BCC_CSM1.1气候模式年代际试验对中国气候的回报能力评估

对比国家气候中心耦合模式BCC_CSM1.1提交CMIP5的历史（Historical）试验和年代际（Decadal）回报试验对中国气候及其年代际变化的模拟。结果表明，Decadal试验回报的中国降水气候分布更接近观测，回报的中国东部气温和降水的年代际距平误差比Historical试验减小明显。对于发生在20世纪70年代末的中国东部降水年代际变化，Decadal试验能回报出长江中下游降水增多的特征，但Historical试验模拟的降水变化与观测相反。由于Decadal试验和Historical试验的区别之一是后者利用观测海温资料进行了初始化，为了探讨观测海温信息的重要性，进一步将Decadal试验与恢复（Nudging）试验（即模式积分过程中，模拟海温始终向观测海温恢复）的模拟结果进行对比。发现Nudging试验能够较好地模拟出“南涝北旱”型降水变化，也能够模拟出相应的东亚急流增强且偏南的特征。这表明气候模式对海温的回报能力是影响其对东亚气候年代际异常模拟的一个重要因素。

Hindcast of China Climate With Decadal Experiment by BCC-CSM1.1 Climate Model

Simulations of China climate and its decadal changes are compared between Historical experiment and Decadal experiment of Beijing Climate System Model version 1.1 (BCC_CSM1.1) submitted to CMIP5. Results show that the climatological distribution of the precipitation in China is better simulated by the Decadal experiment. The error of decadal anoamlies in China simulated by the Decadal experiment is less than the Historical experiment. As for the decadal change of the East China precipitation in the late of the 1970s, the Decadal experiment can simulate the increased precipitation in the Yangtze River basin, while the Historical experiment produce the opposite feature with the observation. The main difference between Decadal experiment and Historical experiment is that the first one initializes the observed sea surface temperature. In order to discuss the importance of simulated sea surface temperature (SST) by the coupled climate model, the Nudging experiment (the modeled SST is nudged to the observed SST during the simulation) and the Decadal experiment were compared. It is found that the Nudging experiment can significantly simulate the “North drought and South Flood” feature in East China. Moreover, it can reproduce the stronger and southward movement of the East Asian jet stream. Those comparisons indicate that the hindcast of the SST by coupled models may have important influence on model’s capability in prediction of East Asian climate and its decadal changes.