Summary Account

　Accurate prediction of hazardous meteorological phenomena such as linear meso-convective systems and typhoons requires precise initial conditions for numerical weather prediction systems. Moreover, even with the best practicable prediction system, these phenomena are intrinsically impossible to predict far into the future in a deterministic manner, necessitating us to produce probabilistic predictions. In this project, we will work on advancing the data assimilation techniques utilizing observational data for better initial conditions. We will also work on developing ensemble prediction systems to support production of probabilistic predictions.