Through the comparative analysis of 457 thunderstorms from 1997 to 2006 in Beijing, the circulation patterns of thunderstorms are classified into 11 patterns as northeast cyclone and shallow groove pattern, Baikal and Mongolian cyclone pattern and coming-west groove pattern. Thirty-three convective parameters, such as convective available potential, lifted index, relative storm helicity index and so on, are calculated using meteorological sounding data of south suburb meteorological observatory. The six convective parameters, BCAPE, BLI, MDCI, KNEW, BIC and SWISS are chosen as key elements of forecast of potential thunderstorm. Using the six parameters and regression estimation of event probabilities (REEP), the probability regression forecast question in eleven weather patterns are established. According to the forecasting fields of WRF model, probability regression forecast of thunderstorm in Beijing in 3 to 36 hours is completed. The experimental result shows that the forecast system is correct and has a certain reference value.%对1997 - 2006年457个雷暴过程的环流形势进行对比分析,将北京地区的雷暴天气分为东北低涡低槽、贝蒙低涡低槽、西来槽等11种雷暴天气型；利用南郊观象台(54511站)的探空资料计算对流有效位能、抬升指数和相对风暴螺旋度等33个对流参数,通过与北京地区SAFAIR3000获取的闪电定位资料进行统计分析,提取BCAPE、BLI、MDCI、BIC、KNEW和SWISS等6个对流参数作为北京地区潜势预报参数；采用事件概率回归(REEP)方法,利用获取的6个对流参数作为变量,形成了11种雷暴天气型下的潜势预报方法.利用WRF模式的预报场,建立适用于北京地区3～36 h雷电潜势预报系统.个例实验结果表明其具有较好准确性.由于该系统建立过程中使用了高分辨率探测资料和中尺度模式的输出结果,实现了雷电潜势预报由点到面,由粗到细的突破,对北京地区雷电预警预报具有一定的应用价值.
展开▼
