Greetings, readers! Our monthly sea ice analyses are returning just in time for the September equinox. In the Arctic, the sea ice extent this summer was tracking 2012 levels until August. While Siberia and the Canadian Archipelago experienced a very warm summer, the high-latitude Arctic Ocean was relatively cool. In the Antarctic, sea ice underwent an unusually large mid-winter decline over a week in August.
Overview of conditions
Arctic sea ice extent for August 2026 averaged 5.56 million square kilometers (2.15 million square miles), the seventh lowest in the satellite record (Figure 1a). This monthly average extent was 840,000 square kilometers (324,000 square miles) above the satellite record low set in 2012 (Figure 1b). Sea ice extent was far below average around most of the Arctic Ocean, with substantial open water particularly in the Beaufort, Chukchi, and East Siberian Seas. The Beaufort and Chukchi Seas experienced rapid change over the month. In early August, the ice edge in the Beaufort and Chukchi Seas was near average, just off the north coast of Alaska. Then, during the first week of August, an open water region developed north of Utqiaġvik at about 75° N latitude. The open water region rapidly expanded, eventually becoming a large embayment in the ice edge. By the end of August, much of the ice in the Beaufort Sea had melted out. On the Atlantic side, relatively low sea ice concentrations were the rule, extending all the way to the pole (Figure 1c).
Figure 1a. Arctic sea ice extent for August 2026 was 5.56 million square kilometers (2.15 million square miles). The magenta line shows the 1981 to 2010 average extent for that month. Sea Ice Index data. About the data — Credit: National Snow and Ice Data Center
Figure 1b. This graph shows Arctic sea ice extent as of September 2, 2026, along with daily ice extent data for four previous years and the record low year. 2026 is shown in blue, 2025 in green, 2024 in orange, 2023 in brown, 2022 in magenta, and 2012 in dashed brown. The 1981 to 2010 median is in dark gray. The gray areas around the median line show the interquartile and interdecile ranges of the data. Sea Ice Index data. — Credit: National Snow and Ice Data Center
Conditions in context
In spring and early summer, low pressure and relatively cool weather persisted over the Arctic Ocean. Average temperatures from May through July at the 925 millibar level (about 2,500 feet above the surface) were up to 2 degrees Celsius (4 degrees Fahrenheit) below the 1981 to 2026 average (Figure 2a). Persistent cold-cored cyclones kept temperatures low, resulting in a pronounced low pressure cell centered over the pole in the three-months average (Figure 2b). Cloudy conditions led to less solar insolation at the surface. This, combined with low temperature, led to late melt onset in the Beaufort and Chukchi Seas (Figure 2c) (Evident in preliminary data derived from passive microwave sensors, provided by Angela Bliss at the NASA Goddard Space Flight Center). Melt began nearly a month later than the 1981 to 2010 average in the southern Chukchi Sea. On the Atlantic side, melt onset was earlier than average or near average. These strong regional contrasts guided the early summer progression of ice loss.
Figure 2a. This plot shows the departure from average air temperature in the Arctic at the 925 hectopascal, in degrees Celsius, for May through July 2026, relative to the 1981 to 2010 average. Yellows to reds indicate above average temperatures; greens to purples indicate below average temperatures. Based on data set ERA5 (European Centre for Medium-Range Weather Forecasts Reanalysis 5), DOI: 10.24381/cds.adbb2d47. — Credit: NSIDC courtesy NOAA Earth System Research Laboratory Physical Sciences Laboratory
Figure 2b. This plot shows average sea level pressure in the Arctic in hectopascal for May through July 2026. Yellows to reds indicate higher air pressure; greens to purples indicate lower pressure. Based on data set ERA5 (European Centre for Medium-Range Weather Forecasts Reanalysis 5), DOI: 10.24381/cds.adbb2d47. — Credit: NSIDC courtesy NOAA Earth System Research Laboratory Physical Sciences Laboratory
August 2026 compared to previous years
The downward linear trend in Arctic sea ice extent through 2026 for August is 69,100 square kilometers (26,700 square miles) per year or 9.6 percent per decade relative to the 1981 to 2010 average (Figure 3). Based on the linear trend, since 1979, August has lost 3.24 million square kilometers (1.25 million square miles) of sea ice. This is equivalent to about two times the size of Alaska.
Ship report from the pole
NSIDC scientist Matt Shupe and other researchers cruised to the pole on the Swedish icebreaker Oden, leaving Svalbard on August 10. Shupe noted the large amount of open water en-route seen in the satellite imagery: “[There was] a remarkable amount of open water, vast expanses sometimes large enough that I could not see the other side of them. These areas of open water extended pretty much the whole way to the pole." He also noted the absence of melt ponds, in contrast to recent years. Despite the open water, he noted that melt ponds had frozen over and were covered by snow (Figure 4). Data and reports from ships and field observations provide valuable detailed information and local context that complements satellite data.
Sea ice down south
In the Antarctic, sea ice is expanding and will reach its annual maximum extent in the coming weeks. This year, extent has been below the 1981 to 2010 average, as has been the case nearly every month since late 2016; however, it has been above the record low levels of 2023 and 2024. The August average extent was 16.46 million square kilometer (6.36 million square miles), 1.26 million square kilometers (486,000 square miles) below the 1981 to 2010 average. Earlier in the year, as late as May, ice extent was extremely low in the Bellingshausen Sea. This area has now largely recovered, but the adjacent Amundsen Sea has low sea ice extent, along with much of the East Antarctic region.
The most notable feature of the austral growth season was a substantial decline in extent over a period of a week in early August. From August 8 to August 14, extent declined 310,000 square kilometers (120,000 square miles). Pauses in growth and short periods of decline are not unusual because of shifting weather patterns that can melt ice near the edge and/or push the ice edge poleward. This is especially true as the sea ice edge reaches the stormy northern latitudes of the Southern Ocean.
Figure 5a. Antarctic sea ice extent for August 2026 was 16.46 million square kilometers (6.36 million square miles). The magenta line shows the 1981 to 2010 average extent for that month. Sea Ice Index data. About the data — Credit: National Snow and Ice Data Center
Figure 5b. This graph shows Antarctic sea ice extent as of September 2, 2026, along with daily ice extent data for four previous years and 2014, the record high year. 2026 is shown in blue, 2025 in green, 2024 n orange, 2023 in brown, 2022 in magenta, and 2014 in dashed brown. The 1981 to 2010 median is in dark gray. The gray areas around the median line show the interquartile and interdecile ranges of the data. Sea Ice Index data. — Credit: National Snow and Ice Data Center
Greenland Ice Sheet
While the other analysis site Ice Sheets Today remains in hiatus, we will include some updates from other sources as we receive them. The Greenland melt season has not been exceptional, but overall melt is similar to 2011 and 2016. The surface mass balance, calculated by the total snow and rainfall minus runoff and evaporation, stands at about -150 billion tons. Low snowfall last winter in several key exposed surface areas led to rapid runoff beginning in mid-July as the previous winter’s snow cover melted off. A few warm events spanned the southern dome of Greenland, but the more central Summit area did not see temperatures exceed -3 degrees Celsius (27 degrees Fahrenheit).
Acknowledgement
We gratefully acknowledge new funding from the Climate Science and Communications Fund. The Fund represents a collaborative group of donors and scientists dedicated to supporting climate scientists and communicators and is administered by Surge Climate Talent. The views expressed in these analysis posts are those of the NSIDC science team and do not necessarily reflect the views of Surge Climate Talent.