Challenges and Advances in Satellite Data Assimilation: From Vertical Covariance Localization to Every-10-Minute Himawari-8 All-Sky IR Radiances

At RIKEN Advanced Institute for Computational Science (AICS), we have been working on a wide range of challenges on satellite data assimilation using the Local Ensemble Transform Kalman Filter (LETKF) with the Nonhydrostatic ICosahedral Atmospheric Model (NICAM) and with the SCALE and JMA-NHM regional models. With NICAM, we explored to run the LETKF with 10240 ensemble members and found that about 1000 ensemble members may be sufficient to remove vertical localization completely for satellite radiance data assimilation. Also, we developed a method to effectively assimilate JAXA’s GSMaP (Global Satellite Mapping of Precipitation) and to improve medium-range forecasts. With reginal models, we explored to assimilate Himawari-8 all-sky IR radiance data for a Typhoon case and a convective case. In this presentation, we will present the most up-to-date results of our research at RIKEN AICS.