- The Arctic: Warm air and ocean temperatures delayed the formation of new sea ice, particularly off the northeast coast of Canada. There were substantial temperature fluctuations on Spitsbergen in December.
- The Antarctic: In December, the sea-ice extent experienced seasonal melting, losing more than half of its area. The monthly mean value was the eighth-lowest in the time series for December.
- Bevor the ice cover in Atka Bay broke up, AWI researchers had the opportunity to observe the inflow of warm water masses into the bay.
The Arctic: Too little sea ice, particularly off the east coast of Canada
2025 will not only go down in history as the third warmest years since the beginning of climate observation;it also ended with a new record low December mean sea-ice extent, the lowest since the beginning of satellite observation in 1979. At 11.32 million square kilometres, it was slightly lower than the mean values from 2016 (11.44 million square kilometres) and 2024 (11.59 million square kilometres).
“The new December minimum confirms the long-term decline in Arctic sea ice. However, we shouldn’t attach too much importance to it; from a statistical standpoint, a difference of 120,000 square kilometres in sea-ice extent isn’t particularly significant. More importantly, new Arctic sea ice formed so slowly in November and December 2025 that the curve for the year remained well below the span of minima and maxima for the period 1981 – 2010 and the trend from the previous winter continued,” says Dr Klaus Grosfeld, a climate researcher at the Alfred Wegener Institute and expert for the Sea Ice Portal (Figure 1).
The new December low is chiefly attributable to the lack of sea ice off the northeast coast of Canada, as can be seen by comparing the sea-ice concentration maps from December 2025, 2024 and 2016 (Figure 2). Conversely, there was more ice than in December 2016 in the Bering Sea and the eastern Barents and Kara Seas. On the basis of the satellite observations alone, it remains unclear why the formation of new sea ice was so sluggish off the northeast coast of Canada.
“In the Labrador Sea and Baffin Bay, not only were sea surface temperatures comparatively high in December 2025; in terms of the monthly mean value, the air was also warmer than the long-term mean. This combination may have slowed the formation of new ice,” says Dr Renate Treffeisen, an atmospheric researcher at the Alfred Wegener Institute and expert for the Sea Ice Portal. “In addition, predominantly landward winds could have helped to compact the existing pack ice just off the coast, which would explain why the sea-ice extent is so low”, she adds (Figure 3).
Figure 1: Development of Arctic sea-ice extent in comparison; the bright blue line is the curve for 2025. For the first three weeks of December, the curve was consistently below the turquoise band, which indicates the span of minima and maxima in the period 1981 – 2010. Only at the end of the month did it come nearer to the band. For comparison, the violet and red lines represent the development of the sea-ice extent in the winters of 2024/2025 and 2016/2017.
Figure 2: Comparison of December mean sea-ice concentration in the Arctic for the years 2025, 2024 and 2016. The circles indicate regions where the satellites detected more (blue) or less (red) pack ice in December 2025 than in the previous years.
Figure 3: Mean sea-surface temperature (left), air temperature (centre) and pressure anomalies at sea level (right) for December 2025 in the Arctic. The red circle indicates the Labrador Sea / Baffin Bay region off the northeast coast of Canada.
Extreme temperature fluctuations – the new normal
December 2025 was also characterised by extreme temperature fluctuations and regional differences in the Arctic (Figure 4). In the course of the month, daily mean temperatures ranging from 2.5 to minus 16 degrees Celsius were recorded at the German-French research base AWIPEV on Spitsbergen. “In December, there were two periods in which the air temperatures were markedly above the long-term mean. Daily mean temperatures above zero were reached in both periods, especially just before Christmas. But if we look at the daily mean temperatures for all of 2025, we can see there had already been such warm periods in February. By now, they’re really nothing out of the ordinary,” explains Dr Marion Maturilli, a meteorologist at the Alfred Wegener Institute and head of the meteorological observatory at AWIPEV.
According to the records, the entire second half of 2025 was unusually warm on Spitsbergen. There was also more precipitation than usual. Especially during the warm periods, it came in the form of rain, not snow. At the beginning of 2026, in turn, it was significantly colder than usual.
Winter temperatures on Spitsbergen are now subject to considerable variability, depending on the atmospheric circulation. The most surprising aspect of the developments in December 2025 is most likely the fact that we researchers are no longer surprised by these extreme temperature fluctuations; we now consider them to be par for the course
Figure 4: Development of daily mean temperatures on Spitsbergen in December 2025 and the first week of January 2026. In the second week of December and right around the Christmas holidays, air temperatures reached or even passed the zero- degree mark. Graphic: Alfred Wegener Institute / Marion Maturilli
Figure 5: In the second half of the year, the monthly mean temperatures on Spitsbergen were clearly above the mean values in the reference period 2006 – 2025. Graphic: Alfred Wegener Institute / Marion Maturilli
Figure 6: Dr Marion Maturilli, head of the meteorological observatory at the German-French research base AWIPEV on Spitsbergen. Photo: Alfred Wegener Institute / Esther Horvath
Annual retrospective in the NOAA Arctic Report Card
The 20th annual NOAA Arctic Report Card, released online in December 2025 (Video 1), offers a comprehensive retrospective on sea-ice development in the Arctic. Once again, experts from the Sea Ice Portal – this time, Dr Stefan Hendricks and Dr Lars Kaleschke – contributed to the chapters on sea ice. The core messages in brief:
- The lowest winter maximum sea-ice extent in the Arctic in the 47 years of satellite observation was recorded in March 2025, followed by the tenth-lowest summer minimum sea-ice extent in September 2025.
- In the first-ever Arctic Report Card, 2005 was listed as the year with the lowest summer minimum sea-ice extent. Today, 2005 is the twentieth-lowest in the 47-year-long record.
- The past 20 years were characterised by lower sea-ice extents and by younger and thinner ice cover than the preceding decades.
- The changing sea-ice extent and thickness are opening the door for more shipping traffic and have prompted reassessments of national security issues.
Released annually, the NOAA Arctic Report Card summarises developments in the most important environmental parameters over the past twelve months in the Arctic. The deadline for submitting content is always the end of September; accordingly, data from October, November and December can only be reflected in the respective subsequent Report Card.
Video 1: Summary of the most important findings from 20 years of research and reporting from the Arctic – original (English) NOAA video to accompany the release of the 20th Arctic Report Card.
Russian tanker trapped in the sea ice
Despite the long-term decline in Arctic ice cover, pack ice remains a hard-to-predict risk for ships traversing Arctic waters – a lesson that the crew of the 293-metre-long Russian LNG tanker “Buran” had to learn first-hand in early December 2025. The tanker is one of four comprising Russia’s “shadow fleet”, which transports natural gas from the Arctic LNG2 transfer port in the Gulf of Ob (Kara Sea) to China.
As the ship was nearing the transfer terminal in the Gulf of Ob between 2 and 7 December, it became trapped in surprisingly thick sea ice near the coast, as reported in The Barents Observer. Four attempts to break free proved fruitless, even after a nuclear-powered icebreaker had attempted to clear a path for it.
The Buran and her sister ships are Ice Class 4, which means they’re not rated for use in ice-covered waters. According to the report, the tanker had previously sustained damage while crossing Arctic waters in the winter of 2024/2025, which necessitated months-long repairs at a Chinese shipyard.
The Antarctic: Trend of low sea-ice extents continues
In the Antarctic, the month of December is characterised by the most intensive seasonal melting. In the last month of 2025, more than half of the sea-ice extent melted away in just four weeks – from 12.11 million square kilometres on the first day of the month to 5.98 million square kilometres on the 31st. Our video shows which regions experienced the most substantial ice-cover loss (Video 2).
Video 2: Development of sea-ice concentration in the Antarctic from 1 to 31 December 2025.
Despite the rapid melting, enough pack ice remained for the sea-ice extent curve to be at the lower edge of the span of minima and maxima for the period 1981 – 2010. “With a mean value of 9.07 million square kilometres, December 2025, too, was markedly below the long-term trendline for the month. As such, these observations confirm the high variability in sea-ice development that we’ve seen in the Antarctic since 2016,” says Klaus Grosfeld (Figures 7 & 8).
Figure 7: Development of Antarctic sea-ice extent in comparison; the bright blue line is the curve for 2025. For the entire month of December, the curve was at the lower edge of the turquoise band, which indicates the span of minima and maxima for the period 1981 – 2010. For comparison, the dashed red line shows the development of sea-ice extent in the previous record-low summer minimum from 2022/2023. Screenshot: Sea Ice Portal
Figure 8: Time series of the mean December sea-ice extent in the Antarctic. The light blue line represents the long-term trend; the mean for December 2025, as the eighth-lowest, was once again below the trendline.
In particular, satellites detected less pack ice compared to the long-term mean in the western Ross Sea, west of the Antarctic Peninsula, in the eastern Weddell Sea, and in the entire Indian Ocean sector. Conversely, there was more ice in the eastern Ross Sea, to the northeast of the Antarctic Peninsula, far off the coast of Queen Maud Land (at the longitude of the Prime Meridian), and off the coast of Adélie Land (Figure 9).
“How the ice was distributed depended both on the long-term trend and on the weather conditions in December 2025. In the Bellingshausen Sea, for example, far too little sea ice formed during the winter months to relieve the situation in the last month of the year. In contrast, the ice distribution in the Weddell and Ross Seas appears to have been shaped by local winds, in some cases very intense ones,” explains Renate Treffeisen (Figure 10).
Figure 9: Difference in the mean position of the ice margin in December 2025, compared to the long-term mean for the years 2003 – 2014. Regions marked in blue had more Antarctic sea ice in the last month of 2025 than in the reference period; those marked in red had less.
Figure 10: Mean pressure anomalies at sea level in the Antarctic in December 2025, compared to the reference period 1971 – 2000. The arrows on the map indicate the wind’s direction; their length is proportional to the windspeed, shown in the map legend with a reference speed of 5 metres per second. Based on the map, it can be assumed that powerful winds especially arose in the Weddell Sea and off the coast of Adélie Land.
Surprisingly warm water masses below the sea ice of Atka Bay
The year 2025 especially went out with a bang over the Weddell Sea. In Atka Bay, the “home bay” of the German research station Neumayer III, the winds were sufficiently powerful to start breaking up the ice cover in the last days of December – not unusual at this time of year.
By then, all sea-ice measurements for the long-term research project Antarctic Fast Ice Network had already been gathered. Beforehand, sea-ice physicists Mara Neudert, Sacha van der Vleuten and Daria Paul had spent six weeks working on the fast ice of Atka Bay. During this time, they had expanded the monitoring programme for the long-term project by taking their readings at additional sites, ones that the overwintering team at Neumayer couldn’t cover in the winter months.
“Fast ice is of critical importance for the climate and ecosystems of the Antarctic. But climate models still don’t reflect it, and many of the processes that shape its energy and mass balance still aren’t fully understood. That’s why we measure the fast-ice thickness and collect samples at more than 20 stations in the summer. The stations are arrayed throughout the bay in the form of a giant ‘hashtag’ and allow us to identify local differences in ice properties,” Sacha van der Vleuten explains (Figure 16).
“These differences can arise, for instance, because the sea ice in the northern part of the bay is influenced by water masses of the open ocean, while below the ice in the southern part, the ‘ice-shelf water’ – that is, extremely cold meltwater from the underside of the Ekström Ice Shelf – is dominant. In addition, thick snow cover accumulates on the fast ice, which of course also affects the ice’s properties and potential thermal flows,” adds Daria Paul (Figures 11, 12 & 13).
Figure 11: In November and December 2025, Mara Neudert, Sacha van der Vleuten and Daria Paul (from l. to r.) continued the sea-ice thickness measurements gathered by the Neumayer overwintering team once a month from the end of June to the end of October. Photo: Alfred Wegener Institute / Lukas Weis
Figure 12: The researchers used Skidoos to move between the 26 measuring stations in Atka Bay. Gathering the data took them nearly six weeks. Photo: Alfred Wegener Institute / Lukas Weis
Figure 13: Snapshots of the fieldwork: Daria Paul took the snow measurements, a lengthy process (left). In contrast, the sea-ice thickness measurements were always a two-person job (centre). In the right-hand photo, Sacha van der Vleuten can be seen holding the new oceanographic probe, which, hanging from a long cable, was lowered into the water through a hole. Photos: Alfred Wegener Institute / Lukas Weis & Sacha van der Vleuten
In addition to the sea-ice thickness, this season the three experts also measured the temperature and salinity of the water masses below the sea ice. With the aid of a new CTD probe, which they lowered into the depths through holes in the sea ice, they were able complete these important measurements throughout the bay. “By now, we have a pretty solid grasp of what properties the water masses below the ice have. Their temperature is normally between minus 1.90 and minus 1.95 degrees Celsius, making it quite stable. So, we were all the more surprised when the probe suddenly started displaying water temperatures of minus 1.70 degrees Celsius on 12 December. From our overwintering team, we know that temperatures dropped again in the second half of the month. Therefore, we surmise that warm water masses unexpectedly flowed into Atka Bay in the first half of December. That being said, we still need to analyse the details,” says Sacha van der Vleuten. The analysis of this data and the long time series of many sea ice parameters in Atka Bay are a central focus of her doctoral thesis at the Alfred Wegener Institute.
And that wasn’t the only surprise in store for the researchers: unlike in past years, an exceptionally large number of icebergs were stuck in Atka Bay in November/December 2025 (Figures 14 & 15). “We don’t yet know why. What we can say: the icebergs gave the ice cover a certain degree of stability – there were far fewer leads in the fast ice than in past years,” Daria Paul relates.
Figure 14: For this aerial photo, Daria Paul sent up a drone far above Atka Bay. The ice-shelf edge and the various icebergs trapped in the fast ice can be clearly seen. Photo: Alfred Wegener Institute / Daria Paul
Figure 15: This map shows the four transects used for fast-ice readings in the summer and the positions of the 26 measuring stations in Atka Bay. The coloured dots indicate which types of readings were taken at the individual stations in November and December 2025. Map: Alfred Wegener Institute / Mara Neudert
Graphics