高级检索
赵丹, 张丽霞, 周天军. 2022. CMIP6模式对中国东部地区水循环的模拟能力评估[J]. 大气科学, 46(3): 557−572. doi: 10.3878/j.issn.1006-9895.2106.21030
引用本文: 赵丹, 张丽霞, 周天军. 2022. CMIP6模式对中国东部地区水循环的模拟能力评估[J]. 大气科学, 46(3): 557−572. doi: 10.3878/j.issn.1006-9895.2106.21030
ZHAO Dan, ZHANG Lixia, ZHOU Tianjun. 2022. Performance Assessment of CMIP6 Model in Simulating the Water Cycle over East China [J]. Chinese Journal of Atmospheric Sciences (in Chinese), 46(3): 557−572. doi: 10.3878/j.issn.1006-9895.2106.21030
Citation: ZHAO Dan, ZHANG Lixia, ZHOU Tianjun. 2022. Performance Assessment of CMIP6 Model in Simulating the Water Cycle over East China [J]. Chinese Journal of Atmospheric Sciences (in Chinese), 46(3): 557−572. doi: 10.3878/j.issn.1006-9895.2106.21030

CMIP6模式对中国东部地区水循环的模拟能力评估

Performance Assessment of CMIP6 Model in Simulating the Water Cycle over East China

  • 摘要: 本文基于观测和再分析资料,采用Brubaker二元模型评估了第六次国际耦合模式比较计划(CMIP6)中19个模式对中国东部季风区气候态水循环过程的模拟能力,并分析了模拟误差来源。结果表明,CMIP6模式集合平均(MME)能够合理再现观测降水和蒸发的年平均气候态空间分布及年循环特征,与观测值的空间相关系数分别为0.92和0.87。较之观测,MME高估了华北地区降水(0.55 mm d−1),低估了华南沿海地区降水(−0.3 mm d−1)。所有CMIP6模式均高估蒸发强度(偏差0.03~0.98 mm d−1),使得模拟的降水与蒸发之差偏少。模式整体能够模拟出我国东部季风区降水再循环率及不同边界水汽来源的贡献率,但低估了由南边界进入季风区的水汽贡献,导致东亚季风区偏干。通过分析模式对影响水汽通量的两个气象要素(风速和大气比湿)的模拟能力,发现研究区南边界的风速大小决定了模式间水汽输送差异。南边界风速越大的模式,由南边界进入的水汽通量越大,模式模拟的降水越多。西北太平洋辐合带的东西位置是影响南边界南风速的重要系统之一,辐合带位置偏东的模式模拟的南风强度较弱,使得水汽输送偏弱、降水偏少;反之,南边界水汽输送偏强、降水偏多。本文通过评估最新一代CMIP6模式在东亚水循环方面的模拟性能,指出了当前气候模式在模拟西太平洋辐合带位置方面存在的偏差及其对东亚水循环的影响。

     

    Abstract: This study evaluates the performances of 19 models from the Coupled Model Intercomparison Project Phase 6 (CMIP6) in simulating the water cycle over East China based on observations and reanalysis data using the Brubaker model. Sources of model bias are also investigated. Results reveal that the CMIP6 multi-model ensemble (MME) can reasonably simulate the climatic distribution and annual cycle of precipitation and evaporation with a pattern correlation coefficient of 0.92 and 0.87, respectively. Compared with observations, MME overestimates the precipitation (0.55 mm d−1) in North China but underestimates the precipitation (−0.3 mm d−1) in coastal areas of South China. All 19 models overestimate the evaporation with biases of 0.03–0.98 mm d−1. Thus, differences between the simulated precipitation and evaporation by most models are smaller than those of the observation and reanalysis data. The MME can well simulate the annual cycle of the contribution of each moisture source to the precipitation but underestimates the contribution of remote moisture via the southern boundary, resulting in a dry bias over the study region. It is found that the southerly wind speed over the southern boundary determines the difference in the water vapor transport among CMIP6 models. The stronger the southerly wind speed is in the model, the higher the water vapor flux incomes via the southern boundary, and the more precipitation the model simulates. The position of the convergence zone over the Northwest Pacific is one of the important systems affecting the southerly wind speed over the southern boundary. The eastward shift of the convergence position in the model results in weaker southerly winds, leading to a weaker moisture transport to the study region and less precipitation, and vice versa. This study systematically evaluates the performance of CMIP6 in reproducing the East Asian water cycle and demonstrates the limitation of the models in simulating the convergence zone over the Northwest Pacific and its impact on the East Asian water cycle.

     

/

返回文章
返回