PATTERNS OF INTERANNUAL VARIABILITY DEFINED BY A CLUSTER-ANALYSIS AND THEIR RELATION WITH ENSO

Results of the GOADS data processing aimed at a classification of the structure of the large-scale low-frequency variability of the sea surface temperature, the sea-level atmospheric pressure and the surface wind-speed components are presented. Using cluster analysis, patterns of interannual variability in the temporal and spatial domains are obtained. Clusters consisting of points possessing similar temporal evolution are compared with the known dipole-type patterns of interannual variability, and special attention is given to selection of the global ENSO-related patterns. Of particular interest are the teleconnection links between the equatorial and the extratropical regions in the Pacific Ocean. Clusters consisting of times with similar spatial patterns, as well as extended canonical correction analysis patterns, show the typical sequence of the spatial patterns during the ENSO cycle, especially for the 'mature' warm and cold phases. For these cases, typical features of the field anomalies, including the extratropical peculiarities in the Pacific Ocean, are considered and summarized.