Impact of AMSU-AB/MHS radiance on prediction of high-impact weather over Korea Peninsula

Effective assimilation of regional observations is one of the important components to improve 0 – 36h numerical weather prediction. For mesoscale models with grid intervals of 1-10km or less, available conventional observations are relatively thin, and these observations are unevenly distributed both in space and time. This is particularly true over the coastal regions, such as over Korea Peninsula, where high-density observations are clustered over the peninsula, China main land and Japan. Assimilating non-conventional observation data such as satellite radiance data over the large surrounding data-sparse sea/ocean environments has been demonstrated to be an effective approach to improve numerical weather model forecast. In this study, we employ two advanced WRF-based data assimilation systems, NCAR's Real–Time Four–Dimensional Data Assimilation (RTFDDA) and WRF-3DVAR hybrid (RTF3H) forecasting system and WRF4DVAR to investigate the impact of incorporating satellite AMSU-AB and MHS radiance measurements on a flood-induced torrential rain event in Seoul areas and two typhoon cases that effect the peninsula. The goal of the research is mainly to gain valuable understanding toward an optimal use of both conventional and satellite data to achieve more accurate 0–36h forecast using the state-of-the-art data assimilation technologies.