A study by the Alfred Wegener Institute (AWI) offers a possible explanation for why the ocean surrounding Antarctica continues to absorb carbon dioxide, contrary to climate model predictions and despite the ongoing effects of climate change.
Climate projections have long indicated that global warming might weaken the Southern Ocean’s capacity to absorb carbon dioxide (CO2). Yet, long-term measurements reveal that this crucial ability has remained largely unchanged in recent decades. A new study by scientists at the Alfred Wegener Institute (AWI) offers insight into why this might be the case.
For many years, low-salinity water near the ocean’s surface has helped trap carbon in the deep sea, preventing it from escaping back into the atmosphere. However, climate change is now disrupting this balance and altering how effectively the Southern Ocean functions as a carbon sink. The findings are detailed in the journal Nature Climate Change.
Oceans collectively take up about one quarter of the CO2 produced by human activity. Of that amount, the Southern Ocean alone accounts for roughly 40 percent, making it one of the planet’s most important regions for slowing global warming.
Its powerful influence stems from the region’s unique circulation patterns. Deep waters rise to the surface, exchange gases with the atmosphere, and then sink again, carrying newly absorbed CO2 into the depths.
During this process, the ocean releases naturally stored CO2 from its interior while simultaneously taking in human-generated CO2 from the atmosphere. The balance between these two processes determines how effectively the Southern Ocean can continue absorbing carbon. If more natural CO2 reaches the surface, the ocean’s ability to take in additional CO2 from the air declines. This delicate exchange depends on both ocean circulation and the layering, or stratification, of different water masses.
The Paradox of Climate Models and Reality
The water that rises from the deep layers of the Southern Ocean is ancient, having remained below the surface for hundreds to thousands of years. Over this long period, it has gathered significant amounts of carbon dioxide (CO2), which gradually makes its way back to the surface through the natural upwelling process.
Climate models indicate that as global warming strengthens the westerly winds circling Antarctica, even more of this CO2-rich deep water will be pushed upward. Over time, this could limit how much human-produced CO2 the Southern Ocean is able to absorb, weakening its crucial role as a global carbon sink.
However, contrary to climate model projections, observational data from recent decades has shown no reduction in its capacity as a CO2 sink. A new study from the Alfred Wegener Institute, Helmholtz Centre for Polar and Marine Research (AWI) now provides an explanation as to why, despite strengthening westerly winds, the Southern Ocean has continued to act as a CO2 sink in recent decades and therefore been able to slow down climate change.
“Deep water in the Southern Ocean is normally found below 200 meters,” says Dr. Léa Olivier, AWI oceanographer and lead author of the study. “It is salty, nutrient-rich, and relatively warm compared to water nearer the surface.” The deep water contains a large amount of dissolved CO₂ that entered the deep ocean from the surface a long time ago. Near-surface water, on the other hand, is less salty, colde,r and contains less CO₂. As long as the density stratification between deep and surface water remains intact, CO₂ from the deeper layers cannot easily rise to the surface.
Cold, low-salinity water keeps carbon-rich water contained – however, climate change brings CO₂ dangerously close to the surface
“Previous studies suggested that global climate change would strengthen the westerly winds over the Southern Ocean, and with that, the overturning circulation too,” says Léa Olivier. “However, that would transport more carbon-rich water from the deep ocean to the surface, which would consequently reduce the Southern Ocean’s ability to store CO₂.” Although strengthening winds have already been observed and attributed to human-made change in recent modeling and observational studies, there is no evidence pointing to the Southern Ocean absorbing less CO₂ – at least at this point.
Long-term observations by the AWI and other international research institutes suggest that climate change may be affecting the properties of surface and deep water masses.
“In our study, we used a dataset comprising biogeochemical data from a large number of marine expeditions in the Southern Ocean between 1972 and 2021. We looked for long-term anomalies, as well as changes in both circulation patterns and the properties of water masses. In doing so, we only considered processes related to the exchange between the two water masses, namely circulation and mixing, and not biological processes, for example,” explains Léa Olivier. “We were able to determine that, since the 1990s, the two water masses have become more distinct from one another.”
The Southern Ocean’s surface water salinity has reduced as a result of increased input of freshwater caused by precipitation and melting glaciers and sea ice. This “freshening” reinforces the density stratification between the two water masses, which in turn keeps the CO₂-rich deep water trapped in the lower layer and prevents it from breaking through the barrier between the two layers.
A Temporary Reprieve
“Our study shows that this fresher surface water has temporarily offset the weakening of the carbon sink in the Southern Ocean, as model simulations predicted. However, this situation could reverse if the stratification were to weaken,” summarizes Léa Olivier.
There is a risk of this happening, as the strengthening westerly winds push the deep water ever closer to the surface. Since the 1990s, the upper boundary of the deep water mass has shifted roughly 40 meters closer to the surface, where CO₂-rich water is increasingly replacing the low-salinity winter surface water. As the transition layer between surface and deep water moves closer to the surface, it becomes more susceptible to mixing, which could be primarily caused by the strengthening westerly winds. Such mixing would release the CO₂ that had accumulated beneath the surface water layer.
A recently published study suggests that this process may have already begun. The result would be that more CO₂-rich deep water could reach the surface, which would in turn reduce the Southern Ocean’s capacity to absorb anthropogenic CO₂ and therefore further drive climate change.
“What surprised me most was that we actually found the answer to our question beneath the surface. “We need to look beyond just the ocean’s surface, otherwise we run the risk of missing a key part of the story,” says Léa Olivier.
“To confirm whether more CO₂ has been released from the deep ocean in recent years, we need additional data, particularly from the winter months, when the water masses tend to mix,” explains Prof. Alexander Haumann, co-author of the study. “In the coming years, the AWI is planning to carefully examine these exact processes as part of the international Antarctica InSync program, and gain a better understanding of the effects of climate change on the Southern Ocean and potential interactions.”
Reference: “Southern Ocean freshening stalls deep ocean CO2 release in a changing climate” by Léa Olivier, and F. Alexander Haumann, 17 October 2025, Nature Climate Change.
DOI: 10.1038/s41558-025-02446-3
Never miss a breakthrough: Join the SciTechDaily newsletter.
Follow us on Google and Google News.
11 Comments
e-8
thanks for this
All models miss something, which is soon found to be everything – because models are abstracts created by tunnel-vision.
This is THE flaw in the sciences. Trying to find a tool, the simpler the better. But all the little things come to bite – which turned out to be the big things – the knee-jerk ideas failed to conceive – because tunnel-vision on abstracts does not address the vast complexity of reality.
There are no simple aspects to the Universe. They are all more complex than you’ve been thinking.
Climate models are missing hell of a lot more, and more and more people are aware of that. But that still doesn’t stop climate fearmongers from keeping on selling their daily dose of doom.
Until all of Manhattan ends up under the sea, I’m not buying anything.
“Oceans collectively take up about one quarter of the CO2 produced by human activity.”
Is there evidence that the oceans take up ~1/4th of anthropogenic CO2, or should they have said, ~1/4th the atmospheric pool of total CO2 from both natural and human sources? I suspect that the latter is the correct description. It is an important distinction because that means if the 4% of the annual flux of CO2 that is attributed to humans were to suddenly disappear, the impact on the atmosphere and oceans would be minor. Just because the annual increase of atmospheric CO2 is about half of the estimated anthropogenic sources, it does not mean that humans are totally responsible for the increase. While some isotopic fractionation occurs with photosynthesis, by and large, the natural sinks can’t tell the difference between natural sources and anthropogenic sources.
“… the Southern Ocean alone accounts for roughly 40 percent, making it one of the planet’s most important regions for slowing global warming.”
The unstated — and therefore unexamined — assumption is that anthropogenic CO2 is solely responsible for the increase in global temperatures. A common refrain is, “What else could it be?” I would point out first of all, that is a logical fallacy. Just because one doesn’t know what is causing something doesn’t mean that it isn’t happening. Just because there is a correlation between two variables, causation is not established. What caused the end of the last major glaciation starting about 20,000-years ago? Surely, anthropogenic CO2 didn’t stop the effects of the Milankovitch Cycles! Studies have shown that cloud albedo has a significant impact on incoming solar radiation. The numerous relationships, largely negative feedback loops, make the dynamic system more complex than can be reduced to a single variable and dismissed with, “What else can it be?”
“…, the ocean releases naturally stored CO2 from its interior while simultaneously taking in human-generated CO2 from the atmosphere.”
The authors fail to point out that when the ocean takes “in human-generated CO2 from the atmosphere”, it is also taking in the more abundant natural CO2. As I stated previously, the natural sinks can’t tell the difference between ‘natural’ and “human-generated” CO2.
“However, that would transport more carbon-rich water from the deep ocean to the surface, which would consequently reduce the Southern Ocean’s ability to store CO₂.”
The atmospheric CO2 is considered to be “well-mixed.” That means there is very little variation throughout the globe. If the up-welling were to provide bottom waters whose CO2 partial pressure is higher than the atmosphere, then the CO2 will out-gas. However, that means any cold water nearby with a lower partial pressure will become a sink for the released CO2. Throughout this article, the authors only cite the processes that exacerbate the situation. The don’t look at the Big Picture or consider opposing processes. It is as though they have an agenda they are trying to support. That is what is known as “cherry picking” data.
“Deep waters rise to the surface, exchange gases with the atmosphere, and then sink again, carrying newly absorbed CO2 into the depths.”
The statement is really a non sequitur. It matters not what the ‘age’ of the CO2 is. The only thing that is important is the net change, if any. CO2 is CO2. It doesn’t have an expiration date and for purposes of IR absorption, there is little functional difference between ‘natural’ and ‘human-generated’ CO2.
“However, this situation COULD reverse if the stratification were to weaken,”
This is speculation. The word “could” only implies a low probability because it is presumed to be not impossible. However, nothing is said about the probability or margin of error, probably because they don’t have any numeric data to support the speculation. It is one thing to propose a demonstrated relationship between two variables might change if the independent variable changes. It is going a lot farther out on the speculative limb to suggest that a relationship that has NOT been observed will suddenly come into existence if a threshold is passed. One doesn’t know what that threshold is until it has been observed.
It has been said that mathematics is the language of science. Where are the numbers for this ‘science’ article? This is one of the least rigorous articles I have seen recently.
We all know how good modeling is, just look at the mess it produced during the Covid Pandemic.