SILVANUS: Operational ML-based Fire Danger Index

Preventing forest fires is crucial to mitigate the significant economic and human losses caused by wildfire outbreaks, which are expected to worsen due to climatic changes. Identifying regions at high risk for forest fires is essential for both preventing wildfire occurrences and optimizing resource management during wildfire season. We have developed an operational pipeline for estimating the daily Fire Danger Index (FDI) using a data-driven approach and machine-learning techniques. This presentation will provide an overview of the pipeline’s architectural framework, detail the machine-learning model utilized, and showcase FDI maps generated for multiple European test sites where the pipeline has been successfully deployed.