September predictions challenge 2019

[Richard Rathbone]

Jaxa: 3.75-4.25, Medium
NSIDC 4.0 - 4.5, Medium

Jaxa 3.25-3.75, M
NSIDC 3.5 - 4.0, M

JAXA 3.5-4.0 H
NSIDC 3.75 - 4.25 H

PIOMAS indicates to me that a record is still just about on, Slater looks 4.25 ish though I won't see the Aug 1st prediction for SIPN till after poll deadlines. Last year I was toying with VH at this stage, this year I'm wondering if I should have stuck with M. Just how massive is the momentum from June that carries the melt for the last month before the minimum? Has it dissipated, does it need a GAC to push the ice out over the hot water?

[Juan C. García]

My June predictions:

JAXA: 3.5 to 4.0, médium confidence
NSIDC: 3.75 to 4.25, médium confidence

My July predictions:

JAXA: 3.25 to 3.75, high confidence
NSIDC: 3.75 to 4.25, high confidence

My August predictions:

JAXA: 3.50 to 4.00, very high confidence
NSIDC: 4.00 to 4.50, very high confidence


P.S. Only ~1.5 days to vote or change your vote...
Am I right? Or we will have more time on this challenge? 

[Paddy]

My June predictions are as follows:
JAXA: 3.75 to 4.25, medium
NSIDC: 4.25 to 4.75, medium

My July predictions are as follows:
JAXA: 3.75 to 4.25, medium
NSIDC: 4.0 to 4.5, medium

Droping my JAXA bin a half step further to 3.5 - 4.0.  Still Medium confidence.

My August predictions are as follows:
JAXA: 3.5 to 4.0, high
NSIDC: 3.75 to 4.25, high

Deadline for entries is the 15th of the month, as per precedent.

[Stephan]

My prediction (JAXA Sep minimum): 3.75-4,25 M km², high confidence
My prediction (NSIDC Sep Average): 4.00-4.50 M km², high confidence

I repeat this prediction, both values with high confidence.

[slow wing]

JAXA:     3.50 to 4.00,  medium (minimum day extent)
NSIDC:   4.00 to 4.50,  medium (September average extent)

[oren]

For July:
Dropping JAXA to 3.25-3.75 and increasing conf to High. I think a finish above 4M has become quite improbable, and below 3M extremely difficult.
Also dropping NSIDC to 3.75-4.25, keeping Medium conf.

Keeping the same for August. My best daily bet is 3.50-4.00 but I still see an option of going below, with very low chances of going above, so I'm skewing my bin accordingly.

[Neven]

My June predictions are as follows:
JAXA: 3.5 to 4.0, medium
NSIDC: 4.0 to 4.5, medium

For July, it's the following, dropping two bins:

JAXA: 3.0 to 3.5, medium
NSIDC: 3.5 to 4.0, high

And for August, it's:

JAXA: 3.25 to 3.75, high
NSIDC: 3.75 to 4.25, high

[Aluminium]

JAXA: 3.5-4.0 high.
NSIDC: 4.0-4.5 medium.

JAXA: 3.5-4.0 high.
NSIDC: 3.75-4.25 high.

JAXA: 3.5-4.0 very high.
NSIDC: 3.75-4.25 high.

"Very high" is as attractive as the red button "do not press".

[Juan C. García]

"Very high" is as attractive as the red button "do not press".

It is the gambler choice: win all or lose all. 

[Richard Rathbone]

Jaxa: 3.75-4.25, Medium
NSIDC 4.0 - 4.5, Medium

Jaxa 3.25-3.75, M
NSIDC 3.5 - 4.0, M

I'm adjusting my August entry to
NSIDC  4.0-4.5 H
on the basis of Slater 50-day showing probable extent back up to 5 by the end of September. Sticking with
JAXA 3.5 - 4.0 H.

[Brigantine]

June:
JAXA 3.75 - 4.25 medium confidence
NSIDC 4.25 - 4.75 medium confidence

July:
JAXA 3.50 - 4.00 high confidence
NSIDC 4.00 - 4.50 high confidence

August: (provisional)
JAXA 3.50 - 4.00 high confidence
NSIDC 3.75 - 4.25 high confidence

[Steven]

JAXA minimum extent:       3.5 to 4.0,  medium confidence
NSIDC September extent: 4.0 to 4.5,  medium confidence

JAXA:     3.25 to 3.75,  medium
NSIDC:  3.75 to 4.25,  medium

August:

JAXA:     3.5  to 4.0,    high
NSIDC:  3.75 to 4.25,  high

[AndyW]

About 4 million

Not that it matters, the long term trends are what matters.  So I will not be losing any sleep over it.

[Paddy]

Now that this is closed, I could potentially calculate the scores resulting from each ultimate outcome... but I think that sounds like a bit too much work. If anyone else wants to, they should feel free, however.

[Juan C. García]

Now that this is closed, I could potentially calculate the scores resulting from each ultimate outcome... but I think that sounds like a bit too much work. If anyone else wants to, they should feel free, however.

Let's wait until September when we can be sure of the 2019 minimum and maybe guess the NSIDC Monthly average.

[Paddy]

Anyone else feeling bemused that their June prediction is looking a lot better than their August one right now?

[Klondike Kat]

Anyone else feeling bemused that their June prediction is looking a lot better than their August one right now?

I did not make an August prediction, but was feeling that my June prediction of NSIDC of 4.0-4.5 was looking rather bleak in early August.  Now, it is looking better as a sub 4.0 extent is appearing less likely.

[Brigantine]

Anyone else feeling bemused that their June prediction is looking a lot better than their August one right now?

I'm feeling you there! But if my August one is out by 1 bin, I'll still be happy.
Happy I didn't pick very high confidence

JAXA under 4.00 is still touch and go, but NSIDC under 4.25 is starting to look distant.

Looking at each of our predictions, it seems we think very alike!

Hypothetical scoring:
Only included contestants scoring over 20 or in the top 3 in the particular or adjacent bin.
I wanted to put this in a spoiler box, but that didn't work so I did this.
JAXA 4.30, NSIDC 4.76: jplotinus 4, Slow Wing 2, Stephan 0, Paddy -2
JAXA 4.26, NSIDC 4.60: Stephan 12, Paddy 5, SW 5, gerontocrat 4, jdallen 4, jplotinus 4, Brigantine 2
JAXA 4.10, NSIDC 4.51: Stephan 24, gc 12, jda 12, Paddy 12, Brig 12, SW 10
JAXA 4.10, NSIDC 4.40: Stephan 36, gc 20, jda 20, Brig 20, Paddy 18, Richard Rathbone 18, SW 16, Steven 13, Aluminium 2, Juan C. Garcia 0
JAXA 3.99, NSIDC 4.40: Stephan 36, Aluminium 30, JCG 30, Brig 28, Paddy 24, gc 24, jda 24, RR 23, Steven 20, SW 19, oren 16
JAXA 3.90, NSIDC 4.24: Al 38, JCG 36, Stephan 36, Brig 30, Paddy 26, Steven 26, gc 24, jda 24, RR 24, oren 20, SW 20, Neven 19
JAXA 3.74, NSIDC 4.10: JCG 40, Al 38, oren 28, Steven 28, Brig 28, Paddy 24, RR 24, Neven 24, Stephan 24, gc 20, jda 20, SW 20
JAXA 3.70, NSIDC 3.99: Al 36, Neven 26, oren 26, Steven 26, Brig 23, Paddy 21, JCG 20, RR 20, SW 18, gc 12, jda 12, Stephan 12

[Richard Rathbone]

August SIPN report is out.

https://www.arcus.org/sipn/sea-ice-outlook/2019/august

Of the ones I tend to track most:
CPOM (Schroeder, melt pond statistics) 4.1
NSIDC (Slater/Barrett, concentration statistics) 4.16
APL (PIOMAS forced by long range weather forecast) 4.4
Navy (HYCOM/CICE model) 3
Nico Sun (Tealight, albedo statistics) 4.13

4-4.5 still looks a good bet to me. I don't hold out much hope for the Navy being within the 2.7-3.2 range they give.

[Juan C. García]

2019 JAXA min, as of September 16^th --> 3.99M km².
      Correct bins: 3.75-4.25   &   3.50-4.00

2019 NSIDC September average, until now (Sept 1^st-15^th) --> 4.29M km². *
     Correct bin:  4.00-4.50
     Possible correct bins: 4.25-4.75   or   3.75-4.25  (depending on values on the rest of Sept.)

* For the average, using daily values (no 5-day trailing average).
The last value is Sept 15^th: 4.16M km² (so the average will decrease on the following days, but certainly the average will increase at the end of the month).

[Neven]

My June predictions are as follows:
JAXA: 3.5 to 4.0, medium
NSIDC: 4.0 to 4.5, medium

For July, it's the following, dropping two bins:

JAXA: 3.0 to 3.5, medium
NSIDC: 3.5 to 4.0, high

And for August, it's:

JAXA: 3.25 to 3.75, high
NSIDC: 3.75 to 4.25, high

Looks like my June predictions will be correct. Which are the hardest, of course. 

[Paddy]

My June predictions are as follows:
JAXA: 3.75 to 4.25, medium
NSIDC: 4.25 to 4.75, medium

My July predictions are as follows:
JAXA: 3.75 to 4.25, medium
NSIDC: 4.0 to 4.5, medium

Droping my JAXA bin a half step further to 3.5 - 4.0.  Still Medium confidence.

My August predictions are as follows:
JAXA: 3.5 to 4.0, high
NSIDC: 3.75 to 4.25, high

Deadline for entries is the 15th of the month, as per precedent.

My JAXA predictions have been lucky enough to be fine throughout, but not so much NSIDC... either I guessed too high in June or too low in August. So probably either 26 or 24 points for me, depending.

[Juan C. García]

My June predictions:

JAXA: 3.5 to 4.0, médium confidence
NSIDC: 3.75 to 4.25, médium confidence

My July predictions:

JAXA: 3.25 to 3.75, high confidence
NSIDC: 3.75 to 4.25, high confidence

My August predictions:

JAXA: 3.50 to 4.00, very high confidence
NSIDC: 4.00 to 4.50, very high confidence

"Very high" is as attractive as the red button "do not press".

It is the gambler choice: win all or lose all. 

I made a gambling guess in August and I was close to losing it, but with the 3.99M km², I am on the winner side again. 20 points just for the August bins.  I will have to wait for the other NSIDC bins.

P.S. NSIDC average can still go above 4.50M km², but I strongly doubt it. Maybe it can go under 4.25M km². A little more chance, even that I also doubt it.

[Richard Rathbone]

Looks like my June predictions will be correct. Which are the hardest, of course. 

I reckon June is the easiest. Its much harder quantifying how much skill I have in later months, but no skill is easy to quantify.

Anyone feeling nervous about JAXA dropping below 3.75 yet?

[oren]

I'm willing to bet it's not gonna happen. But the Arctic can always surprise.

[Stephan]

My prediction (JAXA Sep minimum): 3.75-4,25 M km², high confidence
My prediction (NSIDC Sep Average): 4.00-4.50 M km², high confidence

I repeat this prediction, both values with high confidence.

I think I got it right 

[Stephan]

Anyone feeling nervous about JAXA dropping below 3.75 yet?

It will drop below 3.75 M km², but not in 2019 

[Paddy]

So, now that we can safely call the correct NSIDC bins as being 4.00 - 4.50 & 4.25 - 4.75 M km2, does anyone have the free time and inclination to tot up the final scores?  Apologies for being lazy / distracted by other nonsense...

[Juan C. García]

Hope that these tables are OK:

[Juan C. García]

Hope all the calculations are ok.
(Let me know if there is a correction to make )

[blumenkraft]

Congrats, Stephan!! \o/

[Stephan]

wow, I hadn't expected it to be correct with my estimations right from the start...
Thanks for the congratulations, blumenkraft

"Auch ein blindes Huhn findet einmal ein Korn"

[blumenkraft]

Hihi

[Richard Rathbone]

"If SIPN entries had been here" results:

CPOM (Schroeder, based on melt pond statistics) 4,2,2 = 8

UK met office (dynamic model, seasonal forecaster) -6,-6,none = -12

US Navy (dynamic model, often referred to on these boards as HYCOM) -4,-1,-4 = -9

NSIDC Slater model (concentration statistics) no entry,1,1 = 2 (spot on but no error margin given)

Nico Sun a.k.a. Tealight, (statistics on albedo calculations) 6,2,6 = 14

University of Washington (dynamic model, PIOMAS forced by a seasonal weather forecast) 2,6,6 = 14

Note how PIOMAS crushes the other dynamic models, not only because it was more accurate, but also because the other models have a very poor appreciation of just how inaccurate they are.

CPOM and Tealight are a lot closer than those numbers suggest, one being just the right side and the other just the wrong side of being put in higher value bins.

Statistical models did well, but it was a year that finished close to trend, and it needs a year with markedly good or bad melt to sort them out from one another.

[jdallen]

Hope all the calculations are ok.
(Let me know if there is a correction to make )

How did I miss the August polls?

Well, even without them, I did better 'n Average

[Juan C. García]

My prediction (JAXA Sep minimum): 3.75-4,25 M km², high confidence
My prediction (NSIDC Sep Average): 4.00-4.50 M km², high confidence

Ditto.

You were following Stephan vote, so if you didn't missed the August poll, you would be in first place, like him...

[jdallen]

*That* must be what happened.  I'll take the honorary victory.

[Aluminium]

I made my forecast by calculations for JAXA and by intuition for NSIDC. The difference is obvious.

[Juan C. García]

I was wishing that 2019 could break the 2007 NSIDC September average. That explains my June and July votes (3.75-4.25M km², definitely below the 2007 record of 4.27M km²). But 2007 had a very late refreeze, so it is difficult to break the average, even that 2019 had a lower NSIDC daily mínimum than 2007. Of course, the very low melt/compactation at the end of August 2019 had also an influence. A little more melt/compactation in August and 2019 would break the NSIDC 2007 September average.

Anyway, second place tied with Aluminium, it is good enough. 

[Juan C. García]

It is fun to have this "September predictions challenge" (a real challenge to have a good score), so it will be good if Richard Rathbone or Paddy open this 2020 version thread.

[Richard Rathbone]

It is fun to have this "September predictions challenge" (a real challenge to have a good score), so it will be good if Richard Rathbone or Paddy open this 2020 version thread.

It'll happen. I've been waiting to see what format this year's polls have.