Title: Enhancing the suitability of Sentinel-1 SAR data for the analysis of soil moisture via Google Earth Engine

Abstract: Synthetic Aperture Radar (SAR) is an important remote sensing technique used for estimating soil moisture. Sentinel-1 SAR data are widely used for soil moisture analysis as they are freely accessible and provide high spatial and temporal resolutions at a global scale. The Sentinel-1 user base has increased over the past decade with the launch of the Google Earth Engine (GEE) platform, as it offers Analysis-Ready-Data (ARD) and efficient cloud-based processing capabilities. This study developed an optimised Sentinel-1 GEE processing workflow to extract SAR-derived soil moisture. The optimised workflow was obtained after extracting, evaluating, and finally validating a series of speckle noise and terrain corrected backscatter (σ0) products with the in-situ COSMOS-UK Volumetric Water Count (VWC) data. At ρ = 0.672, the highest linear correlation with the VWC was recorded from the terrain corrected σ0, which linear regression translated into a sensitivity of 0.07 dB / (VWC %). Therefore, the positive results demonstrate that the terrain correction procedure has preserved the soil moisture-induced backscatter from distortions caused by the study site’s tilted terrain and permanent vegetation cover. Furthermore, the JavaScript and Python codes developed in this study are publicly accessible and can be used to extract and evaluate SAR-derived soil moisture at any site.