CMIP5模式对中国年平均气温模拟及其与CMIP3模式的比较

利用CRUT3v和CN05两套观测资料，评估25个CMIP5模式对1906-2005年中国年平均气温变化的模拟能力，并与CMIP3模式对比。结果表明：1906-2005年中国平均温升速率为0.84℃/100a，CMIP5多模式集合平均模拟的增温率为0.77℃/100a。模式对20世纪后期温升模拟好于前期，仅有两个模式能模拟中国20世纪40年代异常增暖。模式对气温气候态空间分布模拟较好，但在中国西部地区存在最大模拟冷偏差和不确定性。1961-1999年，中国北方增暖大于南方。多模式集合平均可以较好地模拟气温变化线性趋势的空间分布，但对南北气温变化趋势的差异模拟过小。总体说来，在中国平均气温变化趋势、气温气候态空间分布和气温变化趋势空间分布三方面，CMIP5模式都较CMIP3模式有所提高。

Assessment of CMIP5 Simulations for China Annual Average Surface Temperature and Its Comparison with CMIP3 Simulations

The simulation for China annual average surface temperature by CMIP5 models was assessed, referring to the observation data from CRUT3v and CN05. The results show that CMIP5 models are able to simulate the warming occurring in China over 1906-2005 with a trend of 0.77℃/100a from the multi-model ensemble mean, close to the observed value of 0.84℃/100a. The simulations of warming for the late 20th century are much better than that for the early 20th century, and only 2 out of the 25 CMIP5 models can capture the anomalous warming period around the 1940s. The simulations for the China-wide distribution of 20-year (1986-2005) averaged surface temperature are broadly well. However, notable underestimations for surface temperature climatology appear almost all over China, and the largest bias and uncertainty occur over western China. On the regional scale, northern China experienced stronger warming than southern China over the period of 1961-1999, for which the CMIP5 multi-model ensemble mean provides roughly realistic simulations, but underestimates the difference in warming trends between the North and the South. To sum up, the simulations from CMIP5 models are obviously improved in the three aspects mentioned above, compared with the CMIP3 simulations.