Propuesta metodológica para la actualización del mapa de amenazas por incendios forestales en el suelo rural del Distrito Capital, elaborado en 2014 por la Universidad Distrital Francisco José de Caldas junto al FOPAE

Abstract

In 2014, the Universidad Distrital Francisco José de Caldas and the then FOPAE developed a wildfire hazard map for the rural areas of Bogotá, a key tool for prevention and response to this environmental risk. However, a decade later, the context has changed significantly: the effects of climate change, urban expansion into rural zones, transformations in vegetation cover, and new socio-economic dynamics have shaped a different territory. This raises the question: what is the predictive capacity of the 2014 wildfire hazard model in relation to the actual distribution and frequency of fires recorded between 2014 and 2024 in Bogotá’s rural areas? A methodological update, based on recent data, new technologies, and rigorous criteria, can make the difference between effective prevention and delayed response. In this context, there is a clear need to critically review and update the wildfire hazard model developed ten years ago. This thesis proposes a methodology to update the map, taking into account fires recorded between 2014 and 2024, as well as environmental, climatic, vegetation, infrastructure, and land-use variables. Spatial and geostatistical analysis techniques were employed using ArcGIS Pro and RStudio, along with statistical methods such as logistic regression, ROC curves, and random forest models. A comparison between the two original indices—one focused on physical factors and the other on vegetation cover—allowed for evaluating their predictive capacity and relevance under the new territorial context. Preliminary results show that, although certain historical patterns persist (such as fire concentration in the south and seasonality), the magnitude and distribution of events have changed, revealing new critical areas and questioning the validity of the original model. The 51% increase in fires during the 2019–2024 period compared to 2014–2018 highlights the urgency of having more robust, adaptive, and updated tools.