Improved Simulations of Atmospheric River Climatology and Variability in High-Resolution CESM
Xue Liu,Ping Chang,Dan Fu,Ramalingam Saravanan,Hong Wang,Nan Rosenbloom,Shaoqing Zhang,Lixin Wu
DOI: https://doi.org/10.1029/2022MS003081
IF: 8.469
2022-01-01
Journal of Advances in Modeling Earth Systems
Abstract:Atmospheric rivers (ARs), referring to long and narrow filamentary bands of intense water vapor transport in the atmosphere, can cause extreme precipitation, floods, and drought events. Their variability has been linked to various climate modes, such as El Nino/Southern Oscillation, Pacific Decadal Oscillation, and Pacific-North America pattern. Understanding and improving simulation of this linkage can provide the potential to predict ARs at subseasonal-to-decadal timescales. Up to now, the extent to which climate model simulations of ARs are dependent on model resolutions has not been fully assessed. Here, we compare and evaluate ARs in a pair of high-resolution (HR) and a low-resolution (LR) Community Earth System Model (CESM) simulations against the observations. The results show that within this CESM framework, climatological AR strength and associated precipitation are severely underestimated by LR compared to the observations, and their relationships with major modes of climate variability are also poorly reproduced. These deficiencies in LR are alleviated to a significant extent in HR. Using a linear regression analysis, we show that HR is able to capture with high fidelity the observed relationships between ARs and major climate modes in the Northern Hemisphere at seasonal-to-decadal time scales. These results suggest that the use of HR CESM may potentially lead to an improved forecast skill of ARs at season-to-decadal time scales. Plain Language Summary Atmospheric rivers (ARs) can trigger extreme rainfall and snowfall events, causing high risk of flooding. Meanwhile, they are sometimes responsible for the occurrence and break of drought events. So, it is of great importance for weather and climate forecast models to accurately represent ARs. Studies have shown that AR variability is linked to various climate modes, such as El Nino/Southern Oscillation, Pacific Decadal Oscillation, Pacific-North America pattern, providing the potential for predicting ARs at subseasonal-to-decadal timescales. In this study, we examine the impact of climate model resolution on AR simulations and their relationship with major climate modes by comparing a pair of high-resolution and a low-resolution (LR) Community Earth System Model (CESM) simulations against the observations. The results reveal major deficiencies in LR that include underestimated climatological AR strength and associated precipitation, and inaccurate representations of ARs relationships with climate modes. Significant improvements are observed by increasing CESM horizontal resolution, suggesting a path for more accurate prediction of ARs.