Estimation of Surface Downward Longwave Radiation and Cloud Base Height Based on Infrared Multichannel Data of Himawari-8

Surface downward longwave radiation (SDLR) is significant with regard to surface energy budgets and climate research. The uncertainty of cloud base height (CBH) retrieval by remote sensing induces the vast majority SDLR estimation errors under cloudy conditions; reliable CBH observation are crucial for determining radiative effect. This study presents a methodology built from 10 thermal spectral data Himawari-8 (H-8) observations, utilizing random forest (RF) algorithm fully account each band’s contribution CBH. utilizes only infrared band data, making it possible obtain 24 h day. Considering some factors that can significantly affect estimation, RF models trained different clouds using inputs multiple H-8 channels together geolocation information target derived CloudSat/CALIPSO combined measurements. validation results reveal new performs well, root-mean-square error (RMSE) 1.17 km all clouds. To evaluate effect on an all-sky based previous predictions proposed. product not has resolution noticeably higher than benchmark products SDLR, such as Clouds Earth’s Radiant Energy System (CERES) next-generation reanalysis (ERA5) European Centre Medium-Range Weather Forecasts (ECMWF), but also greater accuracy, RMSE 21.8 W m−2 hourly irradiance (SDLI).