Southern Ocean's carbon absorption may reverse once emissions stop

The Southern Ocean, the ring of water encircling Antarctica, absorbs more carbon dioxide than any other single ocean region. It has accounted for about 40 to 50% of the total annual uptake of human-caused carbon dioxide by the world's oceans, according to research cited in a study using the Community Earth System Model version 2 (CESM2). A separate analysis published in Nature reports that the region contributed about 40% of global ocean uptake of that carbon between 1870 and 1995.

Before considering how this sink might change in the future, one study questions how well it has even been measured in the recent past. Researchers found that machine learning methods used to estimate the Southern Ocean carbon sink have likely overestimated it over the last three decades, by around 0.4 billion tonnes of carbon per year compared to estimates from ocean biogeochemical models. The cause traces back to when and where measurements of surface ocean carbon dioxide levels are collected: most come from summer, when concentrations are low, while winter measurements, when concentrations run high, are far sparser. In most of the Southern Ocean, existing measurements covered fewer than four separate months across the three decades from 1992 to 2021.

When researchers train prediction models mostly on summer data, the models struggle to accurately estimate winter conditions, and specifically underestimate how much carbon dioxide the water holds during winter. In waters south of 60 degrees latitude, this led to an average underestimation of surface water carbon dioxide levels by 5.77 microatmospheres of pressure when models were trained on all months without correction. Retraining models with a better balance of winter measurements, and adjusting which variables the models use to make predictions, reduced that gap to close to zero. This does not change the physical fate of the Southern Ocean, but it does mean that estimates of how strong a carbon sink it currently is have carried more uncertainty than previously recognized.

Separately from the measurement question, a new modeling study asks what happens to the Southern Ocean's carbon uptake once humanity stops adding carbon dioxide to the atmosphere, or even actively starts removing it. Researchers led by Yechul Shin at Seoul National University ran simulations with CESM2, testing two idealized future paths: one in which emissions fall to zero and stay there, and one in which humanity goes further and actively pulls carbon dioxide out of the air after reaching net zero. Whether carbon dioxide flows into or out of the ocean depends on the difference in carbon dioxide pressure between the air and the surface water; carbon flows from the higher-pressure side to the lower-pressure side. As emissions rise, the atmosphere holds more carbon dioxide than the ocean surface, so the gas flows into the water. The study found that once atmospheric concentrations start falling after net zero, this pressure difference reverses, and the ocean begins releasing carbon dioxide back into the air instead of absorbing it.

In the model, the Southern Ocean transitions from net carbon sink to a substantial source during the zero-emission and negative-emission phases described above. During what the researchers call the restoring period, roughly the years 2300 to 2400 in the simulation, the region emits about 0.33 billion tonnes of carbon per year in the zero-emissions scenario and about 0.35 billion tonnes per year in the scenario with active carbon removal. That is a reversal from the same region's slight uptake of about 0.1 billion tonnes per year at the start of the simulation period, in the early 2000s. Other ocean regions also see uptake weaken as atmospheric carbon dioxide falls, but the Southern Ocean's shift from sink to source was the most pronounced of any region studied.

Two mechanisms drive this shift, according to the CESM2 study. First, the Southern Ocean carries enormous amounts of absorbed heat down into deep water along with the carbon; as that heat gradually resurfaces, warmer water holds less dissolved gas, so more carbon dioxide escapes. Second, as surface waters warm they become less dense, so they mix less readily with the colder, denser, more alkaline water below. Alkalinity is a property of seawater, tied to the shells of certain plankton, that affects how much carbon dioxide the water can chemically absorb; because the study found that reduced mixing brings less alkalinity to the surface, the water's capacity to draw down further carbon dioxide from the air is diminished. The study describes this second effect as a globally uniform decline in surface alkalinity, working alongside the regional warming.

The consequences reach beyond the ocean itself, according to the CESM2 study and outside comment on it. In modeled scenarios where the world reaches net zero and atmospheric carbon dioxide begins to decline, the Southern Ocean's air-sea carbon exchange goes from absorbing about 3 grams of carbon per square meter per year today to emitting 8 to 9 grams per square meter per year by 2400. In the scenario that includes active carbon dioxide removal after net zero, atmospheric concentrations were projected to actually rise again for about 50 years before leveling off around 425 parts per million, close to today's level; without carbon removal, concentrations were projected to level off higher, around 590 parts per million.

Nicolas Gruber of ETH Zurich, who was not involved in the study, described the dynamic as similar to pressing a piston into the ocean while emissions rise, then having the carbon dioxide come back out once the piston is pulled up. He noted that the new research points to two feedbacks, the heat and alkalinity effects described above, that may amplify this release beyond what was previously expected. Alberto Naveira Garabato of the University of Southampton, who also was not involved in the study, said that changes in how the ocean interacts with the atmosphere could have an effect on the global carbon budget comparable to adding a large new source of emissions.

Yechul Shin, one of the study's researchers, said the impact of this transition can last far longer than expected, so that a shift from carbon sink to source in the Southern Ocean continues to work against efforts to reduce atmospheric carbon dioxide even after emissions have been cut. Gruber compared the situation to backing a car out after crashing it into a wall: removing carbon dioxide from the atmosphere does not simply reverse the damage, because the ocean system that emerges is not the same as the one that existed before. The researchers note that the sooner net zero is reached and Southern Ocean warming is halted, the less carbon dioxide these waters will ultimately release.

The scenarios modeled are described by New Scientist as not necessarily realistic: they assume net zero reached in either 2125 or 2025 with atmospheric carbon dioxide as high as 700 parts per million, whereas many countries, including the UK and European Union members, target net zero as early as 2050, and current carbon dioxide removal remains minuscule compared to ongoing emissions. Separately, a funded research project is examining another source of uncertainty: the role of storms. Southern Ocean mid-latitude cyclones, among the deepest in the world, have been observed using underwater gliders to trigger episodes of carbon dioxide outgassing when they mix carbon-rich water up from below. These storms are projected to become more frequent over the coming century due to climate change, and researchers describe their overall role in the region's carbon cycle as not well understood, setting up what they call a potentially large but previously unexplored feedback between storminess and the carbon cycle.

Meanwhile, the finding that existing carbon sink estimates have likely been overestimated due to uneven measurement coverage is a reminder that even the present-day baseline carries real uncertainty. The Nature-published study notes that better data collection, including from sailboats or autonomous floats, alongside improved modeling algorithms, will be needed to narrow this gap. Together, these threads point to a Southern Ocean carbon budget that is harder to pin down, and potentially far more consequential for future climate outcomes, than has generally been assumed.