Advancing Precipitation across the United States through Multi-decadal Convection-Permitting Simulations

Abstract

Abstract Climate simulations at convection-permitting (CP) spatial resolution could offer an improved representation of precipitation. To fully assess the benefits of CP simulations, we need multi-decadal simulations at global and regional scales. This first-in-class study presents results from 20-year simulations using the Weather Research and Forecasting model at the CP scale (ADDA_V2), covering a large domain covering North America and up to Puerto Rico. We evaluated ADDA_V2’s performance in capturing mean and extreme precipitation across the contiguous United States (CONUS). ADDA_V2 can capture the spatial variability and statistics of the mean and extreme precipitation over most of the CONUS. Particularly compared with the reanalysis data, ADDA_V2 exhibits superior performance in orographic-induced (mesoscale-convection-induced) winter (summer) precipitation, adding significant value to more than 40% of the grids in the CONUS. ADDA_V2 clearly improves the representation of diurnal precipitation patterns, the 95th percentile of daily precipitation, consecutive dry days, maximum 5-day precipitation, and very heavy precipitation across all seasons due to improved representation of orographic effects east of the Rocky Mountains and tropical systems in the Southeast. These results provide an incentive for using these simulations for robust assessments of future climate change impacts at regional and local scales.