An analysis of fire dynamics in and around indigenous territories and protected areas in a Brazilian agricultural frontier

Abstract

Sustainable land management is integral to conserving tropical ecosystems and reducing carbon emissions. Human-set fires are part of the land cover change process and are the most common driver of deforestation and carbon emissions in Amazonia. Analyzing the behavior of major fire events provides insight into the effectiveness of current protections. This study aims to quantify the role of Indigenous Territories and Protected Areas (ITPAs) in characterizing anthropogenic fire regimes over the 2020 fire season in Mato Grosso, Brazil. Because of the rapid rate of land cover change and lack of widespread ground validation data in the region, we used a combination of ESA’s Sentinel-5 Precursor Aerosol Index, NASA’s Visible Infrared Imaging Radiometer Suite, and high resolution Planet satellite imagery for the Monitoring of the Andean Amazon Project to analyze spatial and temporal patterns by fire type. ITPAs cover 25% of the study area’s land and contained approximately 20% of significant recorded fires in 2020. Recently deforested areas, forest, grassland, and cropland fires showed varying seasonality and lower frequencies inside ITPAs, but mean fire start dates for all fire types occurred in mid-September. Results suggest that the overall density of major fires is reduced in ITs. PAs only inhibit the density of crop or pastureland fires, but no major fires occurred past 10 km inside their borders. Burn severity of major fires had a weak relationship to distance from ITPAs for some fire types. This study highlights the advantages of near real-time data for individual fire events, provides further evidence of the effect of ITPAs on fire behavior, and demonstrates the importance of adequate protection strategies for mitigating fire activity.

Keywords

• fire
• remote sensing
• geospatial analysis
• protected areas
• indigenous territories
• Amazonia