Can CMIP6 models reproduce the influence of ENSO on the variability of boreal winter Hadley Circulation?
Shuang Wang,Falei Xu,Juan Feng
DOI: https://doi.org/10.1016/j.atmosres.2024.107588
IF: 5.965
2024-07-20
Atmospheric Research
Abstract:The Hadley circulation (HC) is the strongest and largest meridional circulation, playing a crucial role in influencing the global climate. A regime shift is observed in the boreal winter HC variability after 1980, presenting as an equatorially symmetrical pattern. This study examines the performances of 26 CMIP6 models in simulating the variability of boreal winter HC after 1980, analyzes the possible physical processes of the simulation uncertainties, and predicts its future variations. The results reveal that all models can reasonably capture the spatial structure of the climatological boreal winter HC. However, significant inter–model discrepancies are existed in the simulations of the spatial distribution of its linear trend, the dominant mode of HC (EOF1), as well as El Niño–Southern Oscillation (ENSO). By examining the spatial structure of HC EOF1 and ENSO between the simulations and observations, all the models are categorized into three groups, i.e., both the HC EOF1 and ENSO are well–simulated group (WWG), the simulation of HC EOF1 is unsatisfactory and ENSO is well–simulated group (UWG), and neither the simulations of HC EOF1 and ENSO is satisfactory (UUG). The group analysis suggests that the WWG models can well simulate the robust relationship between the HC EOF1 and ENSO, while the UWG and UUG models fail in reproducing the connection between the EOF1 and ENSO. In the WWG models, a greater variability of ENSO would associate with a stronger linkage between the HC EOF1 and ENSO, implying that the amplitude of ENSO variability determines the linkage between HC and ENSO. Conversely, the opposite situation is observed in the UWG and UUG models. The failure of UWG and UUG models in reproducing the relationship of HC EOF–ENSO is mainly due to its unsatisfactory simulation in the ENSO variability. This result shows that a model's ability in simulating the ENSO variability affects its capacity in simulating the relationship between the HC EOF1 and ENSO, which in turn impacting its performance in charactering the variability of HC. Furthermore, the HC EOF1 is projected to maintain an equatorially symmetrical structure towards the end of this century. Compared to the period 1980–2013, the variability of ENSO by century–end has intensified, leading to a stronger association between ENSO and HC EOF1 and a strengthened HC EOF1. These findings further support of the conclusion. Our work focuses on the discrepancies of the simulation of the variability in boreal winter HC and its future prediction in CMIP6, providing valuable insights for the advancement of climate models and understanding the relationship between ENSO and HC.
meteorology & atmospheric sciences