PREDICTING THE DYNAMICS OF LAND USE AND LAND COVER AND FIRE BETWEEN 2018-2023 IN THE STATE OF MARANHÃO, BRAZIL

With fires becoming more frequent, intense and severe, it is important to develop better tools and approaches for improving fire prediction, prevention, and management. The goal of this study was to use modeling approaches to understand fire patterns in Maranhão state, Brazil, from 2018-2023, considering land use changes. We used the Dinamica EGO 7.0 software to simulate the land use and land cover model with a focus on the occurrence of fire, divided into three main steps: calibration, simulation, and validation. In general, the model successfully identified land use and fire changes in Maranhão from 2018 to 2023, but with some limitations: it overestimated fire occurrence and underestimated deforestation and forest cover. The study found that fires predominantly occur in forest areas, especially near protected areas. Due to low levels of similarity indices and aiming to improve model accuracy, future tests should incorporate climate variables like precipitation and soil moisture.