投影寻踪回归方法在长江三角洲地区降水量预报中的应用

PREDICTION IN RAINFALL AT CHANGJIANG DELTA BASED ON PROJECTION PURSUIT REGRESSION MODEL

本文首先利用北半球大气环流、SST场以及环流特征量指数等资料 ,普查了与长江三角洲地区降水量关系密切的预测因子 ,并根据投影寻踪回归方法(PPR)建立了该地区降水量预报模型。试验结果表明 ,该统计方法具有稳健性 。

The relationship of rainfall at Changjiang Delta and general circulation of North Hemisphere、SST in the North Pacific has been studied,by which we get some prediction factors.Then a projection pursuit regression model(PPR) has been established for short term climate prediction in summer rainfall.The test results show that the model is convenient and superior to apply to operational forecast.