Local-Scale Fuel-Type Mapping and Fire Behavior Prediction by Employing High-Resolution Satellite Imagery

Judicial wildland fire prevention and management requires precise information on fuel characteristics and spatial distribution of the various vegetation types present in an area. The aim of this study was to present an integrated approach to forest fire management, combining local-scale fuel-type mapping from very high spatial resolution imagery with fire behavior simulation. The specific objectives were (i) to develop a detail site-specific fuel model in a Mediterranean area that is suitable for fire behavior prediction; (ii) to produce a detailed local-scale fuel-type map with an object-based approach; and (iii) to generate accurate fire behavior maps. The spatial extent of the different fuel types of a forested landscape in northern Greece characterized by heterogeneous vegetation and topography was determined using a Quickbird image. Site-specific fuel models were created by measuring fuel parameters in representative natural fuel complexes. Following necessary preprocessing of the image to compensate for geometric errors, multiscale components of the scene were delineated through a segmentation algorithm. The resulting image objects were assigned to respective fuel types using a CART statistical model with an overall accuracy over 80%. The FARSITE fire simulation model was applied to simulate potential wildland fire growth and behavior. Utilizing the spatial database capabilities of geographic information systems, FARSITE allows the user to simulate the spatial and temporal spread and behavior of a fire burning in heterogeneous terrain, fuels, and weather.