Use of FARSITE simulator to produce fire probability maps in a Mediterranean area

Fire is one of the most significant threats for the Mediterranean forested areas. In the last three decades the occurrence of fires increased significantly, as well as the number of extreme fire seasons. Therefore, the development of fire management policies are required in order to reduce the wildland fire risk by applying methods and models for planning the operational phases of fire management. Although forest fires are predominantly ignited by arson, or by accident, the differences in fire propagation depend on a combined effect of environmental factors (weather, vegetation, and terrain). The type of fire behavior is one of the main factors affecting fire danger and fire risk. Temporal and spatial variations of fire spread and behavior can be predicted using different fire simulators developed over recent years (BEHAVE, FARSITE, etc.), and based on semi physical or empirical fire prediction models (i.e. Rothermel model). The use of a simulator as component of a decision support system for planning the fire management practices involves the assessment of the simulation accuracy under different environmental and vegetational conditions. The main goal of this study was to evaluate the capabilities of FARSITE simulator to provide useful spatial information for fire danger and fire risk assessment in a Mediterranean area predominantly covered by shrubland vegetation. Different distributions of ignition points and weather data were generated using randomization techniques, as Monte Carlo method. FARSITE simulator was used to map several parameters related to fire intensity and severity (rate of spread, fireline intensity, flame length, etc.). After calculating the error matrix, the spatial frequency of fire parameters was defined, and fire probability maps were produced. Sensitivity analysis was performed to determine the effect of different environmental scenarios on simulation results. The study showed the capabilities of the proposed approach to support the production of fire risk maps.