改进的经验正交函数分析方法及其效果验证

An adjusted empirical orthogonal function analysis method and its effect verification

基于中国站点网气候资料,简要介绍了经验正交函数(empirical orthogonal function,EOF)与改进的经验正交函数(adjusted EOF,AEOF)分析方法的关系,给出了验证其分析效果的方案;并将EOF、AEOF分析方法应用于中国160站60 a(1951—2010年)夏季气温距平场序列的分析。结果表明:AEOF分析方法前6个模态的模方拟合率和累积模方拟合率均高于EOF分析方法的相应量;AEOF分析方法前3个模态的典型场对夏季气温距平场序列的模方场的拟合效果明显优于EOF分析方法。由此论证了EOF分析方法中资料均匀化订正的必要性,验证了实际分析中AEOF分析方法的较优效果。

Based on the climatic data of station network in China,this paper briefly describes the relationship between the empirical orthogonal function（ EOF） analysis method and the adjusted EOF（ AEOF） analysis method,and gives schemes of analysis effect verification. Using EOF and AEOF methods,this paper analyses the anomalous summer air temperature series at 160 stations in China during 1951—2010.Results show that the mode square fitting rate and the cumulative mode square fitting rate of the first six modes by AEOF method are higher than those by EOF method. The fitting effect of the typical fields of first three modes by AEOF method on the mode square field of anomalous summer temperature series are better than that by EOF method. Therefore,the necessity of station homogeneity adjustment is demonstrated when EOF method is used,meanwhile the better adjustment effects using AEOF method are verified in the actual analysis.