In the summer of 2026 two significant studies were published on the Atlantic meridional overturning circulation (AMOC), the large-scale circulation of the Atlantic that carries warm surface water northward and cold deep water southward. In Science Advances (opens in a new tab) in June, Oliver Mehling and colleagues added meltwater from the Greenland ice sheet to the EC-Earth3 climate model; in the August issue of Nature Climate Change (opens in a new tab), René M. van Westen, Reyk Börner and Henk A. Dijkstra examined how the stability of the circulation depends on the rate of warming.
Taken together, the two results give a nuanced picture. The weakening of the AMOC is a measured and modelled fact; the scenarios circulating in public debate, of a collapse within a few years and of extreme temperatures, have no scientific basis. This analysis examines what follows from current knowledge for the European climate, and what it means for the thermal protection of the building stock.
Modelled winter extremes in Europe after a complete AMOC shutdown, with 2 °C of global warming
These figures come from a conditional model experiment: they show the winter extremes that would arise after a complete shutdown, in the re-equilibrated climate. They are not forecasts, nor do they refer to the coming years. The effect weakens with distance from the ocean: the model gives no such values for Central Europe.
Not the Gulf Stream: how the overturning circulation works
The Gulf Stream is sustained mainly by the wind system above the ocean and by the Earth’s rotation, so physics gives no reason to expect it to stop. What is weakening is the AMOC, the whole Atlantic overturning circulation, of which the Gulf Stream is only the surface western boundary current. The distinction is not one of wording: the two systems are driven by different processes, and changes in them have different consequences.
The AMOC is driven by density differences. Around Greenland and Iceland, and in the Labrador Sea, cooling and salty surface water becomes denser, sinks and flows south as deep water; the warm water replacing it carries heat towards north-western Europe. Greenland meltwater and increasing precipitation freshen the surface layer. Water of lower salinity is less dense and sinks less readily, so deep-water formation weakens.
What measurements and models show
The circulation has been monitored since 2004 by the RAPID array at 26.5° north; its mean volume transport is about 17 sverdrups, or 17 million cubic metres per second. From reconstructions, Levke Caesar and colleagues concluded in Nature Geoscience (opens in a new tab) in 2021 that the AMOC is currently in its weakest state in the past millennium. According to the IPCC Sixth Assessment Report, further weakening this century is very likely, while an abrupt collapse before 2100 is not expected, with medium confidence.
Estimates of the timing cover a wide range. A 2023 statistical analysis by Peter and Susanne Ditlevsen in Nature Communications (opens in a new tab) placed the tipping point between 2025 and 2095, with a central estimate of 2057. In 2025, Sybren Drijfhout and colleagues, writing in Environmental Research Letters (opens in a new tab), found in CMIP6 model projections a collapse of deep mixing by mid-century, but a shutdown of the overturning only after 2100. In 2026, Mehling and colleagues showed that Greenland meltwater weakens the circulation by about a further ten per cent by 2100, while an abrupt collapse does not occur even by 2300.
The 2026 work by van Westen, Börner and Dijkstra added a new dimension: stability depends not only on the amount of warming but on its rate. Under slow warming the circulation remained stable beyond 5 °C, whereas at a rapid rate similar to today’s it collapsed at around 2 °C of global warming. The critical rate is roughly 0.3 °C per decade, and the world is now approaching it. Cutting emissions therefore directly affects not only the amount of warming but the risk to the overturning circulation as well.
Central Europe: less cooling, a wider seasonal swing
The 2025 model experiment by van Westen and colleagues examined the climate after a complete shutdown, with 2 °C of global warming. It gave winter extremes of around −48 °C in Oslo, −29.7 °C in Edinburgh and −19.3 °C in London, while summers remained only slightly cooler than today. The effect is strongest near the Atlantic and fades inland; no model gives such extremes for Hungary. For comparison, the national cold record is −35.0 °C (Görömbölytapolca, 16 February 1940) and the heat record 42.0 °C (Szécsény, 30 June 2026).
The consequence for Central Europe is therefore above all a wider seasonal swing. According to a 2025 study in Hydrology and Earth System Sciences (opens in a new tab), a collapsed circulation reduces precipitation across Europe and, combined with warming, raises the risk of seasonal drought. As early as 2016, Aurélie Duchez and colleagues showed that cold anomalies in the North Atlantic can intensify European heatwaves through atmospheric blocking. The seasonal amplitude may thus widen in both directions: colder winter extremes, and drier summers burdened with heatwaves.
What this means for the building stock
The design outdoor temperatures and heat-loss calculations of buildings rest on climate data from the past. A wider seasonal swing means that the building stock has to cope at once with higher peak heating demand in colder winter episodes and higher cooling demand in summer heatwaves. Both loads fall on the same energy system: the winter peak is borne by the gas network and the electricity demand of heat pumps, the summer peak by air conditioning.
System-level risk is determined by peak demand, not by the annual average. During a colder winter episode the power demand for heating rises steeply, just when heat pumps are at their least efficient. The thermal protection of the building envelope is therefore a matter not only of energy efficiency but of security of supply: every reduction in heat loss also lowers the peak load on the system and leaves a margin for extreme periods.
Where GWR NANO INSULATION® fits – and the limits of the data
For the winter side, GWR NANO INSULATION® has measurement data taken under external monitoring. According to TÜV SÜD external monitoring report no. 3229268, in the comparative field test at Dunabogdány – measured by MEOLIT Kft. in January and February 2021 on three identically built test houses – the coated house used 43% less heating energy than the uninsulated house and 9.5% less than the reference house with 10 cm of EPS insulation, under the same weather conditions.
The coating’s declared operating temperature range runs from −60 °C to +260 °C, covering even the most extreme modelled European winter minima. The coating remains vapour-permeable (sd = 0.40 m, class V2), so it does not block outward drying of the wall – in colder winters, with greater temperature differences, this matters particularly for moisture safety. For the summer side, however, no independent measurement equivalent to the winter test exists, so we give no summer percentage.
The coating is applied by airless spray, on low- and medium-rise buildings typically without scaffolding, and the building can stay occupied throughout. This allows thermal protection to be improved on articulated, listed or terraced buildings where external board insulation cannot be built. Where a board system can be installed, it delivers the greater and better-documented reduction in heat loss. The report does not regard its field measurement as an internationally standardised procedure.
Engineering conclusions
The current state of AMOC research points to three conclusions for the design of the building stock, each building on the last.
- In climate terms: the weakening of the overturning circulation is a measured and modelled process, but a gradual one; there is no scientific basis for a collapse within a few years. Its risk is also set by the rate of warming, so cutting emissions reduces it directly.
- In design terms: in Central Europe the main consequence is a wider seasonal swing. Design outdoor temperatures and peak demand should therefore be matched to the future climate, not the past one, on both the winter and the summer side.
- In building-physics terms: better thermal protection also lowers peak heating demand and the load on the energy system. For GWR NANO INSULATION®, measurements from the heating season are available, and we stay within their scope.
The future of the AMOC is uncertain today; preparing the building stock is not: the buildings renovated today will serve a climate with a wider seasonal swing for decades to come.
The heat load of the 2026 super El Niño is examined in our analysis of ENSO heat load, and heat-related mortality in the related post. The water–energy–heat nexus is covered in our drought post, and the order of a heat pump retrofit in this article. Our installations in Hungary and abroad are shown on the References page.
GWR NANO INSULATION® is distributed in Hungary and the Central and Eastern European region by Summotive® (Summa Technologiae Kft.). TÜV SÜD test reports, the Declaration of Performance, and technical data sheets are available on request.