MIT Researchers Develop Machine-Learning Model to Accurately Predict Ocean Currents

A team of computer scientists and oceanographers at MIT have developed a new machine-learning model that incorporates knowledge from fluid dynamics to generate more accurate predictions about ocean currents. The new model could help scientists more accurately monitor the transportation of biomass, carbon, plastics, oil, and nutrients in the ocean. By precisely forecasting the weather or approximating how oil will spread after a spill, the researchers could help in search and rescue, track plastic pollution, and combat climate change. The study was presented at the International Conference on Machine Learning.