How Does Regional Convection‐Permitting Modeling Improve the Simulation of the Atmospheric Water Cycle in Spring Over the Tibetan Plateau?

The Tibetan Plateau (TP) provides vital water resources for downstream regions, with spring precipitation contributing considerably to the annual totals over the southeastern TP. The added value of convection‐permitting modeling in simulating the spring climate over the TP is uncertain. Here, we conducted and compared decade‐long regional convection‐permitting (3.3 km) and convection‐parameterized (13.2 km) Icosahedral Nonhydrostatic Weather and Climate Model (ICON) simulations to reproduce the atmospheric water cycle in spring over the TP. Results indicated that 3.3 km mesh ICON (ICON_3.3 km) exhibited notable added value in simulating the spring atmospheric water cycle over the TP. ICON_3.3 km reduced the wet biases of precipitation in the ERA5 reanalysis and 13.2 km mesh ICON (ICON_13.2 km) simulations, and improved the simulation of surface evaporation over the central and eastern TP. The reduction in the simulated precipitation in ICON_3.3 km was primarily followed by a decrease in surface evaporation from March to May, second by a reduction in water vapor flux convergence in May due to decreased water vapor inflow from the southeastern TP. Furthermore, compared to ICON_13.2 km, ICON_3.3 km alleviated the "drizzling" bias, leading to drier surface soils and decreased evaporation, and lead to 3% decrease in the fraction of evaporation converted into precipitation. Sensitivity experiments conducted at resolution of 13.2 km but turning off the convection parameterization demonstrated that both explicit representation of convection and enhanced horizontal resolution were crucial for accurately representing the spring atmospheric water cycle over the TP. Our results highlighted the need to develop kilometer‐scale models for successfully reproducing the climate characteristics across the TP. Plain Language Summary: Climate modeling is evolving toward kilometer‐scale or convective‐permitting model (CPM). The added value of regional CPM in simulating the spring climate over the Tibetan Plateau (TP), known as "Asian Water Tower," is unknown. Here, we have performed and compared two decade‐long regional climate simulations at two horizontal resolutions: convection permitting (3.3 km) and convection parameterized (13.2 km), to reproduce the atmospheric water cycle in spring over the TP. Results indicated that the CPM reduced the wet biases in the latest reanalysis data and the convection‐parameterized model over the central and eastern TP. The reduction of wet biases was mainly followed by the decreased surface evaporation, which was associated with the alleviation of "drizzling" bias in CPM. We further showed evidences that both explicit representation of convection and enhanced horizontal resolution were crucial for accurately representing the spring climate over the TP. Our findings highlight the importance of developing high‐efficiency kilometer‐scale (<4 km) models for reproducing all the climate characteristics across the TP region. Key Points: Regional convection permitting model exhibited evident added value in simulating the spring atmospheric water cycle over the Tibetan Plateau (TP)Reduced precipitation by convective‐permitting model was mainly followed by decrease in surface evaporation, associated with the alleviation of drizzling biasesBoth explicit convection and enhanced resolution were crucial for well representing the spring atmospheric water cycle over the TP [ABSTRACT FROM AUTHOR]