Joint BCAM-UPV/EHU Data Science and Artificial Intelligence seminar: Decision aid models and data science for prevention and Mitigation of wildfires

LOCATION: Maryam Mirzakhani Seminar Room at BCAM and Online

Link to the session here

Abstract:
Disaster risk reduction is a complex task involving important efforts to prevent and mitigate the consequences of disasters. Many countries around the world have experienced devastating wildfires in recent decades, and risk reduction strategies are now more important than ever. Landscape modifications focused on changes in the vegetable fuel load are common activities for prevention and mitigation, avoiding ignitions, spreading fires and facilitating access for response teams. Two approaches will be shown in this presentation. One of them focuses on reducing contiguous areas of high fuel load through prescribed burning. A mathematical programming model is proposed for scheduling prescribed burns on treatment units on a landscape over a planning horizon. The model takes into account the uncertainty related to the conditions to carry out the scheduled prescribed burns, as well as several criteria related to the safety and quality of the habitat. This multiobjective stochastic problem is modelled from a risk aversion perspective, based on the methodology introduced in Leon et al. (2020). The model is applied to a real case study in Andalusia (Spain), Sierra de los Filabres. The other model is a probability-based approach for firebreaks location or fuel management solved with Bayesian networks. Scenarios are also considered, in this case for wind directions, to obtain acyclic directed graphs to be solved with Bayesian networks algorithms. The result is an estimation of fire risk in a landscape that can be used to compare the effect of different strategies on the wildfire risk. This approach is also applied to the Sierra de los Filabres case study.