CMIP6不同分辨率全球气候模式对中国降水模拟能力评估

基于参与CMIP6高分辨率模式比较计划（HighResMIP）9个模式组的18个全球气候模式模拟数据,通过与CN05.1观测资料的对比,评估了不同分辨率气候模式对中国区域1961—2014年降水特征的模拟能力。结果表明：低、高分辨率模式均能模拟出中国区域多年平均降水的总体空间分布特征,以及降水冬弱夏强的季节变化特征,但对降水的模拟都存在系统性偏多的误差;与低分辨率模式结果相比,高分辨率模式对降水空间分布的模拟有明显改善,在青藏高原、华北、华南地区降水模拟的系统性偏差明显减小;与低分辨率模式结果相比,高分辨率模式对年循环变化的模拟效果也更好,多年平均1月及9—12月逐月降水以及年降水的模拟误差均有所减小。对于年际、年代际的前两个主导空间模态,低、高分辨率模式大多无法模拟年代际的第一模态,但对于年际前两个模态以及年代际第二模态,分辨率提高可使半数左右模式组的模拟能力有所改善。

Evaluation on the performance of CMIP6 global climate models with different horizontal resolution in simulating the precipitation over China

Based on the 18 global climate models with different horizontal resolutions from the High Resolution Model Intercomparison Project (HighResMIP) of CMIP6, precipitation over China in the period of 1961-2014 was compared against the observation of CN05.1. Results show that all of the models can reproduce the spatial pattern of mean annual precipitation and the seasonal variation of dry winter and wet summer over China. Compared with the model with low-resolution, the high-resolution model can significantly improve the spatial distribution of precipitation, and the bias in the Tibetan Plateau, North China and South China is significantly reduced. The annual cycle of precipitation is also better simulated by the high-resolution compared to the low-resolution models, with lower bias in January, from September to December, and the annual precipitation. In addition, for the first and second leading modes of interannual and interdecadal, most of the low- and high-resolution models fail to simulate the first mode of interdecadal scale. But for the second mode of interdecadal, the first and second mode of interannual, the performance of about half models are improved with the increase of horizontal resolution.