Mapping the probability of wildland fire occurrence in Central America, and identifying the key factors

Background. Wildland fires are part of the ecology of forests in Central America. Nevertheless, limited understanding of fire probability and the factors that influence it hinder the planning of intervention strategies. Aims. This research combined climatic, anthropogenic and vegetation factors to identify wildland fire probability and determine the most relevant factors. Methods. We performed an exploratory analysis to identify important factors and integrated them with fire observations using random forest. We then used the most relevant factors to predict wildland fire occurrence probability and validated our results using different measures. The results demonstrated satisfactory agreement with the independent data. Key results. Central regions of Honduras, northern Guatemala and Belize have a very high probability of wildland fire occurrence. Human imprint and extreme climatic conditions influence wildland fire probability in Central America. Conclusions. Using random forest, we identified the major influencing factors and areas with a high probability of wildland fire occurence in Central America. Implications. Results from this research can support regional organisations in applying enhanced strategies to minimise wildland fires in high-probability areas. Additional efforts may also include using future climate change scenarios and increasing the time frame to evaluate the influence of teleconnection patterns.