CYGNSS Soil Moisture Estimations Based on Quality Control

In this letter, we proposed a method based on cyclone global navigation satellite system (CYGNSS) for improving the accuracy of soil moisture (SM) estimation through the selection of auxiliary data and a four-step quality control. We investigate the impact of elevation and vegetation, as well as evaluating the quality of Doppler delay map (DDM) and different ground terrains. The program adopts the support vector machine (SVM) algorithm, input CYGNSS, and auxiliary data. With the hourly SM data on Wuhan Baoxie site from January 2020 to August 2020 as an example, the effectiveness of quality control was verified. A substantial improvement in correlation coefficient of ~0.46 for average between CYGNSS reflectivity and
in situ
SM was obtained compared with ~0.03 before quality control, resulting in better SM estimation accuracy compared with that of
in situ
measurements (from
$R = 0.67$
to 0.87).