Progress and Limitations in the Satellite based Estimate of Burned Areas

The detection of burnt areas from satellite imagery seems one of the most straightforward and useful applications of satellite remote sensing. In general, the approach relies on a change detection analysis applied on pre- and post-event images [1]. This change detection analysis is usually carried out by comparing the values of specific spectral indices such as: NBR (Normalized Burn Ratio) [2], BAI (Burned Area Index) [3], MIRBI (Mid Infrared Burn Index) [4]. However, some potential sources of error arise, in particular when near-real-time automated approaches are adopted. An automated approach is needed when the burnt area monitoring should operate systematically on a given area of large sizes (country). Potential sources of error are: clouds on the pre- or post-event images, clouds or topographic shadows [5], agricultural practices, image pixel size, level of damage, etc. [6]Sources of difference between different existing global datasets of burned areas based on satellite images could be related to the spatial resolution of the images used, the land cover mask adopted to avoid false alarms, the quality of the cloud and shadow masks. This paper aims at comparing several burned area databases (EFFIS, ESACCI, Copernicus, FIRMS, etc.) with the objective of identifying and characterizing their approximations. In order to do so, ground survey data provided by CUFA (Comando unità forestali, ambientali e agroalimentari Carabinieri) and CFVA (Corpo Forestale e Vigilanza Ambientale Sardegna) have been used. In using ground data, the accessibility of the burned area has been also considered in order to account for the approximations due to the difficulties in reaching the affected area.