[A-AS05] 気象の予測可能性から制御可能性へ

[AAS05-P07] Employing parameterization schemes to proliferate large-member ensemble for heavy rainfall simulations: Susceptibility to the precipitation and trigger

*Gaurav Tiwari¹、Atsushi Okazaki^1,2 (1. Center for Environmental Remote Sensing (CEReS), Chiba University、2.Advanced Academic Research (IAAR), Chiba University, Japan)

キーワード：Heavy Rainfall Events, Typhoons, Numerical Weather Prediction, SCALE-RM

Japan experiences a lot of Heavy Rainfall Events (HREs), especially during the rainy and typhoon seasons. The geography, location, and closeness to warm ocean currents make the nation susceptible to periods of intense rainfall, which can result in landslides, flooding, and other associated risks. Extreme HREs have been reported in Japan's Kyushu, Chugoku, Kanto, and Tohoku regions, especially in recent years. Predicting these events for operational weather forecasting and disaster preparedness is crucial, which is challenging. To accurately simulate HREs, parameterized physical processes—cloud microphysics, cumulus parameterization, turbulence processes, planetary boundary layer processes, radiation processes, and surface fluxes—are essential components of Numerical Weather Prediction (NWP) models. Our goal in this work is to produce large ensemble forecasts of HREs using a reasonable set of parameterization schemes. For this, we used the regional model of Scalable Computing for Advanced Library and Environment (SCALE-RM). The model was initialized with 0.25^o x 0.25^o NCEP FNL (Final) operational global analysis and forecast data over a large domain. Heavy rainfall simulations were conducted to account for uncertainties and produce probabilistic forecasts involving typhoons, such as Dolphin (2020) and Etau (2015), using varying initial conditions and physics configurations. One of the trickiest parts of numerical atmospheric modelling is convective parameterization. Mass fluxes are parameterized using the Kain-Fritsch scheme in order to examine the sensitivity of the Kessler, Ogura-Cho precipitation functions and Kain, Narita-Ohmori trigger functions. Additionally, simulations were run using a slab ocean model and fixed initial conditions to test how the ocean dynamics behaved. Simulated typhoons have depicted a usual tendency of north-westward propagation. Simulations with a six-class one-moment bulk microphysics scheme, Narita-Ohmori trigger function, Kessler precipitation function, and fixed ocean dynamics yielded the best typhoon trajectory and organized rain bands compared to the observation.