JAXA (ADS) 2021 Arctic SIE September daily minimum: July Poll

[Juan C. García]

The poll is for the minimum Daily Arctic sea ice extent in September 2021, as measured by ADS-NIPR-VISHOP (JAXA).

September mínimums have been:

Date                  Extent
                     (10^6 km2)
Sep 12, 2000      6.04
Sep 17, 2001      6.55
Sep 8, 2002        5.51
Sep 18, 2003      5.93
Sep 11, 2004      5.68
Sep 21, 2005      5.18
Sep 14, 2006      5.63
Sep 17, 2007      4.07
Sep 9, 2008        4.50
Sep 12, 2009      5.05
Sep 17, 2010      4.62
Sep 10, 2011      4.27
Sep 16, 2012      3.18
Sep 12, 2013      4.81
Sep 17, 2014      4.88
Sep 14, 2015      4.26
Sep 7, 2016        4.02
Sep 9, 2017        4.47
Sep 21, 2018      4.46
Sep 17, 2019      3.96
Sep 13, 2020      3.55

Order by lowest to highest:

[Juan C. García]

According to Gero’s latest JAXA post, the minimum September 2021 extent will end on 4.03 million km2. That is in the middle of the 3.75-4.25 range, so it is the safest bet. I agree with that, but I want to ask myself: What is more probable: 4.26 or 3.74 million km2?

Well, we are ending June and it had a weather that was supposed to be good for the ice. Then: Why 2021 is following the years with the lowest September minimum? The true is that I think that the ice is fragile. There has been a lot of talking in the Forum about winds and/or waves and if they are promoting the melting. Storms and specially cyclones are not helping the ice. And I think that the melting consequence is becoming the “new normal”.

If I choose between the 4.0-4.5 or the 3.5-4.0 ranges, I am inclined for the later. And I think that in a near future, human beings will have the surprise (not a surprise for all of us) of breaking the 2012 minimum. So, 3.5-4.0 million km2 will be a better bet. I am staying now on the 3.5-4.0 million km2 range and planning to be my start bet in June polls of the following years (when we really don’t know what is going to happen  ).

I could change my [September 2021] vote on the next 10 days, depending on the results of the storm that is starting on the Atlantic. But I doubt it. I think that it will end up bringing heat all the way from the Atlantic, to Kara, Laptev and ESS, in the following days.

Let’s wait and see.

[Paul]

The storm has probably not slowed down extent because it was a deep compacted low rather than an ice spreading out type of cyclone. It certainly slowed the area numbers down but the cyclone has left its mark with increased dispersion so we may see the area numbers starting to drop again soon I feel.

At the start of the melt season I thought we could match 2018 given the fairly similar winter weather and the distribution of the thickest ice. Sadly weather and early snowcover retreat has meant we probably will finish below 2018 with the melt momentum that has been occurring and the early opening up of the Laptev. Given the upcoming weather conditions which on paper don't look favourable for the ice(even if extent did slow down) because there is a dipole type pattern followed by yet another deepish low then I think extent could finish more around the 2007/2016 mark with the ice potentially looking quite diffused by September.

[Stephan]

I stay with my June poll value of a JAXA minimum between 3.5 and 4.0 M km², probably at the upper end of this interval. High confidence.

[Dreaming of when ...]

Im such a noob. I posted this just now in the freeform chatter without checking if there was a July poll....so apologies in advance for cross posting.

Eyeballed Guestimated Regional Extent Melt ('000km2) in next 8 weeks:
Caa 400
Beaufort 300
Chukchi 400
Hudson 400 meltout
Baffin 350 meltout
Grnlnd 100
Barents 100 melt out
CA 1,200
Kara 400 melt out
Laptev 350 melt out
ESS 550

Total lost 4,550kkm2
Predicted min extent 4,450 (000km2).

My June poll prediction was 4.0-4,5Mkm2 so still looking good.

[Niall Dollard]

Not much between 2016, 2019 and 2007, Jaxa has 2007 recording the 5th lowest. In their table last September, NSIDC has 2007 in 3rd place (lowest daily value).

[oren]

I am upping my July vote one notch to 3.50-4.00, considering the cyclone-driven snag in surface melt / NSIDC area and the lagging Pacific sector. The race is still open though, IMHO.

[KenB]

I stay with my June poll value of a JAXA minimum between 3.5 and 4.0 M km², probably at the upper end of this interval. High confidence.

Same for me (again).

[icefisher]

Weather will be the deciding factor.  My guess is 4.31 somewhat lower than other "recovery" years.

[pauldry600]

My guess is 3.76

[Freegrass]

Between 4.25 and 4.75 million km²

[gerontocrat]

Some more JAXA data as at 7th July on the Daily Arctic Sea Ice extent minimum

The first image shows the projection of the minimum as 3.90 million km2. This is based on the average of the last 10 years remaining extent losses from the 7th July.

The calculation that produces the graph of the minimum in the 2nd image produces a minimum of 3.96 million km2 on 13th September, some 60k above the long-term linear trend value for 2021. This calculation uses the 10 year average daily extent change for each day until the minimum.

Why the difference? It had me scratching my head for a bit, but here is why. Look at the third image attached.

It's all about the date of the minimum in previous years, which varies from year to year. Just before the minimum the 10 year average daily extent change is reduced by extent gains in those years with an early minimum date, e.g. 2016 which arrived at the minimum on the 7th September.

The 10 year averge daily extent change is a gain after the 13th September and thus excludes the sea ice extent losses in years with a late minimum date. This relates especially to 2019 and 2018.

A 60 k difference is no big deal with another 4 million of sea ice extent losses to go, so it has not changed my vote of 3.75 to 4,25 million km2. If it comes in at 3.90 million km2, 3rd lowest in the satellite record, I will be puffed up with pride.

click images to enlarge

To see the third image properly you may also have to click the image again and then press maximise

[Killian]

A 60 k difference is no big deal with another 4 million of sea ice extent losses to go, so it has not changed my vote of 3.75 to 4,25 million km2. If it comes in at 3.90 million km2, 3rd lowest in the satellite record, I will be puffed up with pride.

While I arrive at my predictions in a *very* different way, I said the same a while back: 4M give or take. I am in the middle of formulating my official prediction for this season, so went looking at some comparisons/trends of this year and the rest of the JAXA satellite record.

Based solely on visuals using Worldview, I tried to determine which area in the full Arctic Ocean/basin best correlates with low extent years. I kind of expected it to be the Pacific side, particularly the Beaufort. But looking at Beaufort, ESS/Siberian coast, CAB, and the Atlantic side in isolation and the years 2007, 2010, 2012, 2015, 2016, 2019, and 2020, what stood out was the condition of the ice in the ESS and north of the rest of Siberia. 2021 looks weakest there.

Those of you with the maths at your fingertips might want to check this out if it is not already common knowledge I somehow missed and you already know the answer. Trends often exist, but not yet above the noise, so maybe the correlation will not show up in the stats, or be very weak, but it seems clear from a visual standpoint.

If the correlation is not just my imagination, 2021 might be a good bit lower than the 4M I previously targeted. However, the weather really seems to be favoring that 4M though Fram export has remained more persistent than since 2012.

[Juan C. García]

I am going to keep my vote on the 3.5-4.0 million km2 (between 2nd and 3rd place, 2019-2020).

Only one day left to vote or change your vote. (And make your post  )

[Brigantine]

3.5 - 4.0

Just carrying my June entry forward

[passenger66]

The linear trend is above 3.5 but I still think the models are wrong and the ice is rotten and weak. So I guess 3.5.

[Killian]

Only one day left to vote or change your vote. (And make your post  )

Aaargh! Extension! 2 days!

Oh, wait, the eleventh *where*?

[icy voyeur2]

I went lower before but decided to casst a more conservative opinion.
3.5- 4.0

3.83

The smarter prediction would be up to 3.75 - 4.25 if I was putting money on the line. Nevertheless, I harbor an instinct that the ice is very vulnerable to a wide range of plausible weather outcomes. I tangibly fear some of them will eventuate. I want to be wrong.

[Killian]

Scenario A: We get a GAC or two, tons of sun or dipole shoving wind out the Fram: 3.25 ~ 3.7

Scenario B: Steady as she goes: 3.7 ~ 4.0

Scenario C: Perfect retention conditions: 4.0 ~ 4.3.

Let's call it 3.850 +/- 0.250

[Paddy]

Just as with the June poll, these predictions have really not aged well...