Evaluation of CMIP6 for historical temperature and precipitation over the Tibetan Plateau and its comparison with CMIP5
Evaluation of CMIP6 for historical temperature and precipitation over the Tibetan Plateau and its comparison with CMIP5
复制标题
DOI：
10.1016/j.accre.2020.08.001
复制
发表时间：
2020-08
影响因子：
7.4
通讯作者：
Yuyao Zhu;Saini Yang
中科院分区：
文献类型：
--
作者：
Yuyao Zhu;Saini Yang
在更新的青藏高原耦合模式相互比较项目（CMIP6）中，有必要对历史模拟进行逐模式分析，以确定合理的模式。为了确定CMIP6模式在模拟高原温度和降水方面的能力是否得到了增强，我们将1961 - 2014年期间23个CMIP6模式的输出与观测数据集（CN05.1）进行了比较。结果表明，CMIP6模式在青藏高原的系统性偏差（冷偏差和湿偏差）仍然存在。CMIP6中的大多数模式真实地模拟了地面温度和降水的空间分布，模式相关性超过0.75。多模式集合（MME）模拟的平均表面温度偏差比观测数据低1.08 ° C，与CMIP 5的冷偏差（1.52 ° C）相比有所减少。在季节尺度上，大多数模式在夏季表现出温暖的温度偏差，在冬季表现出寒冷的偏差。与CMIP5相比，CMIP6 MME在干旱地区表现出更高的降水幅度再现性，而在潮湿地区表现出较低的能力。
A model-by-model analysis for historical simulations was necessary for identifying reasonably performing models in the updated Coupled Model Intercomparison Project (CMIP6) over the Tibetan Plateau. To determine whether the capacity of the CMIP6 models in simulating temperature and precipitation over the Plateau has been enhanced, we compared the outputs of 23 CMIP6 models with an observational dataset (CN05.1) for the period 1961–2014. The results suggest the systematic model biases (cold bias and wet bias) in the Tibetan Plateau still exist in CMIP6. Most models in CMIP6 realistically simulated the surface temperature and spatial distribution of precipitation, with a pattern correlation exceeding 0.75. The bias in the mean surface temperature of the multi-model ensemble (MME) simulation was 1.08 °C lower than the observational data, which had been decreased compared with the cold bias of CMIP5 (1.52 °C). At the seasonal scale, most models exhibited a warm temperature bias in summer and a cold bias in winter. The CMIP6 MME displayed a higher reproducibility of the precipitation amplitude over dry regions compared with CMIP5 and a lower ability over wet regions.