Abstract:Based on the reanalysis data of ERA interim and sea ice concentration data of the British Atmospheric Data Centre (BADC), the main spatio temporal variation characteristics of the prevailing Eurasian weather pattern at 500 hPa in the middle troposphere over 3330 days during 1979-2016 and its possible relationship with the anomalies in the mid lower troposphere over the Arctic and the decrease of Arctic sea ice in the recent years are analyzed by using the complex vector empirical orthogonal function (CVEOF). The results show that CVEOF1 accounts for 15.82% of the total anomalous kinetic energy and its two sub patterns are represented by tripole patterns (0° phase and 180° phase) and dipolar patterns (90° phase and 270° phase). The weather patterns of 180° phase and 270° phase were observed in winter when the mid low troposphere over Actic in winter was warmer than normal and the warm Arctic cold Eurasian atmosphere circulation prevailed in the Northern Hemisphere. In addition, sea ice concentration of the Barents Sea east to the Beaufort Sea was less than normal in autumn, which might be one of the impact factors. Furthermore, the prevailing weather patterns (180° phase and 270° phase) have been increasing in the recent years, and are also closely linked to the extreme weather events in winter. Take the winter of 2005/2006 and the winter of 2011/2012 examples. During the cold events of the two years, the frequency of 180° phase and 270° phase was obviously more than normal, accounting for the majority days of winter. Therefore, sea ice of the Barents Sea east to the Beaufort Sea less than normal in autumn, Arctic middle and lower troposphere in winter warmer than normal, 180° phase and 270° phase weather patterns prevailing in winter might be the impact factors for frequent winter extreme weather events.