Prediction of Soil Moisture from Remote Sensing Data

Abstract

This study evaluates the capability of soil water content predicted from remote sensing to indicate the soil/canopy water content at short time and space scale, through comparisons with daily soil moisture data determined in situ, using dielectric devices. Daily aqua moderate resolution imaging spectroradiometer (MODIS) normalized difference vegetation index (NDVI) and the diurnal (daytime and night time) land surface temperature difference (DLST) are employed to retrieving daily volumetric soil moisture content (θ) at Sparta experimental station, during the period June-August, of the years 2010, 2011, 2012 and 2014. Using the concept of apparent thermal inertia (ATI) in the remotely sensed topsoil moisture saturation index, daily θ is obtained from DLST and the volumetric saturated and residual soil moisture content and is compared with the experimental values of volumetric soil moisture content (SM) measured at various depths (10, 20, 30, 40, 60, 80 and 100cm). Simple relationships are also calibrated between SM and ATI or DLST or NDVI during the years 2010, 2011 and 2014 and are tested for predicting θ, during the year 2012. Especially the three first models predict θ satisfactorily as compared with the measured SM and hence they can offer a considerable guidance in irrigated agriculture and other related fields.