【摘要】 利用1961～2006年陇东黄土高原气象观测站的雷暴资料,运用数理统计、函数分解、图表分析方法,分析了雷暴的时空分布特征。(结果表明:陇东雷暴次数的年际变化总体上呈波动状态,60年代为雷暴多发期,70年代波动明显,大致存在4.5年的振荡周期,80年代为增长期,以5年为周期振荡,90年代开始至今以4年为周期振荡并呈现减少趋势。根据41年NCEP/NCAR再分析格点资料,并考虑下垫面热力作用,用统计分析方法得到判别陇东黄土高原雷暴有无的预报函数,建立了雷暴短期短时预报的判别模型。检验表明:雷暴短期短时预报模型是可行的,2005～2006年试报,雷暴预报出现率可达61.7%,在同一时效下模型的预报水平要高于预报员的主观预报。更多还原

【Abstract】 By utilizing the 1961～2006 thunderstorm data from weather stations in loess plateau in the eastern part of Gansu province,this article analyzes thunderstorms’ time and space distribution characteristics.Methods adopted are mathematical statistics,function decomposition and graph analysis.Results indicate that annual border change of thunderstorm days overall appears undulatory with 1960s a frequent period and 1970s fluctuating,the shake cycle is 4.5 years.The growth period starts from 1980s with a 5-year shake cycle and has a decreasing tendency from 1990s,this time the cycle is 4 years.We have analyzed the grid point material again according to 41-year NCEP/NCAR and taken into consideration the pad surface thermal energy function so as to get the prediction function to identify the east loess plateau thunderstorm of Gansu Province via statistical analysis.We have also set up a short-time forecasting model.Test results show that the model is practicable,thunderstorm forecast frequency of occurrence can reach 61.7% in the trial period between 2005 and 2006,which is more effective than predictor’s subjective forecast under identical effectiveness.更多还原