EU Member States, through an EMSA initiative, recognized the need for a tool to support oil spill response at the EU level. EMSA subsequently published a document outlining functional, non-functional, and technical requirements, with a focus on oil spill modeling, the efficiency of response operations, and the importance of accurate metocean data (winds, currents) for effective drift analysis and response strategies.
AI4COPSEC evaluates its real-time capabilities for providing accurate oil spill drift predictions in the context of CSS enhancement. Open-source drift models (GNOME, OpenDrift) are validated against Melodi drifter data and compared with the AI4COPSEC-developed RNN drift model, using live weather and oceanographic data. Three ocean current products are assessed: the Glorys analysis (CMEMS), AIS-derived currents from eOdyn, and a fused product integrating Copernicus data, AIS currents, and the 4DVarnet algorithm. Model validations are performed using NCLS indicators and other Lagrangian diagnostics, with all interactions conducted through the AI4COPSEC integrated system.
A historical analysis focuses on EU waters. AI4COPSEC technologies, including AI/ML methods, data fusion algorithms, and historical environmental data, are applied to simulate and analyse the aftermath of the incident. This scenario demonstrates the potential of AI4COPSEC to enhance predictive accuracy and crisis management capabilities for maritime emergencies.
