Ok.. we’re almost to Nov 7!
The current JAXA number are:
8950000
8949844
8989063
9106719
9255625
-9999
-9999
-9999
This makes the average JAXA NH sea ice extent 9050250.2.
As you can see, at roughly a dozen of us bet low. Not too bad considering a few of the high voters admitted they thought they were betting on the NH max extent, and a few admitted voting high to communicate some message.
Obviously, some people bet high, some low. I bet a bit just slightly above the “established” JAXA trend for these seven days. I lost big time as you can see below:
Evidently, some “others” have suggested the ice – refreeze is nothing to get excited about because the over all trend is down. Well… yeah… duh. See above.
Meanwhile, of us paying attention to the baby ice for the fun, brownies and conversation are watching every reported pixel. Here’s JAXA’s graph, live:
OT, but I just read of the death of Michael Crichton. The article by Dennis McLellan of the LA Times, while otherwise great, did not mention one of Crichton’s most significant novels “State of Fear”, dealing with global warming hysteria. His “Author’s Message” from that book is worth re-reading.
It’s always sad to read of a death. I understand he was battling cancer. My aunt died of that.
With zero additional increase in extent the average would be 9.108 Mm2. If the accumulation is 150000 km2 for each of the next two days the average would be 9.173 Mm2. So the range of likely winners is from chopbox to MartinGAtkins with AndyW or Jared being most likely, I think. The median bet, 9.250 Mm2 is going to be quite close to the true value.
So is this going to be like the Price is Right, with the winner getting close to the actual value without going over or will the smallest absolute delta get the brownies? 🙂
Fred–
No. We’ll go by the absolute value of the difference. If there is an absolute tie, I bake two sets. 🙂
I’m hoping for some big gains in the next two days. Got to get that 9.250! Of course, if steven mosher manages to win this by lowballing my guess he will forver be on my no-christmas-card list. 🙂
Thre cheers for baby ice!
That’ll be three sets of brownies if 9.25 wins. Go baby ice go!
Fred,
You do understand that extent will have to increase by nearly 1,000,000 km2 over the next two days for 9.25 to win, don’t you? Even for 9.226 to win, the increase has to be 343250 km2, which is unlikely at this time of year but is at least in the realm of possibility.
God bless that baby ice. It’s coming home for daddy. :-}
earle… hehe. I had to pick somebody to piss off. My betting proceedure would have been better had I known that some boneheads ( sorry) had guessed max extent or bet to make a point. My point would be that one way to guess is to let the market of guessers guess and then bet late in the game using the information of other guessers to guess. That’s kinda what I did. Same procedure in the hurricane 2008 bet where I just handicapped Gray.
DeWitt Payne,
I wouldn’t go so far as to state that the gains required to reach 9.250 are out of the realm of possibility. Given that we’re looking at extent, some dissemination of existing ice strikes me as being possible. Now, I don’t think it’s very probable and should get started on writing my concession speech.
Latest mean is 9.100 using the early estimate of 0.098594 for day 6. doesn’t look like the mean will reach 9.2.
Preliminary 11/6 extent is 9354219 km2. That puts the average at 9.101 Mm2 and probably eliminates chopbox at 9.109 Mm2. No further increase in extent, highly unlikely, puts the average at 9.137 Mm2. An increase of 200,000 km2, 50,000 more today and 150,000 tomorrow, puts the average at 9.166. At exactly 9.137 AndyW wins. At 9.166, MartinGAtkins wins. Almost anywhere in the middle of that range Jared wins. It’s going down to the wire for a possible photo finish.
This daily average stuff is fun. Much better than waiting around for just one number.
Wow, better start writing my victory speech ending in probably –
” Can I eat those brownies? Yes I can. Yes. I. Can.”
I borrowed a bit of that actually, cough.
To be honest I think I am going to be a touch low and the brownies will be being eaten by Jared or MartinG.
Regards
Andy
recount!
8950000
8949844
8989063
9106719
9255625
9354219
-9999
-9999
Today’s running average: 9100911.667
(Yesterdays: 9050250.2)
That’s odd. It just changed.Now it is:
8950000
8949844
8989063
9106719
9255625
9384688
-9999
Looks like Jared or MartinGAtkins wil win unless some major freeze occurs. Jared gets it if today is are -9999 I think Martin gets it if there’s a measurement. The catch seems to be that yesterday’s extent is quite a bit more than the start of the week (almost half a million more). This seems to be not unexpected either – during October most 6 day periods have seen increses over 600,000
Revised 11/6 extent: 9384688 km2. Average through 11/6: 9105989.833 km2. With no increase in extent (hah), the average will be 9145803.857 km2. If extent increases by 150,000 km2 the average will be 9167232.429 km2. Either would make Jared the winner. MartinGAtkins can still win if the extent on 11/7 exceeds 9543561.002 km2, which would mean an increase of 158,874 km2. That’s somewhat unlikely but far from impossible.
Oh well, looks like I am out of it. My eyeball the 2002-2006 average approach (a decidedly geologist sort of thing) was good until the ice kept coming at a higher than average rate. Good luck to those still in the running.
I forgot to include a significant possibility considering past history. If there is no data for 11/7, chopbox wins. Data losses have been later in the month the past few years, but who knows.
DeWitt Payne (Comment#6388) November 7th, 2008 at 8:45 am
“Either would make Jared the winner. MartinGAtkins can still win if the extent on 11/7 exceeds 9543561.002 km2, which would mean an increase of 158,874 km2. That’s somewhat unlikely but far from impossible.”
Yup! Looks like Jared is gonna win by a head.
Well no data reported is out of the picture. The preliminary data for 11/7 is up: 9456563 km2 for an average of 9156071.714 km2. That’s not the final answer, it will almost certainly be higher, but it does eliminate chopbox.
On a side note, there seems to be some discrepancy in the Arctic ROOS area and extent charts. The top and bottom plots have different 2008 curves. Based on comparison with other data it looks like the top plots of area and extent comparing 2008 to 2006 and 2007 are now correct and the bottom plots comparing 2008 to 2007 and the 1979 to 2007 average are incorrect. The 2008 curves have risen too steeply in the bottom plots.
I will take my cue and bow out as gracefully as I can. If I may take any pleasure out of being so close without actually being able to taste Lucia’s delectable brownies, it is that at least I won’t have to explain how I came up with my own figure (though as best as I can remember, it had something to do with Joy of Cooking’s recipe for making prognostications with Hair of Squirrel Tea).
Lucia, it’s been a fun few weeks. It was a great idea to mark Baby Ice’s Remarkable Comeback in this way and I am surprised by how much I got into it. And who knows, maybe next year’s squirrel tea will be spot on!
If Jared wins then that will be something as his guess was way back on the 13th and he made no corrections!
Regards
Andy
DeWitt Payne (Comment#6409) November 7th, 2008 at 9:12 pm
“Well no data reported is out of the picture. The preliminary data for 11/7 is up: 9456563 km2 for an average of 9156071.714 km2. That’s not the final answer, it will almost certainly be higher, but it does eliminate chopbox.”
And probably AndyW (#5833) October 14th, 9,125,000, another good early
call.
Just in case there are JAXA updates, I will wait until Monday to announce the ‘official’ winner!
Current people ‘near’ the brownies:
Name: Bet amount. Difference:
AndyW 9.13E+06 -3.11E+04
Jared , 9.15E+06 -6.07E+03
Accumulating Average 9.16E+06 0.00E+00
MartinGAtkins, 9.19E+06 3.05E+04
adoucette 9.23E+06 6.99E+04
Jared is just below; MartinGAtkins just above. Of course, the average itself is spot on. 🙂 Unless JAXA “discovers” more pixels of ice, Jared is looking like the winner.
As expected, there was an update from JAXA at 9 AM that increased the extent, but not enough to change the (unofficial at this point) winner, Jared. Unofficial final result: 9160134.143 km2. Congratulations. I’m seriously considering going to the store to buy some chocolate and other supplies to make my own brownies since I won’t be getting any in the mail.
Well…. If you make them yourself, you can eat them warm! 🙂
Although I haven’t been looking at the JAXA data in detail for very long, I’ve yet to see an update past one day.
When researching the reason for the June 1st anomaly I did investigate the algorithm used to determine sea ice.
http://sharaku.eorc.jaxa.jp/AMSR/products/pdf/alg_des.pdf
The algorithm was designed by Josefino C. Comiso at NASA and as he says in the paper:
AMSR Instrument Characteristics
The AMSR is a fourteen channel, eight-frequency total power passive microwave radiometer system that will be launched aboard the ADEOS-II satellite in 2002. It measures vertically and horizontally polarized radiances
at 6.925, 10.65, 18.7, 23.8, 36.5, 50.3, 52.8, and 89.0 GHz. A second instrument, called AMSR-E, is a slightly modified version which will be launched on board EOS-Aqua satellite also in 2000. The ADEOS-II is a
morning satellite while the EOS-Aqua is an afternoon satellite. Data from the two systems thus complement each other and will all be acquired and processed by NASDA’s Earth Observation Center.
Thus I believe what we are seeing in the “9:00 am update” is really just the refinement based upon the inclusion of the second satellite’s data.
As to the bet, (since I don’t believe the 11/1-7 numbers will change)
There were 11 guesses within 100,000 km^2 of the final:
(97,000) 9,060,000 Bob North 11th
(57,000) 9,100,000 FrancisT 5th
(48,000) 9,109,000 chopbox 4th
(32,000) 9,125,000 Andyw 3rd
(7,000) 9,150,000 Jared 1st
30,000 9,187,000 MartinGAtkins 2nd
69,000 9,226,000 adoucette 6th
89,000 9,246,000 Steven mosher 7th
93,000 9,250,000 MarkR, E Williams, F Nieuwenhuis 8,9,10th
Which is pretty good considering this represents 28% of the bets.
Arthur
The extant appears to have changed again now we have
8950000
8949844
8989063
9106719
9255625
9384688
9485000
making the final average 9160134.14
I note that this number also seems suspiciously “round”
In other analysis news – 9485000 is the most ice ever reported on Nov 7 in the years 2002-2008. The second place year is 2004 with 9404063, some 80,000 less
Oops,
I apparently didn’t update my spreadsheet with one of the 9 am updates and so my final average was a tad low, as DeWitt posted the unnofficial final result of 9,160,000 km2 yields 12 bets (30%) within 100,000 km2.
(100,000) 9,060,000 Bob North 12th
(60,000) 9,100,000 FrancisT 5th
(51,000) 9,109,000 chopbox 4th
(35,000) 9,125,000 Andyw 3rd
(10,000) 9,150,000 Jared 1st
27,000 9,187,000 MartinGAtkins 2nd
66,000 9,226,000 adoucette 6th
86,000 9,246,000 Steven mosher 7th
90,000 9,250,000 MarkR, E Williams, F Nieuwenhuis 8,9,10th (by order of bet date)
98,000 9,258,000 Phillip Mulholland 11th
Arthur
FrancisT (Comment#6422) November 8th, 2008 at 9:50 am
I note that this number also seems suspiciously “roundâ€
Why do you think that 60704 pixels is ‘suspiciously round’?
OT, but lines like
“In other analysis news – 9485000 is the most ICE ever reported on Nov 7 in the years 2002-2008. The second place year is 2004 with 9404063, some 80,000 less”
just crack me up. Within “Extent” there’s 1.745 million km square of what’s NOT considered ice as on Nov.7, 2008
CT Arctic Area History:
1-Nov 7132000
2-Nov 7360000
3-Nov 7549000
4-Nov 7721000
5-Nov 7755000
6-Nov 7774000
7-Nov 7740000 < Negative change !!
Anyone offering to do the sums and do a bet on the first 15 days of November: Mean difference between CT Area and JAXA extent.
Enjoy the brownies 😀
“And probably AndyW (#5833) October 14th, 9,125,000, another good early
call”
I’m going to take that compliment and milk it for what it’s worth as obviously I won’t be eating any brownies any time soon, thanks Martin. At least Lucia does not have to send them to the UK. 🙂
I guess now the big competition should be maximum ice extent for 2009? This is a really interesting challenge, Lucia, you up for it again?
Regards
Andy
Sekerob (Comment#6426) November 8th, 2008 at 11:10 am ,
There is an offset in dates for CT. The drop in CT area supposedly for 11/7, corresponds to the drop in extent seen in the JAXA data on 11/2. Prediction: the CT area for 11/8 will show a small increase followed by much larger increases on 11/9, 10 and 11. I’ll have better data on the CT date offset when Uni-Hamburg gets around to reporting their October data, probably next week. You can also look at the previous 365 day Arctic graph on CT and see where the end of the data is compared to the monthly grid lines and compare to the current date.
To explain further: Uni-Hamburg reports area as well as extent from the AMSR/AMSR-E detectors. I have regressed the area data from Uni-Hamburg to the data I’ve collected from the CT iPhone page. The best fit is achieved when a lag of about 5 days is applied to the Uni-Hamburg data. Because I started collecting CT data in late July and Uni-Hamburg data only goes through 9/23 at the moment, the exact lag is less precise than I would like. That should change when Uni-Hamburg updates. I’ll post links to the updated graphs at CA on the current sea ice thread when the data becomes available.
Thanks for clarifying as indeed some of these changes did not look like being in sync as did the 365 day chart always seem to be behind, your grid comment. The 365 day chart should ideally be extended to 367 days for leap years and 366 for the regular so that a point to point comparison can be made. Now, 2 days at the start depending on melt or freeze season over/under pronunciates the difference.
As for collecting the daily, been scraping same since late August. Would be nice for the home charters if that daily data could be tabulated somewhere like IARC-JAXA does.
cheers
Well if the bet had been as of Halloween then I had it nailed! I expected the increase to decrease 😀 and it did, but not at the rate I expected. I didn’t expect the extent to surpass the recent years for this date even though indications are that it is colder than normal in Alaska this year.
hmmm, Why is there the 5 day lag when the NSIDC already published the monthly October figures for Area and Extent on the 3rd? Who wants 5 day old data on their iPhone?
ftp://sidads.colorado.edu/DATASETS/NOAA/G02135/Oct/
Get’s odd: Assuming CT of the 7th, is in fact the 2nd, the ratio Area/Extent goes up to 93.1% for the 2nd. Nowhere ever does the monthly ratio for November get over 88%, not even in the winter months does it go above 89%. Obviously a mean is the result of days above and below, but given that the Arctic is but for the Atlantic side a closed basin, the Bering Strait gap of 85km or so ignored, it’s hard for me to see how 93.1% could be. Explanations?
http://i137.photobucket.com/albums/q210/Sekerob/ArcticSIA-SIERatio2.png
Lucia,
You don’t have to wait till Monday. My sources, ie 2000 specially trained polar bears, have reliably informed me ( by methods that must remain untransparent for the moment) that the ice area and extent is not as the satellites would have us believe. They themselves are firmly of the belief that it is lot colder this winter than last and that there is more ice. Some. if not all, of them say my prediction looks realistic – mind you I’m still working on the statistical interpretation of their unusual and often highly individualistic language.
Dave Andrews,
Those polar bears are a great resource. Now, if we can give them GPS devices and get them to sample 2m air temperature and ice thickness on a grid! With enough polar bears, we could get measurements every 10 miles!
Lucia,
Good thinking. The latest Canadian data confirms previous survey reports that there are in fact way, way more polar bears than the WWF would have us believe. Since we are also up to our eye balls in seals [no small thx to Brigitte Bardot..] we might consider using those as the polar bears’ GPS way points on the grid. Two [Arctic] birds with one ice floe [so to speak]: fewer seals and reasonably reliable data points. 🙂
@tetris
While arguable not about climate as such, your point WRT polar baears and seals is worth pondering. While many consider sealing abhorrent (killing and suffering) they seems unable to realise that choosing to manage ice-bear populations with a goal to maximize (or at least build up the population) is a CHOICE -which implies at least as much suffering for the seals, as well.
Cassanders
IN Cod we trust
By my reckoning, the average of 1 and 7 is currently 8950000 + 9485000 /2 = 9217500, using the new peer reviewed DAYL (Do As You Likey) averaging.
I am also awaiting the inevitable +60000 upward adjustment in the Satellite data. According to my cut through a nearby Bristlecone Pine, (which is teleconnected to the Arctic,) I am sure that it, and not the Satellite will be proven correct. (smiley thing)
MarkR (Comment#6444)
“By my reckoning, the average of 1 and 7 is currently 8950000 + 9485000 /2 = 9217500, ”
All in favor, say Aye!
AYE!
#6439 and #6440
I’m suspicious about Dave’s specially trained polar bears, and wonder if there are many more than he is letting on. Perhaps it helps explain the rapid re-freeze and cold (ish) temperatures – hundreds of thousands of extra polar bears swimming around the Arctic Ocean as a cunning plan to increase the albedo!!! He’s influencing the results of his own experiment damn it!
OT: do polar bears like brownies do you think? (not suggesting they’ve been bribed)
p.s. yes i know it must have been pre-meditated well in advance for the albedo effect to have made much difference before the Arctic winter……..even worse……
Sekerob (Comment#6438) November 8th, 2008 at 3:03 pm ,
I’ve been calculating and plotting ice concentration using CT area and JAXA extent. I know, apples and oranges. Using CT area posted on 11/8 and JAXA 11/2 extent, the concentration is 86.5%, down slightly from 86.9% the previous day. It’s not at all clear to me how you arrive at your 93.1% figure from the chart you posted.
Good for you that you’ve been plotting it and know apples from oranges. Got the explanation for the latter. Picked up the 2nd of JAXA for 2007 ;>), thus the correct 86.5% does fall within range.
CT just dropped their figure, than the 3rd to 7.659.000. Makes it 85.2% which is still healthy compared to a 2007 monthly per NSIDC of 82.6%.
CT also increased their global negative anomaly from 123,000 to 467,000. Pretty steep, so I’d hold off on how CT 7th is going to compare to JAXA for that day. Good for you that you’ve been plotting it.
Remain for me momentarily the big ? why NSIDC can post the monthlies on the 3rd, quite above either JAXA or CT, where latter is 5 days lacking.
ttyl
Comment #6424 Arthur
Thanks for getting my name onto the list by choosing a band of 100,000 sq km about the seven day mean. At the risk of starting an argument (no scrub that statement, arguments are what blogs are for). Let’s see what would happen if we include in the list every estimate of the mean that lies within the recorded range of values for the 1st seven days of November. So every post that lies between 8,950,000 and 9,485,000 sq km is deemed to be the result of a skilful method of assessment (that increases the role of honour to 19 posts). The point is that for a method to compute a seven day mean that lies within this range, some of the 7 estimated daily values must overlap some part of the actual data range. Is this a suitable way of determining skill? Don’t know. Just asking.
Philip
I hope this is not too far off topic. In Sunday’s Wall Street Journal are two Opinion Pieces. One by Ian McEwan saying The Global Climate Crisis is upon us and the other one by Bjorn Lomborg. One sentence from the former is:
That’s it; a single sentence citing a single year. And it struck me that the results for this year weren’t mentioned.
Not too far off Dan!
Yes, I read the WSJ. There is no doubt the ice for 2007 was lower than predicted. Of course, it recovered a bit. We’ll see where it ends up next year. It’s all a bit like 1998, isn’t it?
Chris #6446;
Just a wag, but, how much heat do you think the bodies of the bears would leave in the water if there were enough to actually make an albedo difference??
Kuhnkat..
Who knows? But bear in mind, the bears must be well insulated to survive.
If you want a true engineering solution, you need to deploy the snow white reflective bears further south where they can reflect more sun. 🙂
Lucia,
yup, probably a “wash”.
Sekerob (Comment#6449) November 9th, 2008 at 12:11 pm,
2007 Arctic ice concentration was low compared to the recent average and 2008 is currently somewhat high compared to average. I calculate the recent average from Uni-Hamburg 2003 to 2007 average area data which is then adjusted to match CT average area and divide by JAXA average extent data for 2003 to 2007. Of course the CT average is 1979 to 2000, but the fit is amazingly good. For that fit, I have to lag the Uni-Hamburg data by three days.
There was a big increase today (0.217 Mm2) in CT Arctic ice area. It still looks a lot like the JAXA extent behavior earlier in November. JAXA data is a two day moving average. I don’t know if the CT data is smoothed or not. My guess, based on the fairly large day-to-day changes in the CT data, is that they don’t smooth. Arctic ROOS says that they do smooth.