基于对流参数南宁地区雷暴天气潜势预报方法研究  

A thunderstorm potential forecasting method based on convection parameters in Nanning region

作　　者：梁维亮[1] 屈梅芳[1] 翟丽萍[1] 

机构地区：[1]广西气象台,广西南宁530022

出　　处：《气象与环境学报》2015年第4期107-112,共6页Journal of Meteorology and Environment

基　　金：广西自然科学基金项目(2012GXNSFAA053191);广西自然科学基金项目(2013GXNSFAA019288);中国气象局预报员专项(CMAYBY2014-053)共同资助

摘　　要：利用2003—2012年南宁地区地面气象观测资料和探空资料,统计分析了南宁地区雷暴天气和常用对流参数的关系。结果表明:2003—2012年南宁地区有雷暴、无雷暴天气时对流参数的数值存在明显的区别;整层比湿积分(IQ)、K指数(KI)、抬升指数(LI)与雷暴天气发生概率呈显著的线性关系;总指数(TT)、沙氏指数(SI)和对流有效位能(CAPE)超过一定数值后雷暴天气概率变化较小。对流参数二值化后对雷暴天气的指示作用比实际值更好。将二值化的对流参数作为因子,采用二值Logistic回归法建立雷暴天气概率的预报模型;利用2013年资料进行试预报检验,证明建立的模型预报效果较好,可用于雷暴天气潜势的6 h短时预报。Based on the ground observed meteorological data and radiosonde data from 2003 to 2012 in Nanning region, relationships between thunderstorm weather and convection parameters were analyzed. The results show that the values of convection parameters are significantly different in a day with and without thunderstorm. The probability of thunderstorm is a significantly linear correlation with integral humidity （IQ）, K index （KI） and lifting index （LI）. The probability of thunderstorm is nearly unchanged after total-totals index （ TT）, showalter index （SI） and convective available potential energy （CAPE） exceed a specific value. The binaryzation value of convection parameters is indicative to thunderstorm better than the actual values. A probabilistic forecasting model for thunder- storm is established using a binary logistic regression method, taking binary convection parameters as factors. The probabilistic forecasting model is proved to be effective in 2013, which is available in 6-hour thunderstorm potential short-term forecasting.

关 键 词：雷暴 对流参数 二值化 二值Logistic回归法 概率预报模型 

分 类 号：P456.8[天文地球—大气科学及气象学]