Abstract:
Lack of fire behavior studies and the immediate needs posed by the extent of the fire
problem in forests of Iran require that extensive studies be conducted to develop models
to predict fire behavior in the region. In this study, FARSITE Fire Area Simulator was
applied to simulate spread and behavior of two real fires that had occurred in Northern
Forests of Iran during 2010 summer and fall seasons in a spatially and temporally explicit
manner taking into account the fuel, topography, and prevailing weather in the area.
Spatial data themes of elevation, aspect, slope, canopy cover, and fuel model were
prepared and formatted in GIS along with weather and wind files to run FARSITE fire
behavior model. The effect of weather conditions on the accuracy of FARSITE
simulations was evaluated in order to assess the capabilities of the simulator in accurately
predicting the fire spread in the case study. The WindNinja model was used to derive
local winds influenced by vegetation and topography. The simulations were validated with
the real mapped fire scars by GPS mapping. Kappa Coefficient was used as measure of
the accuracy of the simulation. The Kappa statistic was lower for spatially uniform wind
data (0.5) as compared to spatially varying wind data (0.8) for the two studied events. The
results confirm that the use of accurate wind field data is important in fire spread
simulation, and can improve its accuracy and the predictive capabilities of the simulator.
Keywords: FARSITE, Fuel Model, Weather condition, Wildfire spread, WindNinja
