摘要
本文以ERA5再分析数据下的结果为参考,对比并评估了CMIP6中6个模式(CESM2-WACCM、BCC-ESM1、MRI-ESM2-0、EC-Earth3、CanESM5和NorESM2-MM) Historical实验下无强迫逐日数据对两类SSW (偏移型和分裂型)的基本统计特征的模拟能力。首先,与ERA5下的结果相比,模式数据总是会低估SSW事件的发生频率,尤其是对于分裂型,大部分模式模拟出的频率甚至不到参考值的一半。其次,SSW事件主要发生在1~3月份,而模式对SSW发生月份的模拟则表现出向晚冬(2~3月)的偏移。此外,发生在12月和1月的SSW往往具有更大的振幅,模式对此的模拟效果整体较好。
Simulations of vortex displacement and split stratospheric sudden warmings (SSWs) are evaluated for six models’ daily outputs (CESM2-WACCM, BCC-ESM1, MRI-ESM2-0, EC-Earth3, CanESM5 and NorESM2-MM) from phase 6 of the Coupled Model Intercomparison Project (CMIP6) with the ERA5 reanalysis as a baseline. Statistical characteristics are mainly evaluated. Compared with nearly six SSWs per decade in ERA5, all models underestimate the SSWs frequency, especially for split SSWs (only a half of baseline). SSWs mainly appear in midwinter in ERA5, but one-month climate drift is simulated in the models. Besides, the SSWs in December and January have a lager amplitude, which is well reproduced by most models.
出处
《自然科学》
2022年第1期8-19,共12页
Open Journal of Nature Science
