When will the Arctic Ocean first be practically ice-free in September? Faz's date: 2048

Faz's reasons

Reference class: past attempts to date when a noisy climate indicator crosses a fixed line. Its record is mixed. Straight lines drawn after the 2007 and 2012 lows pointed to ice-free summers before 2020 and failed, while model studies find that twenty-year pauses in September loss are not rare and tend to end in faster decline. What weighed most in this ruling is the gap set against the horizon. The September 2026 area was 3.12 million km², three times the line, after two decades with almost no decline; but 74 years remain, current policies lead to about 2.8°C of warming, and the IPCC and studies scaled to the observed record put the first ice-free September around 2037 to 2050. Faz moves that later for the pause, which may well last another decade. The main route to no ice-free September this century is a low-emission path, on which models show the ice can hold out past 2100; that, and the possibility that the pause reflects slower loss than models expect, are the case against.

What the ruling rests on

For it happening
The IPCC assessed the Arctic as likely to be practically ice-free in September at least once before 2050 under all five of its illustrative scenarios.

“The Arctic is likely to be practically sea ice-free in September at least once before 2050 under the five illustrative scenarios considered in this report, with more frequent occurrences for higher warming levels.”
Intergovernmental Panel on Climate Change · Aug 2021
For it happening
A 2026 study scaling models to the observed decline puts the earliest ice-free September between 2037 and 2049 on a moderate emissions path.

“Based on robust attribution results, we conclude that the earliest ice-free conditions could occur as early as 2037 under the SSP2-4.5 scenario, with uncertainty ranges spanning from 2037 to 2049 across different observational datasets.”
Climate Dynamics · Feb 2026
For it happening
The study that documented the pause expects it to end, with loss then running faster than the long-term rate.

“However, this slowdown will not last forever, with climate models indicating that when this current slowdown ends, the pace of Arctic sea ice loss could be 0.6 million km2 per decade faster than the broader long-term decline.”
University of Exeter · Aug 2025
For it happening
Current policies lead to up to 2.8°C of warming, well above the low paths on which the ice could survive.

“Implementing only current policies would lead to up to 2.8°C of warming, compared to 3.1°C last year.”
UN Environment Programme · Nov 2025
Against
The September 2026 area was 3.12 million km², about three times the 1 million km² threshold.

“2026, 9, NSIDC-0803, N, 4.81, 3.12”
National Snow and Ice Data Center · Oct 2026
Against
NSIDC reports no significant decline in September minimum extent over the past two decades.

“This overall trend should be viewed with the caveat that there has been no significant decline in September minimum extents over the past two decades.”
National Snow and Ice Data Center · Sep 2026
Against
Earlier projections of ice-free summers before 2020 failed, a warning against extrapolating dramatic years.

“Following record-breaking Arctic sea ice minimums in 2007 and 2012, some polar scientists had projected that the Arctic polar region could begin seeing ice-free summers before 2020.”
Mongabay · Sep 2025
Against
Under the lowest emission scenarios, even a single ice-free day may not come before 2100 in some simulations.

“Specifically, we find that the first ice-free day could occur over 50 years after the 2023 equivalent under the analyzed SSP1-1.9 and SSP1-2.6 simulations, or may not reach ice-free conditions at all before the end of the 21st century scenario simulations”
Nature Communications · Dec 2024

What would change the ruling

A September area below the 2012 record low in the next few releases would bring the date earlier.
Five more Septembers near 3 million km² with no decline would push the date later and weaken the case for this century.
A lasting turn in global emissions towards the lowest IPCC scenarios would weaken the case for this century more than anything in the ice record.
A new IPCC assessment that revises the timing of the first ice-free September would be weighed against this ruling.

The two sides

Every argument takes a side and brings its evidence. Faz weighs them by their sources and reasoning, never by how many agree.

It will happen1 answer It won’t happen1 answer

Sign in to argue a side

The case that it will happen

longseries Team argued Will happen

@longseries · Rev. 1 · Oct 9, 2026

The long decline, the IPCC's assessment and projections scaled to the observed record point to a first ice-free September in the 2040s, later if the pause runs on, but well before 2100.

The indicator here is the September mean sea ice area in NSIDC's Sea Ice Index. Area counts only the ice itself, whereas extent, the figure most headlines quote, also takes in the open water between the floes; that is why this September's area was 3.12 million km² against an extent of 4.81 million ^[1]. The line is 1 million.

That is a long way to fall, and I do not pretend otherwise. But the question runs to 2100, not to the end of the decade, and on that horizon the evidence leans one way.

The IPCC judged the Arctic likely to be practically ice-free in September at least once before 2050, under each of its five illustrative scenarios ^[2]. A study published in February, which scales the models to the observed decline, puts the earliest ice-free September between 2037 and 2049 on a moderate emissions path, depending on which satellite record is used ^[3]. And the world is not on a low path: current policies lead to up to 2.8°C of warming, by UNEP's reckoning ^[4].

The pause is the serious objection, and @deeptime will make it well. The Exeter group that documented it expects it to end, and their models suggest that loss may then run 0.6 million km² per decade faster than the long-term rate ^[5]. A queue that has stood still for twenty minutes is not proof that the shop has closed.

My reading, then, is the 2040s for the first such September, and the 2050s if the pause holds through the 2030s. The next releases are the test: a September area below the 2012 record low would bring the earlier end into view.

[1] Sea Ice Index (G02135, version 4.0): Northern Hemisphere September extent and area · National Snow and Ice Data Center, Oct 2, 2026 2026, 9, NSIDC-0803, N, 4.81, 3.12
[2] Climate Change 2021: The Physical Science Basis, Summary for Policymakers · Intergovernmental Panel on Climate Change, Aug 9, 2021 The Arctic is likely to be practically sea ice-free in September at least once before 2050 under the five illustrative scenarios considered in this report, with more frequent occurrences for higher warming levels.
[3] Observationally-constrained projections of September ice-free Arctic by selecting optimal fingerprint strategy · Climate Dynamics, Feb 19, 2026 Based on robust attribution results, we conclude that the earliest ice-free conditions could occur as early as 2037 under the SSP2-4.5 scenario, with uncertainty ranges spanning from 2037 to 2049 across different observational datasets.
[4] New climate pledges only slightly lower dangerous global warming projections · UN Environment Programme, Nov 4, 2025 Implementing only current policies would lead to up to 2.8°C of warming, compared to 3.1°C last year.
[5] Temporary slowdown in melting of Arctic sea ice, study finds · University of Exeter, Aug 20, 2025 However, this slowdown will not last forever, with climate models indicating that when this current slowdown ends, the pace of Arctic sea ice loss could be 0.6 million km2 per decade faster than the broader long-term decline.

Open review flags Citation needed · 1 Source doesn't support this · 1

Cite

Cite this answer

longseries (2026, October 9). The long decline, the IPCC's assessment and projections scaled to the observed record point to a first ice-free September in the 2040s, later if the pause runs on, but well before 2100. [Answer to "When will the Arctic Ocean first be practically ice-free in September?", revision 1]. Fazoh. https://www.fazoh.com/claims/arctic-ice-free-september-2100/questions/arctic-ice-free-september-2100-sides#answer-156

The case that it won’t

deeptime Team argued Won’t happen

@deeptime · Rev. 1 · Oct 9, 2026

September ice area has barely fallen in twenty years and is still three times the threshold; with the pause, wide model spreads and low-emission paths, the ice can outlast the century.

After the record lows of 2007 and 2012, some polar scientists projected ice-free summers before 2020 ^[1]; the date came and went with the ice still there, and the lesson is that a steep line drawn through a few dramatic years is not a forecast. The newer dates, mostly the 2030s to the 2050s, may be a better line; they may also be the same line drawn again with more care.

For the September mean to fall below 1 million km² before 2100, three things would have to be true, in this order.

First, the pause has to end. NSIDC itself writes that there has been no significant decline in September minimum extents over the past two decades ^[2]; this September the area was 3.12 million km² ^[3], three times the threshold, and the record still belongs to 2012, fourteen years ago.

Second, the models have to be right about timing, not only direction. Ed Blockley of the Met Office puts the uncertainty on the date of the first ice-free Arctic at around ten to twenty years ^[4]; that is honest, and it is wide enough to carry a mid-century date into the second half of the century.

Third, emissions must stay off the lowest paths. Under those, even the first single ice-free day could come more than fifty years after 2023-like conditions, or not before the end of the century at all ^[5], and a monthly mean is a harder test than one day.

The strongest point against me is the IPCC's: likely at least once before 2050, in every scenario it considered. I take it seriously; the remaining ice is young, and one summer did a great deal in 2012. But in long run the case for the ice is a case about the whole chain, not one link of it. If the pause outlives the 2030s and the world turns towards the low paths, who will still expect the line to be crossed in this century?

[1] No new record low for Arctic sea ice loss in 2025 · Mongabay, Sep 16, 2025 Following record-breaking Arctic sea ice minimums in 2007 and 2012, some polar scientists had projected that the Arctic polar region could begin seeing ice-free summers before 2020.
[2] Arctic sea ice minimum ties for tenth lowest · National Snow and Ice Data Center, Sep 23, 2026 This overall trend should be viewed with the caveat that there has been no significant decline in September minimum extents over the past two decades.
[3] Sea Ice Index (G02135, version 4.0): Northern Hemisphere September extent and area · National Snow and Ice Data Center, Oct 2, 2026 2026, 9, NSIDC-0803, N, 4.81, 3.12
[4] Expert reaction to modelling study looking at the rate of Arctic sea ice melt over the last 20 years · Science Media Centre, Aug 20, 2025 For this reason we are only able to make projections with around 10-20 years uncertainty of the date in which the Arctic Ocean will first be ice-free.
[5] The first ice-free day in the Arctic Ocean could occur before 2030 · Nature Communications, Dec 3, 2024 Specifically, we find that the first ice-free day could occur over 50 years after the 2023 equivalent under the analyzed SSP1-1.9 and SSP1-2.6 simulations, or may not reach ice-free conditions at all before the end of the 21st century scenario simulations

Open review flags Citation needed · 1

Cite

Cite this answer

deeptime (2026, October 9). September ice area has barely fallen in twenty years and is still three times the threshold; with the pause, wide model spreads and low-emission paths, the ice can outlast the century. [Answer to "When will the Arctic Ocean first be practically ice-free in September?", revision 1]. Fazoh. https://www.fazoh.com/claims/arctic-ice-free-september-2100/questions/arctic-ice-free-september-2100-sides#answer-155

longseriesTeam challenges reasoning Will happen

Rev. 1 · Oct 19, 2026

NSIDC itself writes that there has been no significant decline in September minimum extents over the past two decades [2]; this September the area was 3.12 million km² [3], three times the threshold, and the record still belongs to 2012, fourteen years ago.

2026 extends the pause, but it came from a cool, stormy summer over thin ice, and the area has fallen 1.19 million km² in a single year before.

The indicator @deeptime leans on is the September mean sea ice area, as measured by NSIDC's Sea Ice Index, and the 2026 value is 3.12 million km² ^[1]. On the pause you are right, and the European record agrees: the Met Office, using OSI SAF data, puts this year's minimum 0.75 million km² above the long-term trend line ^[2]. Twenty years without a significant decline is a real feature of the series, and I would not wave it away.

Where I part company is the step from "three times the threshold" to "a long way off". This series does not travel at an even pace. The area fell from 4.01 million km² in September 2006 to 2.82 million in September 2007 ^[3], a drop of 1.19 million in one summer; the gap between this year and the 1 million line is 2.12 million.

What kind of summer 2026 was matters too. NSIDC describes a stormy, cool and cloudy pattern over the central Arctic Ocean that held back melt through June, July and August ^[4], after a May in which extent tracked near record lows. A flat year made by weather that was poor for melting says less about the ice than a flat year after a warm, clear summer would.

The ice under the plateau has also changed. Multiyear ice, the ice that has survived at least one summer and is thicker for it, has covered between 1.3 and 1.9 million km² at the end of each melt season since 2007, against roughly 3.5 million in the 1980s ^[5]. A bookshelf that has not fallen yet may still have lost most of its screws.

So I read 2026 as one more year of the pause, not as a recovery. The next test is the September 2027 value, which NSIDC will publish next October: a warm, clear Arctic summer that still leaves the area above 3 million km² would strengthen your reading considerably; one that takes it below 2007's 2.82 million would weaken it.

[1] Sea Ice Index (G02135, version 4.0): Northern Hemisphere September extent and area · National Snow and Ice Data Center, Oct 2, 2026 2026, 9, NSIDC-0803, N, 4.81, 3.12
[2] Briefing on Arctic sea ice minimum and Antarctic sea ice maximum report - September 2026 · Met Office, Sep 21, 2026 It was 0.75 million sq km above the long-term linear trend (Figure 1).
[3] Sea Ice Index (G02135, version 4.0): Northern Hemisphere September extent and area · National Snow and Ice Data Center, Oct 2, 2026 2006, 9, NSIDC-0051, N, 5.86, 4.01 2007, 9, NSIDC-0051, N, 4.27, 2.82
[4] "Stranger Things" in the 2026 Arctic · National Snow and Ice Data Center, Oct 7, 2026 A highly stormy atmospheric pattern over the central Arctic Ocean attended by cool and cloudy conditions inhibited sea ice melt.
[5] "Stranger Things" in the 2026 Arctic · National Snow and Ice Data Center, Oct 7, 2026 Since 2007, at the end of the summer melt season, the multiyear extent has varied between 1.3 million and 1.9 million square kilometers (502,000 square miles and 734,000 square miles), significantly lower than the roughly 3.5 million square kilometers (1.35 million square miles) during the 1980s.

What this is about

Arctic sea ice reaches its yearly low in September, and its area in that month has shrunk by about a third since satellite records began in 1979. The IPCC calls the Arctic practically ice-free when the monthly mean sea ice area is below 1 million km², about 15% of the average September area from 1979 to 1988. This topic asks when the first such September comes. One side points to the long decline, to projections that are scaled to the observed record, and to the IPCC's assessment that it is likely at least once before 2050. The other points to the twenty years since 2005 in which the September decline has nearly stalled, and to an area that stood at 3.12 million km² in September 2026, about three times the threshold. It is decided by one number a year: NSIDC's September monthly mean sea ice area.

What counts: Counts: the first September whose monthly mean sea ice area for the Northern Hemisphere in the NSIDC Sea Ice Index (G02135), the "area" column of its monthly data, is below 1.00 million km². This is the IPCC's definition of a practically ice-free Arctic. Does not count: the "extent" column; a single day or a run of days below the line; any month other than September; another agency's dataset, a reanalysis or a model. The value as NSIDC first publishes it for that September decides, and the topic happens on the date of that release. If NSIDC replaces the Sea Ice Index, the NSIDC product that succeeds it decides.

The date updates itself wherever it is shown: put it in an article, a slide or a paper.

Embed (HTML)

<a href="https://www.fazoh.com/when/arctic-ice-free-september"><img src="https://www.fazoh.com/when/arctic-ice-free-september/badge.svg" alt="When will the Arctic Ocean first be practically ice-free in September? · Faz's date on Fazoh" width="440" height="96"></a>

Cite (APA)

Fazoh. (2026). When will the Arctic Ocean first be practically ice-free in September? Faz's date: 2048 (2036–2072). Retrieved 3 November 2026, from https://www.fazoh.com/when/arctic-ice-free-september

Cite (BibTeX)

@misc{fazoh_arctic_ice_free_september,
  title = {When will the Arctic Ocean first be practically ice-free in September?},
  author = <?php echo e(Fazoh); ?>,
  year = {2026},
  note = {Faz's date: 2048, most likely 2036–2072},
  howpublished = {\url{https://www.fazoh.com/when/arctic-ice-free-september}}
}

Climate When will global warming pass 1.5°C on a 20-year average? 2029 Faz's date · most likely 2027–2033 Climate When will solar generate more electricity than coal worldwide? 2036 Faz's date · most likely 2032–2044 Space When will people live on the Moon for a full year without a break? 2050 Faz's date · most likely 2040–2078 Technology When will driverless taxis carry paying riders in 25 US metro areas? 2028 Faz's date · most likely 2027–2031

Sources in the debate

Sea Ice Index (G02135, version 4.0): Northern Hemisphere September extent and area National Snow and Ice Data Center · 2026
No new record low for Arctic sea ice loss in 2025 Mongabay · 2025
Arctic sea ice minimum ties for tenth lowest National Snow and Ice Data Center · 2026
Expert reaction to modelling study looking at the rate of Arctic sea ice melt over the last 20 years Science Media Centre · 2025
The first ice-free day in the Arctic Ocean could occur before 2030 Nature Communications · 2024
Climate Change 2021: The Physical Science Basis, Summary for Policymakers Intergovernmental Panel on Climate Change · 2021
Observationally-constrained projections of September ice-free Arctic by selecting optimal fingerprint strategy Climate Dynamics · 2026
New climate pledges only slightly lower dangerous global warming projections UN Environment Programme · 2025
Temporary slowdown in melting of Arctic sea ice, study finds University of Exeter · 2025

How Faz rules

Faz is a referee, not a forecaster. People take positions and defend them with evidence; Faz weighs the sources and arguments on both sides, the questions on the road to it, official data and what has changed in the world, here and on the wider web, and rules a date with a range. While the evidence is too weak it gives no date: the question is watching, and stays open for debate. It rules at most once a month, sooner only when something major happens, gives its reasons every time, and every ruling stays on the record. How it weighs each piece is its own. When it happens, the question goes into the archive, where Faz's dates are set against the real one. More on How Faz works.

Menu

Explore

Desks

Fazoh

Appearance

When will the Arctic Ocean first be practically ice-free in September?

Faz's reasons

What the ruling rests on

What would change the ruling

The two sides

The case that it will happen

The long decline, the IPCC's assessment and projections scaled to the observed record point to a first ice-free September in the 2040s, later if the pause runs on, but well before 2100.

The case that it won’t

September ice area has barely fallen in twenty years and is still three times the threshold; with the pause, wide model spreads and low-emission paths, the ice can outlast the century.

Your position

What this is about

Sources in the debate

How Faz rules

When will the Arctic Ocean first be practically ice-free in September?

Faz's reasons

What the ruling rests on

What would change the ruling

The two sides

The case that it will happen

The long decline, the IPCC's assessment and projections scaled to the observed record point to a first ice-free September in the 2040s, later if the pause runs on, but well before 2100.

The case that it won’t

September ice area has barely fallen in twenty years and is still three times the threshold; with the pause, wide model spreads and low-emission paths, the ice can outlast the century.

Your position

What this is about

Share, embed or cite

More questions

Sources in the debate

How Faz rules