均生函数—最优子集回归在高温极值预测中的应用

Application of Mean Generating Function-Optimal Subset Regressionto the Prediction of High Temperature Extremes

根据南京等城市 1961年以来历年气温极值资料序列 ,采用均生函数—最优子集回归法设计的短期气候预测模型不但能较好地拟合历史实况 ,而且对未来 1～

According to the yearly high temperature extremes from 1961 to 1996 in Nanjing and other three cities, the short range climate prediction models are developed by use of mean generating function and optimal subset regression methods. The results show that they can not only fit the historical sequences perfectly, but also possess predictive capability for coming 1～5 years' changes to a certain extent.