珠三角地区前汛期强对流潜势预报方法研究

利用2004—2006年前汛期探空资料计算的物理量,选取与强对流天气相关性好的大气温湿类(整层比湿积分IntegralQ)、层结稳定度类(K指数)、动力类(潜在下冲气流指数MDPI)、热力动力综合类(瑞士第一雷暴指数SWISS00)作为预报因子,通过对各指数的空间分布特征和数值进行二值Logistic回归分析,得到各指数的参数估算值,建立强对流诊断预报方程,得到前汛期强对流潜势预报因子P,从而制作珠江三角洲(以下简称珠三角)地区未来12小时出现强对流天气的潜势预报。并用此法回报2003—2006年3—6月前汛期的强对流天气。结果表明,P值大于0.9的准确率可达77.5%,P值小于0.5出现强天气的概率仅为3.8%。由于资料有限,对2007年3—4月发生的7次强对流的经验检验效果不明显,但P值小于0.5时不发生强对流的经验检验效果明显。此法对珠三角地区的短时强降水和雷雨大风等强对流天气的临近监测预警有较好的指示意义。

STUDY ON A FORECASTING METHOD FOR SEVERE CONVECTIVE WEATHER POTENTIAL IN THE ANNUALLY FIRST RAINING SEASON OF THE PEARL RIVER DELTA

Based on radiosonde data in the annually first raining seasons from 2004 to 2006,some indices and parameters,including Integral Q,K-index,Microburst-Day Potential Index(MDPI),SWISS00 index,and variables often analyzed for forecasting severe thunderstorms,were selected as predictors.Through analyzing the spatial distributions and the binary logistic regressions of the indices,estimated parameters of the predictors and a convective weather diagnostic prediction equation were established to make 12-h severe weather forecasts for the Pearl River Delta in Guangdong Province.The equation was tested and verified with historical data between March and June from 2003 to 2006.The results indicated that the accuracy of severe convective weather forecast was 77.5% if the equation’s output was greater than 0.9,whereas the probability of severe weather development was only 3.8% when the output was less than 0.5.This method can be used as a very short-range forecasting tool for severe convective weather.