AI & Argo Floats Offer Breakthrough in Atlantic Ocean Circulation Monitoring

Researchers at the GEOMAR Helmholtz Centre for Ocean Research Kiel and Kiel University have demonstrated how Artificial Intelligence (AI) and data from the global Argo float programme can be used to monitor the Atlantic Ocean’s key circulation systems.

The ocean current system known as the Atlantic Meridional Overturning Circulation (AMOC) acts as a vast conveyor belt transporting warm surface water northwards, where it cools, sinks, and flows back south as cold deep water. This system transports substantial heat and heavily influences weather and climate, particularly across Europe. However, direct observations have historically depended on a few fixed measurement series in the Atlantic that are technically challenging and expensive to maintain.

A new study published in the journal Ocean Science outlines how data from the international Argo float programme, comprising roughly four thousand autonomous profiling buoys, can be repurposed to monitor this complex mechanism far beyond its original scope. At regular intervals, these buoys descend to a depth of 2,000 meters and measure parameters including temperature, salinity, and pressure before transmitting the readings via satellite.

By training a machine learning model on ocean simulations, researchers enabled the system to identify correlations between typical current patterns and specific temperature and salinity profiles. When applied to real-world Argo data, this approach allows scientists to infer large-scale current strengths, specifically the geostrophic component of the circulation, from isolated point measurements.

Dr Yannick Wölker, lead author of the study and until recently a PhD student in the Ocean Dynamics research unit at GEOMAR and the research group Archaeoinformatics – Data Science at Kiel University, said, “We wanted to know whether we could use the scattered Argo measurements to gain insights into large-scale circulation systems that, until now, could only be recorded through very extensive measurement campaigns. Artificial intelligence opens up new possibilities here. Combining machine learning with established physical models allows us to get more out of existing measurement data and better understand how key circulation systems work.”

The calculated estimates align closely with established observational series and model simulations, offering an efficient framework for ocean monitoring. Despite these promising results, the authors emphasize clear methodological limitations. Because the technique relies on model assumptions and covers specific time windows, very short-term fluctuations and the overall long-term decline in the AMOC can only be determined to a limited extent. Consequently, the approach is designed to complement rather than replace physical infrastructure.

Yannick Wölker said, “Our approach is no substitute for direct measurements in the ocean. But it can help to make better use of existing data and bridge gaps in observations.”

Integrating data science with oceanography holds strong implications for future observation strategies and international climate monitoring networks.

Prof. Dr Arne Biastoch, a professor at GEOMAR and co-author of the study, added, “Particularly with regard to long-term climate monitoring, we need to consider how to design observations that are efficient, robust and internationally coordinated. Artificial intelligence methods can help to ensure regarding future ocean measurements are made in the best possible way that strategic decisions.”

The study was conducted within the Helmholtz School for Marine Data Science (MarDATA), a doctorate program uniting marine scientists and computer scientists from Kiel and Bremen.