奇异交叉谱分析及其在气候诊断中的应用

Singular Cross-Spectrum Analysis and Its Applicability in Climatic Diagnosis

提出了一种新的用于气候诊断的奇异交叉谱分析方法（记为ＳＣＳＡ）。从理论上证明，它是一类时频域相结合的广义交叉谱分析，也是奇异谱分析（ＳＳＡ）的一个推广。ＳＣＳＡ可获得比经典交叉谱更为强化的耦合振荡信号，并在时域上描述两个系统之间各种耦合振荡信号的时变特征，因而可将频域上的耦合振荡信号在时域上加以合成和分解，包括非线性耦合振荡的弱信号，这对短期气候预报十分有益。文中实例证明，ＳＣＳＡ比经典交叉谱分析有更为优良的特性。

A new singular cross-spectrum analysis (SCSA) is presented for climatic diagnosis It is theoretically demonstrated that the SCSA is a generalized analysis method for cross-spectrum by a joint manner with both time and frequency domains, and the SCSA is also an extension of singular spectrum analysis (SSA) Since the enhanced signals of the coupled-oscillation may be obtained, its features of variation with time for the coupled-signals of different scale between two systems may be described in time domain by way of the SCSA, and the coupled-oscillation signals can be decomposed and composed in time domain Thus, the SCSA method is very useful for short range climatic prediction The real case analyses show that the SCSA has better features than the classical cross-spectrum