Sensitivity of Tropical Oceanic Convection to Horizontal Model Resolution in Idealized Simulations Forced with PRECIP Observations

Recent studies have shown that tropical oceanic phenomena, including equatorial waves and organized rainfall systems, are better represented by models with explicitly resolved convection than with a convective parameterization. However, other studies suggest that typical convection-permitting resolution (i.e., 1-5 km grid spacing) is insufficient to resolve the narrow nature of tropical convective updrafts. Those same studies suggest that a convective parameterization is needed, even at sub-10-km grid spacing, to accurately represent tropical convection including extreme rainfall. To address these issues, this study seeks to investigate the sensitivity of tropical oceanic convection to horizontal model resolution using a hierarchy of model simulations—using grid spacings all the way from 15 km to 50 m. The simulations are initialized with an observed sounding from the 29 July 2022 extreme rainfall event that was sampled in detail during the Prediction of Rainfall Extremes Campaign in the Pacific. Distributions of rainfall rate and simulated reflectivity are compared to understand the similarities and differences amongst different resolutions. A comparison against statistics from radar and disdrometer observations further helps establish the advantages and deficiencies of each model resolution.