A major ocean current system that helps regulate climate across the Northern Hemisphere is likely to weaken far more severely by the end of this century than scientists previously estimated, according to new research published in Science Advances.
The Atlantic Meridional Overturning Circulation (AMOC) is a vast system of currents that transports warm water from the tropics northward, releasing heat into the atmosphere before sinking and returning south.
“It’s basically a loop which transports heat from the equator to the North Atlantic,” lead author of the new study Dr Valentin Portmann, from the Inria Centre de recherche Bordeaux Sud-Ouest in France, told BBC Science Focus.
“This warm and salty water is transported northward. Then it releases its heat, becomes dense and sinks, before heading southward through deep ocean currents.”
According to the study, it’s projected to slow by 51 per cent by 2100. That figure is around 60 per cent higher than the average projection produced by standard climate models and comes with far less uncertainty.
The implications of such an AMOC weakening could be severe. Sea levels along the US northeast coast are already rising faster than the global average partly because of AMOC weakening.
Beyond the US, a weakening is projected to shift the tropical rain belt southward, threatening the monsoons that hundreds of millions of people in West Africa and South Asia depend on for agriculture.
In Europe, such changes are projected to bring colder, harsher winters as the conveyor belt of warm water to the continent slows.
More concerning still is that each additional weakening pushes the system closer to a tipping point – where a full collapse becomes increasingly likely, with consequences that could be catastrophic.
A critical but complicated system
Trying to project what will happen to the AMOC as our planet warms is a notoriously difficult task. The system is vast and complex, and influenced by both local and global factors.
Previous estimates of AMOC’s future vary widely depending on which climate prediction model is used. While most agree that the system is weakening, the extent ranges anywhere from minimal to total catastrophic collapse.
In the new study, the team identified two specific, systematic errors running through many of the best modelling efforts made so far: they simulate the South Atlantic as not salty enough and the North Atlantic as too cold.
Both biases cause the models to underestimate a key process, in which dense, salty water sinks and helps keep the entire current system flowing.
After correcting for both using a statistical technique called ridge-regularised linear regression – rarely used in climate science – the projected weakening rose to 51 per cent, while significantly decreasing the uncertainty associated with the result.
“Generally, in the literature, only one variable – one observation, for example, past AMOC strength – is used in the input,” Portmann said.
“The aim here was to use more and more information using multiple variables simultaneously, which is very important because the AMOC is very complex and relies on multiple processes.”
The current is already measurably weakening. Direct observations from an array of sensors across the Atlantic suggest its strength has dropped by around 10–20 per cent since the mid-2000s, equivalent to hundreds of millions of gallons of water per second no longer flowing north.
A 2025 study by US federal climate researchers found that the current's recent weak phase has already contributed to up to 50 per cent of flooding events along the northeast coast of the US since 2005.
Still, as things stand, it's not yet possible to attribute this decline to human-caused climate change rather than natural decade-to-decade variability. Scientists say we will have to wait until at least 2033 – giving us 29 years of measurements – before they can confidently separate the two.
Not a collapse – but that may not be reassuring
While the results of this latest study are troubling, researchers are eager to point out what it does and doesn’t show.
In its sixth assessment report, the Intergovernmental Panel on Climate Change (IPCC) said that while it was confident the AMOC will decline over the rest of this century, it had “medium confidence” that it will not collapse before 2100.
Such reassurances may offer little comfort, though, given the scale of changes that would result from a collapse, be it before or after the end of the century.
For example, a 2025 study in Geophysical Research Letters found that in such a scenario, cold extremes could reach -20°C (-4°F) in London and -48°C (-54°F) in Oslo. And that’s despite greenhouse gas-driven global warming.
Moreover, any weakening of the AMOC risks us crossing an unknown tipping point threshold. Studies have shown that the AMOC is what’s known as a ‘bistable’ system, meaning it likely has two stable ‘on’ or ‘off’ states which, once flipped, could take thousands of years to change again.
The location of the threshold between them is unknown. A 2025 study in Environmental Research Letters extended climate models beyond their usual 2100 cut-off and found AMOC shutdown occurred in 67 per cent of runs under high emissions and 30 per cent under medium emissions.
“We do not really know how much this threshold is or if this is really the case,” Portmann said. “We can only assume that this even stronger than predicted decline may be closer to a tipping point.”
The emissions window
Portmann's team tested four different emissions scenarios. Three of them – ranging from medium to very high – converged on a similar outcome of around 50 per cent weakening, suggesting that beyond a certain point much of the effects of human-caused climate change will be locked in.
“We have put a lot of heat into the ocean that is going to cool down over centuries,” Portmann said.
But the most optimistic scenario, representing aggressive and sustained emissions reductions, produced a weakening of only around 20 per cent.
“There are these two types of statements: one that says it is a little bit too late because we have already emitted a lot of CO2, and it is going to have a long-term impact,” Portmann said.
“But on the other hand, you can say that before the tipping point, we can avoid a severe weakening with a strong reduction of CO2.”
For now, Portmann is confident his study represents the best bet for where this critical ocean system is heading, though he acknowledges that there could be other key processes yet to be incorporated.
“This is why we have to be cautious about the results,” he said. “There is a huge uncertainty in climate models about what the fate of the future AMOC is. It’s very important we correct this.”
Read more:
