Wildland fires can be responsible for negative impacts on the environment, causing damage to the fauna and flora and increasing the release of greenhouse gases. In the state of Amazonas, wildland fires represent a risk for biodiversity conservation, since more than 95% of the state is covered by Amazon rainforest, one of the largest and most biodiverse tropical forests of the world. This study aimed to analyze the spatiotemporal variation of fire occurrence from 2003 to 2016 in the state of Amazonas, based on data from the AQUA satellite processed by the Brazilian National Institute for Space Research, using the “Collection 5” detection algorithm. The correlation between fire incidence versus anthropogenic and climatic variables was also tested. A significant uptrend was observed in the number of hot spots recorded over the years. About 83% of the wildland fires occurred during the months of August, September and October. The variables that correlated significantly with the number of hot spots for each municipality were deforested area, pasture area, agricultural area, municipality area and mean annual rainfall. The municipality with the highest number of hot spots detected was Lábrea, while Careiro da Várzea presented the highest incidence per km2. The southern and eastern regions of the state were the areas most affected by fire during the analyzed period. The results from this study emphasize the need for implementation of public policies aimed to reduce deforestation and wildland fires in the state, thus ensuring the conservation of the Amazon rainforest and its biodiversity.
hot spots; fire prevention; wildfire; remote sensing