Scientists have developed a new way to track ocean currents using artificial intelligence, revealing movements in the ocean that have been difficult to observe before. The method could help researchers better understand climate change, marine ecosystems and how the ocean stores carbon.
The system, called GOFLOW, uses thermal images from weather satellites and turns them into detailed maps of ocean surface currents. Instead of launching new satellites, researchers used AI to study patterns in water temperature and track how they shift over time.
The research was led by scientists at the Scripps Institution of Oceanography.
Ocean currents do much more than move water. They transport heat around the planet, help regulate the climate, move nutrients that support marine life and pull carbon deeper into the ocean. But some of the most important currents are small, fast-changing and hard to capture with traditional tools.
Many existing satellites can only revisit the same area every several days, which misses changes that can happen within hours. Ships and coastal radar can collect detailed data, but only in limited areas.
GOFLOW helps fill that gap. Using deep learning, researchers trained the system to recognize how temperature patterns on the ocean surface bend, stretch and shift as currents move beneath them. Once trained, the AI could turn satellite image sequences into hourly maps of ocean flow.
“Weather satellites have been observing the ocean surface for years,” researcher Luc Lenain said. “The breakthrough was learning how to turn that time-lapse into hourly maps of currents.”
One area of interest was the Gulf Stream, where the system captured small swirling currents and eddies in far greater detail than older methods. Scientists say these features are important because they drive vertical mixing, the process that moves nutrients upward and carbon downward.
“This opens a range of exciting possibilities,” Lenain said. “We can now measure key signatures of these small, intense currents using real observations.”
Researchers say that matters for studying climate, improving ocean models and even practical uses such as oil spill tracking or search and rescue operations.
Another major advantage is cost. The use of satellites already used for weather monitoring allows the method to be scaled without investing extensively in new infrastructure.
However, there are still challenges. Clouds can block thermal imagery, limiting what satellites can see. Researchers say they hope to combine more data sources to improve coverage.
