Volcanic-induced fire hazard at Stromboli Island.

Forest fire propagation models have been extensively researched and utilized in recent years, using approaches that range from complex physical models to simpler graph or cellular automata-based models. These models take into account several input parameters, such as vegetation flammability, tree types, wind, humidity, radiant capacity, and topography, to accurately simulate fire propagation. Integrated with Geographic Information Systems, these models can also be used to classify soil and map burned areas. The high availability of low-cost high-performance parallel computing hardware has made cellular automata-based simulations more accessible, cost-effective, and capable of simulating fire propagation under various environmental conditions. However, a specific model for volcanic-induced fires is still missing. Leveraging our experience with modeling lava flows and estimating the associated hazard using cellular automata (GPUFLOW), our work aims to develop a similar methodology to generate fire hazard maps on Stromboli Island, using a specifically designed cellular automaton to address the limitations of accessibility, usability, and computational costs of existing applications.