A team of researchers at the Alfred Wegener Institute has overturned a decades-old explanation for how the Antarctic Circumpolar Current, Earth’s most powerful ocean current, first formed. Their coupled climate simulations show that the opening of Southern Ocean gateways alone could not have launched the current. Instead, wind patterns over the Tasman Gateway and the specific arrangement of continents approximately 33.5 million years ago were essential co-triggers, a finding that reshapes scientific understanding of how Antarctica became the frozen continent it is today.
What is verified so far
The core claim rests on a peer-reviewed paper published in PNAS that used coupled climate simulations to reconstruct the geographic setting of the Southern Ocean roughly 33.5 million years ago, during the Eocene-Oligocene transition. The simulations tested whether simply widening the Drake Passage and the Tasman Gateway between Australia and Antarctica was enough to spin up a circumpolar flow. The answer was no. The models showed that winds funneling through the Tasman Gateway, combined with the precise positions of landmasses at that time, were required before a continuous eastward current could take hold around Antarctica.
That result directly challenges a long-standing narrative built on a frequently cited 2005 study that placed the current’s establishment in the late Oligocene and attributed it primarily to gateway expansion. For two decades, that earlier work served as a baseline for paleoceanographic reconstructions. The new PNAS study does not simply refine the older timeline; it argues that the mechanism itself was different, shifting the explanatory weight from plate tectonics alone to a combination of tectonics, atmospheric circulation, and continental geometry.
A separate line of evidence, drawn from neodymium isotope and sortable-silt flow-speed proxy records spanning approximately 31 million years, supports a related but distinct conclusion. That analysis, published in Nature Geoscience, argues that a modern-like, deep-reaching, homogeneous version of the current did not exist before late Miocene cooling, roughly 10 million years ago. Scientists Vasileios Evangelinos and Isabel Cacho, whose work was summarized by the University of Barcelona, framed the current as a consequence rather than a cause of Antarctic glaciation, flipping a cause-and-effect assumption that had shaped climate models for years.
The Nature Geoscience article is accessible through the broader Nature journal index, and readers with institutional access can reach the full text via the publisher’s authentication portal. For ongoing developments, the journal’s RSS feed offers updates on new paleoceanography and climate papers that may further refine the timing and structure of the Antarctic Circumpolar Current.
What remains uncertain
The biggest open question is when the current reached its modern strength. The PNAS simulations model conditions at approximately 33.5 million years ago and demonstrate that a circumpolar flow could have begun forming at that point given the right wind and continental setup. But the Nature Geoscience proxy data, according to Evangelinos and colleagues, suggests that a deep, uniform current matching today’s profile did not appear until the late Miocene, tens of millions of years later. These two findings are not necessarily contradictory, since a shallow or partial current could have existed long before it deepened into its present form, but neither study fully resolves the gap.
A second area of uncertainty involves the sediment-core proxies themselves. The neodymium isotope and sortable-silt records cover an impressive span, but the institutional summaries available do not detail the full primary datasets, spatial coverage, or error margins. Without access to the granular proxy data, it is difficult to assess how sensitive the late-Miocene timing claim is to sampling location or analytical method. The older 2005 study used different proxy-based constraints to reach its late-Oligocene estimate, and no public response from its original authors to the newer challenges has surfaced in available reporting.
There is also a gap in understanding how these ancient findings translate to the modern Southern Ocean. The PNAS paper and the Nature Geoscience analysis both imply that wind patterns are central to the current’s behavior, which raises questions about what happens as climate change shifts the position and intensity of Southern Hemisphere westerly winds. But neither study, based on available summaries, offers quantified projections for how ongoing wind changes might weaken or redirect the current in coming decades.
How to read the evidence
Three tiers of evidence feed this story, and readers should weigh them accordingly. The strongest is the PNAS simulation work, which provides a controlled, reproducible test of the gateway-only hypothesis and finds it insufficient. Climate models are not perfect reconstructions of the past, but coupled simulations allow researchers to isolate variables, in this case testing what happens when gateways open but winds or continental positions are held constant. That kind of experiment cannot be run with sediment cores alone.
The second tier is the Nature Geoscience proxy analysis, which offers direct physical evidence from the ocean floor. Proxy records carry their own uncertainties, including potential diagenetic alteration of isotope signals and uneven spatial sampling, but they provide a check on model outputs that pure simulations cannot. The tension between the two studies is productive rather than damaging: the simulations explain the mechanism, while the proxies constrain the timeline.
The third tier consists of institutional press summaries from research centers such as the Alfred Wegener Institute and the University of Barcelona. These are useful for accessible interpretation and for attributing specific claims to named researchers, but they compress technical detail and omit caveats that appear in the full papers. Readers looking for the strongest grounding should trace claims back to the journal publications rather than relying on summary language alone.
One assumption that deserves scrutiny is the widespread framing of the Antarctic Circumpolar Current as a simple on-off switch for Antarctic glaciation. Both the PNAS and Nature Geoscience teams push back against this idea, but much popular science writing still treats the current’s formation as the single event that isolated Antarctica and triggered ice-sheet growth. The newer work instead suggests a more gradual, feedback-rich evolution: partial currents may have started to encircle the continent as gateways widened and winds intensified, while ice sheets waxed and waned in response to global carbon dioxide levels, orbital cycles, and regional ocean changes.
Under this view, the Antarctic Circumpolar Current is not a solitary trigger but part of a coupled climate system. Winds drive the surface flow, the current reshapes global heat distribution, and the presence or absence of large ice sheets feeds back on both winds and ocean structure. Recognizing this complexity matters for present-day climate questions. If the current’s strength and depth have historically been sensitive to shifts in westerly winds and continental geometry, then rapid modern changes in atmospheric circulation could have outsized effects on how efficiently the Southern Ocean absorbs heat and carbon.
For readers, the takeaway is not that scientists are confused about when the current formed, but that they are progressively narrowing the range of plausible histories. The gateway-only hypothesis, once dominant, now appears incomplete in light of coupled modeling and expanded proxy records. At the same time, the exact timing of a fully modern, deep-reaching current remains under active investigation, with different methods emphasizing different aspects of the system. Following the primary literature through stable identifiers, such as the Nature Geoscience study and the PNAS simulations, offers the clearest window into how this scientific debate is evolving.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
