摘要
为了给防雷工程设计、雷电损害风险评估和建筑物站址选择提供更加科学准确的年平均雷暴日资料,引入主成分典型相关分析(PC-CCA)方法,建立了月雷暴日资料延长模式,用多年人工观测雷暴日资料,对江西省区域内闪电定位网探测的月雷暴日数场序列进行订正延长试验。试验结果表明,模式中提取5个主分量的总体效果最好,PC-CCA方法用于月雷暴日资料的延长是可行的;在此基础上,进行各月累加平均求取年平均雷暴日资料也是可行的。
For thunder protection project design, risk assessment of thunder and lightning damage and choosing building location, datum extension experiment is implemented and datum extension model about monthly number of thunderstorm days detected by lightning localizer in Jiangxi province is established based on canonical correlation analysis （PC-CCA）. Result shows that the whole effect is the best when selecting five principal components. The datum extension method based on PC-CCA is feasible. And then perennial average of monthly number of thunderstorm days is obtained.
出处
《气象与减灾研究》
2006年第4期17-20,共4页
Meteorology and Disaster Reduction Research
基金
江西省科技厅社会攻关基金资助项目(20041B0302300).
关键词
典型相关
雷暴日数
资料序列
延长
分析
Canonical correlation analysis, Number of thunderstorm days, Datum extension ,Analysis.
