BENGUP · Climate and marine-ecosystem predictions in the Angola-Benguela Upwelling System

Climate and marine-ecosystem predictions in the Angola-Benguela Upwelling System

Off the southwestern African coast, the Angola-Benguela Upwelling System (ABUS) is a crucial ecological and economic hotspot. This region, with its robust upwelling patterns, sustains a diverse marine ecosystem and local fishing communities. However, it's been facing challenges in recent decades, with fish stocks fluctuating and some species nearing collapse. Overfishing alone doesn't explain this; it's the interplay of fishing and natural variability. The ABUS ecosystem is affected by interannual climate events, like Benguela Niño and Niña, with far-reaching consequences for regional climate and the local ecosystem. Dr. Bachèlery's BENGUP project in collaboration with experts from the University of Bergen (UiB) aims to develop a pioneering prediction system for these events, providing seasonal forecasts for ABUS variability. Why does it matter? This project has global significance. In a world dealing with population growth and environmental changes, the ABUS is a critical yet vulnerable "marine oasis." Developing a prediction model for this region could transform its management, ensuring the long-term sustainability of marine resources, and set an example for similar ecosystems worldwide. But why have skillful predictions for the ABUS been elusive? It's partly because traditional computer forecasting models used for weather, climate, and ocean conditions have limitations, especially in regions with complex dynamics like the tropical Atlantic and upwelling systems. Additionally, our understanding of what triggers events like Benguela Niño and Niña was incomplete until recently. To predict them, we first needed to grasp their causes. So, how do we tackle this challenge and predict extreme events in the ABUS? The BENGUP project has three key objectives. Dynamical prediction systems play a crucial role here. They provide vast data and serve as digital laboratories to replicate the environment, aiding our understanding of ocean and atmosphere dynamics. The project starts by evaluating the predictive skills of existing models and understanding what makes a model good at forecasting extreme events. By comparing model predictions with real-world observations, we pinpoint where improvements are needed. Reducing uncertainties in predictions is crucial, and we work on this continuously. Finally, the third objective is to develop a new prediction system using cutting-edge machine learning methods, overcoming obstacles faced by dynamic prediction systems. This project aims to provide more accurate predictions for a vital marine ecosystem, benefiting local communities and global conservation efforts.

Текст от CORDIS, на английски · Данни: CORDIS, © Европейски съюз