MODELLING OF FOREST FIRE USING NON-CONVENTIONAL METHODS

Abstract

Forest fires have been an area of global concern since the last few decades. Modeling of forest fires helps the scientists and engineers to understand the propagation and intensity of the fire and plan the fire-fighting strategies accordingly. This paper presents a hybrid model to predict the probability of burning of a regular tree. Stochastic approach is used to understand the forest fire spread and a deterministic approach is used to evaluate the stochastic model parameters. Starting with the temperature distribution around a burning tree using energy balance principles, we proceed to see the effect of multiple burning trees around any regular tree. Although, homogeneous tree species is assumed for simplicity, this model can be extended to the heterogeneous case. Variation of flame-length in different directions due to wind effect is also studied. This model might prove useful in predicting the direction and determining the rate of propagation of forest fires.