Here's an update on the poll widget in the right side bar. A few weeks back I started a new poll to see whether expectations had shifted since the first poll. I have put up a third and final poll for these last few weeks that ends September 1st (so don't wait too long).
This is the result of the second poll:
According to my calculations the average is 4.28 million km2, which is slightly up from the 4.20 million km2 from the last poll. This is probably too low, considering the recent slowdown in extent decrease. But it's still possible.
What isn't possible, is that the minimum extent will be over 5.5 million km2, as IJIS extent currently stands at 5.34 million km2. So some people were wrong, three to be precise. One of them was from downtown Los Angeles, the other one lives real close to Deception Bay, and the third one lives close to Stepaside, which I find pretty funny:
I'm just teasing a bit. There's nothing wrong with being wrong. In fact it's very healthy. I'm wrong at least once a day, or so my wife tells me.
To balance things out a bit I'm showing an area with a high concentration of people who think/thought the IJIS minimum extent will go below 3.5 million km2. Isn't that area part of what they call Tornado Alley? Are people seeing some kind of correlation between sea ice extent and tornado activity?
So this blog's community could be projecting a bit too low (4.3 million km2), the WUWT community could be projecting a bit too high (5.0 million km2). If the average between the two comes about (okay, I know, different data sets, etc.), ie 4.65 million km2, maybe we should do a joint submission to the SEARCH Sea Ice Outlook next year? Now that would be fun!
Forgot to mention my complex method for processing the poll results:
I've calculated the number by multiplying the percentage number with the average of an option. For instance 'between 4.5 and 5 million' = 4.75, and for 'less than 3.5 million km2' I've used 3.25 million km2. I've added up all multiplications and divided the result by 100.
Posted by: Neven | August 21, 2011 at 14:22
My blog visitors bet on the 7 day minimum which will be a litter lower than the September average. Like your visitors, a few have clearly predicted two high. At least one of those who bet too high is a joke. (feabqtcy is a long running joke and he bet the ice minimum would be in the neighborhood of 7-8 millions sq miles.)
Based on my current 'fancy-schmancy' models, the 7-day Jaxa average extent minimum will fall between 3.95-5.03. Precision is required so we can see that 4 and 5 are both officially still "in"!
Posted by: Lucia (The Blackboard) | August 21, 2011 at 15:09
Eyeball technology says that NSIDC, the arbiter, has its 5-day average nearing 5.2, while the UB number is about 5.1.
Posted by: L. Hamilton | August 21, 2011 at 16:31
L. Hamilton-
NSIDC is the arbiter for bets using NSIDC. I have a bunch of people betting quatloos using JAXA. I used CT in 2008, but back realized it was a big mistake when CT didn't report the metric I picked for months.
We've been using JAXA for quatloo betting since 2009. So, to win the quatloos, the arbiter is JAXA. But yes, the values are different.
Posted by: Lucia (The Blackboard) | August 21, 2011 at 17:50
What Larry means, I think, is that NSIDC (numbers) is the arbiter for the SEARCH Sea Ice Outlook.
Posted by: Neven | August 21, 2011 at 18:19
L. Hamilton--
I hadn't looked at the mean differences, but the JAXA 7-day average minimum does seem to have been slightly higher than NSDIC September values. So, presumably, a group predicting JAXA should predict somewhat hither than NSDIC.
Posted by: Lucia (The Blackboard) | August 21, 2011 at 18:34
Neven--
Yes. NSIDC is used in SEARCH. Agreed.
I didn't poll my readers on that. It's too confusing to run bets on different agencies at a single blog. At the end of august, I'm going to go back and see how my various predictors work for NSIDC.
Sorry for the confusion. My main intended point is that some of my bettors have also already lost. It's not many, but some have.
Because my bettors attach names, I'm suggesting people tell me how they came up with their bets. The explanation isn't required, but it could be interesting. Of course Jay/Shoosh just put in a value that exceeded the surface area of the earth. But that's Jay/Shoosh.
Posted by: Lucia (The Blackboard) | August 21, 2011 at 19:50
237 votes, you certainly seem to get more visitors than comentators Neven.
And the old sea ice has foxed us all for yet another year. It is such an astonishingly dynamic enviroment. Her chaning moods seem to have captivated the whole spectrum of the climate debate as we watch the swings in weather and ice.
Posted by: dorlomin | August 21, 2011 at 20:12
How did the standard deviation differ? Are we "closing in" as September approaches?
Posted by: Bfraser | August 21, 2011 at 20:32
I originally guessed 4.5, but given events I think 4.7 to 4.8 is most likely for 2011.
Posted by: Tzupancic1 | August 22, 2011 at 03:38
Tzupanic,
What events?
Posted by: Frivolousz21 | August 22, 2011 at 04:02
This is chris b. And friv. can't get on to comment threw twitter or Facebook on the iPhone. Anyways.
We had a -60.5km2 prelim. That's from a 37.5k loss yesterday. So it was likely around 80km2 single day. The Slp will hammer the ice tomorrow too.
This leaves us at 5,274,500 roughly.
If we lose 40k rest of August thats 400km2 or 4,875,000 on Sept 1st.
2010 and 2008 are out now. Today showed us how thin the Ice is. The right anomaly and we could easily see a record on many fronts
Posted by: D | August 22, 2011 at 05:38
I was just booted off of Steve Goddards blog, and find this one to be MUCH less entertaining, but quite a bit more informative.
I made a prediction there in early July of 4.6 based on a heuristic dynamic model. That seemed the most fun approach, and it seemed the most reasonable at the time, Being an entertainer, I stole the concept of course, but I feel I should still get credit if it is close. No one asked me for a sigma. I am guessing that is a fancy pants way of saying "margin of error" (wow, it is also much shorter). SO I will shamelessly postdate mine and call it 157K.
Posted by: Tony Duncan | August 22, 2011 at 07:41
Ice extent did fall more than 60,000 square km yesterday. I'm pretty sure we'll be down below 5 million by the end of August.
Posted by: Rich and Mike Island | August 22, 2011 at 07:52
Who is Steve Goddard, really? His bio is, um, vague.
Posted by: Rich and Mike Island | August 22, 2011 at 07:55
Rich,
I should let someone else answer that, as I am still too close to the pain of the separation. though maybe nobody knows.
Posted by: Tony Duncan | August 22, 2011 at 07:59
What a shame, Tony. Last time I was at Goddard's, I thought you were doing good work. Did Steve ever write that update to '2011 tracking 2006' blotting out the other trendlines?
Posted by: Neven | August 22, 2011 at 08:05
Tony, I need to say that I have always been deeply impressed by how well you retained your composure out there on Goddard's blog. I posted there several times, and held up alright, but I find the intense bigotry and ad hominem assaults rather draining. Somebody there will find a way to mis-interpret what you say if you say something that does not fit in line with their belief system. Curious to see what made you be booted out of there, but maybe a break will revitalize.
Either way, I just wanted to express my respect to you for standing up to the dark side of debate, and happy to see you here in the real world, where real people are fascinated and impressed and respectfull and I think in general a lot happier and discuss real Arctic developments using ALL data we can get our hands on. Not just a cherry-picked interpretation of reality.
Posted by: Rob Dekker | August 22, 2011 at 10:58
Talking about "ALL data", ArcticRoos suggests that extent as well as area is getting close to the 2008 lines.
http://arctic-roos.org/observations/satellite-data/sea-ice/ice-area-and-extent-in-arctic
Does anyone know why ArcticRoos reports may differ this much from the other (NSIDC and IJIS) numbers ?
Posted by: Rob Dekker | August 22, 2011 at 11:04
Good question, Rob. Don't they use the SSM/I sensor? My problem with Arctic ROOS is the way the 2008 trend line jumps up and down.
Posted by: Neven | August 22, 2011 at 11:11
Posted by: dorlomin | August 22, 2011 at 12:39
Ignore the [fill in the blank]. He's been subjected to Tamino's statistical rigor and after found to have an no body left... the guy who will forever be known as having it snow CO2 in the Antarctic. He's of The Register's "GW Ain't True and certainly not man made" crowd, happily claiming that it has not warmed in the 21st century, and certainly been cooling since 2002 (in that department). Let's not talk about them as it gives them traffic and legitimacy for existing in the discussions. He's got none.
Me and myself, not back-peddling yet on the 'under 2007 for extent' bet aka 4-4.5M range, but the days are starting to get numbered to scale that point. My calculation for August 27 to pass 5M still stands though, unwavering. Second spot is kind of sealed, with lowest volume ever seen.
Posted by: Seke Rob | August 22, 2011 at 13:05
BTW, do I dare, Sept.18/19 is the first lowest 2011 extent point for JAXA, heuristic thumb tapping (The madcap laughs).
Posted by: Seke Rob | August 22, 2011 at 13:15
Hmm, comments from Lucia Liljegren.
Neven, please don't let any of this blog's threads turn into a Pielkean nightmare like her most recent betting thread. The post itself is fair enough, but note how she misunderstands, at length, the answers she gets starting here here, and finally manages to misconstrue even this crystal-clear explication from Crandles.
Posted by: Steve Bloom | August 22, 2011 at 13:33
Don't worry, I won't.
Posted by: Neven | August 22, 2011 at 13:35
Had a conversation recently with my brother, who was bemused that I followed sea ice so closely *as if* I'd bet money.
When betting, you might want to negotiate thresholds carefully, but for non-bets a simple round-numbers approach has appeal. In discussing my Gompertz graphs I referred to conditions as "virtually ice free" once the September mean NSIDC extent fell below 1 million km^2, or the September mean PIOMAS volume below 2 thousand km^3.
Those criteria seem good enough for discussion, especially because they're so easy to state clearly.
Posted by: L. Hamilton | August 22, 2011 at 15:16
Every single year since 2008 I've been like "This is going to be the WORST YEAR EVER!!!1!!" and ever year I've been wrong, so of course I didn't even put forth a guess this year and this year I would have been right.
Sea ice is a fickle mistress.
Posted by: Rlkittiwake | August 23, 2011 at 08:25