Does CMIP6 Inspire More Confidence in Simulating Climate Extremes over China?

Abstract

Based on climate extreme indices calculated from a high-resolution daily observational dataset in China during 1961–2005, the performance of 12 climate models from phase 6 of the Coupled Model Intercomparison Project (CMIP6), and 30 models from phase 5 of CMIP (CMIP5), are assessed in terms of spatial distribution and interannual variability. The CMIP6 multi-model ensemble mean (CMIP6-MME) can simulate well the spatial pattern of annual mean temperature, maximum daily maximum temperature, and minimum daily minimum temperature. However, CMIP6-MME has difficulties in reproducing cold nights and warm days, and has large cold biases over the Tibetan Plateau. Its performance in simulating extreme precipitation indices is generally lower than in simulating temperature indices. Compared to CMIP5, CMIP6 models show improvements in the simulation of climate indices over China. This is particularly true for precipitation indices for both the climatological pattern and the interannual variation, except for the consecutive dry days. The areal-mean bias for total precipitation has been reduced from 127% (CMIP5-MME) to 79% (CMIP6-MME). The most striking feature is that the dry biases in southern China, very persistent and general in CMIP5-MME, are largely reduced in CMIP6-MME. Stronger ascent together with more abundant moisture can explain this reduction in dry biases. Wet biases for total precipitation, heavy precipitation, and precipitation intensity in the eastern Tibetan Plateau are still present in CMIP6-MME, but smaller, compared to CMIP5-MME.

摘要

基于高分辨率的中国区域1961-2005年逐日观测资料,以及参与第6次耦合模式比较计划(CMIP6)的12个气候模式和第5次(CMIP5)的30个模式的结果,评估了模式对中国区域极端温度空间分布和年际变率的模拟能力。结果发现,CMIP6多模式集合平均(CMIP6-MME)能很好地模拟年平均温度、日最高气温最大值和日最低气温最小值的空间分布。但是很难再现冷夜和暖日的空间分布,且在青藏高原上存在很大的冷偏差。对极端降水的模拟性能通常低于极端温度。与CMIP5相比,CMIP6模式对中国区域极端气候的模拟能力得到了一定程度的改善。尤其是极端降水的气候态和年际变率都改善明显。比如,湿日总降水量区域平均的偏差从127% (CMIP5-MME)降低到79% (CMIP6-MME)。其中,最为显着的改善是,在CMIP5-MME中持续且普遍存在的中国南方降水的干偏差,在CMIP6-MME中显著减少。更强的上升运动和更充足的水汽输送可以解释CMIP6中干偏差的减小。青藏高原东部湿日总降水量、强降水和降水强度的湿偏差在CMIP6-MME中仍然存在,但其偏差小于CMIP5-MME。