The system of ocean currents that keeps London milder than Labrador and steers hurricane tracks across the Atlantic is losing power. Decades of measurements now confirm that the Atlantic Meridional Overturning Circulation, a planet-scale conveyor belt that carries warm surface water northward and returns cold, dense water at depth, is weakening across multiple layers of the ocean. Deep-water transport is declining, the Gulf Stream is drifting northward, and the abyss itself is warming. As of spring 2026, the scientific question is no longer whether the AMOC is slowing down but how quickly, how far the decline will go, and whether the system could eventually lurch into an abrupt, irreversible collapse.
What the instruments are recording
Three independent sets of ocean observations, each published in peer-reviewed journals, now point in the same direction.
The first comes from the deep Atlantic. A study in Nature Geoscience found that northward transport of Antarctic Bottom Water at 16°N fell by roughly 12 percent between 2000 and 2020, a loss equivalent to about 350,000 cubic meters of cold water per second. At the same time, temperatures in the abyss have been climbing at roughly one-thousandth of a degree Celsius per year. That may sound trivial, but spread across the vast volume of the deep Atlantic it represents an enormous accumulation of heat, one that contributes to faster sea-level rise in parts of the basin.
The second dataset comes from mooring arrays anchored in the subpolar North Atlantic. Instruments deployed starting in 2014 recorded a 26 percent drop in Deep Western Boundary Current transport over six years. Researchers traced the decline to two causes: the dense water layers that drive the current are getting thinner as fresher water from melting ice dilutes them, and the current itself is slowing. These are not projections from a computer model. They are measurements from sensors bolted to the ocean floor.
The third line of evidence sits closer to the surface. A study published in Communications Earth & Environment combined satellite altimetry data from 1993 to 2024 with subsurface temperature records stretching back to 1965. It documented a gradual northward shift of the Gulf Stream near Cape Hatteras and found a statistical association between that shift and a weakening overturning circulation, though the researchers noted that other factors, including wind forcing and natural variability, also influence the Gulf Stream’s position. In climate-model experiments, the team found that a sudden, step-like jump in the Gulf Stream’s path tends to occur shortly before a simulated AMOC collapse. The observed drift so far remains gradual, but its persistence is a red flag.
Tying these records together is the RAPID-MOCHA-WBTS mooring array at 26°N, operated by a UK-led consortium since 2004. It provides the longest continuous direct measurement of AMOC strength, tracking heat transport, Florida Current flow, and deep-water movement. Its publicly available data form the backbone of direct AMOC monitoring and have been central to nearly every major assessment of the circulation’s health over the past two decades.
Where scientists still disagree
The sharpest debate is over the endgame. Can the AMOC actually shut down under realistic warming, or will it settle into a weaker but stable state?
A multi-model assessment published in Nature in early 2026 tested 34 climate models under strong greenhouse-gas forcing and large freshwater inputs. In every simulation, the AMOC weakened substantially but never collapsed entirely; it stabilized at a reduced strength. The authors noted that some models may underestimate the probability of complete collapse, because the finding applies only to the specific models and forcing scenarios they tested. They cautioned that processes those models handle poorly, such as the pace of Greenland ice-sheet melt and its freshwater pulse into the North Atlantic, could alter the outcome.
The Intergovernmental Panel on Climate Change reached a similar but carefully hedged conclusion in its 2021 Sixth Assessment Report: the AMOC is “very likely” to weaken over the 21st century, but an abrupt collapse before 2100 is “very unlikely.” The panel added that risks grow under high-emission pathways and over longer time horizons. Since AR6 was published, newer observational data, including the studies described above, have reinforced the weakening trend without yet resolving the collapse question.
Early-warning signals add another layer of uncertainty. A study in Nature Communications applied statistical markers known as “critical slowing down” indicators to observation-based AMOC fingerprints. These markers can, in theory, flag a system approaching a tipping point. But the researchers found that the results shifted depending on which dataset they used and how they filtered the data, meaning that different analytical choices led to different conclusions about how close the AMOC is to a threshold. The warning signals are real, but their interpretation remains contested.
A complementary approach uses sea-surface temperature patterns, including the persistent patch of cooler-than-expected water south of Greenland sometimes called the “cold blob,” to infer AMOC weakening stretching back more than a century. This fingerprint method extends the record far beyond what mooring arrays can offer, but it relies on indirect proxies rather than direct flow measurements. Together, the two approaches suggest the recent slowdown is part of a longer trend riding on top of large natural swings from year to year and decade to decade.
What a weaker AMOC means on land
For the roughly 400 million people living around the North Atlantic basin, a slower conveyor belt is not an abstraction. The consequences are already beginning to register in planning documents and risk assessments.
A weakening AMOC is expected to cool parts of northwestern Europe relative to the global average, potentially shaving degrees off winter temperatures in a region that has built its agriculture and infrastructure around a milder climate than its latitude would otherwise allow. At the same time, reduced northward heat transport could shift North Atlantic storm tracks, altering rainfall patterns from West Africa’s monsoon belt to Western Europe.
Along the U.S. East Coast, the effects cut in a different direction. As the circulation slows, water that would otherwise be drawn away from the coastline piles up, raising relative sea levels from Florida to New England. Cities such as Miami, Norfolk, and New York, already contending with tidal flooding, face an additional increment of rise on top of what thermal expansion and ice-sheet melt are delivering. The Gulf Stream’s northward drift also has implications for marine ecosystems: commercially important fisheries in the Gulf of Maine are already experiencing water-temperature shifts that have disrupted cod and lobster populations.
The practical question for policymakers is not whether the AMOC will collapse next year. The weight of evidence says it almost certainly will not. The question is whether a circulation that is measurably losing strength will cross thresholds that amplify flooding, disrupt food systems, or alter weather patterns in ways that outpace adaptation planning.
Why the uncertainty itself is the risk
Holding two ideas at once is uncomfortable but necessary. The AMOC is not on the verge of collapse in any confidently predictable way. Yet the ongoing slowdown is significant enough to matter for climate risk planning, coastal infrastructure investment, and fisheries management on both sides of the Atlantic.
The gap between “very likely to weaken” and “very unlikely to collapse” is not reassuring empty space. It is a zone of genuine ignorance, populated by feedbacks that models struggle to capture: the rate at which Greenland sheds freshwater, the response of deep-water formation sites in the Labrador and Nordic Seas, and the possibility that the system’s sensitivity to forcing is nonlinear in ways current models underestimate.
Continued investment in mooring arrays, satellite monitoring, and next-generation ocean models will be essential to narrow those uncertainties. The RAPID array, now entering its third decade, remains the single most important instrument for tracking the circulation’s pulse. Expanding similar monitoring to other latitudes and depths would give scientists, and the public, a clearer picture of where the system is headed.
In the meantime, the most reliable way to reduce the chance of crossing a poorly understood tipping point remains the same prescription that applies to every other element of the climate system: limiting the magnitude and pace of global warming by cutting greenhouse-gas emissions. The AMOC’s slow fade is a reminder that some of the planet’s most consequential changes are happening out of sight, in cold water, miles below the surface, where no one is watching except the instruments.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
