Broken icebergs in the ocean

Arctic and Antarctic end-of-season report – October 2020


  • Sea ice extent for September 2020 was 3.85 million square km, the second lowest on record (after 2012).
  • The Arctic in 2020 experienced its warmest summer on record. It was also more cyclonic (stormier) than average.
  • There was extreme variation in the rate of ice loss during summer 2020, with two periods of very rapid ice loss separated by a period of very slow ice loss.
  • Predictions of September sea ice extent tended to be on the high side, although some came close to the eventual extent.
  • Antarctic sea ice has likely recorded its seasonal maximum of 18.95 million square km on 27th September, which would represent the 12th highest on record.

September Arctic sea ice extent

Sea ice extent for the month of September 2020 was 3.85 million square km according to the HadISST1.2 dataset (Rayner et al., 2003). This was the second lowest September extent observed in the satellite era (since 1979), after 3.56 million square km in 2012. The 2020 value was 2.54 million square km below the 1981-2010 average, and 0.38 million square km below the long-term linear trend (Figure 1). Extent was exceptionally low in all regions of the Arctic except the Central Arctic and in the Beaufort Sea, where a long ‘arm’ of ice extended along the climatological southern limit of the ice cover (Figure 2). Extent would have been very close to that seen in 2012 (albeit still slightly higher) if this Beaufort Sea ‘arm’ were not present.

Chart, line chart Description automatically generated

Figure 1. September Arctic sea ice extent during the satellite era, according to HadISST1.2, with linear trend indicated.

Map Description automatically generated

Figure 2. Arctic sea ice extent in September 2020, compared to the record low year of 2012 and the 1981-2010 average, with regions named in the text labelled. Data are from HadISST1.2.

Melt season review

Summer 2020 was exceptionally warm in the Arctic, even more so than in 2019 which was itself record-breaking. As measured by the NCEP reanalysis (Kalnay et al., 1996), the May-August average 925 hPa temperature north of 70°N was the warmest on record (since 1948, Figure 3). Temperatures were warm Arctic-wide, but especially in the Central Arctic and in the Laptev and Kara Seas north of Siberia (Figure 4a). The most notable feature of the summer circulation was a high-pressure ridge over Siberia, most prominent from mid-June to mid-July. As well as circulating very warm air into the Arctic Ocean itself, this caused a notable heatwave in Eastern Siberia, with a provisional new record temperature of 38°C observed in Verkhoyansk. There was also a tendency for the weather to be rather stormy and cyclonic in the central Arctic itself (Figure 4b), particularly early in the season.

Chart Description automatically generated

Figure 3. Average May-August temperature anomaly (°C), relative to the 1981-2010 average at 925 hPa over the Arctic Ocean region, for the years 1948-2019. Data are from the NCEP reanalysis.

Diagram Description automatically generated

Figure 4. Anomaly in (a) air temperature (°C) at 925 hPa (about 1km above mean sea level); (b) mean sea level pressure (hPa) for May-August 2020, relative to the 1981-2010 average. Data are from the NCEP reanalysis.

The summer was characterised by dramatic variation in the rate of sea ice melting (Figure 5). The period from 26th June – 16th July saw the highest 3-week rate of ice melting observed in the satellite era. This was largely caused by movement of warm air over the ice due to the Siberian high-pressure ridge. In early July the high pressure extended over most of the Arctic Ocean, blocking the passage of Arctic cyclones that are thought to be help prevent severe sea ice melt (e.g. Schreiber et al., 2020). Most of the ice loss in this period occurred in the Siberian coastal seas, while the Beaufort Sea north of Alaska experienced mainly northerly winds and displayed little ice loss.

Figure 5. 3-week running mean sea ice extent change for 2020 (black), compared to previous years from 1979 (blue) and the 1981-2010 average (dashed).

By contrast, from 22nd July – 11th August the rate of ice loss was unusually slow, the slowest ice loss observed in the satellite era for this time of year. This was associated with a shift in the main high-pressure ridge towards the Atlantic, allowing cyclones to form over the Central and Pacific Arctic, and cooler air from Greenland and the Canadian Arctic to flow over most of the Arctic Ocean.

In mid-August the high-pressure ridge shifted back towards Siberia and the Central Arctic, and severe late-season sea ice melt followed: the rate of ice loss from 16th August – 5th September was the second-fastest observed in the satellite era for this time of year (in addition the rate of ice loss from 31st August – 5th September was the fastest observed for this period). In a change from the earlier period of fast melting, the ice loss was Arctic-wide, as the circulation pattern blew warm temperatures over almost the entire Arctic.

Assessment of September 2020 Arctic sea ice extent predictions

During the summer, the Sea Ice Prediction Network (SIPN) collates predictions of September Arctic sea ice extent from scientific centres around the world and publishes these predictions in a series of Sea Ice Outlooks (SIO), in June, July and August. In 2020, the observed September extent fell towards the lower end of the ranges of predictions for all three outlooks (Figure 6). Notably, while most predictions tended to converge as the season progressed (with an interquartile range of 4.19-4.40 comparing to 4.08 to 4.59 in June), they converged to a value somewhat higher than the eventual September extent. This may be related to the extreme variation in melt rate observed this summer, in particular the unusual increase in melting that took place in mid-August.

Chart, box and whisker chart Description automatically generated

Figure 6. Arctic sea ice extent for September 2020 according to HadISST1.2 (red line), compared to an extrapolation from the long-term linear trend (left) and all the predictions submitted to the three SIO reports (maximum, minimum, median and quartiles from each distribution shown).

Antarctic sea ice extent

The highest Antarctic sea ice extent recorded so far this year was 18.95 million square km on 27th September, since when extent has fallen slightly (Figure 7). If not subsequently exceeded, this would represent the 12th highest maximum sea ice extent on record (since 1979), 0.25 million square km above the 1981-2010 average.

Chart, histogram Description automatically generated

Figure 7. Antarctic sea ice extent for 2020 and other recent years, plus the record low year of 1986, compared to the 1981-2010 average, with 1 and 2 standard deviation intervals indicated. Data are from NSIDC.

Although Antarctic sea ice extent has been at considerably higher levels during 2020 than in many recent years, it remains well below the levels of 2012-2014, when three successive record high winter maxima were observed. In September 2020, below-average extent was largely confined to parts of the Ross Sea and the Indian Ocean sector, with all other regions seeing above-average extent (Figure 8).

Chart, radar chart Description automatically generated

Figure 8. Antarctic sea ice coverage on 28th September 2020, with average September sea ice extent from 1981-2010 indicated in orange and regions referred to in the text labelled. Underlying figure courtesy of NSIDC.


Kalnay, E., and Coauthors, 1996: The NCEP/NCAR 40-Year Reanalysis Project. Bull. Amer. Meteor. Soc., 77, 437–472, doi: 10.1175/1520-0477(1996)077%3C0437:TNYRP%3E2.0.CO;2

Rayner, N. A.; Parker, D. E.; Horton, E. B.; Folland, C. K.; Alexander, L. V.; Rowell, D. P.; Kent, E. C.; Kaplan, A., 2003: Global analyses of sea surface temperature, sea ice, and night marine air temperature since the late nineteenth century. J. Geophys. Res. 108, D14, 4407, doi: 10.1029/2002JD002670

Schreiber, E. A. P. and Serreze, M. C., 2020: Impacts of synoptic-scale cyclones on Arctic sea ice concentration: a systematic analysis. Ann. Glaciol., 1–15. doi:10.1017/aog.2020.23