JAXA extent minimum 2024: August poll

[Steven]

This poll is about the JAXA sea ice extent daily minimum for 2024.  The minimum is usually reached in mid-September.  Below is a list with the minima for previous years (in million km²) ranked from low to high.

year    minimum
2012    3.18
2020    3.55
2019    3.96
2016    4.02
2007    4.07
2023    4.14
2015    4.26
2011    4.27
2018    4.46
2017    4.47
2008    4.50
2022    4.52
2021    4.61
2010    4.62
2013    4.81
2014    4.88
2009    5.05
2005    5.18
2002    5.51
2006    5.63
...

[IceFloe]

I vote for 4.5-5.0. Melting in August is weaker than in July. In August, the ice will melt mainly south of latitude 75. Further north of the latitude, melting will be weak due to the low Sun. ZLab's graphs for regional seas indicate strong melting only in certain seas, such as the Canadian Archipelago and the Kara Sea. A new record is almost ruled out due to the large loss of ice in Antarctica.

P.S. lowered my vote to 3.75-4.25 after Steven's data about SMOS

[oren]

Thanks again Steven.
I kept my 3.50-4.00 vote from July.
Icefloe - Antarctica has nothing to do with this poll.

[IceFloe]

Icefloe - Antarctica has nothing to do with this poll.

There is a clear correlation between the 2012 minimum in the Arctic and the 2014 maximum in Antarctica. Two extreme proximity events that occurred at almost the same time in a 45-year series of satellite observations. I don't think this is an accident.

[Richard Rathbone]

Icefloe - Antarctica has nothing to do with this poll.

Seesaw connections exist between Arctic and Antarctic ice coverage. I don't think they operate fast enough to influence this year, but I could be wrong.

I'd expect whatever year does first beat 2012 to be well ahead at this point, so I'm sticking with 3.5-4.

[Steven]

I'm also sticking with 3.5-4.0 (same as in the June and July polls).  I think there is quite some vulnerable ice especially on the Pacific side, which may or may not melt out depending on the remaining weather. 

FWIW, here a few estimates for the JAXA extent minimum based on linear regressions:

• Gero's 10-year average extent trajectory is currently at 3.96
• Linear regression using the latest NSIDC area numbers would give 3.53
• Linear regression using the SMOS July average pixel count would give 3.81, while AMSR2 July average nonmelting pixel count would give 3.51

[Stephan]

I sticked with my June and July bet on the 4.00 ± 0.25 M km² bin.

[Paul]

Gone down by 1(3.75-4.25) compared to previous votes because even before Steven posted the stats in the previous post, it is suggesting this year could well finish below 4 million on JAXA which is significant as that means 3 of the past 6 years have gone under that figures.

The only real saving grace is whilst the Beaufort will get warmer, the CAB will be cooler and hopefully we get some early cold pooling forming like we did in 2021 and coupled with SSTS being below average across the Chukchi and perhaps a moderation somewhat on the SSTS over the Laptev then we may still finish above 4 million albeit most likely the lower end of it.

[sadmird]

I repeated my 3.75-4.25 vote again as soon as I saw the poll, like in all previous months, but this time I'll leave it open for future change in the next week. Important fact is that NSIDC area is already down to ~3.74M km², which is close to the lowest sea ice area up to this date. As the extent tends to catch up with area (but it's not a given, a good example is 2016 and, to a good extent, the last year, thanks to its massive proportion of sea ice being exported in 2023), it's almost a given that the end extent will be under 4M km².

I am unsure to claim whether if it's really going to be closer or under 3.8M, which would mean that a lower bin would be more appropriate. I still see it as unlikely that the Chukchi sea ice "below" the Wrangel island will survive, despite the solid state of that ice for this date in the melt season. Had it been anywhere else, it'd be safe to say it's largely going to carry on into the next season, especially due to its thickness estimates. But it is as it is, so I am still anticipating that at least 90% of that ice extent will be gone, if not all. I think there's close to 400k km² sea ice extent in Chukotsk Sea. This fact, coupled with lots of weak ice and very low volume, means that we can anticipate top3 score. I don't really feel confident to claim #2 spot and #1 extent is highly unlikely by the end of the melt season, unless something extraordinary happens by the end of the year.

My vote will probably change to the lower bin.

[Steven]

2 more days to vote before this poll closes.

FWIW, here a few estimates for the JAXA extent minimum based on linear regressions:

• Gero's 10-year average extent trajectory is currently at 3.96
• Linear regression using the latest NSIDC area numbers would give 3.53

One week later, gerontocrat's extent trajectory projection is now at 3.93, while linear regression with the latest NSIDC area numbers would give a JAXA extent minimum of 3.78 million km2.

[gerontocrat]

The plume of projections from remaining sea ice extent loss in the last 10 years gives a range of the minimum from just over 4.25 million to just over 3.5 million.

I.e. not one year showing a new record low minimum, but a range of outcomes still as much as 0.75 million km2.

My July forecast now in the dustbin of history (& best forgotten). My vote is for 3.5 to 4.0 million, mostly due to the GFS forecast for a bit more warmth in the High Arctic in the next week or so and that volume loss was quite high last month.

[opensheart]

Arctic Sea Ice will set a new record low when it has exhausted all other possible outcomes.

[Steven]

JAXA extent dropped below 4.25 million km² today.  It's now at 4.236 million km².