Pan-tropical soil moisture mapping based on a three-layer model from CYGNSS GNSS-R data

In this paper, an effective schematic is developed for estimating soil moisture (SM) from CYclone Global Navigation Satellite System (CYGNSS) data. Here, a three-layer model of air, vegetation cover, and soil is considered. In practice, the surface reflectivity (Gamma) along with its statistics derived from the CYGNSS data and the ancillary vegetation opacity (tau) data from Soil Moisture Active Passive (SMAP) are employed. The expression for empirically retrieving SM from tau, Gamma and its corresponding statistics are determined through the linear regression technique. The CYGNSS data collected over the land surfaces within +/- 37 degrees (latitude) during the whole year of 2018 are investigated. The reference SM data are obtained from SMAP, and are regarded as ground-truth in this work. Validation and assessment are performed on the pan-tropical daily data collected throughout the annual circle at a resolution of 36 x 36 km(2). Experimental evaluation demonstrates good consistency between the SM derived from CYGNSS data and the ground-truth, with a correlation coefficient of 0.80 and a root-mean-square error of 0.07 cm(3)/cm(3). This method succeeds in providing SM estimations on a pan-tropical scale that does not rely on ongoing knowledge of SM and merely employs the least ancillary data. Furthermore, the intense temporal and spatial coverages of CYGNSS SM results are also illustrated. The use of CYGNSS SM significantly enhances the pan-tropical coverage of SMAP SM by about 22% on average. The satisfactory outcomes achieved here prove CYGNSS as an efficient complementary tool for pan-tropical SM sensing on a daily basis.