JAXA 2020 Arctic SIE September daily minimum: June Poll

[Juan C. García]

The poll is for the minimum Daily Arctic sea ice extent in September 2020, as measured by ADS-NIPR-VISHOP (JAXA).

September mínimums have been:

 Year               Extent
                  10^6 km2
1980's Avg.     7.19
1990's Avg.     6.49
2000's Avg.     5.41
2010's Avg.     4.33
2000               6.04
2001               6.55
2002               5.51
2003               5.93
2004               5.68
2005               5.18
2006               5.63
2007               4.07
2008               4.50
2009               5.05
2010               4.62
2011               4.27
2012               3.18
2013               4.81
2014               4.88
2015               4.26
2016               4.02
2017               4.47
2018               4.46
2019               3.96

The 2020 minimum is  3,554,798 km², reached on September 13^th.
The right bins are 3.50 - 4.00M and 3.25-3.75M km².

Order by lowest to highest:

[Phoenix]

4.10M km2 (3.75 to 4.25 bucket)

Very strong start in Siberia and Cablantic. Average start in Pacific and CAA holding it down.

[Juan C. García]

I will start with the 3.5 - 4.0 million km2, but we have 10 days to change the vote. 

P.S. I should not include the value of 1978 in the image, because the measurements begin on November 1, 1978. But I will keep it, because I know that if I make a change, then we cannot change our vote.

[Freegrass]

I already said in April we'll be breaking the 2012 record, and drop below 3 million km², and I'm sticking! Between 2.50 and 3.00 million km²

[Phoenix]

Interesting to see the voting pattern. After 14 votes, close to a perfect bell curve around a midpoint expectation of 3.50M km2.

[blumenkraft]

Between 3.25 and 3.75.

13.19% confident level.

[Stephan]

I went for the 4.00 ± 0.25 M km² bin. This is the one that gerontocrat finds when he applies the average loss of the last ten years on the actual extent data (see for details in the data thread).

[jdallen]

Between 3.25 and 3.75.

13.19% confident level.

I like the humor of this.

3.5-4.0 bucket, 50% confidence level.

I suspect my estimate will be high, but much will depend on how the weather plays out.   If June starts off and continues as hot as currently forecast, I may adjust down.

[blumenkraft]

I like the humor of this.

Aww, thank you! 

[Pavel]

I vote between 3,50 and 4,00 considering the remaining ice loss will be average of the last 10 years. The bad start of the season but CAA looks safely. I may change the bet until June 10 if the melt ponds will apper early in the high Arctic

[igs]

How boring, for once I'm with the majority 3.5-4.0


50% C.L.

[Phoenix]

Interesting to see the voting pattern. After 14 votes, close to a perfect bell curve around a midpoint expectation of 3.50M km2.

30 votes now in. Midpoint average is closer to 3.60M km2.

[Tor Bejnar]

2.5 (+/- a quarter) - almost 1% confidence.

[oren]

highest probability is for some result just above 4.00, with some downward trend added, so logic dictates 3.75-4.25 for maximum probablity. However, bearing in mind the massive preconditioning and low albedo, and the large volume anomaly near Svalbard that is at risk of being exported, extreme low results are much more probable than extreme high results. 3.0-3.5 is certainly within the realm of the possible. Thus I went for 3.5-4.0 as some kind of average. The method worked nicely for 2019, which had no GAC though I expected one.

[Phoenix]

4.10M km2 (3.75 to 4.25 bucket)

Very strong start in Siberia and Cablantic. Average start in Pacific and CAA holding it down.

Gerontocrat's charts says 5 year average melt will get us to 4.26M so I'm moving up to the 4-4.5 bucket.

Obviously, the group here senses a big melt event this summer.

[Wherestheice]

3.25-.75

I sense a big melt, however I am not ready to say if we will get a record low. I think we will get to a 2nd lowest though

[jens]

September mínimums have been:

 Year               Extent
                  10^6 km2
1980's Avg.     7.19
1990's Avg.     6.49
2000's Avg.     5.41
2010's Avg.     4.33

I wonder if by going with trends the 2020's average is going to be about 3?

Another thing I have noticed is that usually all years of the following decade have beaten the previous decade's average. For example - the highest of the 2010's is 2014 (4.88), which is less than 2000's average. Only once there was a discrepancy in that 1990's Avg beat 2001.

So going by that every year of 2020's should beat 4.33. That's the very minimum we are looking at. Of course, the beginning of the decade is the most vulnerable in this context and this decade has just started out.

[jdallen]

September mínimums have been:

 Year               Extent
                  10^6 km2
1980's Avg.     7.19
1990's Avg.     6.49
2000's Avg.     5.41
2010's Avg.     4.33

I wonder if by going with trends the 2020's average is going to be about 3?
<snippage>

We have a phrase - "The trend is your friend".

Variability over time averages out, collapses if you will, to the numbers you are looking at.

However the trend itself - or perhaps rather forces embedded in the system which produce it - are non-linear and have functional limits which produce feedbacks and moderate it, changing its slope and potential.

"Around 3" is a reasonable extrapolation for a simple system, but the Arctic is nothing like reasonable nor simple.

[Phoenix]

Excluding the people who are way down there at 3M and below, the bell curve has shifted to a midpoint of 3.75M km2, up from 3.50 km2 early in the process. At least some people are paying attention.

But the bucket to be in is still the 4-4.5 and the lone wolf in the 4.25 - 4.75 bucket has a chance of being the only one to get it right if we go over 4.5.

I'll take this voting trend as a challenge to my ability to persuade people of the logic. The biggest group is still two buckets away from the wise choice. They've migrated north by one bucket so far, let's see if we can engineer another.

[gerontocrat]

3.75 - 4.25 million km2.

The May monthly average volume is 600 km3 above the linear trend.
On the other hand, the Russkies are predicting a long hot Siberian Arctic Ocean Summer, & pretty much ditto for the waters in the far North of North America.

I would have gone 0.25 million km2 lower, but 600 km3 is 2+ years linear trend volume loss.

[Sciguy]

My first entry was 3.75 - 4.25 based on Gerontocrat's outstanding charts in the data forum.

After reviewing the PIOMASS thread with the results for May now out, I revised my entry to 4.00 - 4.50.  The Beaufort, CAA and central Arctic have a lot of thick ice now, and that will be difficult to melt out over the course of the season.

With the other peripheral seas opening up early and absorbing heat, an August GAC could of course lead to another 2012.  GACs are unpredictable months in advance, so my June guess will assume that cyclone activity will be near normal in the Arctic this year.

[jdallen]

Similar - moved up a bucket to 3.75-4.25.  Convinced by the drop in momentum.

Could still very much be wrong, but very much doubt the ice will remain above 4.25.

[Phoenix]

There is no bucket which contains both 3.99 and 4.26. So, which is more likely?

In order to get down to 3.99, we need 6.75 loss in the rest of the season. This much loss from this point in the season has been accomplished only once since 2008 when 2012 came with the crusher.

Betting on the bucket which requires the second best performance in over a decade when dealing with high volume anomalies in CAB and CAA is perhaps not a good idea. Move up.

[oren]

This thread is for voting and explaining your vote, and is not intended as an opportunity to criticize others' choices or to convert them. If there was just one right choice, we wouldn't be needing a poll. Some people vote for their hope, some for their feared scenario, some for the most probable. It's up to them.

[Phoenix]

This thread is for voting and explaining your vote, and is not intended as an opportunity to criticize others' choices or to convert them. If there was just one right choice, we wouldn't be needing a poll. Some people vote for their hope, some for their feared scenario, some for the most probable. It's up to them.

In this case, the poll is anonymous so nobody is being singled out. I am running the risk of having egg on my face if the prediction is way off.

The purpose of the comments is to draw attention to the factors which help people predict the outcome and perhaps give people an opportunity to reflect on why they have an expectation which is not in synch with anything predictable.

Tools which are available to help us predict the future of the ice may be important to decision making. If there is some cultural influence here which is causing people to be more negative about ice outcome than an objective inquiry might yield, maybe it should be open for exploration. It is noteworthy that such a high proportion of the voters predict an outcome which involves a rare late melt season event.

The intention here is educational, not to put anyone down. The hope is for the world to move to the most objective methods of appraising risk and maintain the credibility of those of us who fight for proactive policy like a Green New Deal. It is counter productive to have an overly alarmist message about certain threats which do not materialize and give deniers ammunition to accuse us of unfounded alarmism.

[oren]

give people an opportunity to reflect on why they have an expectation which is not in synch with anything predictable.

I repeat, No.

If there is some cultural influence here which is causing people to be more negative about ice outcome than an objective inquiry might yield, maybe it should be open for exploration. It is noteworthy that such a high proportion of the voters predict an outcome which involves a rare late melt season event.

Take that up in "accuracy of poll predictions" thread.

The hope is for the world to move to the most objective methods of appraising risk and maintain the credibility of those of us who fight for proactive policy like a Green New Deal. It is counter productive to have an overly alarmist message about certain threats which do not materialize and give deniers ammunition to accuse us of unfounded alarmism.

Irrelevant here. I repeat, No.

[binntho]

The minimum is obviously not predictable other than that it is going to fall somewhere fairly close to the linear downward trend.

The JAXA annual minimum downward  trend is as near to 4 million as makes no difference. So by far the safest bet is 3.75 - 4.25 (which was my choice this time).

Nothing so far indicates anything other than that the melt is following the standard trajectory of very fast vs. very slow vs. nothing really happing vs. all the indicators point to above average / below average, at least according to the august ensemble of prognasticors posting on this forum.

So I'll be boring and stick to the obvious and safe bet.

[Phoenix]

Nothing so far indicates anything other than that the melt is following the standard trajectory

I agree with that. I also think we should consider the beginning point.

Minimum = Today's Ice - Melt From Today To Minimum

We have a load of information about ice as of 5/31.

[binntho]

Nothing so far indicates anything other than that the melt is following the standard trajectory

I agree with that. I also think we should consider the beginning point.

Minimum = Today's Ice - Melt From Today To Minimum

We have a load of information about ice as of 5/31.

I'm pretty sure that it has been very well established that there is no correlation between ice at max and at min. Annual meltout has been growing and last year was in second place for Max-Min. So predicting whether the minimum is going to be half a million km2 more or less based on the winter maximum is futile.

[Paddy]

Alright, binntho has convinced me - going for the 3.75 to 4.25 bracket as well

[Phoenix]

I'm pretty sure that it has been very well established that there is no correlation between ice at max and at min. Annual meltout has been growing and last year was in second place for Max-Min. So predicting whether the minimum is going to be half a million km2 more or less based on the winter maximum is futile.

I don't suggest trying to use total Arctic volume at the max as the barometer for predicting total volume at the min. I would suggest using CAB or CAB + CAA (core) volume at 6/30 as a better indicator of total volume at 9/15.

It's the relative advantage in the seas that will retain ice that matters. If someone will produce a chart showing the relationship between core volume and the minimum, I bet we get a different story. Of course, the total volume at the max will contain a lot of noise related to seas which will contain no ice at the minimum. Let's try to strip the the parts out which are known to contain noise and find out if the core story is a good predictor.

I don't have all of the numbers. If anyone has them, I would suggest the following for last ten years.

CAB and CAA volume at 6/30 and minimum
Total volume, area and extent at minimum

7 columns, 10 rows or 5 columns if one excludes CAA volume.

[The Walrus]

Alright, binntho has convinced me - going for the 3.75 to 4.25 bracket as well

I was convinced beforehand, but his graphs have just added more credibility (as if I had any) to my vote.  The trend is so close to 4.0 that unless one has a very compelling reason to vote otherwise, it is the safe bet.

[Phoenix]

unless one has a very compelling reason to vote otherwise, it is the safe bet.

I believe that 2020 is a distant 9th in CAB volume at 5/31. This is where most of the ice will be at the minimum.

4.0M km2 represents 3rd all-time in extent.

Will 2020 move from 9th according to this indicator to 3rd or better in the balance of the season? It's certainly possible.

[The Walrus]

unless one has a very compelling reason to vote otherwise, it is the safe bet.

I believe that 2020 is a distant 9th in CAB volume at 5/31. This is where most of the ice will be at the minimum.

4.0M km2 represents 3rd all-time in extent.

Will 2020 move from 9th according to this indicator to 3rd or better in the balance of the season? It's certainly possible.

It is certainly possible.  It is also possible to remain in 9th place, which is currently 4.5M km2.  I viewed that possibility as less likely, so refrained from voting for the higher bin.

[Wildcatter]

3.25 - 3.75

Conservative end of my range was really more 3.5 - 4.0, but what's the fun in that? Pieces seem to be falling into place, so I'd rather have more room on the lower range.

[binntho]

I'm pretty sure that it has been very well established that there is no correlation between ice at max and at min. Annual meltout has been growing and last year was in second place for Max-Min. So predicting whether the minimum is going to be half a million km2 more or less based on the winter maximum is futile.

I don't suggest trying to use total Arctic volume at the max as the barometer for predicting total volume at the min. I would suggest using CAB or CAB + CAA (core) volume at 6/30 as a better indicator of total volume at 9/15.

We are not talking volume here, only extent. Volume is of course the only "real" measure that matters, but I don't think you have much change of making predictions there either.

[binntho]

It's the relative advantage in the seas that will retain ice that matters. If someone will produce a chart showing the relationship between core volume and the minimum, I bet we get a different story.

Why don't you do it yourself? We all have the same access to the data, and spreadsheets are easy.

And I'm not at all sure that we'd get a "different" story. There is no guarantee that the ice in the "core" at maximum was the same as that  in the "core" at minimum. And this is not at all likely, the ice pack is highly mobile, even the thick ice along the Canadian shoreline has been know to take off and hurry away.

Again, this thread is about extent which is the headline grabber.

[Phoenix]

We are not talking volume here, only extent. Volume is of course the only "real" measure that matters, but I don't think you have much change of making predictions there either.

Yes, we are talking extent. I am suggesting mid-season CAB volume as the best interim predictor of extent at the minima. I also started a PIOMAS prediction thread.

I am eschewing the routine purchase of a personal computer as unnecessary so can't download Wip's volume data files. Have a simple device for internet access and try to keep a minimal environmental footprint.

[binntho]

They have libraries where you live?

[binntho]

Yes, we are talking extent. I am suggesting mid-season CAB volume as the best interim predictor of extent at the minima.

So show us some numbers. Start by finding the "mid-season CAB volume" and the "extent at minima", both a freely available at a large number of sources. Scratch a grid on a sandy surface, with one set of numbers on one axis and the other set of numbers on the other axis, and then plot each year as a point on the grid. If the points approach anything looking vaguely like a trend, take a picture of it and post it to the forum.

[oren]

This thread is going OT. Stop.

[Phoenix]

This thread is going OT. Stop.

Which is the correct thread for this kind of discussion?

[oren]

Certainly not here. You can use your recently opened PIOMAS thread, change the title to "Relation of PIOMAS volume to area minimum" or whatever, and go on from there. But if you are truly intent on not doing some of the data work yourself, effectiveness will be very limited.
In any case, leave this thread alone. You've stated your prediction, let others do the same in peace.

[blumenkraft]

When in doubt, there is always the 'The off topic off topic thread' to follow up on discussions.

https://forum.arctic-sea-ice.net/index.php/topic,2995.msg246051.html#msg246051

[pauldry600]

3.5 to 4 I'll be boring

Think 3.8 million will be the minimum

[jens]

3.75 - 4.25 seems like kind of a safe bet to me, as "safe" as you can make a prediction. This means 2020 could end up being anywhere from 2nd to 5th lowest on record, which is a wide enough range for an outcome.

More than 4.25? Well, that would be a whimper of a melt season, and we have to consider it's already 2020, not 2015 or 2010 any more, so the expectations to go "down" are higher.

Less than 3.75? Maybe, but it seems like it needs some stars to align for that to happen. Which means at some time during summer the daily loss must really start sprinting compared to other years (apart from 2012). I don't have the know-how to call, how likely is that to happen.

[Juan C. García]

Only 5 hours to vote or change the vote.

Good day to all! 

[igs]

How boring, for once I'm with the majority 3.75-4.25


75% C.L.

Changed my vote 5 hours before closure one bin up in this but left NSIDC & Piomas where they were thus far.

[JNap]

Using a simple linear trend model, the extent declines approximately 100-125k per year on average.  But "simple" is a word that I am not inclined to use for all the dynamics that go on in the Arctic. 

But given the current extent, and with the forthcoming dipole predicted for the next several days, it seems that the probability is that we could reach the 2nd lowest total.  Thus I am in the 3.5 - 4.0 bin and have a 75% CL.

[jens]

But given the current extent, and with the forthcoming dipole predicted for the next several days, it seems that the probability is that we could reach the 2nd lowest total.  Thus I am in the 3.5 - 4.0 bin and have a 75% CL.

2nd lowest seems indeed on the cards. It looks like 2020 may start to live up to its hype and expectations as they have pretty much entered the lead group now in June. Basically what 2020 has to do is to keep up with the leaders (2019 & 2012) in July. In August 2019 slowed down, so in the end they barely beat 2016 and 2007. So if 2020 could keep up the momentum better in August, it would easily end up 2nd lowest.

I'm having second thoughts about not going for 3.5-4.0 really... It was a close call.