中国冬季气温对ENSO的非线性响应特征

Nonlinear Response Characteristics of Winter Temperature to ENSO

运用一双排套装的非线性神经网络统计模型,讨论了中国冬季气温对ENSO事件的非线性响应特征。结果表明,响应具有较明显的非线性特征,且这种非线性主要表现在振荡强度的差异上,即气温对El Nino事件的响应程度强于对La Nina事件的响应。El Nino位相下,气温的正距平中心主要出现在东北、江淮、福建以及新疆南部;La Nina位相下,上述地区气温出现负距平中心,但振荡强度较弱。气温对ENSO的响应可分为线性响应部分和非线性响应部分,其中非线性响应解释的方差约占线性响应的2/5,因此非线性响应的作用不可忽视。

By a double-barreled nonlinear NN statistical model, the nonlinear response characteristics of China＇s winter temperature to ENSO are discussed. The results show that the response has distinct nonlinear characteristics and the nonlinearity is mainly represented by the difference of oscillation intensity, namely, the response of temperature to El Nino event is much stronger than that to La Nina event. During the El Nino phase, the positive anomalies are mainly centered over Northeast China, including Jianghuai, Fujian and the south part of Xinjiang; while during the La Nina phase, the negative anomalies occur over the above area, however, the oscillation intensity is much weaker. Also, the response of winter temperature to ENSO can be separated by a linear part and a nonlinear one, the nonlinear response accounts for about 2/5 as much variance as the linear part, so the impact of nonlinear response can＇t be ignored.