摘要
通过计算全球1951～2008年100hPa、500hPa高空资料、74项大气环流特征量、地面气压、太平洋海温以及中国160个测站的降水、气温等6种环流场和特征值与雨季开始期的相关性，找出了雨季开始期变化的高影响关键区，对关键区进行因子合成，筛选出了5种环流场的33个关键因子，分析因子的最佳界值和正贡献指数，经线性回归计算，建立分类环流一次预报方程和综合环流多因子集成预报二次方程，最后通过二次方程和正贡献指数方程建立三次集成预报模型，投入实际气候预测应用，4年来准确预测了玉溪雨季开始期。
By computing the correlation between circulation fields, characteristic values （including global high altitude data of 100 hPa and 500 hPa, 74 items of atmospheric - circulation characteristics, ground air pressure, sea surface temperatures of the Pacific during 1951 -2008 as well as precipitation, air temperature of 160 meteorological stations in China etc. ） and rainy season onset, then finding out the key area that has heavy influence on the variety in rainy season onset. Doing factor synthesis on the key area, and screening out 33 key factors in 5 kinds of circulation fields. Analyzing the best critical value and positive contribution index number of factors, then building up the linear forecast equation of classified circulation and the quadratic equation of comprehensive circulation multiple factors integrating forecast, finally building up the cubic integration model of forecast which predict the rainy season onset of Yuxi accurately in the actual usage for four years.
出处
《云南地理环境研究》
2009年第3期63-67,共5页
Yunnan Geographic Environment Research
关键词
多因变量
物理集成
雨季开始预测
multiple factors
physical integration
prediction of rainy season onset
