Final Arctic minimum sea ice prognostications
The final guesses are shown above.
In the comments of the last post, Crandles was nice enough to put a data table up. With minor corrections, that table is included below. Unfortunately, due to the lousy table support in blogger, I've simply screen captured an excel spreadsheet. Comment if you want the numbers.
Finally, here is a comparison of the aggregate guess to the Wegener Institute forecasts. 
Feel free to post links to other professional forecasts in comments, and I will add them. From a statistical point of view, our 'alarmist plateau' has a statistically strange shape.
As discussed in the rules, the prize is a blog post on the topic of your choice. If you're feeling cocky, post your topic now. Otherwise, continue to trash-talk amongst yourselves. And keep watching IARC-JAXA's latest numbers. We should have a winner by the end of September.
Thanks y'all for playing, and feel free to suggest ways of over-interpreting everybody's estimates.
19 comments:
Can you give a legend for the colors on the second graph. Our collective prognostication is easy to pick out, but what are the other two lines?
It would seem to me that only Penguin and possibly jyyh have put in suitable levels of uncertainty. Everyone else has reduced their uncertainty estimate to silly levels with the aim of claiming their spot at their favoured estimate (or claiming a wide range knowing the uncertainty is high).
Now if someone can win this three years running that will be very impressive and I may have to eat my words about "reduced their uncertainty estimate to silly levels".
Re "From a statistical point of view, our 'alarmist plateau' has a statistically strange shape"
Can you know that without running many of these competitions? Perhaps you already have?
I'm feeling pretty cocky, so I'll post a 'defend the indefensible' blog request.
Choose from: "Why Lord Monkfish is spot on about global warming" - or "Indesputable evidence that Al Gore is fat"??
- no less than 500 words mind ;)
I'm feeling pretty good about my position from 5262 to 5494. If it just does the average of what it did in prior years from this point forward to the low point, I might be right on the money.
Of course, plenty of time for an unusual weather pattern to pop up and send it somewhere else. But even so, the odds look pretty good it will remain above 5000.
Which Al, Gore or Yankovic?
Ah well, time to give up. Looks like my guess is somewhat off. There developed no significant blocking high pressures in Siberia this summer, could have guessed it for the PDO is still negative (isn't it?). Assumed the Chinook-winds would turn northwards during the whole summer melting the ice pushed there by a longstanding high in N Siberia.
Well I don't think it is going below 5m now, so it would seem to me that the most likely remaining contenders are James Annan, Hypocentre and Divalent.
better delete that last comment before anyone sees it :-)
James, it had crossed my mind to bid a pound then see if I could sue for 1000 pounds damages if I won and you are unable to deliver. But I guess I will let you off ;)
IJIS currently has 5254219 below Divalent's winning range.
Another few days of falling 41k km^2 per day and James may be able to express his earlier (over?) confidence again.
Anyway, I suggest a new competition where people have to submit their central estimate and uncertainty range for each of the next three years before 31 Oct 2009.
"IJIS currently has 5254219 below Divalent's winning range."
Not so fast! I think you read it wrong (there is a 5554219 in the table). According to the data, the minimum is 5,278,594 (its present value as of 9/11), above my low point of 5,262,000, so I am still in the money. Still have a 16,000 Km^2 buffer (nearly 0.3%!).
One Krakatoa explosion tonight and I might be able to hold on for the win.
I don't think I read it wrong - the value is updated twice a day but only the later one is saved.
The rules said
"The function with the highest value for x=minimum daily measured ice extent (from IARC-JAXA) wins."
'daily' not 'twice daily' as I suspect the saved later data point will be the one used.
So you are right you still have a chance and perhaps I should have known that when I posted but I hadn't looked up the rules nor did I know the later data. There have been earlier extent minimums so 16000 km^2 could be enough without needing a Krakatoa explosion but I think the odds are against it.
BTW I thought I gave enough of a clue with the 41k Km^2 fall. 41k plus the 5254k is 5295k which is the value for 10 Sept. ;o)
Crandles, three years is a long time in the land of blogs, and I certainly don't want to commit to doing this through 2012.
I'll probably do this again next year- but with an earlier closing date. I'm thinking solstice to perihelion. That's late enough for people to think they have an idea, but early enough for them to be wrong.
Once we have two or more seasons of guesses, you can take the arithmetic, geometric, and harmonic means of everyone who has played multiple times and tirelessly argue about which of those is a more accurate predictor of "skill".
Sciencewoman: sorry for the late reply, but the other two curves are model predictions from a German climate institute. The link is way back in a weekly update from early July/ late June somewhere.
Well, it's a moot point now. Update as of Sept 12 is 5,259,375, so it's officially dropped below my low end.
The NSIDC declaration is for a 5.10 million square km minimum, so it looks like James is the winner. Fitting, I suppose, given his occupation.
Congratulations James!
Oh, I don't think even I can claim that as a win. This pool was clearly based on IARC-JAXA which has higher values.
Well I think that is enough to call it as a victory for Hypocentre.
An increase of 98.5k Km^2 from minimum of 5250k Km^2 on 13 September and 4 new maximums since then makes that likely to be the minimum. A reduction to 5121k Km^2 to allow James to win would require a reduction of extent of more than three times the extreme rate shown in 2005. That seems pretty unlikely. So congratulations Hypocentre.
Does JAXA adjust its numbers weeks to months after the fact? I wouldn't want to award the prise, just to have an upward adjustment whisk it away...
Chuck,
I found an old saved spreadsheet of data to 9 Aug 2009. On the 2262 dates I didn't find a single change in the numbers between 9th August 2008 data and todays data.
So I think you can reasonably safely say that it is unlikely.
Post a Comment