Forecasting oil spills in the Caribbean Sea. The Amuay Refinery Incident

The Colombian Caribbean harbors promising offshore exploitation zones, but they are also highly susceptible to oil spills, which lead to impactful environmental, social, and economic consequences. In this context, this study analyses the development of a new ecosystem of climate services for Colombia, including maritime oil spills, based on different observations and the use of NOAA's General Operational Modelling Environment (GNOME). To showcase its performance, the study focuses on the oil spill event in the Colombian and Venezuelan Caribbean Sea: The Amuay refinery incident of October 31st, 2017. Simulations were conducted using the GNOME model with various atmospheric forcings –from the Climate Forecast System version 2 (CFSv2) and the Centre ERS d'Archivage et de Traitement (CERSAT)– and oceanic forcings –the Copernicus Global Ocean Physics Reanalysis (GLORYS), the Hybrid Coordinate Ocean Model (HYCOM), and the Navy Coastal Ocean Model (NCOM). The results were compared with ground-based data and satellite imagery, incorporating optical and radar satellite data (Planet, Landsat 8, and Sentinel-1). The use of the system in the Colombian-Venezuelan Caribbean effectively replicated the spill's arrival date on the Colombian coast and the affected areas, providing a trajectory closely aligned with satellite observations. Nevertheless, the simulations' efficacy was found to be contingent on the spill's initial conditions and the utilized forcing data. The results indicate that, at least for cases like the one analyzed in the Colombian-Venezuelan Caribbean Sea, GNOME is a useful analytical tool for immediate decision-making in hydrocarbon spill incidents. It has the potential for extended forecast horizons at sub-seasonal timescales. However, ongoing efforts to enhance prediction accuracy are essential to strengthening the capacity to respond to oil spill emergencies in the future.