Soil Moisture Retrieval from the CyGNSS Data Based on a Bilinear Regression

Soil moisture (SM) has normally been estimated based on a linear relationship between SM and the surface reflectivity (?) from spaceborne Global Navigation Satellite System (GNSS)-Reflectometry, while it usually relies inputs of data without considering vegetation optical depth (VOD/?) effects. In this study, new scheme is proposed for retrieving soil Cyclone GNSS (CyGNSS) data. The variation CyGNSS-derived ?? modeled as function both variations in VOD (?SM ??). For SM, ancillary ? can be obtained Moisture Active Passive (SMAP) mission. addition to option, model simulating ?? suggested an alternative. Experimental evaluation performed time span August 2019 July 2021. Excellent agreements final retrievals referenced SMAP products are achieved training (1-year period) test duration) sets. On whole, overall correlation coefficients (r) 0.97 0.95 root-mean-square errors (RMSEs) 0.024 0.028 cm3/cm3 models using simulated ??, respectively. generates r RMSE 0.031 cm3/cm3. efficiency necessity thus confirmed by its enhancement against one ?, usefulness approximating sinusoidal functions also validated. Influences statistics terms mean variance retrieval accuracy evaluated. This work unveils interaction CyGNSS data, demonstrates feasibility integrating approximation into bilinear regression obtain results.