Evaluation of spatial–temporal features and circulation patterns of Northeast China cold vortex in CMIP6 AMIP simulations

Northeast China Cold Vortex (NCCV), known as cut-off low, frequently occurs and induces low temperatures, rainstorms, and even catastrophic regional weather. This study evaluated the performance of Beijing Climate Center-Climate System Model-Medium Resolution (BCC-CSM2-MR), Chinese Academy of Sciences-Earth System Model version 2.0 (CAS-ESM2), Community Earth System Model version 2 (CESM2) and Finite Volume 2 (CESM2-FV2), Flexible Global Ocean–Atmosphere–Land System model, versions f3-L (FGOALS-f3-L) and g3 (FGOALS-g3) in capturing the spatial and temporal features of NCCV during summer, using CMIP6 AMIP and JRA-55 daily data over 1979‒2012. The biases of the six models were discussed from the perspective of the mean flow and Rossby wave trains. The six models generally underestimated the occurrence and duration of NCCV, but most of them overestimated its extent and intensity. Since these models featured a dipole of 500-hPa height anomalies with positive (negative) anomalies equatorward (poleward) of the westerly jet, they generated NCCVs associated with modest blocking-type circulations over northeastern Asia. Moreover, this dipole was linked to a more northward and pronounced East Asian jet, which deformed and conveyed NCCVs eastward more quickly. Therefore, the models produced fewer populations of NCCVs with a more northward distribution and shorter duration. To offset the eastward advection by the stronger westerly, the models required more powerful Rossby wave trains to generate NCCVs. Therefore, the models tended to produce more extensive and intense NCCVs, particularly BCC. However, the models were unable to overpopulate stronger Rossby wave trains to produce comparable numbers of NCCVs with respect to JRA-55.