NSIDC 2020 Arctic SIE September average: August poll

[Juan C. García]

ATTENTION: There are two ASI extent polls on the ASIF. This one is for NSIDC sea ice extent monthly minimum or September average (which is also used for the SIPN sea ice outlook), the other is for JAXA sea ice extent daily minimum [provided by JAXA (ADS-NIPR-VISHOP)].
Make sure you are aware of the difference before voting.

These are the September averages on 2000-2019 (in millions km2):

   Year          Extent
                10^6 km2
1980's Avg.   7.21
1990's Avg.   6.66
2000's Avg.   5.65
2010's Avg.   4.68
2000             6.25
2001             6.73
2002             5.83
2003             6.12
2004             5.98
2005             5.50
2006             5.86
2007             4.27
2008             4.69
2009             5.26
2010             4.87
2011             4.56
2012             3.57
2013             5.21
2014             5.22
2015             4.62
2016             4.53
2017             4.82
2018             4.71
2019             4.32

Edit: Final 2020 result 3.92.

From lowest to highest:

[Juan C. García]

For today, my bet is between 3.75 and 4.25 million km^2.

[oren]

Dropped one bin to 3.50-4.00. This is the hardest achieve of all my predictions, upside risk much higher than downside risk.

[icefisher]

+1 Juan.  My range is 4.22+/- .18.  Probably 2nd but 3rd also possible.

[Stephan]

I repeated my guesses from June and July with the bin centered at 4.25 M km².

[gerontocrat]

Some data to help confuse the issue.

Despite the last 7 days extent losses being way below average, the July 2020 Monthly Average Extent is a record low 7.50 million km2, some 304k less than July 2019, which itself was a record low.

7.500 million km2 is 437k less than the value from the linear trend, which is a reduction of 70k per annum. i.e. July 2020 is more than 6 years less than that linear trend.

IF remaining extent losses to minimum turn ou to be at the average of the last 10 years remaining melt, the September Monthly Average would be 3.96 million km2, 412k above the record low of 3.55 million km2 in 2012, but 317k less than the 2007 (2nd lowest) value.

3.96 million km2 is 412k less than the value from the linear trend, which is a reduction of 83k per annum. i.e. July 2020 is about years less than the value from the linear trend.

i.e. Average extent loss from now puts the September monthly average pretty much slap in the mid-point of 2012 (lowest) and 2007 (2nd lowest).

Like all projections / predictions / forecasts there is only one certainty. It won't happen that way, and even that is uncertain.

I'm waiting for PIOMAS volume, a few days more data and Mystic Meg before I cast my vote.

[Juan C. García]

For today, my bet is between 3.75 and 4.25 million km^2.

IF remaining extent losses to minimum turn ou to be at the average of the last 10 years remaining melt, the September Monthly Average would be 3.96 million km2, 412k above the record low of 3.55 million km2 in 2012, but 317k less than the 2007 (2nd lowest) value.

Like all projections / predictions / forecasts there is only one certainty. It won't happen that way, and even that is uncertain.

I'm waiting for PIOMAS volume, a few days more data and Mystic Meg before I cast my vote.

So, my prediction is in corcondancia with your numbers, for now.
I am also waiting for PIOMAS and also, to see how much extent/area ASI melting we will have in the following 7 days.

[Juan C. García]

Just 2 more days to vote or change your vote. 

[Juan C. García]

I am going up one bin, to 4.0-4.5 M km2.
There is only one day left to vote or change your vote.

[glennbuck]

3.0 m km2- 3.5 m km2

[slow wing]

Dropping down half a bin as the high stayed high & the sun kept shining, & then the wind blew in the Beaufort.

August poll for NSIDC: 3.75-4.25M km2 VHC (Very high confidence)

 I always estimate (guess) the JAXA daily minimum then add 0.5M km2 to predict the NSIDC.

[The Walrus]

I am going up one bin, to 4.0-4.5 M km2.
There is only one day left to vote or change your vote.

Doing the same.  The recent melt slowdown cause me to rethink the final position. 

[Juan C. García]

I am going up one bin, to 4.0-4.5 M km2.
There is only one day left to vote or change your vote.

Doing the same.  The recent melt slowdown cause me to rethink the final position.

The middle of our bet is right on what NSIDC is forecasting.
I am not sure is that is good or bad! 

NSIDC has done a mid-August update. https://nsidc.org/arcticseaicenews/

It includes a projection of the 2020 minimum extent - tends to just above 4 million km2.

Given JAXA data tends to be a bit lower, it corresponds well to a projection for JAXA extent of a bit under 4 million km2.

[igs]

All know and frequently cited algorithms will fail by a huge margin this year.


This si one of situations where machines don't come even close to the human brain and eye.


Stats and algorithms are useful and powerful tools but we humans have to recognize at which moments in time we have to take the helm exclusively.

[Juan C. García]

This is NSIDC Chartic with data until August 26th, 2020.

[Juan C. García]

Sea Ice Outlook: 2020 August Report
31 August 2020

The August Outlook is based on a total of 39 forecasts (Figure 1), of which 25 are new August submissions while the remaining 14 are carried over from June and July. The median August Outlook value for the September 2020 sea-ice extent is 4.30 million square kilometers, with quartile values of 4.1 and 4.5 million square kilometers. Of the 39 August 2020 contributions, 16 are based on dynamical models, 17 are based on statistical methods, 4 are based on heuristic approaches (qualitative analyses), and 2 used machine learning-based methods. The median of the August submissions is close to that from July (4.36 million square kilometers) and June (4.33 million square kilometers). The 2020 August Outlook median is higher than 2019 (4.22 million square kilometers) and lower than 2018 (4.57 million square kilometers).

https://www.arcus.org/sipn/sea-ice-outlook/2020/august

[Juan C. García]

September 8^th NSIDC 5-day trailing average at 3.8M km².
It looks that is sure that 2020 September average will be 2^nd lowest, on the range 3.75-4.25M km². Not sure if it will be above or under the 4M km².

[gerontocrat]

Juan's post reminded me to have another look at that.

The Plume attached shows how the September monthly average most likely depends on the speed of extent gain in mid-late September. You can see from the plume that this varies greatly - from a very fast refreeze in 2016, compared with in 2018 a further significant extent loss and a very slow refreeze.

Hoever, if further extent loss and then extent gain is at the average for the rest of September, the monthly average will be just under 3.9 million km2, close on 0.5 million km2 or 5+ years ahead of the linear trend.

[Juan C. García]

the monthly average will be just under 3.9 million km2, close on 0.5 million km2 or 5+ years ahead of the linear trend.

Right (it is close enough).

According to my calculations, the NSIDC daily extent has to increase 140K km2 on the following 6 days to reach an average of 4M km2 on September. That is a huge average increase, so it is not going to happen. The NSIDC September average is going to be on the 3.85-4.00M km2 range.

                         ASIE (10^6 km2)
Avg. Sep 1-24             3.848
Sept 24th.                  4.118
Avg. increase to 4M     0.140 

The right bins on this poll are 3.75-4.25M km2 and 3.50-4.00M km2.

I finally voted on the 4.0-4.5 km2, so on the August poll I missed this one.

[gerontocrat]

I attach the monthly September Average graphs for NSIDC extent and area.
Note that the calculation is the simple average of the daily NSIDC (5 day trailing average) data.

I will be surprised if the NSIDC calculation is much different.

The 2020 September Extent average is 3.92 million km2, 0.45 million km2 below the long-term linear trend.

The 2020 September AREA average is 2.79 million km2, 0.22 million km2 below the long-term linear trend.

[Richard Rathbone]

I will be surprised if the NSIDC calculation is much different.

The 2020 September Extent average is 3.92 million km2, 0.45 million km2 below the long-term linear trend.

NSIDC September report is out https://nsidc.org/arcticseaicenews/

Prepare NOT to be surprised.

Its 3.92 by their calculation too.

[gerontocrat]

I will be surprised if the NSIDC calculation is much different.

The 2020 September Extent average is 3.92 million km2, 0.45 million km2 below the long-term linear trend.

NSIDC September report is out https://nsidc.org/arcticseaicenews/

Prepare NOT to be surprised.

Its 3.92 by their calculation too.

I am surprised - score one for the home team

[Juan C. García]

Arctic sea ice extent averaged for September 2020 was 3.92 million square kilometers (1.51 million square miles), the second lowest in the 42-year satellite record, behind only September 2012. This is 350,000 square kilometers (135,000 square miles) above that record low, and 2.49 million square kilometers (961,000 square miles) below the 1981 to 2010 average.

http://nsidc.org/arcticseaicenews/2020/10/lingering-seashore-days/

The September average is 3.92 million km². The on-target bins are 3.50-4.00M and 3.75-4.25M km².
There is a mistake difference on the total members that voted. Should be 62, not 59. The total members that vote on the right bins are 29, which is the 47%.

[Steven]

There is a mistake on the total members that voted. Should be 62, not 59.

It's a bug in the forum software.  It means that 62 people have voted in this poll, but 3 of them have deleted their user account in the meantime.  Those votes are removed from the "Total members voted", but are still included in the individual bins.

[Juan C. García]

There is a mistake on the total members that voted. Should be 62, not 59.

It's a bug in the forum software.  It means that 62 people have voted in this poll, but 3 of them have deleted their user account in the meantime.  Those votes are removed from the "Total members voted", but are still included in the individual bins.

Good explanation.
Thanks.