(István Bencsik, scientific resarcher, RS3 Informatika,
November 7, 2026)
Multi-year ice now accounts for less than 5% of the Arctic ice volume at its September minimum; however, the volume of ice that freezes annually has, to date, shown hardly any decline (along the northern coast of Greenland and within the Canadian Arctic Archipelago). Since 1979, the Arctic has lost approximately one-third (about 30–33%) of its winter ice volume, primarily because 95% of the thick (3–4 meter) "native" multi-year ice has vanished. The year 2026 marks a tipping point: the amount of ice freezing annually in the Arctic will fall below the amount melting. By the end of the century (2100)—at which point the ice-covered area will have shrunk by approximately 1 million km²—the ice in the remaining areas will thin to a seasonal layer of only about 1 meter in thickness. According to the models, an additional 20–25% of the winter sea ice volume will disappear by 2100. Furthermore, the models indicate a linear relationship; consequently, the AMOC is projected to weaken by 40–50% by 2100. This near-linear relationship is driven by the large volume of meltwater and the fact that less salt is expelled from the seawater as it freezes. Currently, the annual amount of salt expelled during the freezing process in the Arctic remains relatively stable—a factor that ensures the intensity of the Denmark Strait overflow (and the AMOC as a whole) remains nearly constant, despite a slight decline.
The estimate mentioned in the title is a rough approximation; according to a precise physical model, the intensity of the AMOC is expected to decrease by 51% ± 8% by 2100. This estimate is based on the fact that, while multi-year Arctic sea ice has already melted away, the amount of ice that freezes annually is currently still roughly equal to the amount that melts each year. By 2100, it is projected that only about a quarter of the current annual volume of ice will form; since polar ice is salt-free, this will result in a reduction in the amount of salt expelled into the surrounding water. The strength of the AMOC current is approximately linearly proportional to the amount of salt expelled during freezing. The slowdown of the AMOC is driven by freshwater influx and a decrease in water density resulting from global warming. According to a report on global climate change indicators, warming reached 1.37°C in 2025. Data from Copernicus/ERA5 indicate an observed annual average of 1.47°C for 2025. In 2026, the figure was approximately 1.5°C globally, and approximately 2.5°C in Europe. Over the past few decades, the cooling effect of the Arctic has diminished by about a quarter, driven by a reduction in sea ice extent and a decrease in albedo. Multi-year ice has largely disappeared: while thick ice older than four years accounted for 33% of the Arctic ice cover in 1985, this proportion had fallen to 4.4% by 2020.
On a positive note, the sea ice cover in the Arctic Ocean reached its winter maximum on March 15, 2026, spanning 14.29 million square kilometers—a figure only slightly lower than the 14.31 million square kilometers recorded in 2025. Both values are roughly in line with the 1981–2010 average. The ice cover reaches its minimum extent every year in September (on September 12, 2026, the 10th lowest minimum in the history of satellite measurements was recorded); the autumn refreezing period was significantly weaker than usual due to the record-warm October of 2026, and the growth of the ice cover in November is falling well short of the long-term average.

(Image source: https://www.arcticdeathspiral.org/)
Carbon dioxide causes warming, winds are weakening, and meltwater dilutes the currents. However, in the "Denmark Strait overflow"—a phenomenon akin to an underwater waterfall—cold water that has accumulated in the Nordic basins spills over the submarine ridges between Greenland and Scotland due to gravity. This process accounts for roughly two-thirds (60–70%) of the AMOC’s deep-water return flow. Another ~20% is contributed by open-ocean convection—the direct, "vertical" sinking of saline water in the Labrador and Greenland Seas. If the surface water fails to become sufficiently dense (for instance, due to an influx of fresh water), open-ocean sinking slows down; yet, the "Denmark Strait overflow" continues to function, driven by the cold water already present in the basins and by wind forcing. This is why the anticipated shutdown of the AMOC would occur only gradually. (Warming increases the rate of evaporation, which makes the surface water saltier and denser, while slightly diminishing the impact of wind patterns.) Research indicates that there is also a wind-driven component, which accounts for approximately 20% of the AMOC’s current intensity.
Up to a certain (tipping) point, the AMOC can adapt linearly to changes in polar ice and freshwater influx; however, once a critical threshold is crossed, a sudden and drastic slowdown—or even a complete collapse—may occur. The intensity of the AMOC is not driven solely by the freezing of ice. It is also influenced by other critical factors, such as the melting of Greenland’s glaciers (which releases freshwater), increased precipitation, the rate of global temperature rise, and wind patterns.
The latest climate models and research based on observational data indicate a more drastic and rapid slowdown of the AMOC by 2100 than what was predicted in earlier consensus reports. While the previous IPCC report (AR6) considered the probability of a complete collapse within this century to be relatively low, the most recent models - refined using actual observational data - suggest that the risk is far more serious.
Key findings from the latest simulations indicate that the AMOC will slow down by 42–58% by 2100. The results show that even under a moderate greenhouse gas emission scenario, the AMOC could weaken by approximately 51% (±8%) by 2100. If the volume of ice that freezes annually decreases by only 25%, since the long Arctic night as long as half a year, and since the multi-year ice has already melted, the amount of salt that freezes out annually is reduced by 25 %.
Conclusion: According to recent research, by the end of the century, cc. 25 % less ice will freeze annually at the North Pole, thus by the end of the century the Intensity of the AMOC will weaken by 40 - 50 %.
