基于主分量法的雷达回波资料的雹云识别及其效果检验  

HAIL CLOUD RECOGNITION AND ITS RESULTS VERIFICATION BASED ON RADAR ECHOES DATA BY USING PRINCIPAL COMPONENT ANALYSIS

作　　者：李祚泳[1] 刘少侬[1] 邓新民[1] 张辉军[1] 

机构地区：[1]成都气象学院,四川省成都市610041

出　　处：《高原气象》1993年第1期84-89,共6页Plateau Meteorology

基　　金：国家气象局气象科学基金资助课题

摘　　要：本文用主分量分析法建立了雷云等强对流天气的综合判别式。该方法首先由某地区的雷达回波历史资料,建立判别指标之间的相关系数矩阵,然后用主分量分析法确定因子荷载,并由荷载矩阵。相关系数矩阵确定判别指标权重,从而建立雹云等强对流天气的综合判别式。该方法用成都市的雷达回波历史资料所建立的判别式对成都市试报和对沪州市进行了效果检验,并将判别结果与综合概率法比较,效果尚好。表明该方法用于雹云判别是可行的。This paper presents a comprehensive recognition formula of hailstorm cloud by using principal component analysis. The approach establishes correlation coefficient matrix of recognition indexes with historical data of radar echoes in a certain location. Then factor loads can be determined by using principal component analysis and the weights of recognition indexes are derived by load matrix and correlation coefficient matrix. Finally, the discrimination formula of hailstorm cloud is derived. The discrimination formula established based on historical radar echoes data in Chengdu by this method is applied to forecasting test of Chengdu and results verification of Luzhou. The recognition results are better than comprehensive probability method. It shows the method is suitable for hail recognition.

关 键 词：雹云 主分量分析 雷达 回波 

分 类 号：P426.5[天文地球—大气科学及气象学]