Evaluation of CMIP6 HighResMIP models in simulating precipitation over Central Asia

• 作者: Liang-Liang LI Jian LI Ru-Cong YU
• 来源期刊: 气候变化研究进展（英文版）
• 年，卷（期）: 2022,（1）
• 所属期刊栏目: Changes in climate system
• 页码范围: 1-13
• 页数: 13页
• 语种: 英文

摘要

The High Resolution Model Intercomparison Project (HighResMlP) experiment within the Coupled Model Intercomparison Project Phase 6(CMIP6),for the first time,has provided an opportunity to evaluate the performance of climate models over complex topographies.Based on the HighResMIP's historical simulations of atmospheric general circulation models,the performances of global high-resolution models,with a horizontal resolution finer than 50 km,in representing precipitation over Central Asia were evaluated using rain gauge observation datasets.All the models successfully reproduce the large precipitation regions that are located over the mountainous areas and Northern Central Asia.However,nearly all the models overestimate precipitation frequency over Central Asia and large overestimations of precipitation amount and frequency are located over the mountainous areas.Although the HighResMIP multi-model ensemble mean performs better than all individual models at simulating the spatial pattern of precipitation frequency,it is inferior to HadGEM3-GC31-HM and ECMWF-IFS-HR at simulating that of precipitation amount.The simulation performance exhibits remarkable regional differences.Over the Qilian Mountains,the relationship between precipitation and elevation is totally captured by climate models.In contrast,over the Tianshan Mountains,the models fail to simulate the decrease in precipitation frequency after elevation higher than the maximum precipitation elevation.Most models successfully reproduce the annual cycle shape of precipitation amount over the Southern Central Asia,Qilian Mountains and Tianshan Mountains,but fail to reproduce it over the Northern Central Asia.More than half of the high-resolution GCMs have a reduced bias relative to the corresponding low-resolution GCMs.The performances of most high-resolution GCMs in simulating precipitation pattern are well over the Tianshan Mountains.