Assistant Professor Tohoku University, Miyagi, Japan
Abstract Submission: In Japan, line-shaped Mesoscale Convective Systems, recently named “Senjo-kousuitai” in Japanese, often result in catastrophic disasters. Significant efforts have been spent on accurately modeling such localized heavy rainfall events using Numerical Weather Prediction (NWP) models to better understand rainfall mechanisms, to improve forecast performances and to estimate design rainfall under changing climate. However, accurately modeling localized heavy rainfall events is still known to be challenging since it requires a comprehensive understanding of the effects of physical parameterizations, initial and boundary conditions, and model domain configurations on simulated rainfall amounts. This study examined the effects of physical parameterizations on the simulation performance of the recent Senjo-kousuitai events in the Tohoku and Hokuriku regions in Japan to find the optimal model configuration. Using the optimal model configuration, this study attempted to estimate maximum precipitation scenarios towards Probable Maximum Precipitation (PMP) estimates for the national river watersheds in the region: Akagawa River Basin, Arakawa River Basin, and Mogami River Basin. This study proposed the novel method considering climate change-induced thermodynamic changes in the model-based precipitation maximization, to examine the effects of climate change on maximum precipitation scenarios. The estimated maximum precipitation scenarios under the current, middle future (2050s), and far future (2090s) were compared to the operational estimates of the national government to discuss the validity of the estimates. Our findings can provide important insights for updating operational PMP estimates.
