The Benguela Current, which flows down the west coast of southern Africa, and the El Niño phenomenon in the South Atlantic significantly impact African climatic systems, marine ecosystems, and the El Niño Southern Oscillation. Predicting El Niño events in this region has been challenging, and until recently, no one had been able to forecast these warm episodes.

According to Marie-Lou Bachèlery‘s research in Science Advances, projections might now help with community readiness and traditional model creation for the first time. Two forms of El Niño episodes occur in the tropical Atlantic between the coastlines of Brazil and West Africa: the Eastern Atlantic Niño, which is closer to West Africa, and the Middle Atlantic Niño, which has warming in the middle equatorial zone.

These warnings significantly impact ocean life and local weather. Strong warming trends have been seen in the South Atlantic, endangering ecosystems and fisheries. Bachèlery clarified that her first strategy was forecasting these occurrences using climate models. After 18 months, she deemed this strategy ineffectual and said the project was in a “deadlock.” She is currently employed at the Euro-Mediterranean Center on Climate Change in Italy, although at the time, she was stationed at the University of Bergen‘s Geophysical Institute.

AI Breaks the Modeling Barrier

Due to their incapacity to replicate upwelling dynamics when wind drives colder water to the surface, traditional climate models frequently fail to forecast warm episodes in the tropical Atlantic. This lack of precision skews forecasts of temperature and atmospheric interactions.

Bachèlery claimed that her familiarity with the area prompted her interest in machine learning. With the correct data inputs, she developed a model to forecast warm episodes two months ahead. According to her supervisor, Professor Noel Keenlyside, many have attempted and failed to predict Atlantic Niño occurrences for decades. He said that this new approach was successful because it avoided the well-known flaws in traditional models.

The study was carried out as part of the TRIATLAS project, which aims to comprehend the marine ecosystems of the South and Tropical Atlantic and forecast their future over a period of months to decades. Keenlyside clarified that these projections will aid in fishery management by directing choices during extreme occurrences.

Also Read: IMD Issued Heat Wave Alert For Delhi And Northern India Till April 18

Looking Ahead

Bachèlery claimed that she first questioned the model’s output and kept checking the results for accuracy. The machine learning system examined maps of sea surface temperatures to find patterns in actual data based on physical causes.

She continued that positive responses have been received, with individuals asking questions about the algorithms utilized. The group is working with Angola’s National Institute for Fisheries to improve the data for real-world applications and is developing a dashboard to make projections available. According to Keenlyside, the growing relevance of basic research to society is positive. Predicting El Niño events this way could greatly improve climate forecasts and decision-making.

Also Read: California’s $59 Billion Agricultural Industry Faces Uncertainty Amid Trump’s Wavering Tariffs