A pioneering forest fire decision support system enhances resilience in the face of climate change
Forest fires are increasing in frequency and intensity, often causing irreparable damage to ecosystems, people and livelihoods. The 2022 EU wildfire season was the second worst on record since 2000. Although the total area burnt in 2023 was much less than that in 2022, more than 40 % of areas burnt in the EU were part of Natura 2000, the EU’s biodiversity protection network. Current forest fire management approaches are challenged to keep pace with the rapidly evolving nature of wildfires exacerbated by climate change. The EU-funded SAFERS project created an integrated platform featuring a forest fire decision support system to enhance societies’ resilience against forest fires.
Real-time high-resolution data integration
SAFERS leverages an extensive array of data sources. Images from in situ cameras detect signs of fire in near-real time. According to project coordinator Claudio Rossi of LINKS Foundation: “Earth observation data from the EU’s Copernicus programme provides satellite observations and topographic data. Sentinel-2 satellites are key to SAFERS because they are used to compute updated fuel and burned-area delineation maps at a 10-metre resolution, using advanced AI based on deep neural networks.” This satellite data is combined with crowdsourced information from social media or the SAFERS communication tool implemented as a chatbot. The latter enhances situational awareness with information exchange between command-and-control centres and those in the field regarding an ongoing emergency. In this way, SAFERS actively involves citizens and first responders, empowering them with knowledge and tools for better self-protection and response strategies.
AI algorithms support intelligent services and actionable information
“Our AI algorithms digest this wealth of data to produce automatic alert generation and decision support information such as early warning maps and fire propagation simulations, enabling prompt and informed responses to emerging threats. They also produce burned area and impact assessments, as well as vegetation recovery monitoring. This is a significant leap forward, showcasing how technology can be leveraged to protect our communities and natural environments more effectively,” adds Rossi. SAFERS was tested in four countries in the Mediterranean, where fires are increasingly devastating. Iosif Vourvachis, head of the Hellenic Rescue Team and part of the Greek demonstration case, notes that "SAFERS is a platform that delivers actionable results to decision-makers," encapsulating SAFERS’ aims and ultimate success. The team has made sophisticated data analysis accessible and practical, empowering those responsible for making life-saving decisions with the best possible information at a crucial moment.
Open-source platform, business-to-business intelligent services
“The platform's open-source release invites adaptation and expansion, potentially revolutionising how we approach emergency management across Europe and beyond. Although focused on forest fires, the methodologies and technologies we've developed have potential applications in managing other types of crises, such as floods, extreme weather events, and landslides. In particular, the probabilistic sub-seasonal weather forecasts, social media monitoring and chatbot can be reused with minor modifications; the other services will require some adaptation,” Rossi explains. In addition to the adaptable open-source platform, interested parties can purchase intelligent services via a business-to-business model targeting organisations seeking tailored solutions and advanced functionalities including companies, government agencies and other entities. SAFERS demonstrates what can be achieved when diverse partners from across Europe – including research institutions, technology companies and emergency services – come together with a common goal. Its unified web-based emergency management platform sets a new standard in emergency response.
Keywords
SAFERS, AI, forest fire, fires, wildfire, emergency management, intelligent services, decision support system, Copernicus, emergency response, emergency services
