Impacts of Resolution on Heavy-Precipitating Storms in Climate Model Hindcasts

Abstract

The present study investigates the impact of horizontal resolutions on heavy-precipitating storms using the Energy Exascale Earth System Model version2 (E3SMv2) at low (∼100 km, LR) and high (∼25 km, HR) resolutions through short-range hindcasts. The short-range hindcast approach ensures a faithful comparison of model resolution in simulating the same storm events under a controlled large-scale environment. Using a phenomenon-based framework, we attribute precipitation to specific storm types: tropical cyclones (TCs), extratropical cyclones, atmospheric rivers, and mesoscale convective systems (MCSs). Our findings show that E3SM hindcasts with both HR and LR configurations significantly underestimate storm-associated precipitation intensity but overestimate precipitation from other sources. Furthermore, both HR and LR hindcasts face significant challenges in accurately simulating extreme precipitation events, particularly over MCS hotspots. Nevertheless, HR simulations capture more detailed and intense precipitation patterns with an improved representation of storm dynamics. HR hindcasts produce 16% more storm precipitation compared to LR. For precipitation extremes, HR simulates a 33% higher 99th percentile precipitation magnitudes compared to LR, and most of the increment comes from these four heavy-precipitating storm types. The increase in precipitation mainly comes from stratiform precipitation rather than convective precipitation. The improvement in HR simulations varies across different storm types with TC showing the largest improvement. The phenomenon-based approach provides important insights into precipitation simulations especially for extremes. Our results emphasize the need for further refinement in high-resolution models to improve the accuracy of precipitation predictions, which is crucial for better understanding and mitigating climate change impacts.