Researchers published a study in Nature Geoscience on April 20, 2026, detailing a new artificial intelligence framework named GOFLOW (Geostationary Ocean Flow). This technique enables the measurement of ocean surface currents at a resolution and frequency previously unattainable with conventional satellite altimetry. By applying advanced optical flow algorithms to thermal infrared sequences captured by existing geostationary weather satellites, GOFLOW provides a continuous, high-definition view of ocean dynamics across the global tropics and mid-latitudes.
The GOFLOW system utilizes data from the current generation of geostationary orbiters, including the National Oceanic and Atmospheric Administration (NOAA) GOES-16 and GOES-18 satellites, the Japan Meteorological Agency’s Himawari-8 and Himawari-9, and EUMETSAT’s Meteosat Third Generation. These satellites capture sea surface temperature data every 10 to 15 minutes. The AI model tracks the movement of thermal patterns—small-scale temperature gradients—to derive surface velocity vectors. This approach bypasses the limitations of traditional nadir-pointing radar altimeters, which typically offer spatial resolutions of approximately 100 kilometers and temporal repeats of several days to weeks.
Technical specifications released in the study indicate that GOFLOW achieves a spatial resolution of approximately 2 kilometers. This allows for the observation of submesoscale eddies and fronts, which are critical for understanding how heat, carbon, and nutrients are transported within the ocean. The researchers noted that the AI was trained on massive datasets to distinguish between actual water movement and atmospheric interference, such as cloud drift or water vapor fluctuations, which have historically hindered the use of infrared data for current mapping.
The implementation of GOFLOW does not require the launch of new orbital hardware, representing a significant optimization of existing multi-billion-dollar satellite constellations. According to the lead authors, the technique fills a critical data gap in the submesoscale range, where much of the ocean's kinetic energy is dissipated. The study highlights that the system can generate real-time current maps, a capability that has immediate applications for maritime operations, including search and rescue coordination, oil spill trajectory modeling, and the tracking of plastic debris.
Validation of the GOFLOW outputs was conducted using a global array of drifting buoys and ship-borne acoustic Doppler current profilers. The results demonstrated a high correlation with in-situ measurements, particularly in regions with strong thermal gradients like the Gulf Stream and the Kuroshio Current. The researchers stated that the framework is now being integrated into experimental data streams for further testing by meteorological and oceanographic agencies.