Retrieval of Soil Moisture from FengYun-3D Microwave Radiation Imager Operational and Recalibrated Data Using Random Forest Regression

Three Microwave Radiation Imagers (MWRI) were carried onboard the FengYun-3B/C/D satellites and have collected more than 10 years of data since 2010. To create a robust climate quality data, MWRI level one reprocessed with new calibration. This study evaluates performance retrieving global soil moisture from recalibrated (RCD) quantifies difference retrieved between operational calibration (OCD) RCD. Soil Moisture Operational Products System (SMOPS) products NOAA on four days different seasons collocated brightness temperatures, then used for training an algorithm through machine learning. The using OCD RCD evaluated against independent SMOPS products, in situ networks SMAP product. It is shown that random forest suitable FY-3D coefficient determination (R2) 0.7223, mean bias ?0.0062 unbiased root square (ubRMSD) 0.0476 m3 m?3 compared over period 12 July 2018 to 31 December 2019. spatially heterogeneous. Both temporal spatial coverage accuracy existing are significantly improved.