AMOC weakening shows faster odds
- A new Science Advances study led by Valentin Portmann says the Atlantic Meridional Overturning Circulation is likely to weaken more than expected this century after combining observations with CMIP6 climate models. - The paper’s best estimate is a roughly 51% decline by 2100, about 60% stronger weakening than model-only projections, based on sea-surface temperature and salinity patterns across the Atlantic. - That lands in a live scientific dispute: other 2025 studies said collapse this century remains unlikely even as weakening is expected, while the IPCC still says abrupt collapse is unlikely. (ipcc.ch)
The Atlantic Meridional Overturning Circulation is the Atlantic’s heat conveyor, and a new study says it may weaken much more this century than many models projected. (science.org) (phys.org) The current moves warm, salty surface water northward, where it cools, gets denser, sinks, and returns southward at depth. That loop helps shape temperatures in Europe, rainfall in Africa and the Americas, and sea levels along the U.S. East Coast. (climate.mit.edu) In April 2026, Valentin Portmann and co-authors reported in Science Advances that observational constraints point to about a 51% AMOC decline by 2100. The paper says that is about 60% stronger weakening than estimates based on climate models alone. (science.org) (phys.org) The authors used present-day sea-surface temperature and salinity patterns to reweight CMIP6 model projections. Their estimate spans roughly 43% to 59% weakening by 2100, rather than the smaller decline often cited from unconstrained model ensembles. (science.org) (yahoo.com) That does not mean scientists now agree an AMOC collapse is more likely than not by mid-century. The new paper projects stronger weakening by 2100; it is not the same as a formal forecast of a shutdown before 2050. (science.org) That distinction matters because the field is split on how close the system is to a tipping point and how well current methods can detect one. Several recent papers have pulled in opposite directions on how much confidence to put in worst-case outcomes. (aos.wisc.edu) (nature.com) A January 2025 paper highlighted by the University of Wisconsin–Madison said statistical “critical slowing down” methods may not reliably predict whether or when the AMOC will collapse. Lead author Clark Zimmerman said the same slowdown signal can arise for reasons other than an approaching tipping point. (aos.wisc.edu) A May 2025 Nature Geoscience paper reached a different kind of constraint, concluding the AMOC would weaken by about 3 to 6 sverdrups, or roughly 18% to 43%, by 2100 regardless of emissions scenario. That is still substantial, but lower than the new Science Advances estimate. (nature.com) A February 2025 Nature study led by Jonathan Baker at the Met Office and University of Exeter went further, arguing an AMOC collapse over the next 75 years is unlikely, even though weakening remains very likely. Rowan Sutton at the Met Office said that result did not change expectations of important climate impacts from a weaker circulation. (phys.org) (metoffice.gov.uk) The Intergovernmental Panel on Climate Change’s standing assessment is still that the AMOC will very likely weaken over the 21st century, while a collapse is very unlikely, with medium confidence. The IPCC also says collapse becomes much more plausible by 2300 under high emissions. (ipcc.ch 1) (ipcc.ch 2) So the update here is sharper than the viral posts but narrower than some of their claims: one new paper raises the expected size of AMOC weakening by 2100, while the argument over a near-term collapse is still unresolved. (science.org) (nature.com)
