中文摘要：

应用Lamb-Jenkinson大气环流分型方法,根据由1970—2012年NCEP/NCAR逐日海平面气压场计算得到的环流指数,对淮河流域的环流进行分型。分析了冬季主要环流型和气温的分布特征及两者的联系,并以环流指数和主要环流型为预测因子,结合经验正交函数分解（EOF）方法和逐步回归方法,建立了淮河流域冬季气温距平的预测模型。结果表明,淮河流域冬季的主要环流型是东北风型、东风型、反气旋环流型以及东北风、东风配合下的反气旋性环流型,划分的环流型符合实际情况,这些环流型具有显著的年际和年代际变化特征。通过对预测模型进行后报试验和独立预报试验,表明该模型具有一定的预报技巧。

英文摘要：

Using of Lamb-Jenkinson atmospheric circulation typing method and daily sea level pressure data from NCEP/NCAR, six circulation indices were calculated to analyze the circulation catalogue over Huaihe valley. And the circulation indices and dominant atmospheric circulation patterns were introduced to construct the winter temperature anomaly prediction model as predictors by using the Empirical Orthogonal Function （EOF） method and the Stepwise multiple regression method over Huaihe valley. The results show that main circulation catalogues impacting on climate of Huaihe valley are northeasterly type, easterly type, anticyclone type, anticyclone northeasterly type and anticyclone easterly type, respectively. These circulation types present obvious interarmual and interdecadal characteristics. According to the results of hindcasts and independence prediction experiments, it indicates that the forecast model presents a certain degree of forecast skill.