Poll: ASI JAXA Minimum Extent range of possibilities

[Tor Bejnar]

All 'guess the minimum' polls to-date have been to guess the minimum (or maximum) of something.  Here, we are to exclude the ranges (bins) we consider not possible this year.  There are 12 bins, and you can vote 'against' up to 11 bins.

For example, if you think the Arctic Sea Ice Extent (ASIE) will be 4.0 million km^2 [= km²] at minimum, you'll accept it might be a little more or a little less, but definitely not a lot more or less, definitely not a blue ocean event (BOE, for this poll's purpose = under 1m km²), therefore you will select (vote for) [something like] the top 3 bins (5.0 and up) and the bottom 4 bins (2.5 and under, leaving 4 bins you consider possible (2.5 - 5.0).  The more confident you are, the more bins you will select.  E.g., if you 'know' there will be a BOE, you'll select the top 10 bins.

[Tor Bejnar]

The value of a poll like this is to help us capture what we see as the true possibilities during mid-July at what the minimum might be.  Those with high confidence and those with low confidence get combined 'organically', no extra record keeping required!

If the weather I predict happens, and ice melts as I expect it to, the ASIE will be X m km^2.  If the weather is worse for ice, it will be Y, and if it's good for ice, Z.  X might be the average of Y and Z, but a skewed result is certainly possible.  One might think X is just under 4.0 m km^2, over 4.5 is not possible, and 2.6 is possible 'with a GAC'.  Some might say anything from 1 to 6 m km^2 is still possible, for all they know (or don't know); no other poll would allow for such a range of honesty with actual votes.

with 15 voters casting opinions, every bin is seen as a possibility by at least one of us, and yet a clear picture is already emerging.  Vote your opinion and see what I see!  Changing your mind is allowed. 

At the end, some voter's range will not include the reported value:  food for thought!

A 'winning-est winner' would be the voter who votes out all but one bin, and that bin includes the reported value.

[Rich]

I suggest a tiebreaker. In the event of a tie, the person who submits their prediction first is the winner.

[RikW]

I'd say we get below 4.5 for sure, excluding a miracle (supervolcanoe exploding en filling the air with stuff, a nuclear war starting a nuclear winter etc.), seeing the state of the ice. I think we will drop below 4M m2 for the 2nd time ever with average melting conditions, unless we get very unfavourable melting conditions next weeks/months.

But if we have a perfect GAC early/mid august with favorable melt conditions for the next couple of weeks, ice will crash and in that case I fear the 2012 record will be smashed.

[oren]

Range of possibilities is 2.5-4.5 but the outer bins are a 5% chance together.

[kynde]

I ruled out less than 3 and abovce 4.5. Thus far I've been guessing little less than 4, but I'm thinking there's more chance to go considerably below rather than above, and more accurately I would've put my likely 95% estimate inside [3.0,4.25].

[Klondike Kat]

I went a little narrower; eliminating below 3.5 and above 4.5.

[RealityCheck]

Earlier today I voted and 'ruled out' below 3.5 or above 4.5. Why? Just the trends in the data. Could 2012 be exceeded? Of course. Is it likely? I think not... although the state of the ice is woeful, judging by the commentary be the experts hereabouts. Now, if this poll was for min Area, not Extent, it might be a different story. I think we'll see very distributed ice come Sept.

[Tor Bejnar]

It has just dawned on me that someone could be thinking: 

• Under expected conditions the ASIE minimum will be between 3.5 and 4.0
• A GAC, however, will definitely clear out a lot of scattered ice, resulting in 2.5 to 3.0
• Therefore excluding the 3.0-3.5 bin along with the too-high and too-low ones.

I like the shape of the resulting bars that show after one votes. 

In the future, we might try a "What is your 99%ile likelihood range look like?"  Here we may have some contributors using 95% confidence with others using 99.9% confidence which would give different spans.  If we asked for a 4 S.D. confidence, would we still have some people's range not include what the actual final results will be, as we have now?

[Stephan]

I have ruled out all bins below 3.5 M km² and above 4.5 M km². This is in-line with my bets on the JAXA minimum and the NSIDC Sep average.

[Paddy]

I've ruled out everything from 5 and above and everything from (EDIT) 2.5 and below. I'm happy to agree woth those who'd advocate for tighter boundaries that there's a less than 1% chance of, say, an extent less than 3, but such rarities do sometimes happen.

[Tor Bejnar]

This poll closes in a few days.  Remember, vote "for" all the bins you're sure will not contain the JAXA September minimum.

Even folks new to Arctic sea ices can vote, as excluding even one bin "that's not possible this year" puts you in - your vote will be valued.  Obviously, if you can exclude more than one bin, that's better.  You can change your vote after you see how others voted, if you need to.

[oren]

Thanks for the reminder Tor.
Recklessly, I have edited my vote and added the 4.0-4.5 box to those I've marked as impossible.
With all the preconditioning, the Inner Basin situation and the AWP, I simply don't see how 2019 can finish behind 2016/2011.
If you're gonna be wrong, be spectacularly wrong.

[Rich]

 :Anything below 4 5 is still theoretically possible including a BOE. 

If you get some kind of atmospheric conveyor belt that brings 5C heat to the pole for 2 weeks, the whole thing could implode. Not likely, but possible.

It's puzzling to me how some are eliminating below 3.5 when this year is the volume leader to date.

[Richard Rathbone]

2019 is not the volume leader by much, and 2012 was exceptional by high melt year standards, in how low its volume went in August, and how low its extent went in September. The mid July PIOMAS and July SIPN entries might persuade me that it could be narrowed but for the moment I have to leave 3.49 and 4.51 in there.

[Tom]

I've gone with 3 to 4.5.  I was tempted to include another low bin, but since this is extent and not area I think the chances of a highly compacted finish are low, and this would be necessary for a really low extent figure.

It's also 0.75 either side of my actual jaxa Sept min prediction of 3.75 in a solid 2nd place

[be cause]

I managed to eliminate 6 bins .. leaving 6 .. I feel I have almost certainly covered the outcome .. b.c.

[crandles]

Just over a day to change vote or make one.

Jul 18 2019: 7.19

Minimum decline to minimum from Jul 18th is 2.12 in 2001 and this is only year that declined less than 2.19.

So would expect most people to eliminate at least top 2 probably 3 bins and most have. 53 voted 48 eliminated top bin so we have 5 voters who seem hopelessly optimistic thinking it is possible to lose less than half the minimum loss from this date when the trend is toward higher losses in this period. They using wishful thinking?


At the other end, largest losses have been 4.12 well ahead of 3.77 then 3.51. With trend being towards higher values and if think this seasons is/will be exceptional might be reasonable to think under 3 is just about possible. Seems like most should eliminate bottom 4 bins and most have.
7 people not eliminated bottom bin.

5 deniers 7 alarmists? Or maybe some didn't understand?

[Tor Bejnar]

It is always possible, Crandles, that a few folks voted 'backwards', selecting bins they expect to be credible outcomes.  (I know, this means they cannot follow directions, but the directions are antagonistic to how we've always been taught to vote.)

[Paddy]

I feel happy leaving it on the relatively open 2.5 to 5. I feel it would be very unlikely to go below 3 or above 4.5, but I couldn't with hand on heart say that I was 99.9% confident it wouldn't doone of these things. And even then, events with a 1 in 1000 chance still happen that one time.

[Tor Bejnar]

14 hours and some; counting down. No excuses (except, "I don't want to.") to not vote.  There's universal enfranchisement here (until voting closes..). The more votes, the merrier!

Remember, vote "for" all the bins you're sure will not contain the JAXA September minimum.

And every visitor to the ASIF should be sure of something!

[Tor Bejnar]

2 hours left for voting. 

... No excuses (except, "I don't want to.") to not vote.  There's universal enfranchisement here (until voting closes..). The more votes, the merrier!

Remember, vote "for" all the bins you're sure will not contain the JAXA September minimum.

And every visitor to the ASIF should be sure of something!

And thanks!
[No animals were harmed while typing this message - Houdi the cat's litter box is clean and fresh food was presented to him within the half hour.]

[Tor Bejnar]

Thanks all who joined this experiment by voting!  By getting folks to 'vote away the chaff', and allowing different degrees of narrowness in their prediction range, I hoped to obtain a 'better average' than when we just vote for the 'best' [most likely] bin.  We'll find out, of course, how well we did in September/October.

We've created a fairly normal looking graph, centered on 3.5 - 4.0 million km², predicting the JAXA September daily minimum. 

Given the number of "Above 6 and under 1 votes, along with the four 3.5-4.0 votes, I wonder if a few people voted 'backwards'.  It is hard to teach us old dogs to do new tricks.

Would we have achieved the same graph if folks voted for 'all reasonably possible outcomes - 95%ile' - instead of the 'not-possible outcomes'?  Did you like getting to vote for a range of your own choosing?

Edit:  the chart's footer says "3.3" left.  That should be "4.3" bins left as possible (per average voter).

[Stephan]

The most likely range is the bin between 3.5 an 4 M km². I wonder whether the four voters who have excluded this central bin went for lower than 3.5 M km² or higher than 4 M km² with their choice.

[BenB]

Jaxa extent now 4,768,552 km2, so moving into the territory that some people considered possible as a minimum.