Degradation by logging, fire, and fragmentation is increasing as a spatially distinct disturbance in the Amazon landscape.

Predicting the spatial distribution of wildfires is an important step towards proper wildfire management. In this work, we applied two data-mining models commonly used to predict fire occurrence – logistic regression (LR) and an artificial neural network (ANN) – to Brazil’s Federal District, located inside the Brazilian Cerrado. We used Landsat-based burned area products to generate the dependent variable, and nine different anthropogenic and environmental factors as explanatory variables. The models were optimised via feature selection for best area under receiver operating characteristic curve (AUC) and then validated with real burn area data. The models had similar performance, but the ANN model showed better AUC (0.77) and accuracy values when evaluating exclusively non-burned areas (73.39%), whereas it had worse accuracy overall (66.55%) when classifying burned areas, in which LR performed better (65.24%). Moreover, we compared the contribution of each variable to the models, adding some insight into the main causes of wildfires in the region. The main driving aspects of the burned area distribution were land-use type and elevation. The results showed good performance for both models tested. These studies are still scarce despite the importance of the Brazilian savanna.

Over the past several decades, the Brazilian State of Rondônia has been the destination of many rural migrants drawn from Brazil's middle southern regions by massive government colonization projects. Factors such as explosive population growth, logging, mining, small-scale farming and ranching have synergistically fuelled deforestation in the state. The total area deforested in Rondônia in 1978 was 4200 km². In 1988, the area increased to 30 000 km², in 1998 to 53 300 km² and by the year 2003, a total of 67 764 km² of Rondônia was deforested. In response to the high rate of deforestation observed in Rondônia and other Amazonian states, state and federal agencies worked to create a network of conservation units (CUs) in Brazil during the 1990s that was signed into law(Law 9985/00) in 2000. The ability of these CUs to reduce the rate of deforestation was analysed. Remotely-sensed data from Landsat and thematic coverages were used to measure deforestation inside all CUs in Rondônia. A more detailed analysis of CUs with the highest levels of deforestation, including an analysis between soil types and deforestation and a forecast of potential future deforestation, was conducted. The creation of conservation units in Rondônia has been useful in curbing deforestation within their boundaries; however, many CUs face pressure from the combined activities of illegal loggers, cattle ranchers and small-scale farmers seeking new sources of timber and agricultural land. For example, an exponential increase in the amount of deforestation was observed in Rondônia's Bom Futuro National Forest between 1992 and 2000. A regression model indicated a total of 20 500 ha deforested by 2002, while measurements from 2002 imagery showed an actual total deforestation of 20 720 ha. Should this trend persist, Bom Futuro National Forest could be completely deforested by 2017. CUs in Rondônia must be developed and implemented jointly by all stakeholders through the creation of partnerships between local communities, non-governmental organizations and government agencies.