First, what actually collapses
The AMOC is the ocean's large-scale conveyor that carries warm surface water northward in the Atlantic and returns cold, deep water southward. It is treated as a "tipping element" — a part of the climate system that can, past a threshold, flip into a different state rather than drifting smoothly [4]. In model experiments, adding enough freshwater to the North Atlantic (for instance from melting ice) weakens the circulation until it shuts down, and it then stays shut down until the forcing eases [1].
The mechanism behind that flip is the salt-advection feedback: the circulation itself carries salty water into the Atlantic, and salt makes water dense enough to sink, which keeps the circulation going. When the overturning stops importing that salt, the Atlantic freshens further, which weakens the sinking further — a self-reinforcing loop [4]. One study finds this feedback only switches on when the AMOC is carrying net salt into the Atlantic, and that many current climate models are biased toward an AMOC that is too stable, which would mean they understate the risk of collapse [4].
The tropical Pacific would calm down, not boil over
The most concrete modeled consequence concerns El Niño–Southern Oscillation (ENSO), the year-to-year pattern of warm and cool tropical Pacific water that drives droughts and floods worldwide. In a climate model forced into an AMOC collapse, the tropical Pacific mean state cools and the Walker circulation — the east–west atmospheric loop over the Pacific — strengthens. That weakens the air–sea coupling that lets El Niño grow, so ENSO variability drops by about 30% [2].
The character of El Niño would change too, not just its size. Extreme El Niño events become roughly 95% less frequent, and the pattern shifts so that central Pacific El Niños become more common than eastern Pacific ones, with less of the lopsided asymmetry that normally distinguishes strong warm events from strong cool ones [2]. The authors present this as a mechanism for how past and future ENSO might behave under a much weakened circulation, not as a forecast of a specific date [2].
Carbon uptake and regional winners and losers
The collapse would also touch the carbon cycle. One Earth System Model study examining a strong AMOC weakening under both low- and high-emission scenarios reports that the response of atmospheric CO2 depends on the emission pathway, and that some of the local effects of a collapse could be beneficial in specific places [3]. That is a reminder that "effects" are not uniformly bad everywhere — but the same study shows the outcome is not a fixed number; it shifts with the background climate and emissions trajectory [3].
Beyond the physics, the collapse is treated as a security and governance problem, not just a scientific one. A policy analysis notes that observations of the AMOC only began in 2004, so projections rely on proxies, models, and simulations that produce widely varying results and high uncertainty [5]. Even so, in late 2024 dozens of climate scientists warned publicly that the risk and its consequences should not be underestimated, and by 2026 the Nordic Council of Ministers had adopted a precautionary action plan while acknowledging that whether the AMOC will cross a threshold remains highly uncertain [5].
How much should you trust these numbers?
The 30% and 95% figures come from a single global climate model experiment in which freshwater was added to the North Atlantic to force a collapse [2]. They describe what that model did, and the authors frame them as mechanistic insight into possible past and future behavior rather than a prediction [2]. The broader literature is explicit that model and simulation results vary substantially even when scientists use the same observations, and that measurement coverage is too sparse and too recent for computing power to close the gap [5].
So the honest answer has two layers. The direction of some effects — a calmer, more central-Pacific ENSO, a reorganized carbon cycle, regionally uneven consequences — is supported by model evidence [2][3]. The magnitude, timing, and whether a collapse happens this century are not settled, and the finding that many models may make the AMOC look too stable means the risk could be larger than standard projections suggest [4][5].
Sources used in this answer
Effects of the Bering Strait closure on AMOC and global climate under different background climates
In model experiments, strong enough buoyancy forcing causes the AMOC to collapse and remain collapsed until the forcing weakens again.
Reduced ENSO Variability due to a Collapsed Atlantic Meridional Overturning Circulation
In a climate model forced by North Atlantic freshwater input, an AMOC collapse cools the tropical Pacific and strengthens the Walker circulation, reducing ENSO variability by about 30%, cutting extreme El Niño events by roughly 95%, and shifting El Niño toward the central Pacific.
Response of atmospheric pCO to a strong AMOC weakening under low and high emission scenarios: AA Boot et al.
The atmospheric CO2 response to a strong AMOC weakening depends on whether emissions are low or high, and a collapse can have local effects that are beneficial in some places.
Feedback Processes Causing an AMOC Collapse in the Community Earth System Model
The destabilizing salt-advection feedback dominates AMOC tipping and activates when the circulation carries net salt into the Atlantic; many models' overly stable AMOC bias may understate collapse risk.
The Nordic Council and the Atlantic Meridional Overturning Circulation (AMOC)
AMOC projections carry high uncertainty because direct observations only began in 2004, yet the Nordic Council of Ministers treats the collapse risk as real and significant and has proposed a precautionary action plan.