Seasonal soil moisture variation analysis using RADARSAT-1 satellite image in a semi-arid coastal watershed

Abstract

This study presents multi-temporal soil moisture using RADARSAT-1 Synthetic Aperture Radar (SAR) satellite imagery in Choke Canyon Reservoir Watershed (CCRW). Soil moisture is a critical element of hydrological cycle that drastically impacts humans’ activities in semi-arid area. Point measurements of soil moisture across different geographical landscapes are impossible to comprehend the soil moisture variations temporally and spatially. RADARSAT-1 is a promising tool for measuring the surface soil moisture over seasons with its all-weather capability and the short-period return of its orbiting. Time constraint is almost negligible since the RADARSAT-1 is able to capture surface soil moisture over a large area in a matter of seconds, if the area is within its swath. The CCRW was selected as the study area contributing to the reservoir, which is mostly agricultural and range land in a semi-arid coastal environment, South Texas. RADARSAT-1 images presented at here were captured in three acquisitions in 2004, including April, September and December. Essential radiometric and geometric calibrations of the multitemporal SAR images were performed to improve the accuracy of information and location, with the aid of five corner reflectors deployed by Alaska Satellite Facility (ASF). The horizontally spatial errors were reduced from initially 560 m down to less than 5 m at the best trial-and-true. Slope data, land cover data, aspect data, and soil type data were incorporated into the regression models, derived from genetic programming algorithm, to predict soil moisture using SAR data. It is necessary to use slope data and aspect data together to represent the effect of the geological slope to the radar backscatter because the slope data only represents the magnitudes of elevation change, while the aspect represents the direction of the slope. The soil moisture estimations show that soil moisture wholly varies in space and season. Keywords-component; RADARSAT-1, SAR, soil moisture, multi-temporal remote sensing, Ecohydrology