- The Arctic: In November, there was unusually little sea ice in the Barents Sea, western Kara Sea, Hudson Bay and, to the west of Greenland extending high into the Canadian Arctic Archipelago.
- Atmospheric research: During the CONTRASTS expedition, for the first time an AWI team set up sensor towers that operated autonomously for weeks at a time – an experiment that yielded great successes and an unexpected surprise.
- The Antarctic: The sea-ice extent remained low in November. Far off the coast of East Antarctica, there are clear signs of an ocean heatwave.
The Arctic: Focus on the Barents and Kara Seas
The Sea Ice Portal’s ice-concentration maps, available online, extend back to 2002. As such, the 30th of November 2002 is the oldest last day of November for which we can show the sea-ice cover in the Arctic. Let’s take a closer look at the sea-ice situation 23 years ago in the European part of the Arctic (Figure 1).
Back then, the northern Barents Sea was mostly frozen over; the east coast of Svalbard and the Russian islands Nowaja Semlja and Franz Josef Land were surrounded by sea ice. The sea-ice cover on the Kara Sea was virtually seamless, except for a few small areas along the Siberian coast. All in all, this marginal region of the Arctic Oceans was covered in sea ice – just like it always was at this time of year, we would have said back then.
Figure 1: Map of the sea-ice concentration on the Arctic Ocean on 30 November 2002 (left) and on 30 November 2025 (right). The yellow circle indicates the northern Barents Sea and western Kara Sea.
Figure 2: Map of mean sea-ice concentration in the Barents Sea / Kara Sea region in November of selected years. The annual comparison shows the steadily declining sea-ice concentration in the region since 2005 and the increasing presence of extensive ice-free areas since 2015.
Back to the year 2025. The sea-ice conditions on 30 November 2025 could hardly be more different (Figure 1). The northern Barents Sea and the waters surrounding Svalbard are now ice-free. Scattered pack ice drifts off the coast of Franz Josef Land. Except for some windblown pack ice off its east coast, the waters off the island Nowaja Semlja are also ice-free. The same is true for the entire western part of the Kara Sea. There’s no sea ice – on the last day of November!
“The northern Barents Sea and the neighbouring Kara Sea are among those marginal zones of the Arctic Ocean where we can most clearly see the progression of climate change and resulting sea-ice retreat. The winter sea-ice cover in the region has been declining for two decades, making the large ice-free areas no surprise. But the direct comparison with the sea-ice conditions on 30 November 2002 underscores once again the scope of sea-ice changes,” says Dr Renate Treffeisen, an atmospheric researcher at the Alfred Wegener Institute and co-initiator of the Sea Ice Portal.
Far too much heat in the atmosphere and ocean
To understand the causes of this drastic sea-ice retreat, all it takes is a look at the air and sea-surface temperatures in the northern Barents Sea and the Kara Sea.
In the course of November 2025, the surface waters in both marginal seas were up to 3.5 degrees Celsius warmer than in the reference period, as the data from selective dates shows (Figure 3). For anyone interested in tracking the development over the month in time-lapse form, we recommend checking out the preliminary data from NOAA’s Climate Reanalyzer. However, NOAA uses a different reference period than the Sea Ice Portal. We calculate our sea-surface temperature anomalies in comparison to the long-term mean for the years 1971 to 2000, whereas NOAA uses the years 1991 to 2020.
Figure 3: Sea-surface temperature anomalies on 9 November 2025 compared to the long-term November mean for the reference period 1971 – 2000. On this November day, the temperature anomalies in the Barents Sea and western Kara Sea were especially pronounced. As the colours indicate, many parts of the ocean’s surface were 2 to 3.5 degrees Celsius too warm.
Figure 4: Mean monthly sea-surface temperature anomalies in the Arctic in November 2025, compared to the monthly mean values for the reference period 1971 – 2000. As the data shows, surface temperatures were up to 3 degrees Celsius warmer than in the reference period in the waters northeast of Svalbard. The combination of this warmth and high air temperatures slowed surface cooling, and with it, the formation of new sea ice.
Air temperatures were also far too high, especially in the first week of November. The air-temperature data we apply was gathered at an altitude of ca. 764 metres above sea level. Here, satellites recorded anomalies of up to 10 degrees Celsius on 15 November 2025 (Figure 5). In terms of monthly mean values, the air over the Barents and Kara Seas was up to 4.5 degrees Celsius warmer than in the reference period 1971 to 2000 (Figure 6).
“In November, we saw the continuation of a trend that was first hinted at in summer. Both the ocean and atmosphere in this part of the Arctic have been far too warm for months,” says Dr Klaus Grosfeld, co-initiator of the Sea Ice Portal, with regard to the latest satellite readings.
Figure 5: Air temperature anomalies on 15 November 2025, compared to the long-term November mean for the reference period 1971 – 2000. Over the Kara Sea, the air was up to 10 degrees Celsius warmer than in the reference period at a mean altitude of 925 hPa (ca. 764 metres above the surface).
Figure 6: Mean air temperature anomalies in November 2025, compared to the long-term November mean for the reference period 1971 – 2000. Over the Barents Sea / Kara Sea region, the air was up to 4.5 degrees Celsius warmer than in the reference period at a mean altitude of 925 hPa (ca. 764 metres above the surface). The temperature anomalies were even more pronounced over Greenland and northeast Canada.
Second-lowest November sea-ice extent
The combination of excessively warm surface waters and unusually high air temperatures didn’t just slow the formation of new ice in the Barents and Kara Seas; a similar trend could be observed in the sea-ice region west of Greenland, in Hudson Bay, and somewhat farther north in the Canadian Arctic Archipelago (Figures 5 and 6). In terms of monthly means, all three regions had substantially less sea ice than the long-term mean for the years 2003 to 2014 (Figure 7).
There was only more pack ice than the long-term mean in November 2025 in the Chukchi Sea, which was most likely due to the prevailing wind direction: here, winds chiefly blew to the south. Therefore, the pack ice in the Chukchi Sea was probably scattered by the winds.
Sea-ice retreat in four regions of the Atlantic-influenced part of the Arctic, together with only minor growth in the Pacific-influenced part, yielded a monthly mean sea-ice extent of 9.1 million square kilometres – the second-lowest mean value for November since the beginning of satellite observation; only in November 2016 (Figure 8) was there less ice. Yet back then, the sea-ice extent rapidly rebounded later. It remains to be seen whether the pattern will be repeated in December 2025 (Figure 9).
Figure 7: Difference in the mean position of the ice margin in November 2025, compared to the long-term mean for the years 2003 – 2014. Regions marked in blue had more Arctic sea ice in November 2025 than in the reference period; those marked in red had less. The latter was particularly the case in Hudson Bay, west of Greenland, the northern Barents Sea and in Kara Sea.
Figure 8: Time series of the mean November sea-ice extent in the Arctic. The light blue line represents the long-term trend; the mean for November 2025 remained well below it.
Figure 9: Development of Arctic sea-ice extent in comparison; the blue line is the curve for 2025. For the entirety of November, the curve was below the turquoise band, which indicates the span of minima and maxima in the period 1981 – 2010. The solid red line represents the sea-ice extent development in the winter of 2016/17. Clearly recognisable: the rapid increase in sea-ice extent in the second half of November and in December 2016.
Atmospheric research during CONTRASTS: Equipped with drones, fuel cells, and wooden blocks
The atmospheric researchers from the AWI Potsdam had already joined in various previous Polarstern expeditions. Whenever the ship dropped anchor at a floe, they would set up their monitoring towers for air temperature, moisture and radiation, or launched weather balloons. But, until the CONTRASTS expedition in the late summer of 2025, Dr Sandro Dahlke and his team never would have dreamed of leaving their instruments behind on a floe for weeks at a time.
Figure 10: Atmospheric researcher Dr Sandro Dahlke from the AWI Potsdam led the atmospheric fieldwork during the CONTRASTS expedition. Photo: Alfred Wegener Institute / Mario Hoppmann
“For the first time, the CONTRASTS expedition gave us the chance to accompany the summer melting on three highly diverse floes – assuming our instruments could continue taking readings autonomously for several weeks. It was clear to all of us that, unlike on other expeditions, we wouldn’t be able to venture onto the ice every day to check the latest data, straighten any slumping towers, or change out batteries,” Sandro Dahlke explains.
Therefore, in early summer and under immense time pressure, the physicists began developing their autonomous “Met-City.” The system consisted of a 10-metre-tall sensor tower and a smaller counterpart. The larger of the two was equipped with temperature and moisture sensors 2, 6 and 10 metres above the ice; the smaller of the two formed the Surface Energy Balance Station (Video 1).
According to Sandro Dahlke: “We used it to measure incoming solar radiation, reflection from the ice’s surface, thermal radiation and downward radiation, and sensible and latent thermal flows over the ice. We need these parameters to calculate the surface energy balance, since they’re what’s collectively responsible for the sea ice melting in summer.”
Video 1: Drone flight over the monitoring station that Sandro Dahlke and his team set up on the third floe, which was characterised by particularly old and thick sea ice. The 10-metre-tall tower for measuring air temperature and moisture and the smaller one for various radiation parameters can both be seen. Video: Alfred Wegener Institute / Sandro Dahlke
Figure 11: Setting up the Met-City on Floe 1. Photo: Alfred Wegener Institute / Sandro Dahlke
Figure 12: How can you anchor tethers on a surface that’s melting? Sandro Dahlke and his team tried using wooden blocks, which they inserted under the ice through drilled holes and then turned them perpendicular to the hole, allowing the tether to be drawn taut. Due to the melting ice, this solution couldn’t last forever either, but worked far better than conventional ice stakes. Photos: Alfred Wegener Institute / Sandro Dahlke
Each Met-City was supplied with power from special-purpose batteries and a fuel cell. To anchor the tethers for the sensor towers, the team had brought with them long wooden blocks. They attached each block to the end of the tether, inserted it below the sea ice through a drilled hole and then turned it so it was perpendicular to the hole. “Unfortunately, it wasn’t a perfect solution, but we knew that trying to hammer stakes or nails into the ice of the melting floes wouldn’t get us anywhere,” adds Sandro Dahlke.
In retrospect, he sees potential for improvement when it comes to gathering data: “Due to the lack of prep time, we had to save our monitoring data on the SD cards inside our measuring instruments. Unfortunately, we couldn’t set up a data transfer by satellite,” the AWI expert recalls. Consequently, the atmospheric researchers had to look on helplessly, from a hundred miles away, as the pack ice under one of their stations suddenly collapsed, taking both the instruments and the latest data down with it.
“At the third station, home to especially old and thick ice, we had set up our towers on 2.7-metre-thick pack ice. We were all convinced the old ice would survive the summer. But just one day after Polarstern had weighed anchor, the flow broke into several pieces, and a few days later, our two towers were miles apart,” Sandro Dahlke reports.
Figure 13: During their third visit to Floe 2, the ice edge had retreated so close to the monitoring towers that the experts were forced to dismantle them and set them up elsewhere – physically demanding work. Photo: Alfred Wegener Institute / Phillip Eisenhuth
Figure 14: Snow makes the difference. Once it had snowed on Floe 1, its surface albedo rose rapidly. Anyone who sees the pictures from 12 and 25 July side by side will understand why. Photos: Alfred Wegener Institute / Sandro Dahlke
For the first time, we monitored radiation and thermal flows during the summer melting over a period of several weeks – and simultaneously in multiple regions. Thanks to the various cameras on the sea ice, we can precisely compare our monitoring data with surface images, minute by minute. We can see, for instance, that even light snowfall is enough to rapidly change the ice’s reflectivity (albedo), while rain turns the ice’s surface blue, seriously reducing its reflectivity
In addition, he’s very satisfied with the images and readings taken by drones up to 1,000 metres above the surface. “We were able to complete several monitoring flights and now have both the data and the simultaneously made video recordings, which is great,” he enthuses. He’s selected some of the best images and videos for the Sea Ice Portal (Figures 15, 16 & Video 2).
Figure 15: Photo or painting? Drone image of the icebreaker Polarstern, moored at Floe 3 for research fieldwork. Photo: Alfred Wegener Institute / Sandro Dahlke
Figure 16: Drone image of the German research icebreaker in the pack ice of the northern Fram Strait. Old, thick sea ice drifts in the foreground. Photo: Alfred Wegener Institute / Sandro Dahlke
Video 2: Flight over the third floe and adjacent pack ice under excellent weather conditions. A beautiful and inspiring sight. Video: Alfred Wegener Institute / Sandro Dahlke
The Antarctic: Trend of low sea-ice extents continues
In November, the sea-ice extent in the Antarctic declined much faster than the long-term mean. Whereas the curve was only just below the span of minima and maxima in the period 1981 – 2010 in October, it dropped markedly lower in November (Figure 17).
At the beginning of the month, the sea-ice extent was still 16.19 million square kilometres. By the last day of November, it had fallen to 12.31 million square kilometres – a loss of more than 3.88 million square kilometres, or an area nearly eleven times the size of Germany. The monthly mean value of 14.52 million square kilometres was the fourth-lowest November mean since the beginning of satellite observation (Figure 18).
Especially striking: the drastic sea-ice loss off the coast of Enderby Land and MacKenzie Bay on the Amery Ice Shelf (30 – 90 degrees East) (Figure 19). Both regions are influenced by the Indian Ocean and were home to both high air temperatures and unusually high sea-surface temperatures in November. On some days, it was up to 10 degrees Celsius warmer than in the reference period 1991 to 2020 over Enderby Land, as NOAA air-temperature data from the Climate Reanalyzer shows. At the same time, the Southern Ocean in this part of the Antarctic was unusually warm. Our map of mean monthly sea-surface temperature anomalies shows individual areas with deviations of up to 2 degrees Celsius (Figures 20 and 21).
On the basis of the available data, it can’t currently be said to what extent these two heat signals drove sea-ice melting in the region. Here, local interactions between the sea ice, ocean and atmosphere would need to be investigated in more detail. However, there are several indications that the high air and water temperatures played an important part.
Ozone hole and mysterious patterns on the seafloor: Good news from the Antarctic
In the past few weeks, there’s been nothing but good news from the Antarctic. First of all, this year’s ozone hole over the Southern Hemisphere was thankfully small and closed again much faster than in the six previous years. As the European climate service Copernicus reports, this gives grounds for optimism that Earth’s ozone layer can recover completely. In the years 2020 to 2023, a large and prolonged ozone hole had formed.
British polar researchers have now solved a riddle that had puzzled them since an expedition to the western Weddell Sea in 2019. During said expedition, they had discovered marked indentations in the seafloor in the region of the former Larsen C Ice Shelf. They appeared in six different patterns – e.g. as lines, crescents or U-shaped (Figure 22).
Six years later, they have now published the explanation: the depressions in the seafloor were nesting sites for the yellowfin notie (Lindbergichthys nudifrons), an up to 19-centimetre-long orange-brown Antarctic fish known for males that defend their nests.
Figure 22: The experts identified six patterns for the nesting sites (from top left to bottom right): cluster, crescent, line, oval, sharp U, and singular. Collage: Connelly RB, Woodall LC, Rogers AD and Taylor ML (2025) A finding of maintained cryonotothenioid nesting sites in the Western Weddell Sea. Front. Mar. Sci. 12:1648168. doi: 10.3389/fmars.2025.1648168, CC-BY
The fact that the fish make their nesting sites in certain patterns would appear to be a communal measure to protect them from potential predators like brittle stars and priapulids, the researchers surmise. As such, every male defends not just his own nest, but to some extent those of his neighbours. The researchers also found stones next to some nests, which would certainly offer protection on the otherwise bare seafloor.
The discovery of these sites in the Weddell Sea is not only of scientific importance. Such studies are also relevant when the Commission for the Conservation of Antarctic Marine Living Resources (CCAMLR) deliberates on the proposed Marine Protected Area for the Weddell Sea. Accordingly, the experts have provided a further argument for protecting the Sea and its fauna.
Thanks and happy holidays!
The sea-ice update for November 2025 will be our last one for this year. Thank you for your interest and we hope you’ll stay with us in the new year. Until then, we wish you and yours a peaceful Advent season and happy holidays. We at the Sea Ice Portal will be back in January 2026 – when we’ll review the sea-ice conditions in December 2025.
Graphics