Widespread in the subtropics and tropics of the Southern Hemisphere, savannas are highly heterogeneous and seasonal natural vegetation types, which makes change detection (natural vs. anthropogenic) a challenging task. The Brazilian Cerrado represents the largest savanna in South America, and the most threatened biome in Brazil owing to agricultural expansion. To assess the native Cerrado vegetation (NV) areas most susceptible to natural and anthropogenic change over time, we classified 33 years (1985–2017) of Landsat imagery available in the Google Earth Engine (GEE) platform. The classification strategy used combined empirical and statistical decision trees to generate reference maps for machine learning classification and a novel annual dataset of the predominant Cerrado NV types (forest, savanna, and grassland). We obtained annual NV maps with an average overall accuracy ranging from 87% (at level 1 NV classification) to 71% over the time series, distinguishing the three main NV types. This time series was then used to generate probability maps for each NV class. The native vegetation in the Cerrado biome declined at an average rate of 0.5% per year (748,687 ha yr⁻¹), mostly affecting forests and savannas. From 1985 to 2017, 24.7 million hectares of NV were lost, and now only 55% of the NV original distribution remains. Of the remnant NV in 2017 (112.6 million hectares), 65% has been stable over the years, while 12% changed among NV types, and 23% was converted to other land uses but is now in some level of secondary NV. Our results were fundamental in indicating areas with higher rates of change in a long time series in the Brazilian Cerrado and to highlight the challenges of mapping distinct NV types in a highly seasonal and heterogeneous savanna biome.

Fire is a significant agent of landscape transformation on Earth, and a dynamic and ephemeral process that is challenging to map. Difficulties include the seasonality of native vegetation in areas affected by fire, the high levels of spectral heterogeneity due to the spatial and temporal variability of the burned areas, distinct persistence of the fire signal, increase in cloud and smoke cover surrounding burned areas, and difficulty in detecting understory fire signals. To produce a large-scale time-series of burned area, a robust number of observations and a more efficient sampling strategy is needed. In order to overcome these challenges, we used a novel strategy based on a machine-learning algorithm to map monthly burned areas from 1985 to 2020 using Landsat-based annual quality mosaics retrieved from minimum NBR values. The annual mosaics integrated year-round observations of burned and unburned spectral data (i.e., RED, NIR, SWIR-1, and SWIR-2), and used them to train a Deep Neural Network model, which resulted in annual maps of areas burned by land use type for all six Brazilian biomes. The annual dataset was used to retrieve the frequency of the burned area, while the date on which the minimum NBR was captured in a year, was used to reconstruct 36 years of monthly burned area. Results of this effort indicated that 19.6% (1.6 million km²) of the Brazilian territory was burned from 1985 to 2020, with 61% of this area burned at least once. Most of the burning (83%) occurred between July and October. The Amazon and Cerrado, together, accounted for 85% of the area burned at least once in Brazil. Native vegetation was the land cover most affected by fire, representing 65% of the burned area, while the remaining 35% burned in areas dominated by anthropogenic land uses, mainly pasture. This novel dataset is crucial for understanding the spatial and long-term temporal dynamics of fire regimes that are fundamental for designing appropriate public policies for reducing and controlling fires in Brazil.

Background: Fire significantly transforms ecology and landscapes worldwide, impacting carbon cycling, species interactions, and ecosystem functions. In the Brazilian Cerrado, a fire-dependent savanna, the interaction between fire, society, and the environment is evident. Given that wildfires significantly contribute to greenhouse gas emissions, our study aimed to analyze four decades of burned area data to understand changes in fire dynamics, using Collection 2 of annual MapBiomas Fire maps (1985 to 2022). Our study examined spatiotemporal patterns, fire recurrence, fire distribution across land uses, temporal changes in fire scar size, burned area variations across ecoregions, and their correlation with farming areas. Results: From 1985 to 2022, fire impacted 40% (792,204 km²) of the Cerrado biome, with 63% burning more than once. Natural vegetation was the most affected, primarily due to human-driven ignition during the dry season. A noticeable trend of later peaks in fire activity, concentrated towards the end of the dry season, along with an increase in patch size over time, characterized a clear shift in the Cerrado fire regime. Recently, the MATOPIBA region and the northern biome exhibited significant fire clusters, with burned areas rising alongside farming expansion. The ecoregion-based analysis identified fire hotspots, with the "Bananal" ecoregion, the largest wetland area in the biome, exhibiting increased fire recurrence and larger patch size over time. Conclusions: Our four-decade analysis of fire dynamics in the Cerrado revealed human-induced changes in the fire regime, originally shifting from July to September to a new fire season from August to October. This shift poses several environmental threats given their overlap with the driest months of the year. This study improved our understanding of changes in fire patterns and their impacts on each ecoregion and land use. Wetlands experienced the highest relative burned area, highlighting their ecological importance and increased vulnerability. In the southern Cerrado, where farming is established and natural vegetation more fragmented, fire events tend to decrease; while in the north, with recent farming expansion, fire susceptibility rises. Conservation-oriented strategies, like the Brazilian Integrated Fire Management (MIF), are crucial for mitigating impacts while enhancing the Cerrado’s resilience to climate change.

Savannas play a key role in estimating emissions. Climate change has impacted the Cerrado savanna carbon balance. We used the burned area product and long-term field inventories on post-fire vegetation regrowth to estimate the impact of the fire on greenhouse gas emissions and net carbon dioxide (CO₂) emissions in the Cerrado savanna between 1985 and 2020. We estimated the immediate emissions from fires, CO₂ emissions by plant mortality, and CO₂ removal from vegetation regrowth. The burned area was 29,433 km²; savanna fires emitted approximately 2,227,964 Gg of CO₂, 85,057 Gg of CO, 3010 Gg of CH₄, 5,103 Gg of NO_x, and 275 Gg of N₂O. We simulated vegetation regrowth according to three fire regime scenarios: extreme (high fire frequency and short fire interval), intermediate (medium fire frequency and medium fire interval), and moderate (low fire frequency and long fire interval). Under the extreme and intermediate scenarios, the vegetation biomass decreased by 2.0 and 0.4% (ton/ha-year), while the biomass increased by 2.1% under a moderate scenario. We converted this biomass into CO₂ and showed that the vegetation regrowth removed 63.5% of the total CO₂ emitted (2,355,426 Gg), indicating that the Cerrado savanna has been a source of CO₂ to the atmosphere.

The semi-arid Caatinga biome is particularly susceptible to fire dynamics. Periodic droughts amplify fire risks, while anthropogenic activities such as agriculture, pasture expansion, and land-clearing significantly contribute to the prevalence of fires. This research aims to evaluate the effectiveness of a fire detection model and analyze the spatial and temporal patterns of burned areas, providing essential insights for fire management and prevention strategies. Utilizing deep neural network (DNN) models, we mapped burned areas across the Caatinga biome from 1985 to 2023, based on Landsat-derived annual quality mosaics and minimum NBR values. Over the 38-year period, the model classified 10.9 Mha (12.7% of the Caatinga) as burned, with an average annual burned area of approximately 0.5 Mha (0.56%). The peak burned area reached 0.89 Mha in 2021. Fire scars varied significantly, ranging from 0.18 Mha in 1985 to substantial fluctuations in subsequent years. The most affected vegetation type was savanna, with 9.8 Mha burned, while forests experienced only 0.28 Mha of burning. October emerged as the month with the highest fire activity, accounting for 7266 hectares. These findings underscore the complex interplay of climatic and anthropogenic factors, highlighting the urgent need for effective fire management strategies.

The Cerrado biome faces profound changes in its fire regime driven by anthropogenic pressures and climate change, threatening ecosystem resilience. Reliable monitoring is essential, as underestimation or overestimation of burned areas can misguide management and obscure fire-regime trends. However, satellite-derived products often yield divergent estimates due to differing sensors and methodologies. Assessing these differences is essential for identifying data gaps and uncertainties, which contributes to informing future sensor design and algorithm development while guiding decision-makers toward appropriate product applications. This study provides a systematic comparative assessment of three major burned area datasets (2001–2024): the regional MapBiomas Fire (Collection 4), the global GABAM (both 30 m Landsat-based), and the moderate-resolution MODIS MCD64A1 (500 m). Results reveal significant discrepancies, as only 23% of the total burned area was detected simultaneously by all three products. MODIS accounted for the largest share of exclusive detections (30%), while MapBiomas and GABAM identified 16% and 6%, respectively. Even between comparable 30 m products, convergence remained limited, indicating that discrepancies are driven by methodological and validation choices beyond sensor characteristics. While all datasets identified savanna formations as the most fire-prone class, they diverged in anthropogenic landscapes; MODIS and GABAM attributed higher fire activity to agricultural areas (6–11%) than MapBiomas (2%), whereas MapBiomas identified a greater share in pastures (22%). These findings demonstrate that the products are complementary: MapBiomas provides long-term thematic depth, MODIS offers high-frequency monitoring for large-scale trends, and GABAM provides high-resolution global coverage. Integrating these datasets reduces uncertainties and better informs fire management, conservation planning, and climate mitigation strategies in the Cerrado.