Collapse of the Atlantic Meridional Overturning Circulation Would Lead to Substantial Oceanic Carbon Release and Additional Global Warming
- Apr 15
- 6 min read
A 2026 study finds that AMOC collapse would trigger deep Southern Ocean convection, releasing 47–83 ppm of CO₂ and adding ~0.2°C of global warming.
The Atlantic Meridional Overturning Circulation (AMOC) — the vast system of ocean currents that redistributes heat across the globe — has long been recognized as one of the climate system's most consequential tipping elements. Most research into a potential AMOC collapse has focused on its dramatic regional temperature effects: severe cooling across the Northern Hemisphere and a modest warming in the Southern Hemisphere. But one critical and underexplored dimension of an AMOC collapse has received far less attention: what happens to the ocean's carbon.
Nian et al. (2026) fill this gap directly in their study, using a computationally efficient Earth system model to systematically quantify how an AMOC collapse would affect the global carbon cycle across a range of background warming levels — and find that the carbon consequences are substantial and largely irreversible.
Key Findings
Experimental Design: Isolating the Carbon Cycle Response
The study uses CLIMBER-X, a fast Earth system model that includes a 3D ocean, statistical-dynamical atmosphere, sea ice, ocean biogeochemistry, and a land surface model with dynamic vegetation — enabling full simulation of the global carbon cycle and atmospheric CO₂ evolution.
To systematically isolate the roles of different components, the authors design three model configurations: a fully coupled version (interactive ocean and land carbon cycles), a land-carbon-off version (ocean carbon cycle only), and a fixed-CO₂ version (pure physical climate response, no carbon cycle feedbacks).
Freshwater hosing is applied to the North Atlantic between 50°N and 70°N to force AMOC collapse, with experiments initiated from equilibrium states at CO₂ concentrations ranging from pre-industrial (280 ppm) to more than double pre-industrial (600 ppm). Each baseline is preceded by a 10,000-year spin-up to ensure true equilibrium, and simulations run for 7,000 years after hosing ends to allow the system to re-equilibrate.
At CO₂ levels of 280 ppm, the AMOC shows monostable behavior — it collapses under hosing but recovers once the forcing is removed. At 350 ppm and above, the AMOC becomes bistable: once it collapses, it cannot recover and remains permanently shut down.
The Carbon Cycle Response: Large and CO₂-Dependent
In the fully coupled model, an irreversible AMOC collapse increases atmospheric CO₂ by 47–83 ppm (equivalent to 100–175 GtC of additional atmospheric carbon) across experiments with baseline CO₂ levels from 350 to 600 ppm. Crucially, the magnitude of this CO₂ increase grows with higher baseline CO₂ levels.
When the land carbon cycle is disabled (land-carbon-off setup), the CO₂ increase is even larger: 72–130 ppm (153–275 GtC). This reveals that terrestrial carbon sinks act as a partial buffer against the oceanic carbon release, absorbing a fraction of what the ocean releases — but cannot neutralize it.
Across both setups, the additional atmospheric CO₂ from AMOC collapse scales consistently at roughly 20–22% of the baseline CO₂ level in the land-carbon-off setup, and around 13% in the fully coupled version, underscoring the compensating role of land carbon uptake.
These numbers far exceed what has been inferred from glacial paleoclimate records, where AMOC-related CO₂ changes were on the order of ~15 ppm. The study explains why: the ocean's carbon inventory is substantially larger at higher CO₂ background levels, meaning more carbon is available to be released when convective ventilation occurs.
The Mechanism: Southern Ocean Deep Convection
The central physical mechanism driving the oceanic carbon release is the triggering of deep convection in the Southern Ocean — a direct and robust response to AMOC shutdown in the model.
As the AMOC weakens and North Atlantic deep water formation declines, the Southern Ocean destratifies and eventually develops widespread deep convective overturning, particularly in the 70°S–80°S band around Antarctica. This "bipolar convection seesaw" ventilates carbon-rich deep waters that have been sequestered for millennia, releasing them to the atmosphere in a large and rapid burst of CO₂ outgassing.
Support for this mechanism comes from paleoclimate proxy records: during Heinrich events (when large North Atlantic freshwater discharges from iceberg calving collapsed the AMOC in glacial times), Antarctic ice core records show abrupt increases in both temperature and atmospheric CO₂ in the centuries following the collapse — consistent with a sudden onset of Southern Ocean deep convection. CLIMBER-X has been independently validated to reproduce this response.
In warmer background climates, the ocean's larger carbon inventory means that the same convective mechanism produces a proportionally much larger CO₂ release than observed during glacial periods — making past glacial CO₂ changes an inadequate analog for future AMOC-collapse scenarios.
Temperature Response: Global Warming Despite Physical Cooling
In the fixed-CO₂ setup — where carbon cycle feedbacks are disabled — AMOC collapse produces a net global cooling of 0.2–0.3°C, confirming that the physical ocean dynamics alone (redistribution of heat, sea ice expansion, albedo feedbacks) have a cooling influence on global mean temperature.
When the ocean carbon cycle is activated (land-carbon-off setup), the resulting CO₂ increase drives a net global warming of 0.41–0.51°C — overcoming the physical cooling entirely and producing a net warming effect.
In the fully coupled version, terrestrial carbon uptake partially offsets the oceanic carbon release, reducing the net additional global warming to approximately 0.17–0.27°C above the pre-collapse baseline. This is roughly 0.2°C of additional warming on top of whatever greenhouse warming was already present.
The relative contribution of AMOC collapse to global temperature change decreases at higher CO₂ levels: at 350 ppm baseline, the AMOC collapse contributes approximately 19% of the total warming above pre-industrial levels; at 600 ppm, this drops to around 8%. This means AMOC collapse is proportionally more consequential and more dangerous at lower warming levels.
Pronounced Regional Temperature Asymmetry
While the global mean warming effect is modest (~0.2°C), the regional temperature consequences are dramatic and strongly asymmetric between hemispheres.
Arctic regions (60°N–90°N) cool by approximately 7°C following AMOC collapse, driven by the loss of northward heat transport, amplified by sea-ice expansion and associated albedo feedbacks. North Atlantic cooling is the most intense of all regions.
Antarctic regions (60°S–90°S) warm by approximately 6°C, driven by the deep Southern Ocean convection that ventilates warm deep waters and simultaneously releases CO₂. This Southern Hemisphere warming is substantially stronger than what has been found in previous pre-industrial AMOC collapse studies.
When the 450 ppm background greenhouse warming is superimposed on the AMOC collapse response, the combined warming in some regions south of 60°S can exceed 10°C relative to pre-industrial conditions — an exceptional regional climate signal.
Because the model does not include dynamic ice sheets, these temperature responses do not account for additional feedbacks from Greenland or Antarctic ice sheet melting, suggesting the regional estimates may be conservative.
The Carbon Feedback Nobody Was Accounting For
The broader significance of this study lies in what it reveals about a largely overlooked feedback loop in climate risk assessments. AMOC collapse has traditionally been framed as a regional cooling threat — severe for Europe and the North Atlantic, but partially offset globally. This study demonstrates that the picture is fundamentally more complex:
A CO₂ amplifier hidden in the deep ocean: The deep ocean stores vast quantities of carbon accumulated over centuries. An AMOC collapse that triggers Southern Ocean deep convection would ventilate this reservoir rapidly, releasing carbon that no mitigation policy can intercept. The ocean becomes, in effect, an uncontrolled carbon emitter.
Scale with background warming: The higher the CO₂ concentration at which an AMOC collapse occurs, the larger the subsequent carbon release. This creates a deeply concerning feedback: higher emissions make an AMOC collapse more likely, and a collapse at higher CO₂ levels releases more carbon, pushing concentrations higher still.
Terrestrial sinks are not a complete safeguard: Land carbon uptake absorbs part of the ocean's carbon release, reducing additional warming from ~0.45°C to ~0.2°C. But terrestrial carbon sinks are themselves vulnerable — already weakened in many regions by land-use change and climate-driven stress — meaning the compensating effect may be smaller in the real world than in these simulations.
Policy implications for carbon budgets: The 47–83 ppm CO₂ increase projected from an AMOC collapse at near-future CO₂ levels represents a major unaccounted-for addition to global carbon budgets. If AMOC collapse were to occur, it would consume a substantial fraction of any remaining carbon budget consistent with Paris Agreement temperature targets.
Between Two Poles
The study shows that the collapse of the AMOC would not simply cool Europe and disrupt monsoons — it would quietly unlock one of the ocean's largest carbon stores and warm the planet from below, through a mechanism playing out thousands of meters beneath the Southern Ocean's surface. The Arctic would freeze; Antarctica would warm. And a world already struggling to hold warming below 1.5°C would face an additional, self-reinforcing push in the wrong direction — one that no emissions reduction program could quickly reverse.
Whether or not an AMOC collapse occurs this century, this study makes clear that the carbon consequences of such a collapse must be included in climate risk frameworks, tipping point analyses, and the construction of global carbon budgets. The ocean's carbon is not safely locked away — it is, under the right (or wrong) conditions, waiting to be released.
Sign up for our newsletter or connect with us on social media to stay up-to-date with our latest posts and permaculture inspiration.
Explore our inspiring series and posts:
Love the post? Share it with your circle, inspire your people.