Seasonal characteristics of the long-term sea surface temperature variability in the Yellow and East China Seas

Abstract

A multi-decadal sea surface temperature (SST) in the Yellow and East China Seas (YECS) has been widely believed to increase persistently associated with the variation of atmospheric circulation in the North Pacific. The AVHRR-based optimum interpolation SST anomalies after removing seasonality, however, exhibit that warming trends are quite localized with the characteristic temporal pattern of that recent cooling follows two peaks occurred at the end of 1990s and middle of 2000s. To explain the localized trends and the two peaks in the SST, spatial patterns and principal component time series of the SST anomalies are analyzed for the period of 1982&#8722 2014. Cyclostationary empirical orthogonal function (CSEOF) analysis separates three principal modes for the SST anomalies. Among modes, the first and the third are related to the long-term SST changes, while the second captures the interannual variability. The first mode, explaining 25% of the total variability in the SST anomalies, seems to link with the first SST peak at the end of 1990s through the meridional wind anomalies for winter and spring seasons. During the warming period before the first peak, the principal component of the first mode is highly correlated with the Pacific decadal oscillation index, and then its correlation sharply drops to an insignificant level. The third mode is likely to explain the second SST peak at the middle of 2000s, which could interpolation SST anomalies after removing seasonality, however, exhibit that warming trends are quite localized with the characteristic temporal pattern of that recent cooling follows two peaks occurred at the end of 1990s and middle of 2000s. To explain the localized trends and the two peaks in the SST, spatial patterns and principal component time series of the SST anomalies are analyzed for the period of 1982&#8722 2014. Cyclostationary empirical orthogonal function (CSEOF) analysis separates three principal modes for the SST anomalies. Among modes, the first and the third are related to the long-term SST changes, while the second captures the interannual variability. The first mode, explaining 25% of the total variability in the SST anomalies, seems to link with the first SST peak at the end of 1990s through the meridional wind anomalies for winter and spring seasons. During the warming period before the first peak, the principal component of the first mode is highly correlated with the Pacific decadal oscillation index, and then its correlation sharply drops to an insignificant level. The third mode is likely to explain the second SST peak at the middle of 2000s, which could