基于对流参数的雷暴潜势预报研究

Thunderstorm Potential Trend Forecast Based on Convection Parameters

为提高雷暴天气的潜势预报能力,在统计分析安徽省雷暴形成天气条件的基础上,利用2003—2004年T213资料,选取与雷暴相关性好的对流参数作为预报因子。并在考虑因子季节变化特征的基础上,分别用判别分析法和指标叠加法制作雷暴潜势预报,结果表明指标叠加法优于判别分析法。最后用指标叠加法试报了2005年3—8月的雷暴,临界成功指数CSI=69.4%,命中率POD=89.5%,虚假报警率FAR=24.4%,分区预报准确率也较高。另外,试报期间的区域性冰雹、雷雨大风全部命中,可见该方法对冰雹、雷雨大风也有较好的指示意义。

In order to improve thunderstorm potential forecast ability, on the basis of statistics and analysis of thunderstorm-forming weather conditions in Anhui Province, the convection parameters were selected from the T213 data during 2003 to 2004 as forecast factors. By considering the seasonal variation characteristics of the factors, the thunderstorm trend is forecasted with the discrimination analytical method and index accumulation method. The result indicates that the latter method excels the former. Finally, the index accumulation method was used to forecast the thunderstorm from March to August in 2005, the result is that Critical Success Index （ CSI ） is 69.4 %, Percent Of Doom （POD） is 89.5 %, False Alarm Rate （ FAR ） is 24.4 %, and the accuracy of regional forecast is better. Besides, regional hail and thunderstorm with strong wind during the testing period are all forecasted correctly. It is obvious that this method has better indicative significance to forecast the hail and thunderstorm with strong wind.