- In the Arctic, young and thin ice is melting rapidly and on a large scale; in the Antarctic, unusually warm air masses are slowing the formation of new ice in key regions.
- Remote sensing of sea ice: In the winter 2024/25, satellites detected more than 13,800 cubic kilometres of seasonal ice growth in the Arctic.
- Research icebreaker Polarstern: Arctic expedition takes a closer look at the past, present and future of sea ice.
The Arctic: Young and thin sea ice melting rapidly
In May, the Arctic sea-ice extent shrank by an area more than five times the size of Germany: Whereas the extent was still 13.41 million square kilometres on Thursday, 1 May 2025, by Saturday, 31 May 2025 it had dropped to 11.52 million square kilometres – a difference of 1.89 million square kilometres (Figure 1). The monthly mean value was 12.48 million square kilometres, coming in eighth in the long-term statistic.
In terms of the monthly mean, the sea-ice extent was above the long-term trend and closer to the mean value for the past 20 years. But we can also see that it declined comparatively quickly over the month, meaning a great deal of sea ice melted in just a short time. That in turn tells us that the ice must have been relatively thin
According to our satellite data, the Gulf of Bothnia, Gulf of Finland, and Sea of Okhotsk were ice-free in the last week of May (Figure 2). Satellites detected significantly less sea ice compared to the long-term mean for the years 2003 – 2014 in the northern Hudson Bay, western Bering Sea, eastern Laptev Sea, northeast Barents Sea and in the Kara Sea (Figure 3) in May. The rapid sea-ice retreat in the Russian marginal seas of the Arctic Ocean could be due in part to air temperatures clearly above the long-term mean. Our map of mean temperature anomalies in May shows temperatures 2 to 4 degrees Celsius above the mean for the reference period 1971 – 2000 in these regions (Figure 4). Conversely, temperatures were below the long-term mean over Hudson Bay and parts of the Canadian Arctic Archipelago.
Figure 1: Development of Arctic sea-ice extent (blue line) in comparison. The turquoise band indicates the span of minima and maxima in the period 1981 – 2010. In May, the blue curve for 2025 dropped steeply, indicating that large amounts of sea ice melted during the month.
Figure 2: Sea-ice concentration on the Arctic Ocean on 31 May 2025. The bright green line marks the mean May sea-ice extent for the years 1981 – 2010.
Figure 3: Difference in the mean position of the ice margin in May 2025, compared to the long-term mean for the years 2003 – 2014. Regions marked in blue had more Arctic sea ice than the reference period in May 2025; those marked in red had less.
Pack-ice levels were above average off the eastern coast of Canada, in the eastern Bering Sea, and in Fram Strait (Figure 3). The higher sea-ice extent in Fram Strait could be due to the fact that a great deal of old sea ice from the Arctic drifted into the North Atlantic in the first months of the year. The Sea Ice Portal had reported on this increased ice export in the sea-ice updates for March and April. The current sea-ice surplus in Fram Strait could be the direct result of this trend.
Figure 4: Air temperature anomalies in May 2025 compared to the long-term May mean for 1971 – 2000. The map shows anomalously high air temperatures across the Arctic Ocean, extending to the Siberian shelf zone, northeast Canada, and the North Atlantic.
The bottom line after winter: seasonal sea-ice growth of 13,800 cubic kilometres
Since 2010, sea-ice physicists from the Alfred Wegener Institute have analysed sea-ice data from the two satellites CryoSat-2 and SMOS on behalf of the European Space Agency (ESA). They calculate e.g. how much sea ice forms on the Arctic Ocean over the winter (ice volume); they also use satellite data to seamlessly derive the sea-ice thickness and create anomaly maps for the Arctic Ocean.
Their analysis for the winter 2024/2025 has been available for the past few days. According to the analysis, there were more than 18,300 cubic kilometres of sea ice in the Arctic at the end of winter. If you put it in a giant ice-cube tray with each hole measuring one kilometre on a side, you’d come up with 18,300 enormous ice cubes – an almost unimaginably large amount of ice, of which 13,800 cubic kilometres formed this winter alone (Figure 5).
“The latest Arctic was nothing to write home about in terms of the ice volume. We’re not seeing any direct influences of the low sea-ice extent, either, because the pack ice was thicker than in past years in many regions, which had a very positive effect on overall volume,” says Dr Stefan Hendricks, a remote sensing specialist at the AWI.
Figure 5: Development of sea-ice volume since the two satellites CryoSat-2 and SMOS were launched (2010). In the winter 2024/2025, the volume rose from 4,509 cubic kilometres in October 2024 to 18,371 cubic kilometres in March 2025. Graphic: Stefan Hendricks, Alfred Wegener Institute
At the end of winter (April 2025), CryoSat-2 and SMOS detected particularly thick sea ice chiefly off the north coast of Greenland, in the Chukchi Sea, and in parts of the Russian shelf seas. The pack ice was thinner than in the reference period 2010 – 2021 in the inner waters of the Canadian Arctic Archipelago, in parts of the Central Arctic, and in the Laptev Sea and northern Barents Sea (Figures 6 and 7). “Over the winter, a great deal of pack ice accumulated off the north coast of Greenland. Accordingly, we assume the markedly high ice thicknesses in the region are primarily due to ice deformation – in other words, to floes colliding and in many cases stacking up,” the expert explains.
Figure 6: The sea-ice thickness measured by CryoSat-2 and SMOS in April 2025 (end of winter) and difference in comparison to the monthly mean for April in the period 2010 – 2021. In the right-hand map, all areas with thinner sea ice than in the reference period are shown in red; all areas with thicker ice are marked blue. Graphic: Stefan Hendricks, Alfred Wegener Institute
Figure 7: The sea-ice thickness measured by CryoSat-2 and SMOS in October 2024 (beginning of winter) and difference in comparison to the monthly mean for October in the period 2010 – 2021. As can be clearly recognised, the sea ice in the region north of Greenland was still unusually thin in October 2024. Graphic: Stefan Hendricks, Alfred Wegener Institute
Figure 8: Time series on sea-ice thickness in the Arctic, measured by the two satellites CryoSat-2 and SMOS. The winter 2024/2025 was neither particularly negative nor positive in comparison to previous years. Graphic: Stefan Hendricks, Alfred Wegener Institute
The Antarctic: Striking parallels between sea-ice distribution and air-temperature anomalies in comparison to the long-term mean
In the Antarctic, due to the winter the sea-ice extent climbed in May – by more than 2.7 million square kilometres, an area nearly eight times the size of Germany (Figure 9). The monthly mean was 9.41 million square kilometres, putting May 2025 in seventh place in the long-term statistics. The sea-ice extent has only been lower at this time of year in 1980, 1981, 1986, 2017, 2019 and 2023 (Figure 10).
Further, a comparison shows that the development of sea-ice extent in May 2025 followed a similar course to that in May 2024; in fact, the sea-ice extent was the same at the end of both months. But this year, large areas of pack ice drifted in other regions than in May 2024 (Figure 11). For example, there were more floes year on year in the eastern Weddell Sea, off the coast of East Antarctica, in parts of the Amundsen Sea, and far from the coast of Marie Byrd Land. In contrast, the satellites reported less sea ice than in May 2024 off the northern tip of the Antarctic Peninsula (western Weddell Sea), in the Bellingshausen Sea, and off the coast of Adélie Land and of Queen Maud Land (Figure 11).
Figure 9: Development of Antarctic sea-ice extent (blue line) in comparison. The turquoise band indicates the span of minima and maxima in the period 1981 – 2010.
Figure 10: Development of mean sea-ice extent in the Antarctic for the month of May. The light blue line represents the long-term trend.
Figure 11: Difference in the mean position of the ice margin in May 2025, compared to the mean position in May 2024. Regions marked in blue had more Antarctic sea ice than the reference period in May 2025; those marked in red had less.
With one exception, the distribution of sea ice in the Antarctic corresponded to the pattern of temperature anomalies in May 2025 (Figure 12). The air masses over e.g. the Bellingshausen Sea and the southern part of the Antarctic Peninsula were up to 10 degrees Celsius warmer than in the reference period 1971 – 2000. Unusually warm air temperatures were also recorded over Queen Maud Land and off the coast of Adélie Land. Only over the Ross Sea and the adjacent Ross Ice Shelf in East Antarctica were temperatures lower than in the reference period in May.
Figure 12: Air temperature anomalies in May 2025 compared to the long-term May mean for 1971 – 2000. Substantially warmer air masses can especially be seen over the Antarctic Peninsula and the Bellingshausen Sea.
The sea ice of yesterday, today and tomorrow: RV Polarstern departs on new expedition to the Arctic
For the AWI’s Sea Ice Physics Section, May 2025 was a very busy month: They needed to prepare for a major summer sea-ice expedition on board the research icebreaker Polarstern and load up all their instruments and gear by Friday, 29 May, the day the ship left Bremerhaven, bound for the Arctic.
The experts themselves won’t board the ship until early July (in Tromsø, Norway), when the second segment of this year’s Arctic season begins. It will take cruise leader Dr Marcel Nicolaus and his team to the Central Arctic, where they will investigate the sea ice of yesterday, today and tomorrow at numerous sites.
“Today, we’re already seeing three different sea-ice regimes in the Arctic. Off the north coast of Greenland, there are still remnants of the old, thick pack ice, which can easily be three to four years old. In contrast, one- to two-year-old ice dominates the Transpolar Drift, while the sea ice in the Russian shelf zones melts again after just a few months. These three regimes illustrate the past, present and future of the Arctic sea ice. That’s why it’s essential for us to understand how they differ in terms of their physical characteristics, and what the changes in the pack ice will mean for the Arctic Ocean, its biotic communities, and Earth’s overall climate,” said Marcel Nicolaus at a talk in the last week of May.
He’ll soon share more about the advanced observational systems that he and his team will use to address these questions, and explain why this Polarstern expedition, titled “Contrasts 2025”, can rightly be considered revolutionary – right here on the Sea Ice Portal. Stay tuned!
Figure 13: Dr Marcel Nicolaus, sea-ice physicist at the Alfred Wegener Institute and cruise leader for the 2nd segment of this year’s Arctic season. Photo: Marcel Nicolaus, Alfred Wegener Institute
Grafiken
