California AI Redefines Ocean Mapping

Ocean currents, with their changing patterns, play a crucial role in regulating climate and weather conditions. Understanding their behavior remains a major challenge. However, a team of researchers from the University of California, San Diego has developed an artificial intelligence tool, named GOFlow, which promises to revolutionize the mapping of these currents with unmatched precision.

On April 13, their work was published in the journal Nature Geoscience. The team trained an AI network using thermal images from weather satellites, in an approach they call GOFlow (Geostationary Ocean Flow). Luc Lenain, an oceanographer at the Scripps Institution of Oceanography at UC San Diego and the lead author of the study, explained to CNET that this technology now allows for the observation of rapidly changing small ocean currents from space with much greater detail and frequency than before. These currents are crucial as they help control how heat, carbon, nutrients, and pollutants move through the ocean.

A New Observation Method

A few years ago, Luc Lenain noticed visual patterns in the temperature variations of major currents, such as the Gulf Stream, while examining satellite thermal images of the North Atlantic Ocean. This observation led him to envision a new method for measuring ocean currents using artificial intelligence. The researchers trained the GOFlow neural network on simulated ocean currents and then applied it to real satellite images. The AI tool used these images to track surface temperatures, which vary due to the underlying ocean currents. By following these temperature changes, GOFlow was able to deduce which current caused them.

The team verified the accuracy of its results by comparing them to data collected by ships in the Gulf Stream region. They also tested GOFlow's results against more traditional satellite methods, which rely on tracking variations in ocean surface height. The researchers found that GOFlow's results aligned with those obtained through other methods but offered a level of detail on ocean currents that had previously only been documented in computer models. "These types of AI-driven approaches do not replace physics," Lenain stated. "On the contrary, AI helps us extract physical information that is already present in satellite observations but has been difficult to retrieve with traditional methods until now."

Challenges and Future Perspectives

Despite the advancements brought by GOFlow, the researchers acknowledge certain limitations, including cloud cover that can obstruct satellites' view of the ocean. To overcome these obstacles, they plan to integrate additional satellite data into their future research.

The computer code developed for GOFlow will be made public, allowing the scientific community to reproduce and extend this work. "We wanted to make this work transparent, reproducible, and useful to the broader community," Lenain said. "We see GOFlow as a step towards a more routine use of large remote sensing data sets combined with machine learning."

Using satellite images to better understand ocean currents is an example of Earth observation. This data is essential for governments, military organizations, as well as for farmers and insurance companies, who rely on it for decision-making. The GOFlow project is part of a broader trend where AI is used to accelerate and improve the accuracy of data analysis. Organizations such as NASA, the European Space Agency, and private space companies have begun to develop and test AI tools capable of analyzing these types of data.