Abstract
In this paper, the potential of space-borne Synthetic Aperture Radar (SAR) sensors combined with optical ones has been exploited by analyzing datasets collected on two vegetated areas in Italy, by using COSMO-SkyMed X-band and Sentinel-1 C-band SAR, PRISMA hyperspectral and Sentinel-2 multispectral imagery, combined with field measurements acquired with spectroradiometers. On the mountain area in Alto Adige, a biomass estimation approach was developed by combining Sentinel-1 SAR and spectroradiometer hyperspectral data. On Val d’Elsa area in Tuscany, COSMO-SkyMed StripMap HIMAGE and Sentinel-1 Interferometric Wide swath mode SAR data have been integrated with Sentinel-2 imagery for improving the classification of agricultural crops. Convolutional Neural Networks (CNN) have been used for the classification of agricultural areas using these three sensors.