Retrieving Surface and Rootzone Soil Moisture Using Microwave Remote Sensing

Soil moisture is one of the least monitored of all the hydrologic variables. It is greatly influenced by unpredictable and intermittent precipitation, varying evapotranspiration rates, heterogeneous soils, land cover, topography, and is extremely changeable in both space and time. The aim of this study is to retrieve surface and rootzone soil moisture in fallow land at a field scale using Sentinel-1A SAR data. The study explores the potential of obtaining surface soil moisture over fallow land at two different soil types from C-band SAR data. The study area consists of two plots having different soil types. The study area was divided into 80 grids, each measuring 10 × 10 m, to collect soil samples which are synchronized with Sentinel-1A passes. The soil moisture which are retrieved from plot 1 were used to develop the model. The developed model was validated in plot 2. In order to study the impact of soil moisture and dielectric constant on backscattering coefficients, a multiple regression analysis was used to create a semi-empirical model. Rootzone soil moisture retrieval model was developed by considering the backscattered coefficient, volumetric surface soil moisture as an independent variable and volumetric rootzone soil moisture as dependent variable. The predicted surface soil moisture using the regression model were identical to in-situ observed surface soil moisture, with R² of 0.77, RMSE of 1.31 m³/m³, and NSE of 0.75. The estimated rootzone soil moisture matches the in-situ observed rootzone soil moisture identically with R² = 0.74, RMSE = 1.23 m³/m³, NSE = 0.73. This study aids local farmers in their irrigation water management.