The Atlantic carries heat towards our shores. What happens if that circulation loses strength—and how close are we to finding out?
Britain’s weather can feel chaotic enough without the ocean changing the rules beneath it. Yet one of the systems helping to shape our climate is expected to weaken as the planet warms.
Its name is unwieldy: the Atlantic Meridional Overturning Circulation, or AMOC. Its consequences are easier to grasp. Less heat carried north. Changes in rainfall. Pressure on ecosystems and food production. Under a much more severe collapse scenario, substantial cooling around the North Atlantic.
Could global warming leave Britain colder? Yes, under some scenarios. But the difference between gradual weakening and collapse is the difference between a useful explanation and a misleading headline.
An ocean moving heat
The AMOC is a large circulation system: warmer upper-ocean water moves northwards, while colder deep water returns southwards. Cooling and changes in salinity help water become dense enough to sink in northern regions. Mixing and winds elsewhere—including over the Southern Ocean—also help sustain the system. The familiar conveyor-belt image is a simplification, not a single pipe running through the sea.
It is also not interchangeable with the Gulf Stream. That current is part of the Atlantic’s upper-ocean circulation and has a major wind-driven component. An AMOC collapse would not mean every Atlantic current simply stopped. Nor does the ocean alone explain Britain’s relatively mild climate: atmospheric circulation matters too. Met Office background.
Why warming can weaken it
Warmer surface water is less dense. Freshening from rainfall, runoff and melting ice can also reduce density. These changes can make northern waters less able to sink, weakening the overturning. That does not make Greenland melt the system’s only control; the whole ocean and atmosphere are involved.
The IPCC’s 2021 assessment judged a decline during this century very likely across the emissions scenarios assessed. It expressed medium confidence that abrupt collapse would not occur before 2100. Those are different statements: confidence in weakening is stronger than confidence about the absence of collapse. Neither supplies an expiry date. IPCC assessment, section C.3.
Weakening is not the same as Britain freezing
Under gradual weakening, reduced ocean heat transport can offset some greenhouse warming around western Europe. The Met Office still describes the overall UK trend in that situation as warming. “Cooler than it would otherwise have been” does not necessarily mean “colder than today”.
A collapse would be a much larger disruption. It could produce strong North Atlantic cooling and major changes in rainfall, winds and storm risks. Agriculture would face changes in growing conditions and water availability, not merely a different number on the thermometer. Effects would extend well beyond Europe, including shifts in tropical rainfall and monsoons.
Ocean circulation also affects regional sea level. That is a further reason to take weakening seriously, but it does not give us a single defensible AMOC-only flood height for the UK coastline. Local outcomes depend on other ocean changes, land movement, tides and coastal defences.
Our Thwaites Glacier article examines another part of the coastal-risk story. Ice loss and circulation change are distinct processes whose consequences can overlap.
Why the research does not speak with one voice
A 2025 Met Office-led modelling study found a source of resilience: Southern Ocean winds helped maintain overturning even under extreme forcing. It supported the view that complete collapse this century is unlikely, while still projecting weakening. Read the researchers’ explanation.
Other research raises sharper concerns. An April 2026 paper used observational constraints to project roughly 50% weakening by this century’s end. That is a constrained model projection, not an observation that half the circulation has already disappeared. Science Advances study.
An August 2026 study investigated whether the rate of climate forcing matters as well as the eventual warming level. Faster carbon-dioxide increases produced collapse at lower warming levels than a slow increase in its experiments. Crucially, the main simulations began from a freshwater-adjusted state selected to investigate a vulnerable circulation. The authors discuss model biases and the implications of that setup. These are experiments exploring mechanisms, not a universal temperature threshold or a dated forecast. Nature Climate Change study.
The disagreement is a reason to keep testing assumptions. Model structure, initial conditions, freshwater treatment and the way observations constrain projections all affect the answers. We should neither select only the most frightening result nor turn a more reassuring one into a guarantee.
Watch: the circulation that helps keep Europe warm
The Met Office’s July 2025 explainer introduces the circulation and its climate effects. It is background material and predates the 2026 studies discussed above.
No countdown, and no comfort in uncertainty
Continuous observations across the subtropical Atlantic began in 2004. That record reveals substantial variability, but remains short beside the timescales needed to separate some long-term changes from natural fluctuations. Scientists also use other observing systems, indirect evidence and models; none removes every uncertainty.
The defensible concern is not that Britain will wake beneath an ice sheet tomorrow. It is that we are changing a climate system whose regional responses may be disruptive, uneven and difficult to reverse.
Reducing emissions limits the pressure we put on that system. Monitoring it improves our chance of recognising change. Planning for severe outcomes acknowledges their consequences without pretending to know their arrival date.
The Atlantic does not need to stop overnight to change the conditions under which we live.