Forecasting Benguela Niño & Niña events
Why?
The Angola-Benguela Upwelling System (ABUS) located off the coasts Angola, Namibia and South Africa (Fig. 1a - green circle) is characterized by a strong upwelling dynamics that sustains a diversified marine ecosystem and supports subsistence and commercial fishing industries. However, in the last 50 years, fish catches in the Angolan and Namibian waters have shown large variations and the decline (to the verge of collapse) of several small pelagic fish species. While overfishing has undeniably contributed to the destruction of the ABUS ecosystem, variations in the fish stocks cannot be solely attributed to overexploitation, but most likely result from the interplay between the increasing fishing pressure and ​ natural variability at interannual-to-decadal timescales​. Among the environmental factors, the occasional manifestation of pronounced interannual (~14-18 months) warm and cold events in the Angola-Benguela Area (ABA; 10oS-20oS) known as Benguela Niño and Niña (Fig. 1).​
These anomalous conditions have substantial implications for regional climate (such as southern African rainfall) and can drastically impact the distribution and abundance of ABUS marine resources through significant changes in coastal upwelling intensity, and oxygen content along the continental shelf.
Given the strong socio-economic and ecological implications,​ the development of a new prediction system capturing the predictability of the environmental factors (Benguela Niño and Niña events, associated oxygen events) and their impact on marine ecosystems, is highly relevant to support local ecosystem management and to ensure the long-term sustainability of the living marine resources. This is even more relevant in the context of a world growing population and a changing climate, in which this “​ marine oasis ​ ” is likely to become more and more crucial and vulnerable.
How?
Despite the increasing need and demand for a forecasting capability, skillful predictions useful for the fisheries sector have not been developed yet. There are several key reasons for this. Firstly, the potential for forecasting the large-scale environmental variability depends mainly on the understanding of the processes driving it. However, the mechanisms controlling Benguela events ​ have not been fully understood until recently. In fact it was showed that Benguela Niño and Niña events are significantly related with equatorial dynamics : Zonal wind anomalies in the western – central part of the basin trigger equatorial Kelvin Waves (EKW) that travel eastward along the equator. When they reach the African continent, a substantial part of their energy is transmitted poleward along the African coast as Coastal Trapped Waves (CTW). As they propagate, EKW and CTW affect the currents, stratification, and entrainment, and thereby influence the temperature, and oxygen along the continental shelf​. It is now widely accepted that the equatorial waves are the main cause of Benguela events, explaining ~70% of the interannual temperature variability in the ABA​. The timing of the anomaly seems to be tied to the fluctuations of local atmospheric forcing characterized by variation of the along-shore wind in ABA. ​ The novel discovery that Benguela Niño and Niña events occur roughly one/two months after the development of the wave signal at the equator, open a possibility to predict those events using in-situ and remote sensed data, as well ocean reanalysis. Coupled ocean-atmosphere dynamics could potentially extend the prediction horizon further.
Although there is a potential for prediction, a second main reason why useful predictions for the fisheries sector have not emerged is due to the fact that climate prediction models have in the past suffered from major errors in this region. In this regard, new possibilities have also emerged. ​ Recent studies and projects (as the international project PREFACE, and TRIATLAS) have contributed significantly to the improved prediction skill in the Tropical Atlantic. Statistical prediction tools have also been recently applied to predict marine-ecosystem changes based on mechanistic understanding and observational data. With recent advances in understanding and modelling for the Benguela region, it is now possible to apply such statistical approaches to augment the historic usage of purely dynamical approaches.
In summary, seasonal climate and ecosystem predictions are unprecedented for this region and of highly relevant for a broad sector of domains from science to fisheries industry. Our recent understanding of the dynamics of this system as well as the novel predictions tools now available make the development of a skillful prediction model and reliable forecasts a possibility.