Sentinel-1 SAR Workflow for Estimating Soil Moisture

Abstract

Soil moisture is key for sustaining rainfed agriculture in Morocco, where most farmland depends on rainfall. Precipitation variation and poor agronomy practices are making this problem worse by reducing how much water the soil holds. This study used cloud-based Sentinel-1 SAR satellite data to track soil moisture and see how crop choice and farming practice change the availability of soil moisture. Sentinel-1 SAR scenes were pre-processed using Google Earth Engine. The main tool was the Dual-Polarization Radar Vegetation Index (DpRVIc), which makes it easier to measure soil moisture even when vegetation is present. A change detection method estimated the soil moisture variation in both agronomy practices over three growing seasons, confirmed by in-situ data from a nearby irrigated field. The soil moisture maps from SAR data well matched with seasonal wet and dry periods from the measured data. Details has been presented presented in Anwari (2025). This approach, combining remote sensing and practical field work, offers solid guidance for farmers and planners, making agriculture more resilient against unpredictable climate and supporting sustainable food security. The main goal of this technical report is to show a clear step-by-step process on how to estimate and monitor field scale soil moisture using Sentinel-1 GRD SAR data. This methodology implementing a modified Change Detection Approach (CDA), critically incorporating the SAR-derived Dual-Polarization Radar Vegetation Index (DpRVIc) to effectively reduce vegetation effects and enhance the precision of soil moisture estimation over vegetated soil surfaces Bhogapurapu et al (2022). The process utilizes Google Earth Engine for cloud-based SAR data preprocessing and deriving Dual-Polarization Radar Vegetation Index (DpRVIc), R for statistical calibration and analysis, and QGIS for soil moisture mapping. To implement this workflow, you can get the source code from here.