Impact of Spatial Resolution on Modeling the Relationship Between Summertime Precipitation Extremes and Temperature Over Land in E3SM

Abstract Despite the well‐documented impact of spatial resolution on the model extreme precipitation intensity (EPI), it is unclear whether and how this might influence the relationship between EPI and meteorological conditions. In this study, we assess how this relationship is represented in the Energy Exascale Earth System Model (E3SM) at low (110 km), high (25 km), and ultra‐high (3.25 km, SCREAM) resolutions, using the Stage IV and IMERG data as observational references. The convection‐permitting SCREAM performs well in capturing the EPI variations with temperature and saturation deficit under a wide range of atmospheric saturation levels. When the near‐surface atmosphere is close to saturation, observations show a continuous increase in EPI with temperature, a pattern qualitatively well reproduced by SCREAM, although its short simulation of 40 days limits the range of temperature variation needed for a quantitative scaling rate comparison. In contrast, E3SM at 110 and 25 km resolutions produces an increase of EPI with temperature, followed by an erroneous negative scaling at high temperatures, even in saturated conditions. These comparisons are robust across all sample regions covering the midlatitudes, tropics, high altitudes, and mountainous areas, revealing that unless the spatial resolution is high enough to resolve deep convection, model physics dominates over spatial resolution in improving the performance in capturing the emergent relationship between EPI and the environmental drivers. Among other factors, our analyses indicate that the biased EPI scaling with temperature in the low‐ and high‐resolution E3SM simulations might be related to deficiencies in parameterizing the dependence of atmospheric convection on relative humidity.