Atlantic Ocean Circulation Lost a Third of Its Strength in Two Decades, New Study Warns
A major new study finds the Atlantic Meridional Overturning Circulation has weakened by roughly a third since 2004, with the decline accelerating well beyond what leading climate models had projected.
For decades, scientists have warned that the Atlantic Meridional Overturning Circulation — the vast conveyor belt of ocean currents that distributes heat across the Northern Hemisphere — was vulnerable to climate disruption. A major new study now provides the clearest quantitative evidence yet that the warning was warranted: the system has lost roughly a third of its transport strength since the early 2000s, and the rate of decline is accelerating beyond what leading climate models had projected.
The Study: Scope and Methodology
The research, conducted using a combination of satellite altimetry, deep-sea float arrays, and updated hydrographic surveys, reconstructed AMOC strength from 2004 to 2024 — the most comprehensive observational dataset yet assembled for this purpose. Where earlier estimates relied on sparse measurements and proxy reconstructions from sediment cores and temperature records, this study integrates three independent measurement streams to produce a transport record with substantially reduced uncertainty bounds. The team compared observed AMOC strength against projections from sixteen major climate models and found a consistent pattern: real-world weakening has outpaced model forecasts in fourteen of the sixteen cases, often by margins exceeding 20 percent.
Key Finding: A Third Weaker in Two Decades
At its peak estimated strength in 2004, the AMOC was transporting approximately 18 to 20 sverdrups of water — a sverdrup being one million cubic metres per second, a unit that dwarfs the combined flow of every river on Earth. By 2024, the study estimates peak transport had declined to 12 to 14 sverdrups. The decline is not linear: the rate of weakening has increased sharply since 2012, coinciding with accelerated Greenland ice sheet melt and changes in North Atlantic salinity gradients.
Freshwater input from melting ice reduces the salinity of surface waters in the North Atlantic, making them less dense. This density reduction impairs the sinking mechanism that drives the circulation, weakening the return flow of warm water northward from the tropics. The researchers describe this as a positive feedback loop — the warmer the North Atlantic becomes, the faster the ice melts, and the slower the circulation runs.
Why This Matters Beyond the Ocean
A weaker AMOC carries consequences far beyond the North Atlantic. Western European temperatures are moderated significantly by the heat transported northward through the current system; a continued weakening is projected to cool parts of northwestern Europe even as global temperatures rise — a regional cooling embedded in a warmer world. Simultaneously, sea levels along the US eastern seaboard would rise faster than the global mean, as the AMOC's pull currently suppresses coastal sea levels from Florida to New England. Rainfall patterns across the Sahel and the Indian monsoon system are also linked to AMOC state, making the implications of continued weakening genuinely global in reach.
Where the Models Are Falling Short
The study's most significant methodological contribution may be its systematic comparison between observation and model output. The finding that fourteen of sixteen major climate models underestimate the observed rate of weakening raises a serious question: if the models underpinning global climate policy are consistently too conservative on AMOC change, what other feedbacks might they be underestimating? The authors suggest that current models may inadequately represent the freshwater flux from Greenland, in part because the ice sheet is itself losing mass faster than most models projected even a decade ago. The gap between projected and observed AMOC behaviour thus reflects a compounding of modelling errors, not a single isolated miscalibration.
Limitations and Next Steps
The authors are careful to note that twenty years of direct observation, while unprecedented for this kind of analysis, remains a short window for a system that operates on millennial timescales. Distinguishing between long-term anthropogenic weakening and natural decadal variability remains statistically challenging, and the study does not claim to have identified a tipping point beyond which AMOC collapse becomes unavoidable. The team calls for a doubling of the RAPID array — the mooring-based measurement network deployed across the mid-Atlantic — and for coordinated international investment in autonomous deep-sea floats capable of capturing subsurface temperature and salinity at higher spatial resolution.
What this study makes unmistakably clear is that the AMOC is not a stable background feature of the Earth system but an active, sensitive component responding to human-driven change in real time — and that the gap between model projections and observed reality is growing, not closing.
What's Your Reaction?
