水、锋面降水的预报效果,同时利用WRF高分辨率中尺度模式同化常规观测资料和雷达资料,对此次过程进行数值模拟试验。结果表明:NCEP和ECMWF的全球集合预报产品都能预报出北京市"7·21"特大暴雨过程,但在暖区降水阶段和锋面降水阶段存在6 h左右的时间滞后,且降水量偏小;BJ-RUC区域模式预报出了整个强降水过程,且较好地预报了暖区降水,优于NCEP和ECMWF预报,但锋面降水较之实况锋面阶段降水偏南,预报的降水量小于实况。对于此次特大暴雨过程的模拟,暖区降水和锋面降水的预报要优于业务预报,且暖区降水接近实况降水,但整个锋面降水过程存在3 h的时间滞后。

Based on several kinds of predictions on"7.21"heavy rainfall in Beijing,using ECMWF,NCEP global forecasting products and BJ-RUC area forecasting product assess the predictability of their corresponding models on warm sector precipitation,frontal precipitation and past frontal precipitation.Further more,using WRF high resolution model and WRFDA-3DVAR which is assimilated conventional observational data and radar data to simulate the"7.21"heavy rainfall.The results indicate that both global ensemble forecast of NCEP and ECMWF catch the pattern of the heavy rain.However,their forecast on warm sector precipitation and frontal precipitation is about 6 hours time lag and less than observation.BJ-RUC predicts the whole heavy rainfall process and the warm sector precipitation which are better than the NCEP and ECMWF,but the frontal precipitation is further south and the accumulated rainfall is less than observation.Assimilating conventional observation data and radar data can improve the warm sector precipitation and frontal precipitation.