中文摘要：

This study evaluates the ability of the global coupled climate models in hind casting the Arctic Oscillation(AO)and Antarctic Oscillation(AAO).The results show that the models can well simulate the spatial distribution of AO with better results in winter than in spring.In the troposphere in spring,the simulation of AO on the whole is still relatively good with a comparatively high correlation with the NCEP/NCARre analysis.The models can also well reproduce the spatial distribution of AAO throughout the year at all levels of the troposphere,and the spatial simulation is better at 850 hPa than at the surface.Although the simulation is better in winter than in other seasons,the seasonal variation is not so significant andthe differences among different models are relatively small.In addition,the capability of the models for "predicting" the AO and the AAO index time series is limited,because only a few models can capture their observed interannual variability at the 95% significance level.

英文摘要：

This study evaluates the ability of the global coupled climate models in hindcasting the Arctic Oscillation （AO） and Antarctic Oscillation （AAO）. The results show that the models can well simulate the spatial distribution of AO with better results in winter than in spring. In the troposphere in spring, the simulation of AO on the whole is still relatively good with a comparatively high correlation with the NCEP/NCAR reanalysis. The models can also well reproduce the spatial distribution of AAO throughout the year at all levels of the troposphere, and the spatial simulation is better at 850 hPa than at the surface. Although the simulation is better in winter than in other seasons, the seasonal variation is not so significant and the differences among different models are relatively smaU. In addition, the capability of the models for ＂predicting＂ the AO and the AAO index time series is limited, because only a few models can capture their observed interannual variability at the 95% significance level.