Desenvolvimento de uma plataforma web para simulações de incêndios florestais

Abstract

This work addresses the development of an accessible and efficient platform for the study and simulation of forest fire behavior, aiming to facilitate both the development and use of simulation models based on cellular automata. The system developed, Forest Fire Simulation Web App (FFSWA), allows the generation of grids from georeferenced image maps, the visualization of fire simulations and the download of raw data from these simulations for more detailed analysis. Thus, it is expected that this tool will effectively assist in decision-making during the planning of actions to prevent and combat the problem of forest fires, especially in the context of the Brazilian Cerrado biome. By integrating fire propagation models based on cellular automata developed by the Bioinspired Computing Laboratory (LCBio) of the Federal University of Uberlândia (UFU), the developed system enables the study and simulation of fire behavior, aiming to provide useful data that assist in the prevention and management of these devastating events. After development, a case study was conducted on the Pau Furado State Park, using one of the fire simulators developed by the research group. Based on the tests performed, it was observed that the system was capable of generating grids of different sizes and levels of precision, satisfactorily mapping the types of vegetation represented in the maps to the grid cells. The integration of the web system with the fire propagation model allows simulations of fire spread over the generated grid, generating data that can be used to support decision-making during the planning of public policies for fighting and preventing fires. The simulator and region management module enables the system to be adapted to different simulators and maps, making the developed system more independent and flexible and allowing it to be used by different research groups.