Observing inland water bodies by means of X-band polarimetric SAR imagery from second generation COSMO-SkyMed constellation

This study aims at demonstrating the ability of full-polarimetric X-band synthetic arperture radar measurements provided by the COSMO-SkyMed second generation satellite constellation to extract the waterline of inland water bodies and to classify the land use/cover of the surroundings. The former task is undertaken using a state-of-the-art unsupervised method that exploits a global threshold constant false alarm rate method together with morphological filters and a Sobel edge detection algorithm; while the latter goal is pursued considering an unsupervised scattering-based approach the relies on the Wishart distribution and the eigendecomposition parameters of the coherency matrix.Experiments are carried out on satellite data set collected over the San Giuliano reservoir, in the South of Italy, where a full-polarimetric COSMO-SkyMed second generation image and a Sentinel-2 optical image are collected. Results show that the adopted methodologies allow effectively extracting the profile of the reservoir and providing classification outputs of the surroundings according to the dominant scattering mechanisms.