赤道太平洋次表层海温距平的EEOF分析及与厄尔尼诺的关系

The Extended EOF analysis of subsurface temperature anomaly in the equatorial Pacific Ocean and its Relation to El Nino

通常气候变量场在时间上存在着显著的自相关及交叉相关,扩展的经验正交函数(ExtendedEOF,以下记作EEOF)分析是在经典的EOF基础上发展而来的,它同时考虑了要素的空间和时间相关性,可以得到变量场的移动性分布结构。本文通过对赤道太平洋次表层海温距平(SOTA)场的EEOF分解,发现第一特征向量是关于ElNino的模态,它反映了ElNino的发生持续消亡的整个过程,对应的时间系数(第一主分量)与Nino3指数有很好的同时相关。第二特征向量是关于西太平洋暖池的模态,它反映了西太平洋暖池从暖位相到冷位相(同时东太平洋从冷位相到暖位相)的过程,第二主分量与滞后6～10个月的Nino3指数有很好的相关性。这两个主分量不但有助于了解赤道太平洋海温异常的过程,而且为厄尔尼诺的预报提供了重要线索。

Climate variables usually have remarkable auto-correlation and cross-correlation in temporal fields. Extended EOF （empirical orthogonal functions） analysis based on classical EOF takes spatial and temporal correlation into account in the same time, and can get moving distributions of variables. By the EEOF analysis of subsurface ocean temperature anomaly（SOTA） in the equatorial Pacific Ocean, It is founded that, the first eigenvector is the mode about El Nino and it reflects the entire generation, persistence and diminishment process of E1 Nino phenomenon, there is good contemporary correlation between the corresponding time coefficient （first principal component） and the Nino3 index. The second eigenvector is the mode about Western Pacific Warm Pool（WPWP） and it reflects the process from warm phase to cold phase for WPWP （at the same time, from cold to warm for eastern Pacific）, there is also good correlation between the second principal component and the Nino3 index of 6 -10 months lag. This two principle components not only help to learn the process of SOTA in the equatorial Pacific Ocean, but also give us an important clue to forecasting El Nino.