FIRE PATTERNS IN BRAZILIAN MUNICIPALITIES: RELATIONSHIP WITH THE FIRE PROBABILITY FORECAST

Brazil contains some of the most fire-prone regions in the world, and this motivates the development of a fire probability forecast system. CEMADEN has collaborations with the states of Acre (AC), Goiás (GO), Maranhão (MA) and Mato Grosso do Sul (MS), which use the CEMADEN fire probability forecast data to guide prevention and mitigation actions. In this study, we aim to identify the fire pattern by municipality and the relation between the fire occurrence and fire probability forecast product for each state, using the 2022 January-September period as a case study. Our results demonstrate that AC state has the highest average burned area. GO and MA exhibit similar fire patterns and fire probability forecasts. MS municipalities reach the highest values in three of the fire patterns. We identified that some adjustments could be applied to the fire probability forecast data, aiming to improve its adherence to the spatial gradient.