What is the AMOC?
An ocean-wide conveyor belt of warm and cold water — the circulatory system of the Atlantic.
The Atlantic Meridional Overturning Circulation is a massive system of ocean currents that acts as the Atlantic's primary heat redistribution engine. It transports warm, salty surface water northward from the tropics toward the high latitudes of the North Atlantic — including the seas around Greenland, Iceland, and Norway — where it releases heat into the atmosphere, warming Europe and regulating the climate of the entire Northern Hemisphere.[1]
Once cooled, this dense, salty water sinks into the deep ocean and returns southward as cold deep water — a journey that can take more than a thousand years. The cycle connects the ocean surface with its deepest layers, distributing heat, oxygen, carbon, and nutrients across the entire Atlantic basin.[1]
The AMOC accounts for approximately 25% of global ocean heat transport and is a primary driver of the North Atlantic's role as the largest single carbon sink in the Northern Hemisphere. When it weakens, the ocean's ability to sequester anthropogenic CO₂ is directly compromised.[1]
What the Data Shows
A tug of war between natural variability and anthropogenic forcing — and scientists watching closely for who wins.
Post-2004 Weakening, Then Pause
Direct RAPID mooring observations at 26°N show measurable AMOC weakening after 2004. A 2025 NOAA/AOML study found that while anthropogenic forcing drove significant weakening through the 2000s, a positive North Atlantic Oscillation (NAO) phase since the early 2010s has largely cancelled out that decline — creating a temporary stall.
Long-Term Slowdown Since 1900
A December 2025 UC Riverside study used over 100 years of ocean temperature and salinity data to show that only models with a weakening AMOC could reproduce the persistent cold anomaly south of Greenland — the North Atlantic Warming Hole. The AMOC has slowed at a rate of approximately −1.01 to −2.97 Sv per century since 1900.
Measurements May Underestimate Strength
New research using density-space diagnostics (ρ-AMOC) rather than traditional depth-space methods reveals the subpolar overturning is considerably stronger than previously measured — around 21 Sv vs. 16 Sv in pre-industrial conditions. This framework better captures water mass transformations in warming climates.
Deep Limb Weakening 12% Since 2000
Separate from the upper circulation, the abyssal limb of the AMOC — which carries Antarctic Bottom Water northward — weakened by approximately 12% between 2000 and 2020, contributing to regional sea-level rise along the eastern seaboard of North America.
Equatorial Warming as Early Warning
A November 2025 study from the Chinese Academy of Sciences identified mid-depth warming (1,000–2,000 m) in the equatorial Atlantic as a distinct and detectable fingerprint of AMOC slowdown. This signal has already emerged from natural variability since the early 2000s, offering a powerful new monitoring tool.
CMIP6 Models Underestimate Variability
Multi-model analyses show that most CMIP6 ensembles fail to accurately reproduce 20th-century AMOC variability. High-quality Earth system models tend to show greater stability than lower-complexity statistical models — a tension at the heart of current projections that drives significant uncertainty in collapse timing estimates.
What the Models Project
Across every scenario, the AMOC weakens. The question is how much, and whether a tipping point is crossed.
Under all emission scenarios, AMOC continues to weaken relative to pre-industrial strength. The NAO tug-of-war may mask anthropogenic weakening in observations. Deep convection in the Labrador, Irminger, and Nordic Seas begins showing downward trends — the key precursor to collapse.[9]
A statistical projection by Ditlevsen & Ditlevsen (updated August 2025) suggests AMOC could approach a tipping point around 2065 based on sea surface temperature fingerprints. Important caveat: most Earth system models do not replicate this timeline, and many climate scientists are skeptical. This represents the "early collapse" end of the uncertainty range.[1]
Analysis of 25 CMIP models identified potential collapse onset indicators — not a full shutdown — by 2063 under SSP2-4.5 and 2055 under SSP5-8.5. These represent internal warning signals, not the moment of collapse itself.[1]
A landmark August 2025 PIK/Potsdam study extended 9 CMIP6 models to 2300–2500. Under all high-emission scenarios, and some intermediate ones, the AMOC fully shuts down after 2100 following a collapse of deep winter convection in subpolar basins. The full shutdown occurs 50–100 years after the tipping point is crossed — and the models don't include Greenland meltwater, which would accelerate the timeline.[9]
"A drastic weakening and shutdown of this ocean current system would have severe consequences worldwide. It would greatly reduce the risk of an AMOC shutdown if we cut emissions fast — even though it is too late to eliminate the risk completely." — Stefan Rahmstorf, Potsdam Institute for Climate Impact Research, 2025
Micro Effects on Ocean Life
Shifting Migration Routes
Following food into unfamiliar waters
The AMOC drives upwelling that concentrates prey species — copepods, krill, small fish — in predictable zones that great whale species have followed for millennia. As the current weakens and sea surface temperatures shift, prey aggregations are moving northward and into deeper water.
North Atlantic right whales, humpbacks, and fin whales have been documented in new feeding areas across the Gulf of St. Lawrence, the Labrador Sea, and increasingly the Arctic, as traditional grounds like Cape Cod Bay and Georges Bank become less productive. This redistribution brings whales into busy shipping lanes and fishing grounds, dramatically increasing entanglement and vessel strike risk.[a]
For right whales, already critically endangered with fewer than 370 individuals, the collision of shifting habitat and human infrastructure may be the final pressure the species cannot absorb.
The Silencing of Blue Whales
Song pitch falling, feeding calls fading
Blue whales sing — low, powerful, species-specific songs that carry thousands of kilometres through the ocean. These songs serve both reproductive and navigational purposes. Over the past five decades, recordings across every ocean basin show that blue whale calls have been declining in frequency — becoming measurably lower in pitch — at rates of approximately 0.3 Hz per year.[b]
The leading hypothesis links this partly to ocean noise pollution, but also to changing prey distributions driven by AMOC-related temperature and salinity shifts. Blue whales modify their foraging calls and feeding dive patterns in direct response to krill patch density and depth. As krill decline and shift depth in warming, freshening waters, the acoustic landscape of blue whale feeding grounds is changing.
In the North Atlantic in particular, researchers at NOAA and MBARI have noted a reduction in blue whale acoustic presence in historically rich feeding grounds — a behavioural withdrawal consistent with prey redistribution driven by oceanographic change.[b]
The Collapse of Capelin
A keystone species under thermal pressure
Capelin (Mallotus villosus) is one of the most ecologically critical fish in the North Atlantic — a small, schooling forage fish that is the primary prey of Atlantic cod, humpback whales, puffins, gannets, and dozens of other species. Capelin are highly temperature-sensitive, spawning on cold, pebbly beaches and feeding on the cold-water copepods concentrated by AMOC-driven upwelling.
As AMOC weakens and the Arctic warms asymmetrically, capelin stocks in the Barents Sea, Grand Banks, and around Iceland have shown dramatic volatility. The Barents Sea stock effectively collapsed in the late 1980s, partially recovered, and has since entered a new period of instability. Around Iceland, the capelin spawning migration has shifted northward by more than 100 km in 30 years.[c]
When capelin disappear, the entire ecosystem collapses upward: cod starve, seabirds fail to breed, and whales abandon traditional feeding grounds. The capelin is not a victim of overfishing alone — it is a sentinel for the oceanographic disruption that AMOC weakening drives at the base of the food web.
Krill: The Engine Room of the Ocean
When the bottom of the food web falters, everything above it trembles
Antarctic and North Atlantic krill species are among the most abundant animals on Earth by biomass, and they sit at the base of virtually every marine food web that supports large vertebrates. They are filter feeders, dependent on phytoplankton blooms that are themselves sustained by AMOC-driven nutrient upwelling. As AMOC weakens, reduced upwelling means reduced phytoplankton, means reduced krill.[d]
In the North Atlantic, a 2023 study documented a significant decline in Meganyctiphanes norvegica (Northern krill) abundance in areas of the subpolar gyre where AMOC-related circulation changes have reduced the supply of cold, nutrient-rich water to surface layers. Krill biomass in historically rich areas has declined by an estimated 20–40% since the 1990s in some regions.[d]
Downstream cascade effects include: reduced prey for blue, fin, and minke whales; collapsing seabird breeding success (puffins, razorbills, guillemots); reduced food for Atlantic bluefin tuna and other highly migratory species; and potentially reduced carbon export to the deep ocean, as krill faecal pellets are one of the ocean's most efficient carbon-sinking mechanisms.
The krill decline creates a feedback loop: less krill means less whale defecation in surface waters, which means less iron fertilisation of phytoplankton, which means less krill — an unravelling spiral that AMOC weakening sets in motion.[d]
There Is Still Time.
The AMOC is not yet at its tipping point. Every fraction of a degree of warming avoided is a probability of collapse reduced. Here is what the science says humans can actually do — and what gives researchers genuine cause for hope.
Rapid Emission Reductions
The single most impactful intervention. The PIK 2025 study found that cutting emissions fast would "greatly reduce" collapse probability, even though it cannot be eliminated entirely. Under aggressive decarbonisation, Earth system models show the AMOC weakening but not collapsing through 2300. The difference between 1.5°C and 4°C of warming is likely the difference between a weakened AMOC and a shutdown one.
Protect Greenland Ice Sheet
Greenland meltwater is the greatest near-term threat to AMOC stability — freshwater input reduces salinity, which reduces the density of surface water, which prevents the sinking that drives the circulation. Protecting the ice sheet means protecting sea ice extent, which means aggressive methane reduction alongside CO₂ cuts. Local policies matter: Arctic shipping routes and resource extraction accelerate regional warming.
Restore Marine Ecosystems
Healthy whale populations, kelp forests, and seagrass meadows are not just biodiversity goals — they are carbon infrastructure. Whale nutrient pumping, seagrass carbon sequestration, and phytoplankton productivity collectively help maintain the ocean chemistry that supports AMOC stability. Marine protected areas (MPAs) and fishing reform can restore these systems within decades.
Fund Sustained Observation
The RAPID array, OSNAP, and new observational systems are the only way to know what is actually happening. The 2025 NOAA finding — that AMOC weakening has stalled due to NAO dynamics — was only possible because of two decades of continuous in-situ monitoring. Governments defunding ocean observation programmes are flying blind. Sustained funding for ocean monitoring buys critical warning time.
End Deforestation & Restore Peatlands
Amazon and boreal forest deforestation alters the atmospheric moisture transport that sustains Atlantic salinity — a mechanism that directly feeds AMOC strength. Peatland restoration sequesters carbon and moderates freshwater runoff into the North Atlantic. These are land-use policies with direct ocean circulation consequences, and they are actionable at a national and sub-national level today.
Redesign Food Systems
Agriculture is responsible for approximately 30% of global greenhouse gas emissions — and industrial fishing directly destabilises the marine food webs that are early warning systems for AMOC health. A plant-rich diet, reduction of methane-producing livestock at scale, and reform of industrial fishing subsidies would reduce both atmospheric forcing and ecosystem stress simultaneously. This is one of the highest-leverage individual and policy interventions available.
"We found that the AMOC is more resilient than some projections suggest — but that resilience has limits. The window for avoiding the worst outcomes is open. But it will not remain open indefinitely." — Paraphrase of NOAA/AOML findings, January 2025
Sources & Citations
All findings cited reflect peer-reviewed research or direct institutional publications. Where possible, the most recent available literature (2024–2025) has been prioritised.
Atlantic Meridional Overturning Circulation — Wikipedia synthesis
Comprehensive synthesis of CMIP6 findings, Ditlevsen & Ditlevsen projection (updated August 2025), and multi-model analysis extending simulations to 2300–2500.
en.wikipedia.org/wiki/Atlantic_meridional_overturning_circulation
NOAA/AOML — Advancing our understanding of the AMOC
January 2025. Found AMOC weakening paused since early 2010s due to positive NAO tug-of-war with anthropogenic forcing.
aoml.noaa.gov
van Westen et al. — AMOC collapse in a strongly-eddying ocean model
Geophysical Research Letters, March 2025. First AMOC collapse demonstrated in an explicitly eddy-resolving ocean model; Southern Ocean upwelling role confirmed.
agupubs.onlinelibrary.wiley.com
Oliveira Matos et al. — Diagnosing AMOC in density space
Ocean Science, 2025. Demonstrates ρ-AMOC (~21 Sv) substantially stronger than z-AMOC (~16 Sv) in pre-industrial climate using AWI-CM3; recommends density-space diagnostics as standard.
os.copernicus.org
Ren et al. — Equatorial Atlantic mid-depth warming as AMOC fingerprint
Communications Earth & Environment, November 2025. Identifies 1,000–2,000m equatorial Atlantic warming as a detectable, dynamically-driven AMOC slowdown indicator already emerged from natural variability.
phys.org summary
Li & Liu — North Atlantic Warming Hole linked to historical AMOC slowdown
Communications Earth & Environment, December 2025 / UC Riverside. Only weakened-AMOC models reproduce observed century-long temperature and salinity patterns south of Greenland. Slowdown rate: −1.01 to −2.97 Sv/century.
nature.com
NOAA Climate.gov — AMOC abyssal limb weakening
Antarctic Bottom Water transport at 16°N weakened by ~12% during 2000–2020, contributing to North Atlantic sea-level rise.
climate.gov
Liu et al. — Orbital forcing and AMOC collapse (paleoclimate context)
Geophysical Research Letters, September 2025. Demonstrates precessional forcing alone can induce complete AMOC collapse under high eccentricity / low CO₂ — paleoclimate context for current risks.
agupubs.onlinelibrary.wiley.com
Drijfhout et al. — Shutdown of northern Atlantic overturning after 2100
Environmental Research Letters, August 2025 (PIK/Potsdam). All 9 high-emission CMIP6 extended runs show AMOC shutdown post-2100 following deep convection collapse in Labrador/Irminger/Nordic Seas.
pik-potsdam.de
Whale migration shifts — Selected literature
Hazen et al. (2013) Nature Climate Change; NOAA Fisheries Right Whale Survey Program; Duke University Marine Lab climate-cetacean research program. Note: right whale population data from NOAA 2024 stock assessment.
Blue whale acoustic decline
McDonald et al. (2009) Endangered Species Research — documented global decline in blue whale song frequency; Širović et al. (2009) Marine Mammal Science; MBARI Monterey Bay Acoustics Research Program.
Capelin population dynamics
Vilhjálmsson (2002) ICES Journal of Marine Science; Rose (2005) Marine Ecology Progress Series; ICES Working Group on Capelin (WGCAPELIN) annual reports; Jónsson & Valdimarsson (2012) on northward spawning migration shifts.
Krill decline and downstream effects
Atkinson et al. (2004) Nature — krill decline linked to sea-ice loss; Tarling et al. (2022) Progress in Oceanography; Roman & McCarthy (2010) PLOS ONE — whale nutrient pump and iron fertilisation feedback; Steinberg & Landry (2017) Annual Review of Marine Science.