Physical Address
304 North Cardinal St.
Dorchester Center, MA 02124
Changing patterns of ocean currents shape our climate and weather. Even today, understanding how ocean currents behave remains difficult. But researchers have developed Artificial intelligence tools It can map currents across large swaths of the ocean at a level of detail that was previously unattainable.
A team from the University of California, San Diego, Publish their work In Nature Geoscience on April 13. They trained the AI network using thermal images from weather satellites in an approach they called GOFlow (geostationary ocean flow).
“We can now observe small, rapidly changing ocean currents from space with much greater detail and frequency than before,” Luke Lenin, an oceanographer at the Scripps Institution of Oceanography at the University of California, San Diego, and first author of the study, told CNET. “Those currents are important because they help control how heat, carbon, nutrients and pollutants move through the ocean.”
ebb and flow
A few years ago, while Lenin was looking at thermal satellite images of the North Atlantic Ocean, he noticed visible patterns in temperature changes caused by major currents, such as the Gulf Stream. He had an idea for a new way to measure ocean currents by taking what his eye could see and putting it into an artificial intelligence tool.
The researchers trained the GOFlow neural network to simulate ocean currents, then used it on real images from a weather satellite. The AI tool used satellite images to track surface temperatures, which change due to underlying ocean currents.
By tracking temperature changes in the images, GOFlow deduced what current caused them.
The team checked their work against data collected by ships in the Gulf Stream area. They also tested GOFlow’s results against traditional satellite methods that rely on tracking elevation changes in the ocean surface.
The researchers found that their outputs are consistent with results from other ships and satellites, but they say GOFlow also provided greater detail about ocean currents than had previously been documented in computer models alone.
“These kinds of AI-based methods do not replace physics,” Lenin said. “Instead, AI helps us extract physical information that already exists in satellite observations, but has been difficult to recover with traditional methods until now.”
The view from the top
Despite GOFlow’s accomplishments, researchers note that the tool has some limitations, such as cloud cover: cloudy days can obscure a satellite’s view of the ocean. They say future work will include additional satellite data to fill these gaps.
The computer code developed by the researchers will be made available to the public to assist in further work.
“We wanted to make this work transparent, replicable, and useful to the broader community,” Lenin said. “We see GOFlow as a stepping stone toward more routine use of large remote sensing datasets combined with machine learning.”
Collecting satellite images to learn about ocean currents is an example of Earth observation. Governments and militaries, as well as farmers and insurance companies, rely on this data to make decisions.
The GOFlow project is part of a larger AI trend, where AI tools can speed up and improve the accuracy of data analysis. NASAthe European Space Agency Private space companies have begun building and testing artificial intelligence tools capable of analyzing such data.