东亚冬季气温除了季节平均外，其显著的季内起伏也对国民生活及经济活动有着深远影响。本文利用1959-2018年台站及再分析资料，使用S-EOF(season-reliant Empirical Orthogonal Function)方法提取东亚冬季气温季内起伏的主要年际变化模态，其主要模态表现为前冬暖（冷）、后冬冷（暖），即为前、后冬反相，其方差贡献达到31.1%。这种前后冬反相的特征并非局地现象，在北半球大尺度均存在。环流场上它表现为欧亚遥相关型波列（Eurasian teleconnection, EU）从前冬12月的负位相（正位相）向后冬2月正位相（负位相）的转变，相伴随的是低层西伯利亚高压与阿留申低压的强度在前、后冬转折，高层副热带急流的变化也与之匹配。分析发现，欧亚遥相关型的季内转向可能与北大西洋涛动（North Atlantic oscillation, NAO）在前冬12月与后冬2月的转向有关，后者通过北大西洋热通量作用进而影响下游EU波列的转向。此外，宽窄厄尔尼诺—南方涛动（El Ni?o-Southern Oscillation, ENSO）事件也有一定贡献，当厄尔尼诺(El Ni?o)发生时，经向上更宽（窄）的海温异常利于前冬气温偏高（低）向后冬气温偏低（高）的转向；而当拉尼娜(La Ni?a)事件发生时，情况与厄尔尼诺年相反。
In addition to winter mean temperature, the conspicuous intraseasonal oscillation of East Asia winter air temperature (EAT) exerts a large influence on human activities and the economy. Based on the station data and the reanalysis data during the period of 1959-2018, we analyzed the intraseasonal spatiotemporal variability for East Asian winter temperature by using the method of season-reliant Empirical Orthogonal Function (S-EOF). The results show that the spatiotemporal of EAT is mainly characterized by the intraseasonal phase inversion variation pattern, which means warmer (colder) than normal in early winter and colder (warmer) in late winter. This mode accounts for the variance contribution by 31.1% and also represents the intraseasonal winter temperature over the entire Asian continent. The circulation evolutionary of this mode mainly characterized by the phase inversion of Eurasian teleconnection pattern (EU) from a positive (negative) phase in early winter (December) to a negative (positive) phase in late winter (February), accompanied by the of turning of lower Siberian high and Aleutian low intensity, and the intraseasonal evolution of the subtropical jet. The possible influencing factors which affecting the EU phase reversal are as follows: first, the circulation over the North Atlantic reverse between early winter and later winter and which influences the EU through the North Atlantic heat flux. Second, wide and narrow SST anomalous of ENSO events. In El Ni?o years, the events of wider (narrower) SST anomalies are more likely to inverse from the phase of warmer (colder) in the early winter to colder (warmer) in late. When La Nina events occur, cases on the opposite.