Multi-Polarization Methods to Detect and Classify Burned Areas using Sentinel-1 Sar Data

• Published in: 2021 IEEE 6th International Forum on Research and Technology for Society and Industry (RTSI) pp. 217 - 221
• Main Authors: Ferrentino, Emanuele, Nunziata, Ferdinando, Buono, Andrea, Sarti, Maurizio, Migliaccio, Maurizio
• Format: Conference Proceeding
• Language: English
• Published: IEEE 06-09-2021
• Subjects: C-band; change detector; Fires; Industries; Optical polarization; Polarization; Radar cross-sections; SAR; Satellites; Sensitivity; wildfire
• ISSN: 2687-6817
• DOI: 10.1109/RTSI50628.2021.9597245

In this study, multi-polarization Synthetic Aperture Radar (SAR) features extracted from Sentinel-1 C-band SAR measurements are used to identify wildfires and to classify burn severity. SAR features include co- and cross-polarized normalized radar cross sections and the total backscattered power, namely the SPAN. The test case refers to the wildfire that affected about 10 km 2 in Tuscany region (Central Italy) during September 2018. Experiments, undertaken on actual SAR data, collected before and after the considered wildfire, demonstrate the soundness of the proposed approach and the different sensitivity of the multi-polarization backscattering features to burn severity.