石家庄市人工防雹预警方法的研究

Study on Early Warning Method of Artificial Hail Prevention in Shijiazhuang City

通过对石家庄市2007年—2019年出现的冰雹云和雷雨云雷达回波特征参量统计分析后,筛选出冰雹云雷达早期识别预警指标,采用直观对比法、综合概率判别法和Fisher判别法建立了冰雹云识别预警数学模型,并以此进行了历史回报检验和样本预报检验。结果表明:基于3种判别法建立的冰雹云早期识别预警数学模型在业务中均具有应用价值,尤其是Fisher判别法在冰雹云识别的准确度、漏报率和空报率方面更具有一定的优越性,平均预警时长≥40 min。

Through the statistical analysis of the hail cloud and thunderstorm radar echo characteristic parameters that appeared in our city from May to August in 2007 to 2019, the early recognition and early warning indicators of the hail cloud radar selected by the intuitive comparison method, comprehensive probabilistic discrimination method and Fisher discriminant method established a mathematical model of hail cloud identification and early warning, and used this to conduct historical return test and sample forecast test. The results show that the mathematical model of early recognition and early warning of hail cloud based on the three discriminant methods has certain application value in the business, especially the Fisher discriminant method is more accurate in hail cloud identification accuracy, missed report rate and empty report rate. It has certain advantages, and the average warning time is greater than or equal to 40 min.