A landmark study has just moved the deadline for AMOC collapse closer than climate models ever predicted - and the consequences would reshape every continent on Earth.
On April 15, 2026, a paper published in Science Advances quietly reset one of the most consequential timelines in climate science. Using a methodology that filters out a known bias in the models scientists have relied on for decades, the study concluded that the Atlantic Meridional Overturning Circulation - the vast system of ocean currents that keeps much of the Northern Hemisphere habitable - is on course to weaken by approximately 51 percent by the end of this century. That is 60 percent worse than what the average of all climate models had been projecting. And it may cross a point of no return as early as the middle of this century.
The paper, led by researchers in Europe and published in Science Advances, did something deceptively simple: it corrected for a systematic error that had been skewing climate model projections for years. Most climate models - the same ones that feed the IPCC's reports and inform national policy - contain a bias in how they simulate sea surface salinity in the South Atlantic. The models tend to underestimate how much fresh, low-density water flows into the South Atlantic basin, which in turn makes them underestimate how sensitive the circulation is to disruption.
The researchers used ridge-regularized linear regression - a statistical technique that identifies signal in noisy data by penalizing overfitting - to integrate a large set of observable variables from real-world ocean measurements. They then applied those observational constraints to the full suite of climate model projections. The result narrowed the margin of error dramatically: from plus-or-minus 37 percentage points (the range spanned by unadjusted models) down to plus-or-minus 8. The South Atlantic salinity bias alone accounts for 47 percent of the gap between old projections and the corrected figures.
Simultaneously, a complementary study from the University of Miami analyzed real-world data from four ocean monitoring moorings along the western boundary of the North Atlantic, which have been continuously measuring water temperature, salinity, and current velocity since 2004. Their findings: AMOC has been measurably weakening at four different latitudes over the past two decades. Together, these two studies - one using better statistical methods on model projections, one using direct observation - point in exactly the same direction.
The new projection is not the product of a more alarming model. It is the product of a less biased one. Stripping out the South Atlantic salinity error reveals that the circulation is significantly more vulnerable than we had calculated - not because the physics changed, but because we were measuring them incorrectly.
The Atlantic Meridional Overturning Circulation is one of the most important systems on Earth, and also one of the least visible. It has no surface you can see from space, no shoreline you can photograph. It is a three-dimensional engine of moving water that spans the full depth of the Atlantic Ocean and runs continuously for thousands of years at a time - until it doesn't.
The system works through a basic principle of fluid physics: density drives flow. Warm water is less dense than cold water. Salty water is denser than fresh water. The AMOC exploits these differences to create a perpetual conveyor. Warm, relatively salty surface water flows northward in the upper Atlantic - you know part of this system as the Gulf Stream, the current responsible for the mild winters of western Europe. As this warm water travels north, it loses heat to the atmosphere (warming Norway, Scotland, and Iceland in the process) and some of it evaporates, leaving behind water that is colder and saltier - and therefore denser.
In the Norwegian Sea and the Labrador Sea, this cold, dense water does something remarkable: it sinks. Not gradually. It plunges - down hundreds, sometimes thousands of meters into the deep ocean, becoming what oceanographers call North Atlantic Deep Water. That sinking creates a kind of suction at the surface, pulling more warm water northward to replace what just descended. The deep water then flows slowly southward along the ocean floor, eventually upwelling in the Southern Ocean and completing a circuit that takes roughly 1,000 years per lap.
The critical vulnerability is that sinking zone. It requires the arriving water to be dense enough to sink. If the North Atlantic is flooded with freshwater from melting ice - freshwater is less dense than saltwater - the arriving surface water may no longer be dense enough to descend. The sinking slows. The suction weakens. The warm northward flow slows with it. This is the mechanism by which melting Greenland ice becomes a threat not just to sea levels, but to the entire circulation system.
The AMOC currently transports roughly 17 million cubic meters of water per second - a volume equivalent to about 100 Amazon Rivers flowing simultaneously. It moves heat northward at a rate of approximately 1.3 petawatts, or 1.3 million billion watts. To put that in perspective, that is more than 80 times the total electricity generating capacity of the entire United States. This is the amount of thermal energy being redistributed across the Atlantic every second. Weaken it by half, and that redistribution slows proportionately - with consequences for every weather system and ecosystem connected to the Atlantic basin.
The concern about AMOC is not new. Paleoclimate records have long shown that the circulation can collapse - and that when it does, it can happen with terrifying speed. The clearest example is the Younger Dryas, an event that occurred approximately 12,700 years ago when Earth was warming out of the last ice age. A massive outflow of meltwater from retreating glaciers flooded the North Atlantic with freshwater, disrupted the sinking mechanism, and triggered a rapid return to near-glacial conditions. Within decades - possibly within years - temperatures in Greenland dropped by 10 degrees Celsius or more. The cold lasted 1,200 years. Then, just as abruptly, the circulation restarted and temperatures snapped back.
Modern direct measurement of the AMOC only became possible in 2004, when the RAPID array was deployed across the Atlantic at 26.5 degrees North - a line of undersea sensors, cables in the Florida Strait, and deep-ocean velocity instruments that have been running continuously ever since. Twenty years of RAPID data have shown considerable natural variability in the circulation's strength, which initially led some scientists to conclude the system was stable. But that 20-year window is short relative to the timescales on which AMOC operates, and newer analyses - including the University of Miami study published alongside this week's Science Advances paper - are finding coherent weakening signals at multiple latitudes.
"The models were not wrong about the physics. They were wrong about the starting conditions. Correcting that single bias shifts the entire projected trajectory."
- Lisa Pedrosa, lisapedrosa.comProxy evidence from sediment cores and ocean temperature records extends the picture further back and paints a less comfortable picture. Analyses of fingerprint patterns in sea surface temperatures - regions that should warm as AMOC strengthens and cool as it weakens - suggest the circulation has been slowing since roughly the mid-20th century, consistent with the beginning of accelerated greenhouse warming. A 2021 study in Nature Climate Change put the current AMOC strength at its weakest point in over 1,000 years.
A 51 percent weakening of the AMOC is not a collapse. It is a profound and partially irreversible disruption with consequences that would play out across every continent and every ocean on Earth. But the new study also raises the specter of a full collapse - a state in which the sinking mechanism shuts down entirely - by highlighting how much faster the system is approaching the tipping point than we thought. Some models now place the tipping point risk as early as the 2050s.
The effects on Europe would be the most dramatic. Despite being at latitudes equivalent to Alaska and Siberia, cities like London, Amsterdam, and Paris enjoy mild winters largely because AMOC delivers tropical heat to the eastern North Atlantic. Lose that heat delivery, and European winters would no longer be anomalously mild - they would align with what the latitude actually implies. Research projects drops of 5 to 15 degrees Celsius across northwestern Europe. London could see sustained winter cold at minus 2 Celsius. Oslo could see temperatures plunge toward minus 55 Fahrenheit at the extremes. Arable farming across much of Britain and Northern Europe would be compromised. Water supply systems designed for a temperate climate would be stressed by a suddenly subarctic one.
North America would not be spared. Sea levels along the US East Coast would rise significantly faster than the global average - AMOC currently acts as a partial counterweight to sea level rise in the western North Atlantic. Shut it down and that counterweight disappears. New York, Miami, and Boston face accelerated inundation risk independent of the global warming signal. Cooling in the North Atlantic would also shift weather patterns across the continent, altering precipitation and drought cycles in ways that are difficult to model precisely but clearly negative.
The tropics would see some of the most consequential shifts. The Intertropical Convergence Zone - the band of rainfall that drives the monsoons of West Africa, the Amazon basin, and South and Central America - would shift southward as the northern Atlantic cools. This means prolonged drought in the African Sahel, where hundreds of millions of people depend on monsoon rains for agriculture. The Amazon basin, already under pressure from deforestation, would face a changed precipitation regime on top of its terrestrial stresses. The potential for cascading failures across interconnected Earth systems - the Amazon, the West Antarctic ice sheet, the Saharan and Indian monsoons - is part of what makes AMOC not just a regional problem but a planetary one.
What the April 2026 study adds to this picture is urgency - and something arguably more unsettling than a worse number. It reveals that the models feeding our policy decisions have been systematically underestimating the risk. This is not a matter of nature behaving unexpectedly. It is a matter of our measurement tools being calibrated incorrectly, and every policy decision made on the basis of those uncorrected projections having been made on the basis of an overly optimistic assessment of how much time we have.
The RAPID array continues to run. Greenland continues to melt - losing mass at an accelerating rate as Arctic warming outpaces global averages. The freshwater is already in the North Atlantic. The question the new study puts with uncomfortable precision is not whether the conveyor belt is slowing, but how much slower it will get before it stops - and whether anything we do between now and the tipping point will be enough to pull it back from the edge.
© 2026 Lisa Pedrosa · lisapedrosa.com
All articles cited to primary institutional or peer-reviewed sources
Buy me a coffee