中文摘要：

亚洲-太平洋涛动是夏季欧亚大陆东部（15°-50°N，60°-120°E）与北太平洋上空（15°-50°N，180°-120°w）温度场反相变化的现象。亚洲-太平洋涛动指数由对流层上层（500--200hPa）温度定义，反映了亚洲-太平洋纬向热力差异。基于一个全球海-气耦合模式FGOALS_gl的20世纪气候模拟试验结果，讨论了其对20世纪亚洲-太平洋涛动指数变化的模拟能力。结果表明，较之ERA-40再分析资料（1960-1999年），模式很好地刻画出上层温度场的平均态和主导模态的空间型。从趋势上看，模式对北太平洋上空温度的年代际变化和趋势模拟较好，但未能模拟出亚洲东部陆地上空的降温趋势。从频谱分析结果看，模拟的亚洲-太平洋涛动指数2—3。a的年际变率与再分析资料相当，5—7a周期的变率较弱。模式能够较好地模拟出与亚洲-太平洋涛动指数相关的亚洲季风区气候异常。在20世纪模拟中，外强迫因子会改变耦合系统的年际变率，在自然因子强迫下亚洲-太平洋涛动指数的功率谱向低频方向增强，人为强迫因子的作用则相反。自然强迫因子和人为强迫因子在不同时期对亚洲-太平洋涛动年际和年代际变率的作用不同。在年际变率中人为强迫因子能够控制亚洲-太平洋涛动的变率使其不致过大；在年代际变率中人为强迫因子会增强自然强迫下亚洲-太平洋涛动的变率。模式上层温度的主导模态受ENSO调制，可能影响亚洲-太平洋涛动的年际变率。因此，模式对ENSO模拟能力的缺陷是制约模式对流层上层温度及亚洲-太平洋涛动指数变率的重要因素。

英文摘要：

The Asian-Pacific Oscillation （APO） is a phenomenon in which the temperature changes out of phase over the East Asia continent （15°-50°N, 60°-120° E） with the North Pacific （15°-50°N, 180°-120°w） in the upper troposphere. The APO index reflects the Asian-Pacific zonal thermal contrast. The performance of the fast coupled version of the LASG/IAP climate system model FGOALS_gl in simulating the upper troposphere temperature and the APO index over the 20th century is evaluated. Compared with the ERA-40 reanalysis data, it is shown that the model performed well in simulating the climatology and the dominant modes of the upper troposphere temperature. However, the results show that the simulated APO index failed to capture the descent trend after !960s over the East Asia continent as indicated in the ERA-40 data. Based on the power spectrum analysis, the 2 - 3 a variability of the model APO index is equivalent with that in the reanalysis but the 5-7 a variability is weaker. Despite several regional departures, the large-scale circulation over Asian monsoon section related with the APO index is well reproduced in the model. A comparison among the 20th century simulations shows that external forcing could change the interannual variability of a couple system. The natural forcing causes a spectrum shift to low frequency and the anthropogenic forcing does inversely. Natural forcing and anthropogenic forcing can play different roles in different periods. It seems that anthropogenic forcing could limit the interannual variability of APO and enhance the interdecadal variability. The dominant mode of the upper troposphere temperature in the model is modulated by ENSO and further impacts the interannual variability of APO. The defect of the model in the ENSO simulation may be an important limitation to reproducing the upper troposphere temperature and the variability of APO index.