The annual competition between ice and water in the Arctic ocean has reached the maximum for water, which typically occurs mid September.  After that, diminishing energy from the slowly setting sun allows oceanic cooling causing ice to regenerate. Those interested in the dynamics of Arctic sea ice can read numerous posts here.  This post provides a look at mid September from 2007 to yesterday as a context for understanding this year’s annual minimum.

The image above shows Arctic ice extents on day 260 (lowest annual daily extent on average) from 2007 to 2024 yesterday.  Obviously, the regions vary as locations for ice, discussed in more detail later on. The animation shows the ice deficits in years 2007, 2012, 2016, and 2020, as well as surplus years like 2010, 2014, 2022 and 2024.

Note that for climate purposes the annual minimum is measured by the September monthly average ice extent, since the daily extents vary and will go briefly lowest on or about day 260. In a typical year the overall ice extent will end September slightly higher than at the beginning.  2024 September ice extent averaged 4.6M over the first 16 days, and is likely to end the month with at least that amount for the entire month. For comparison, the 17 year average for Sept. 1-16 is 4.7M.

The melting season to mid September shows 2024 tracked lower than average but ended the period slightly above.

The graph above shows September daily ice extents for 2024 compared to 18 year averages, and some years of note. Day 260 has been the lowest daily ice extent on average for the last 18 years.

The black line shows on average Arctic ice extents during September decline 358k km2 down to 4.5M Km2 by day 260. The average increase from now on is 490k km2 up to 5.0M km2 end of September.  2024  tracked a little lower than the 18-year average in the second week reaching a low of 4.49M km2 on day 255, before going above average on day 260.

SII was reporting deficits as high as 0.5M km2 (half a Wadham) compared to  MASIE early in September.  For some reason, apparently data access issues, that dataset has not been updated for the last five days.  2023 bottomed out at 4.1M while 2007 daily minimum hit 4.0M, ended ~ 0.5M km2 in deficit to average and 535k km2 less than MASIE on day 260.  2020 ice on day 260 was ~740k km2 in deficit to average.

The main deficit to average is in CAA with a smaller loss in Chukchi, overcome by surpluses almost everywhere, especially in Central Arctic along with Laptev and Greenland seas. And as discussed below, the marginal basins have little ice left to lose.

The Bigger Picture 

We are close to the annual Arctic ice extent minimum, which typically occurs on or about day 260 (mid September). Some take any year’s slightly lower minimum as proof that Arctic ice is dying, but the image above shows the Arctic heart is beating clear and strong.

Over this decade, the Arctic ice minimum has not declined, but since 2007 looks like fluctuations around a plateau. By mid-September, all the peripheral seas have turned to water, and the residual ice shows up in a few places. The table below indicates where we can expect to find ice this September. Numbers are area units of Mkm2 (millions of square kilometers).

The table includes some early years of note along with the last 4 years compared to the 17 year average for five contiguous arctic regions. BCE (Beaufort, Chukchi and East Siberian) on the Asian side are quite variable as the largest source of ice other than the Central Arctic itself.   Greenland Sea and CAA (Canadian Arctic Archipelago) together hold almost 0.5M km2 of ice at annual minimum, fairly consistently.  LKB are the European seas of Laptev, Kara and Barents, a smaller source of ice, but a difference maker some years, as Laptev was in 2016 and 2023.  Baffin and Hudson Bays are inconsequential as of day 260.

2024 extent of 4.58 is 1.3% over average, mainly due to surpluses in Chukchi and East Siberian seas.

For context, note that the average maximum has been 15M, so on average the extent shrinks to 30% of the March high (31% in 2022) before growing back the following winter.  In this context, it is foolhardy to project any summer minimum forward to proclaim the end of Arctic ice.

Resources:  Climate Compilation II Arctic Sea Ice

Figure 1. Distribution of SIO contributors for August estimates of September 2024 pan-Arctic sea-ice extent. No Heuristic methods were submitted in August. “Sun” is a public/citizen contribution. Image courtesy of Matthew Fisher, NSIDC.

2024: August Report from Sea Ice Prediction Network

The August 2024 Outlook received 24 pan-Arctic contributions (Figure 1). This year’s median
forecasted value for pan-Arctic September sea-ice extent is 4.27 million square kilometers with
an interquartile range of 4.11 to 4.54 million square kilometers. This is lower than the 2022 (4.83
million square kilometers) and 2023 (4.60 million square kilometers) August median forecasts
for September. . .This reflects relatively rapid ice loss during the month of July, resulting in August
Outlooks revising estimates downward. The lowest sea-ice extent forecast is 3.71 million square
kilometers, from the RASM@NPS submission); the highest sea-ice extent forecast is 5.23
million square kilometers, submitted by BCCR.

These are predictions for the September 2024 monthly average ice extent as reported by NOAA Sea Ice Index (SII). This post provides a look at the 2024 Year To Date (YTD) based on monthly averages comparing MASIE and SII datasets. (18 year average is 2006 to 2023 inclusive).

The graph puts 2024 into recent historical perspective. Note how 2024 was slightly above the 18-year average for the first 5 months, then tracked slightly lower to average through August. The outlier 2012 provided the highest March maximum as well as the lowest September minimum, coinciding with the Great Arctic Cyclone that year.  2007 began the period with the lowest minimum except for 2012.  SII 2024 started slightly higher than MASIE the first 3 months, then ran the same as MASIE until dropping in August 400k km2 below MASIE 2024 and also lower than 2007 and 2012.

The table below provides the monthly Arctic ice extent averages for comparisons (all are M km2)

The first two data columns are the 2024 YTD shown by MASIE and SII, with the MASIE surpluses in column three.  Column four shows MASIE 2024 compared to MASIE 18 year averages, while column five shows SII 2024 compared to SII 18 year averages.  YTD August MASIE and SII are below their averages, SII by nearly half a Wadham. The last column shows MASIE 2024 holding surpluses over 2007 most of the months, and nearly the same in August.

Summary

The experts involved in SIPN are expecting SII 2024 September to be much lower than 2023 and 2022, based largely on the large deficits SII is showing in July and August. The way MASIE is going, this September looks to be lower than its average, but much higher than SII.  While the daily minimum for the year occurs mid September, ice extent on September 30 is typically slightly higher than on September 1.

Footnote:

Some people unhappy with the higher amounts of ice extent shown by MASIE continue to claim that Sea Ice Index is the only dataset that can be used. This is false in fact and in logic. Why should anyone accept that the highest quality picture of ice day to day has no shelf life, that one year’s charts can not be compared with another year? Researchers do this, including Walt Meier in charge of Sea Ice Index. That said, I understand his interest in directing people to use his product rather than one he does not control. As I have said before:

MASIE is rigorous, reliable, serves as calibration for satellite products, and continues the long and honorable tradition of naval ice charting using modern technologies. More on this at my post Support MASIE Arctic Ice Dataset

MASIE: “high-resolution, accurate charts of ice conditions”
Walt Meier, NSIDC, October 2015 article in Annals of Glaciology.

The images above come from AARI (Arctic and Antarctic Research Institute) St. Petersburg, Russia. Note how the location of remaining ice at late August varies greatly from year to year.  The marginal seas are open water, including the Pacific basins, Canadian Bays (Hudson and Baffin), and the Atlantic basins for the most part.  Note ice extent fluctuations especially in Eurasian seas (lower right) and in Can-Am seas (upper right).  Notice the much greater ice extent in 2021 compared to 2018. As discussed later on, some regions retain considerable ice at the annual minimum, with differences year to year. [Note: Images prior to 2009 are in a different format.  AARI Charts are (here)

The annual competition between ice and water in the Arctic ocean is approaching the maximum for water, which typically occurs mid September.  After that, diminishing energy from the slowly setting sun allows oceanic cooling causing ice to regenerate. Those interested in the dynamics of Arctic sea ice can read numerous posts here.  This post provides a look at end of August from 2007 to yesterday as a context for anticipating this year’s annual minimum.  Note that for climate purposes the annual minimum is measured by the September monthly average ice extent, since the daily extents vary and will go briefly lowest on or about day 260. In a typical year the overall ice extent will end September slightly higher than at the beginning.

The melting season mid July to mid August shows 2024 melted at nearly the average rate, while retaining more ice extent at the end than some other recent years of note.

Firstly note that on average August shows ice declining 1.8M km2 down to 4.9M km2.  2024 started 288k km2 below average and on day 244 was only 98k km2 or 2% in deficit to average. The extents in Sea Ice Index in orange  were considerably lower during August, meaning that SII August 2024 monthly average will be ~400k km2 lower than MASIE., nearly half a Wadham.

The table for day 244 shows how large how the ice is distributed across the various seas comprising the Arctic Ocean.

The largest deficit to average is in Beaufort Sea, followed by smaller losses in Greenland Sea, CAA and Central Arctic.   Hudson Bay and Barents Sea are mostly open water. The offsetting surpluses are in Chukchi, East Siberian and Laptev seas.

For context, note that the average maximum has been 15M, so on average the extent shrinks to 30% of the March high before growing back the following winter. Presently 2024 is at 32% of last March maximum.  In this context, it is foolhardy to project any summer minimum forward to proclaim the end of Arctic ice.

Resources:  Climate Compilation II Arctic Sea Ice

Research ship drifting along with Arctic ice, May 2019 US Naval Institute

In May, most of the Arctic ocean basins are still frozen over, while the melting of ice extent is underway in the marginal regions.   During May, on average according to MASIE, Arctic ice extents lose 1.7 M km2, and 2024 matched that.  The few basins where open water appears this time of year tend to fluctuate and alternate waxing and waning.  Unusual were the much greater extents estimated by SII (Sea Ice Index, the satellite dataset)

The graph below shows for the month of May patterns for ice extents on average, this year and some other years of note.

The graph shows the 18-year average loss for April is 1.7M km2.  2024 tracked nearly average this month throughout. Remarkably, SII showed higher all month, ~200k km2 on average and 315k km2 higher than MASIE yesterday. Other recent years have been nearly average, while 2006 ended with a large defict of ~300k km2 below average.

The table shows regional ice extents in km2.  Note that Hudson and Baffin Bays have started melting, and Hudson is ahead of normal and will likely go to open water in a few weeks. Sea of Okhotsk on the Pacific side is 66k in deficit, with little ice left to lose.  Note the huge surplus in Barents sea on the European side. Everywhere else is mostly in surplus, especially the seas of Barents, Greenland and Kara.  2006 had 329k km2 less ice extent than 2024 (one third of a Wadham).

The polar bears had a Valentine Day’s wish for Arctic Ice.

And Arctic Ice loves them back, returning every year so the bears can roam and hunt for seals.

Footnote:

Seesaw accurately describes Arctic ice in another sense:  The ice we see now is not the same ice we saw previously.  It is better to think of the Arctic as an ice blender than as an ice cap, explained in the post The Great Arctic Ice Exchange.

Sunrise over frozen Bering Sea

Research ship drifting along with Arctic ice, May 2019 US Naval Institute

In May, most of the Arctic ocean basins are still frozen over, while the melting of ice extent is underway in the marginal regions.   During the last 30 days, on average according to MASIE, Arctic ice extents lose 1.4M km2. The few basins where open water appears this time of year tend to fluctuate and alternate waxing and waning.

The graph below shows the mid April to mid May patterns for ice extents on average, this year and some other years of note.

The graph shows the 18-year average loss for April is 1.4M km2.  2024 started this period with a slight deficit and ended 136k km2 above average.  SII showed higher throughout, and much greater extents in May (still awaiting the number for Day 136). Other recent years have been nearly average, while 2006 ended with a large defict of ~400k km2.

The table shows regional ice extents in km2.  Note that Hudson and Baffin Bays have started melting, and Hudson will likely go to open water in a few weeks. Sea of Okhotsk on the Pacific side is down 81k, offset by a similar surplus  in Bering sea. Note the huge surplus in Barents sea on the European side. Everywhere else is mostly in surplus, especially the seas of Barents, Greenland and Bering.  2006 had 582k km2 less ice extent than 2024 (more than half a Wadham).

The polar bears had a Valentine Day’s wish for Arctic Ice.

And Arctic Ice loves them back, returning every year so the bears can roam and hunt for seals.

Footnote:

Seesaw accurately describes Arctic ice in another sense:  The ice we see now is not the same ice we saw previously.  It is better to think of the Arctic as an ice blender than as an ice cap, explained in the post The Great Arctic Ice Exchange.

Sunrise over frozen Bering Sea