The present work is devoted to the development of a hybrid model for predicting the rate of spread of wildland fires at a large scale, taking into account the local heterogeneities related to vegetation, topography, and meteorological conditions. Some methods for generating amorphous network, representative of real vegetation landscapes, are proposed. Mechanisms of heat transfer from the flame front to the virgin fuel are modeled: radiative preheating from the flame and embers, convective preheating from hot gases, radiative heat losses and piloted ignition of the receptive vegetation item. Flame radiation is calculated by combining the solid flame model with the Monte Carlo method and by taking into account its attenuation by the atmospheric layer between the flame and the receptive vegetation. The model is applied to simple configurations where the fire spreads on a flat or inclined terrain, with or without a constant wind. Model results are in good agreement with literature data. A sensitivity study is conducted to identify the most influential parameters of the model. Eventually, the model is validated by comparing predicted fire patterns with those obtained from a prescribed burning in Australia and from a historical fire that occurred in Corsica in 2009, showing a very good agreement in terms of fire patterns, rate of spread, and burned area.
