Responses of Convective Heavy Rainfall to Atmospheric Moisture Amplification: Implications for Probable Maximum Precipitation (PMP) Estimation

Abstract Mesoscale convective systems (MCSs) can produce extreme precipitation and flash flooding, emphasizing their importance for probable maximum precipitation (PMP). To estimate PMP using a numerical weather prediction model, the relative humidity maximization (RHM) method has been most commonly used, in which atmospheric grids in the modeling domain are fully saturated to maximize moisture availability for the storm. However, such drastic saturation practice may weaken thermodynamic gradients that are essential for MCS development, potentially reducing rather than maximizing precipitation. This study investigated the response of MCS-induced heavy rainfall to various atmospheric moisture maximization strategies using the Weather Research and Forecasting (WRF) Model, focusing on an extreme rainfall event in August 2022 in Japan. Applying RHM to all grids in the atmosphere resulted in weakened convective organization, reduced precipitation magnitude, and altered spatial rainfall distribution due to atmospheric stabilization and diminished low-level convergence. To address this, we selectively applied RHM to lower atmospheric layers and high-humidity regions to maintain a strong thermodynamic gradient, based on our understanding of the storm mechanism. This selective RHM approach successfully increased atmospheric moisture during the storm while maintaining (even enhancing) convective organization. The RHM practice below 900 hPa with the threshold of relative humidity > 90% preserved strong vertical and horizontal thermodynamic gradients and moisture flux convergence, leading to a 25.8% increase in 48-h maximum precipitation over the Arakawa basin. Thus, both vertical and horizontal gradients in thermodynamics matter for maximizing precipitation of MCSs. Further studies should apply similar methodologies to a broader range of MCS events to generalize our findings and to develop an effective PMP estimation methodology for MCSs.