中文摘要：

用MICAPS平台中的T213产品场与暴雨发生的初始、中间、结束过程相对应的三维空间物理量场的相似程度,根据场相似的数量综合判别分析作出有无区域暴雨的预报。预报因子选取涡度、散度、垂直速度和水汽通量散度,空间使用850、700、500hPa,时间上选取12、24、36h预报场。该方法实现了区域暴雨预报的客观化和程序化。根据天气过程分类建立的两类模型以场平均距离为依据,作出的暴雨预报检验效果令人满意,2004—2010年的5—10月7年平均预报准确率超过33%,具有较好的实践参考指导作用。

英文摘要：

By virtue of the similarity between the dynamic processes from the T213 products on the MICAPS plat{orm corresponding to the three-dimensional physical element fields relative to the formation, development, and extinction of heavy rainfall, the yes/no prediction of regional heavy rainfall is made according to the comprehensive examination and analysis of the number of similar fields. Forecast factors include vorticity, divergence, vertical velocity, and moisture flux convergence at 850 hPa, 700 hPa, and 500 hPa, at the time scales of 12-, 24- and 36-hour forecast fields, respectively. The programming, automation, and objectivity are realized. The two types of models constructed by means of weather processes are based on the field-mean distance. The verification indicates that the prediction results of heavy rainfall are satisfactory, with the mean TS of 2004 to 2010 （May to October） being over 33%.