Temporal assessment of fire severity: a case study in Brazilian savannas

Abstract

Wildfires have become a significant environmental challenge, exacerbated by climate change. Remote sensing technologies have advanced, but their use for monitoring fire frequency and severity remains limited, particularly in resource-constrained regions. To address this, we developed an open-source tool on Google Earth Engine that generates fire regime maps, combining fire frequency and severity data, thus allowing a temporal assessment of fire severity. The tool utilizes freely available satellite data, including Mapbiomas Fire data (1985-2020) and fire spectral indices (NBR, dNBR) from Sentinel-2A imagery (2017-2023). Our methodology provides high temporal and spatial resolution, enabling continuous analysis of fire patterns over time. It offers a cost-effective, scalable solution for monitoring fire dynamics, identifying areas most impacted by wildfires, and supporting informed decision-making. This capability helps prioritize conservation actions. The main advantages of this approach are its accessibility, adaptability, and processing time. This tool represents a significant advancement in remote sensing for conservation, aligning with global strategies to mitigate the impacts of climate change. It provides a practical, efficient, and replicable solution that not only supports local management efforts but also contributes to broader scientific and policy initiatives aimed at wildfire mitigation and sustainable ecosystem management.