An improved soil moisture retrieval method considering azimuth angle changes for spaceborne GNSS-R

Abstract

The soil moisture (SM) retrieval of spaceborne GNSS-R is significantly affected by azimuth angle. In this paper, the K-means algorithm for clustering is used to weaken the effects caused by azimuth angle changes on spaceborne GNSS-R reflectivity. The matched Cyclone Global Navigation Satellite System (CYGNSS) and Soil Moisture Active and Passive (SMAP) observables are categorized differently using the K-means algorithm, and subsequently employed in SM retrieval based on the Reflectivity-Temperature-Vegetation (R-T-V) model. The SM retrievals are verified with the SMAP as well as the International Soil Moisture Network (ISMN). In the verification with SMAP, the results show that, compared with traditional method, the root-mean-square error (RMSE) is 0.040 cm3/cm3 with a reduction of 9.09 % and the correlation coefficient (R) is 0.946. In the verification with ISMN, the results demonstrate improvement in both RMSE and R at the selected sites compared with the traditional method. Specifically, RMSE is decreased from 0.086 cm3/cm3 to 0.054 cm3/cm3, while R is increased from 0.316 to 0.840 at Chalender. Therefore, the improved method can further strengthen the ability of spaceborne GNSS-R for SM retrieval by weakening the effects of azimuth angle changes.