Abstract: Eurasian winter sea level pressure anomalies during 1948-2007 were investigated by applying a nonlinear principal component analysis (NLPCA) method. The results show that there was a distinct nonlinear characteristic in Eurasian winter sea level pressure anomaly changes. The spatial distribution was obviously asymmetrical when the first nonlinear principal component (NLPC1) took various positive or negative anomalous values. It is found that there was an evident interdecadal change of the intensity and period of interannual oscillations for the first nonlinear principal mode in the early 1970s, afterwards the nonlinear interannual oscillation was significantly enhanced.

Key words: sea level pressure, nonlinear spatial mode, Eurasia, winter, interdecadal variation