Refinement of NOAA AMSR-2 Soil Moisture Data Product—Part 2: Development With the Optimal Machine Learning Model

Advanced Microwave Scanning Radiometer-2 (AMSR2) is a successor of AMSR for Earth-Observation System (AMSR-E), while the third generation (AMSR3) will be launched in near future. The AMSR2 soil moisture product also an important component Soil Moisture Operational Products (SMOPS) datasets that are operationally produced by National Oceanic and Atmospheric Administration (NOAA). Refinement NOAA data can not only benefit past AMSR-E upcoming AMSR3, but improve SMOPS quality. In this second paper two-part series, Extreme Gradient Boosting (XGB) model was trained using 6.925 GHz, 10.65 18.7 GHz 36.5 brightness temperature (Tb) measurements dual polarizations, ancillary maps vegetation index datasets, turn used to predict daily global retrievals from 2012 2021. Validation results show refined (AMSRr) overwhelming advantage accuracy over currently operational (AMSRc). Comapred AMSRc, developed AMSRr presents significant improvement on availability. Results indicate comparable with latest version Active Passive product. Based study, higher quality NOAA, eventually blended its users.