A first Evaluation of the Prisma Hyperspectral Data Potentialities for Civil Protection Applications. Case Study: Burned Area Mapping

The objective of this work was to carry out a first evaluation of the PRISMA hyperspectral data potentialities for Burned Area (BA) mapping in an operational context related to wildfire risk management (i.e., simulating a civil protection scenario). To this aim, an automatic processing chain was developed to map the area burned by the wildfire that occurred in Pantelleria Island (Southern Italy) during the Summer of 2022. The differenced Normalized Burn Ratio (dNBR) index was used to perform this task and a simple, yet effective, criterion was defined to identify the spectral bands most suitable for the analysis. Results showed the potential retained by PRISMA data to map BA ca. one month after the fire outbreak, in an operational context (Dice Coefficient: 0.88). Their discussion also provided valuable insights related to the exploitation of hyperspectral remote sensing data for similar purposes.