利用中尺度数值模式产品建立暴雨落区潜势预报

The Rainstorm Potential Predictability for Rain Area Using the Mesoscale Numerical Model

统计2005—2006年AREM模式输出相关物理量与暴雨的关系,采用模糊逻辑方法建立暴雨潜势预报多参数组合方程,然后分别代入AREM模式预报场相关物理量24 h平均值和最大值,进行暴雨落区潜势预报。对比分析2007—2009年潜势预报情况和暴雨Ts评分,结果表明:最大值方法潜势预报强中心偏多,暴雨空报较多;平均值方法能够较好反映24 h内整体降水分布情况,其Ts评分较高,虽然对强降水中心落区预报仍有偏差,但其总体预报效果要优于最大值方法。

By studying the relation between the physical quantity output from AREM and rainstorm from 2005 to 2006 and using fuzzy logic method,a potential predictability combination equation with multiple parameters was setup and applied to forecast rainstorm with the average and maximum values of 24hr AREM forecast field.Through contractive analysis of the potential predictability and rainstorm Ts scores from 2007 to 2009,the conclusion is that more strong centers and the false alarms were forecasted by the potential predictability with maximum value.The potential predictability with average value can reflect the whole precipitation condition in 24 hours and its Ts scores are higher,and its whole effect is better though it has deviation to forecast the central rain area of severe precipitation.