Extraction of Soil Moisture Based GNSS-R Considering Vegetation Factors

One of the keys to retrieve soil moisture (SM) using the Spaceborne Global Navigation Satellite System-Reflectometry (GNSS-R) technique is to correct for the influence of vegetation. In this paper, the surface reflectivity is calculated using the Cyclone Global Navigation Satellite System (CYGNSS) data, and combine the Vegetation Water Content (VWC) provided by Soil Moisture Active Passive (SMAP) data to establish linear regression model to retrieve SM products with a temporal resolution of 3 days and a spatial resolution of 36 km on a pan-tropical scale, and each of the models is parameterized pixel-by-pixel to allow for tuning in accordance with regional variations. According to the experimental findings, CYGNSS may offer useful SM estimations across regions with moderate vegetation, and the correlation coefficient (R) with SMAP reference data is up to 0.7. However, in the arid and densely vegetated regions, the retrieval performance is degraded, and the R is 0.4 and 0.3 in the forest and bare soil areas, respectively. The overall root mean square error (RMSE) is 0.042 cm3/cm3. In addition, a time-series comparison of in-situ data from the International Soil Moisture Network (ISMN) and the CYGNSS SM revealed a good correlation. The study proves the necessity of considering vegetation effect in SM retrieval, which is of positive significance for the promotion of the operational application of Spaceborne GNSS-R SM retrieval.