Researchers at the University of California, San Diego, have created an AI tool called GOFlow that maps ocean currents across vast ocean areas with unprecedented detail using satellite thermal imagery. The team published their findings in Nature Geoscience on April 13. The system tracks small, fast-changing currents that influence the movement of heat, carbon, nutrients, and pollutants.
Luc Lenain, an oceanographer at UC San Diego's Scripps Institution of Oceanography and lead author of the study, developed GOFlow after spotting temperature patterns in North Atlantic satellite images, particularly from the Gulf Stream. The neural network was trained on simulated ocean currents and then applied to real thermal images from weather satellites. By monitoring shifts in surface temperatures, GOFlow infers underlying current activity.
