Desenvolvimento de um Modelo Baseado em Autômatos Celulares para Simular a Propagação de Incêndios Florestais

Abstract

Wildfires in the Brazilian Cerrado, although part of its natural cycle, have been intensifying each year, mainly due to human activities, preventing these ecosystems from adequately recovering. This phenomenon poses a constant concern for the preservation of Brazilian flora and fauna. Computational tools can aid in understanding fire propagation, supporting the development of effective control strategies for environmental preservation. This work presents the development of a computational model based on Cellular Automata (CA) to simulate wildfire spread in the Brazilian Cerrado. The implemented model is strongly based on a reference study, incorporating environmental factors such as different vegetation types, air humidity, and wind. Additionally, the terrain’s topography was included as a factor influencing fire dynamics. Experiments demonstrated that the model can reproduce behaviors similar to those generated by the reference model. Simulations including topography showed faster fire spread on slopes, reflecting the physical behavior of fire in inclined terrains. Although terrain influences the fire’s propagation speed, wind remains the predominant factor in fire dynamics. The results also highlighted the model’s high sensitivity to parameter variations, enabling the simulation of diverse wildfire scenarios and providing a useful tool for research and fire prevention.