【摘要】 为了研究初值不确定性对我国南方暴雨数值预报的影响,依据历史降水资料的统计分型,选取华南暴雨型、江淮暴雨型和黄淮暴雨型等三类型中具有代表性的暴雨过程,采用AREM模式以不同资料分析场(常规观测资料、NCEP和ERA40同化分析资料)作为初值进行暴雨预报对比试验。试验结果表明,对于各类型暴雨,不同资料初值的差异都会引起暴雨数值预报结果的较大差异;在此基础上,采用经验正交函数(EOF)分析方法,分析了各类暴雨初值误差及其变化信息,从中提取出主要的误差分量,并对这类误差的数值预报影响进行了敏感性试验和预报验证,结果显示,寻找和消除主要的初值误差,对改进暴雨数值预报结果具有重要作用。更多还原

【Abstract】 This paper focused on studies on the effect of initial uncertainties on heavy rainfall forecast. Based on statistical heavy rainfall patterns from historical rainfall data for fifteen years, typical heavy rainfall processes were chosen from heavy rainfall patterns of Huanan, Jianghuai and Huanghuai, and then were simulated using AREM with different types of initial data including conventional observations, NCEP data and ERA40 Reanalyses data. The result suggested that for all cases from three kinds of the heavy rainfall patterns, difference between ingested initial data resulted in difference between heavy rainfall forecast. And then EOF analysis method was used to analyze initial errors of all types of heavy rainfall and their evolution to obtain major error components. At last, sensitivity experiments were performed to analyze and validate the effect of those major errors on numerical forecast, which suggested, finding and eliminating those major initial errors is important for improvement of heavy rainfall forecast.更多还原