A system of deep-ocean currents that carries tropical heat toward northern Europe appears to be losing strength faster than scientists expected just a few years ago, according to research published in spring 2026 that identifies a new potential early-warning signal: the northward creep of the Gulf Stream itself.
The study, published in the journal Communications Earth and Environment, analyzed satellite altimetry and subsurface ocean temperature records spanning 1993 to 2024. It found that the Gulf Stream’s path has shifted measurably northward over that period, and that in model experiments, abrupt jumps in the current’s position preceded a broader collapse of the Atlantic Meridional Overturning Circulation, or AMOC. If the pattern holds in the real ocean, monitoring the Gulf Stream’s position could give researchers years of advance notice before the circulation reaches a critical threshold.
The finding lands in the middle of an intensifying scientific debate. For decades, climate researchers have asked whether the Atlantic’s vast conveyor belt, which pulls warm surface water north and sends cold, dense water back south at depth, is drifting toward a tipping point or simply riding out natural fluctuations. The new paper argues the answer may be closer to the former, and that the clues are already visible from space.
What direct measurements show
The most reliable window into AMOC’s health comes from the RAPID monitoring array, a line of instruments stretched across the Atlantic at 26 degrees North. Operated jointly by the UK’s National Oceanography Centre, NOAA’s Atlantic Oceanographic and Meteorological Laboratory, and the University of Miami, the array has tracked the circulation’s strength continuously since 2004. Its two-decade record captures seasonal swings and longer shifts, providing the closest thing oceanographers have to a real-time pulse check on the overturning.
“The RAPID array has fundamentally changed how we understand Atlantic overturning variability,” said Ben Moat, a scientist at the UK National Oceanography Centre who works with the monitoring data. “Before 2004 we were relying on occasional ship-based snapshots. Now we can see the circulation breathing in real time.”
A separate reconstruction published in Nature (Caesar et al., 2021, DOI: 10.1038/s41561-021-00699-z) used sea-surface temperature fingerprints and proxy records to look further back in time. That analysis concluded AMOC has weakened significantly compared with the past several centuries, placing the modern circulation near its lowest point in roughly a millennium. The RAPID data and the proxy record tell a consistent story: the system is not what it used to be.
Building on those observations, a 2024 study in Science Advances proposed a physics-based diagnostic centered on freshwater transport by the overturning circulation at roughly 34 degrees South. The metric, known as FovS, tracks how much freshwater the circulation imports across that latitude. When the import shifts in a particular direction, models show the system becomes less stable and more prone to abrupt shutdown. The framework offers a quantitative way to gauge proximity to a tipping point, though it depends partly on reanalysis data rather than a dedicated instrument array in the South Atlantic.
The IPCC’s Sixth Assessment Report, published as part of its 2021 Physical Science Basis volume, stated it is “very likely” AMOC will weaken over the 21st century. That language, calibrated to the models available at the time, set the consensus baseline. Both the Gulf Stream path study and the FovS framework push beyond it by proposing real-time, observable indicators that could detect acceleration toward collapse before models alone would flag the danger.
Where scientists still disagree
Not everyone reads the trend the same way. A 2024 study in Nature Communications reported that the weakening trend appears to have paused since the early 2010s, at least relative to the steeper decline measured in the years before. If that pause reflects natural decadal variability, the long-term slide could resume or accelerate once the cycle shifts. If it reflects something more structural, some of the more urgent projections lose near-term force.
“We need to be careful not to over-interpret a two-decade record,” said Laura Jackson, a climate scientist at the UK Met Office who studies AMOC variability. “Decadal fluctuations can mask or mimic longer-term trends, and we are still learning to tell them apart.”
Collapse timelines are similarly contested. A statistical early-warning analysis published in Nature Communications estimated a potential shutdown window stretching from 2025 to 2095, with a central estimate near mid-century. But a separate modeling study in Nature reached the opposite conclusion, finding a full 21st-century collapse unlikely under the conditions it simulated. A Nature Geoscience paper, summarized by the University of Washington, used observational constraints to argue that future weakening will be more limited than some dramatic model projections suggest. Reporting by the Associated Press on these competing results highlighted researchers who stress that severe weakening remains plausible even if a complete shutdown does not materialize this century.
One persistent blind spot is the Southern Hemisphere. The FovS indicator at 34 degrees South relies on reanalysis products and model output because no monitoring network comparable to the RAPID array exists there. Until oceanographers deploy dedicated instruments in the South Atlantic, the early-warning framework will rest partly on indirect data, making it harder to separate short-term noise from a sustained drift toward instability.
Greenland’s accelerating ice loss adds another layer of uncertainty. Meltwater pouring into the North Atlantic can freshen the surface layer, reducing the density contrast that drives deep-water formation and, by extension, the overturning itself. But models handle freshwater plumes, small-scale mixing, and stratification differently. Two simulations running the same greenhouse-gas scenario but using different freshwater-forcing schemes can produce starkly different outcomes: one showing rapid slowdown and potential collapse, the other a gentler weakening spread over centuries.
Sorting strong evidence from open questions
Three tiers of evidence run through this debate. The strongest is the RAPID array’s direct instrumental record: geographically limited to one latitude but highly reliable for tracking seasonal-to-decadal changes in overturning strength.
The second tier consists of proxy reconstructions, like the Caesar et al. fingerprint study, that extend the timeline centuries beyond any instrument. These are valuable for context, suggesting today’s circulation is weaker than during most of the preindustrial era, but they carry assumptions about how temperature patterns translate into circulation strength and cannot pinpoint the timing of any future tipping event.
The third tier is model-based work, from the FovS diagnostic to the competing collapse and no-collapse simulations. Models can isolate mechanisms that observations alone cannot, such as how a given rate of Greenland meltwater interacts with shifting wind patterns. But their conclusions hinge on structural choices about freshwater forcing, vertical mixing, and feedbacks between sea ice and stratification. When models disagree, the divergence usually traces back to those choices rather than to randomness.
The 2026 Gulf Stream study sits between the second and third tiers. Its observational backbone, satellite altimetry from 1993 to 2024 and subsurface temperature records reaching back to 1965, is solid. But its central claim, that abrupt Gulf Stream path shifts act as a precursor to AMOC collapse, emerges from model experiments, not from a documented historical collapse. The proposed early-warning signal has not been tested against a real-world tipping episode, and false alarms remain possible.
What weakening would mean on the ground
Even under moderate weakening scenarios, the practical consequences are broad. A slower AMOC would deliver less heat to northern Europe, potentially cooling winter temperatures in the United Kingdom, Scandinavia, and surrounding regions. Along the U.S. East Coast, the redistribution of ocean mass would contribute to faster sea-level rise, compounding flood risks in cities from Miami to Boston. The so-called North Atlantic “cold blob,” an area of anomalously cool surface water south of Iceland that has already appeared in observational data, offers a visible hint of reduced heat transport and has been linked to shifts in European weather patterns.
Agricultural growing seasons, commercial fisheries, and Atlantic storm tracks all depend on how heat moves through the basin. Changes in storm paths could alter wind energy potential and accelerate coastal erosion in new areas. Marine ecosystems, from plankton productivity to fish migration routes, would reorganize in ways that ripple through food chains and fishing economies on both sides of the Atlantic.
Why closer monitoring of the Gulf Stream matters now
The IPCC’s assessment that further weakening is very likely already informs adaptation planning in coastal communities and national climate strategies. The newer studies sharpen the question of speed and severity but have not yet been incorporated into an institutional consensus update. Until the next IPCC assessment cycle revisits AMOC with these indicators and fresh observational data, the responsible reading is neither complacency nor alarm. The emerging picture supports treating AMOC as a high-impact risk with low-to-moderate probability of abrupt change in the coming decades, one that warrants closer monitoring, continued investment in ocean observation networks, and contingency planning even as researchers work to pin down how close the system truly is to a tipping point.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
