Akshay DineshThe Southern Ocean is one of the most important places on Earth for moving carbon out of the atmosphere, yet scientists have far fewer measurements there than they would like. The problem is not a lack of interest. Winter brings darkness, drifting ice, towering waves, and winds strong enough to make a long research voyage dangerous and expensive. A new five-year project is sending wind-powered, uncrewed Saildrone vehicles around Antarctica to keep measuring when ships rarely can. Their goal is not simply to prove that the ocean absorbs carbon dioxide, but to learn where, when, and how quickly that exchange changes.
The project is called Constraining Ocean Carbon with Optimized Observing, or COCO2. Led by researchers at the National Oceanic and Atmospheric Administration, Columbia University, and Australia’s Commonwealth Scientific and Industrial Research Organisation, it began with a vehicle departing Hobart, Tasmania, in September 2026. Three more are due to join it during 2027. Each one is expected to circle Antarctica repeatedly, bringing a missing season into view and testing how autonomous instruments can improve estimates of the global carbon budget.
Why the Southern Ocean matters so much
Oceans take up roughly a quarter to a third of the carbon dioxide released by human activity each year. That does not make the extra carbon harmless: absorbed carbon changes seawater chemistry and contributes to ocean acidification. Still, without this ocean sink, much more carbon dioxide would remain in the atmosphere and warming would advance faster.
The Southern Ocean has an outsized influence because it connects the Atlantic, Pacific, and Indian oceans. Rather than being boxed in by continents, it is wrapped by the Antarctic Circumpolar Current, the only major current that flows all the way around the planet. Strong winds push surface water northward and help draw deep water upward. Elsewhere, cold surface water sinks and carries dissolved carbon toward the ocean interior, where some of it can remain isolated from the atmosphere for decades or centuries.
This circulation makes the region both a carbon gateway and a source of uncertainty. Cold water can hold more dissolved carbon dioxide than warm water, which favors absorption. Microscopic phytoplankton also take up carbon during photosynthesis, and some of that organic material sinks after the organisms die or are eaten. But upwelling can bring old, carbon-rich water back to the surface, where carbon dioxide may escape into the air. The Southern Ocean is therefore not a uniformly operating sponge. It is a restless patchwork whose behavior changes with winds, currents, biology, temperature, and season.
The winter gap in ocean measurements
Research ships remain indispensable because they carry scientists, laboratories, and instruments that can sample far below the surface. They can collect water, identify organisms, and make careful calibrations that a small vehicle cannot. But a ship is also costly to operate, limited in where it can safely go, and usually present for only a short campaign. Those limits matter most during the Antarctic winter, when the most severe conditions coincide with some of the largest gaps in carbon observations.
Satellites provide broad views of sea-surface temperature, color, winds, and ice, but they cannot directly measure every part of the exchange between seawater and the air above it. Profiling floats can descend through the water column and reveal changes at depth, yet they do not measure the atmosphere at the same point in the same way as a surface vehicle. Moorings produce valuable long time series, but each one stays in a fixed location. No single observing tool can solve the coverage problem by itself.
The missing winter data can bias an annual estimate. Imagine trying to calculate a household’s yearly electricity use after observing it mainly in mild weather. Even excellent spring and summer readings may miss winter heating. Ocean carbon exchange has its own seasonal swings, and researchers need measurements from the difficult months to learn whether apparent annual absorption is too high, too low, or simply in the wrong place.
How a Saildrone turns wind into a research platform
A Saildrone Explorer looks more like a narrow orange surfboard with a rigid vertical wing than a conventional ship. Wind supplies its propulsion, while solar panels power sensors, computers, cameras, and satellite communications. Operators can change its course remotely, but nobody is aboard. That combination allows the vehicle to remain at sea for months without carrying fuel for propulsion or exposing a crew to every storm.
The design still faces punishing engineering demands. Saltwater corrodes electronics. Waves strike the structure repeatedly. Ice can damage equipment, solar power is limited during polar winter, and satellite contact must work far from land. The project’s vehicles use a strengthened wing developed after an earlier Saildrone completed the first uncrewed surface-vehicle circumnavigation of Antarctica in 2019. That 196-day voyage covered 11,879 nautical miles and showed that the platform could collect useful data in hostile water.
Each COCO2 voyage is expected to cover about 12,000 miles around Antarctica before the vehicle returns to Hobart for repair, sensor replacement, and redeployment. The fleet will not simply repeat one fixed track. Researchers at Columbia will use models and existing observations to locate areas where an additional measurement is most likely to reduce uncertainty. Routes can then be adjusted so that the vehicles sample those high-value regions, a strategy known as optimized observing.
Measuring carbon exchange, not just carbon
To tell whether carbon dioxide is entering or leaving the ocean, researchers need paired measurements. Sensors compare the amount of carbon dioxide dissolved near the sea surface with its concentration in the air. They also record wind, because stronger turbulence generally speeds gas transfer across the surface. Temperature and salinity help determine how readily carbon dioxide dissolves, while barometric pressure, humidity, and solar radiation describe the atmosphere surrounding the measurement.
The vehicles also measure dissolved oxygen and chlorophyll. Chlorophyll is a useful indicator of phytoplankton abundance, though it is not a complete count of life in the water. Oxygen and chlorophyll help researchers distinguish changes driven by physics from those connected to photosynthesis, respiration, and ecosystem activity. A cold patch may absorb carbon because of solubility, for example, while a phytoplankton bloom can lower carbon dioxide in surface water through biological uptake.
Scientists combine these observations to estimate the direction and rate of carbon dioxide flux across the air-sea boundary. A large difference between air and water does not automatically mean a large flux; calm water can slow the exchange, while strong winds and breaking waves can accelerate it. The calculation is sensitive to the details, which is why a platform that measures several conditions at the same place and time is so valuable.
Measurements are transmitted in near real time to NOAA’s Pacific Marine Environmental Laboratory. From there, automated systems can share suitable data with research users and the World Meteorological Organization’s Global Telecommunication System, which supports weather and climate services. The carbon observations are also intended to improve the Global Carbon Budget, the international assessment that tracks human emissions and the amount taken up by land and ocean sinks.
What five years of repeated voyages could reveal
One trip around Antarctica offers a snapshot; repeated circuits begin to show a system. NOAA’s 2019 mission found that some winter waters released carbon dioxide when carbon-rich deep water reached the surface. That result did not mean the entire Southern Ocean had stopped being a net carbon sink. It showed why sparse summer sampling can miss important episodes and why researchers should not treat a vast ocean as if it behaved the same everywhere.
With four vehicles operating across seasons, the COCO2 team can compare one winter with another and watch how shifting winds, sea ice, currents, and biological activity alter the exchange. The observations should help test the maps produced from ships, floats, satellites, and statistical models. Where those estimates disagree, targeted measurements can reveal which assumptions need work. Better observations can also show whether a surprising signal is a durable regional pattern or a short-lived event.
There are limits. Four surface vehicles cannot observe the entire depth of the ocean, identify every biological process, or replace research ships and floats. Their paths will still leave large spaces unsampled, and interpreting carbon storage requires models that connect the surface to deeper circulation. The mission’s strength lies in adding persistent, coordinated surface measurements where the observing network is weakest.
That makes the project larger than a story about a clever ocean robot. Climate projections depend on knowing how much carbon natural sinks absorb and whether that uptake is changing. The Southern Ocean exerts enormous influence from a place most people will never see, and its harshest season has long hidden part of the evidence. By sending instruments into that gap year after year, researchers can replace a broad assumption with a clearer record of when the ocean breathes carbon in and when it releases some back.
