Continued from here .
On August 19, 2009, NSIDC published the following August forecast of sea ice minimums by the leading climate modelers around the world. The majority of modelers predicted that 2009 sea ice minimums would be below 2008 and one (Arbetter et al) even predicted that 2009 would break the 2007 record. The range was 4.2 to 5.0 million sq km. Detailed report is here.
While I don’t usually get involved in guessing outcomes of various climate situations (where no one has any real basis for their guess), I do occasionally. I called a low 2006 hurricane season very early. And on August 7, two weeks before the publication of the modelers’ forecasts, I observed:
2009 is now slightly behind 2008. My prediction is that 2009 will end up over 500,000 sq km behind 2008.
That prediction didn’t look all that great a couple of weeks later, but right now it looks pretty much right on the money. As of today, 2009 is 470,000 sq km behind 2008 and the chances of 500,000 seem pretty realistic.
That my guess was so close was due more to good luck than acumen, but there were some reasons for it. Canada has some exposure to northern weather and it has been a cool summer here and very cool in northern Ontario. 2008 had not been as big a melt as 2007 and presumably there was presumably a bit more two-year ice in 2009 than in 2008. While 2008 and 2009 were about even at the time, the trajectories looked different and it seemed to me that 2009 might stabilize at a higher level than 2008.
And yet in early/mid August, these factors didn’t seem to be on the minds of the official agencies since, as noted above, EVERY official agency substantially over-estimated the melt.
Climate Audit 
770 Comments
I saw a post on WUWT about the weather being more amenable to early freeze since last week. Hence the slow down in melt.
Funny how the weather can affect climate. 😉
re agency estimates – what we are seeing is the consensus at work, and not models plus data plus acumen. They need to answer why a layperson can more often better predict these things than them.
Re: pete m (#1),
That’s because they’re not guessing, they’re trying to develop a scientific methodology to be able to predict ice coverage. The Sea Ice Outlook group was set up to bring together methods and data to better describe the evolution of the sea ice because existing models greatly underestimate the loss of sea ice. Their estimates are based on using the previous months data and making an estimate based on their respective methodologies, in June more than half of the group were estimating more than 2008. What we are seeing is ‘models plus data and acumen’ not consensus! The current extent value is below the 30yr linear trend, combined with the continued loss of multiyear ice this is not a ‘recovery’. The problem with the models used at present (eg CCSM3 or HADGEM3) as Maslowski has pointed out is that their projections underestimate the ice loss in the western arctic and Siberian seas.
Re: Phil. (#6), That’s because they’re not guessing, they’re trying to develop a scientific methodology to be able to predict ice coverage.
Really? When I was young, it seemed that the distinguishing feature of science was its ability to do better than a couple farmers leaning over the back fence. Here we have the greatest minds of Arctic climate science, expending untold petas of calculations and with an initial go-round and two chances to make corrections this summer based on actual data, not one is going to successfully reflect the size of the extent at minimum.
Whereas the two farmers, looking at the graph, conclude that
(a) 2007 and 2008 broke out of the pack during the summer in an anomalous fashion, despite having been in the crowd on June 1, and
(b) anomalies being anomalous, and 2009 starting much like other years, then 2009 would end much like other years, and not like 2007 or 2008.
With those two stunning insights, the two farmers (and a good many followers of this blog) will have trumped the climate scientists with their PhD’s.
Now if the climate scientists were not following a herd instinct, one would have expected there to be a range of results centering around the obvious one. Instead, there was a range of results centering around the preferred headline.
Re: Rob Spooner (#27),
Really? When I was young, it seemed that the distinguishing feature of science was its ability to do better than a couple farmers leaning over the back fence. Here we have the greatest minds of Arctic climate science, expending untold petas of calculations and with an initial go-round and two chances to make corrections this summer based on actual data, not one is going to successfully reflect the size of the extent at minimum…….
Now if the climate scientists were not following a herd instinct, one would have expected there to be a range of results centering around the obvious one. Instead, there was a range of results centering around the preferred headline
That’s where you’re wrong, those making predictions here are doing so out of their emotions not science.
The scientists are trying to develop scientific prediction methods, since the recent history is of declining sea ice it is reasonable to expect that the current outlook would reflect that. The statistical approaches would obviously project further losses those methods involving lifetime of ice of known age/thickness will reflect the state of the ice at the beginning of the season which was the lowest recorded so these also will tend to show a continuation of the previous losses. One thing that is conveniently forgotten here is the published uncertainty bounds of those estimates, e.g. Kauker et al. estimated a probability of 80% that the mean September ice extent in 2009 will be in the range between 4.52 and 5.52 million km^2. The authors’ did have a chance to update their estimates based the changing data, however they didn’t change their methods. Of course those on here who changed their guesses in the last week or so are likely to be close, plenty of those who guessed here are already out of the picture because they guessed too high! It’s worth noting again that more than half of the estimates of the panel in June were for a slight increase compared to 2008.
Re: Phil. (#37),
Any chance we can get access to the code?
Re: steven mosher (#38),
Apparently it’s available here
Re: Phil. (#45),
Thanks. I assume they are running the stand alone version?
http://www.ccsm.ucar.edu/models/ccsm3.0/csim/
Re: Phil. (#45),
Looks like that code predates the big 2007 ice minimum.
http://www.ccsm.ucar.edu/models/ccsm3.0/
2004 for the original release, and 2006 for an update.
Re: steven mosher (#48), so what sort of data do these models use? Do they use ocean temperatures at (say) 50N, and do they use sunspot data which show we are at the deepest minimum for 100 if not 200 years?
What really amazes me about that graph which heads the thread, is how little change there appeared to be from the earlier predictions, given that the early August melt did not follow 2008’s and the July melt had not followed 2007’s. Really – two farmers leaning over a fence could and did do better. If I was one of those modellers (and I do modelling of a different kind) I’d be ashamed.
Rich.
Re: See – owe to Rich (#51),
The data are all listed in the associated files. WRT to sun spots. A GCM is a physics model. As such it would be driven by TSI ( for the past). For “projections” some models ( I recall) use a constant TSI and some ( I recall) may use a presumed 11 year cycle with TSI varying peak to peak. If you wanted to drive a climate model
(GCM) with sun spots you would need to.
1. Have a physical model that explained sunspots of the past and predicted them for the future. ( ha)
2. Supply a physics based model for how sun spots ( other than changing TSI ) impacted the climate.
Basically, the sun spot stuff I’ve seen is barely correlations and not causation. If we held that stuff to the same standard that we hold the team to they’d not survive the gauntlet of CA.
Please don’t turn this into a sun spot thread. there is plenty of space to comment at other places.
Re: steven mosher (#53),
They have no idea how the Sunspots affect Arctic ice.
Yet to be determined as opposed to the science is settled.
Science has barely scratched the surface of climate.
When predicting the best thing is to look at history and expect it to repeat. If the PDO turns cold the Arctic ice increases.
Re: steven mosher (#53), thanks for your answer. I won’t turn this into a sunspot thread – I just like to remind people occasionally. Re the TSI, Leif Svalgaard has data on that so it would be easy to put it into a GCM, and if you’re only predicting 4 months ahead (summer sea ice) use a currently constant value.
The harder part is that the correlations to which you allude suggest something stronger than TSI, like modulation of clouds for instance. But we don’t need/want to go there now and here, as it’s OT once one has checked what’s in the models – but for modellers who are good at twiddling knobs to fit data, I just suggest cranking up your solar sensitivity dial a bit!
Thanks,
Rich.
Re: Phil. (#45), CCSM3.0 is not a very good model in terms of temperature trends:
Re: Phil. (#37),
Phil. as always raises an interesting point. As he noted above the “models” had underestimated the artic sea ice
loss. Observations were out of wack WRT the models. In this case things were “worse” than the models thought.
I saw a graphic the other day ( FWIW.. posted by a AGW proponent, but otherwise I have no knowledge of it’s
Accuracy.
Now, when the models are out of wack with the observations, as Phil. notes, in this case people start to work on the
models. And they of course will look for factors that lead to this “bias”. They don’t question the data. They don’t
say “it’s just the weather” they don’t say ” anomalies happen.” they “work” the models. And of course they work
them in the “right” direction.
Now, contrast this behavior with the behavior you see when models get the temperature wrong. When the Models predict a warmer troposphere, a warmer surface. What you get is this. A lecture on how 8 years of data for temperature doesnt mean anything.
The scientists are not driven by emotion, but they can be driven ( or funded) by a desire to confirm a certain storyline.
Models that underpredict the impact of AGW ( on sea ice) must be fixed. This is a bad thing. Models that over predict warming, well cut them some.. they err on the side of caution.
For now we can only speculate on the changes made to models. That’s bad. Free the data, free the code and you will
free the debate from discussions of agenda driven bias.
Re: steven mosher (#41),
You know how they’re going to respond. It will be “See, it’s worse than we thought!”
Re: Phil. (#37), I don’t know where you get your insights into my emotional state, but I actually don’t care a whit about Arctic ice extent, the plight of polar bears, the bravery of the Inuit, or so forth. This is purely an intellectual point. The graphs could have been graphs of anything.
So getting away from the science vs layman squabble, as someone has suggested, what does this say about the statistical approach? Basically, that it is highly suspect if you’re using the data to decide whether you have a linear time trend. With respect to the thin-ice-driven models, one would expect that following a couple anomalously low minimums, it would take a while to return to normal. Bringing us back to the over-5-million default prediction.
If the scientists were unemotional, then it’s hard to understand why every one of them turns up low in a year with no major exogenous events. Same caveat as 29 above.
Here are today’s figures
9/5/2009 5340156 -25625 -27678
9/4/2008 4868906 -58125 -42031
9/5/2007 4484531 -43594 -25759
9/5/2006 5934531 -782 -3259
9/5/2005 5670625 -11875 -8214
9/4/2004 5899531 12500 -10290
9/5/2003 6116406 -32344 -31629
The difference wrt 2008, 2007 and 2005 are now respectively 471k, 855k, -330k. Looks like 200ç is going for 3d lowest extent. The ice started increasing again on 9/10/08 (which was quite early), so the difference with 2008 could change a lot in the following days.
Re: Flanagan (#2),
Noticing the latest picture today from the “pole” cam which was at 84.542 N 2.244 W yesterday. Camera internal temp was -2.5C and it runs a smidge warmer than the outside air. The atmospheric temperature instrument has stopped working. The ponds from the rain event a couple of weeks ago look frozen over though they don’t look snow covered.
It looks like my gamble of a September 8 minimum was off the mark but it isn’t over till it’s over.
Re: crosspatch (#3),
Actually it’s at 83.740°N 2.997°W, close to the Fram Strait.
Re: Phil. (#7),
The coordinates I got were for buoy 2009A at here which was a day earlier than the current picture.
Re: Phil. (#7),
Oh, DUH! That was a MONTH ago! Sorry. 08/05, not 09/05
Re: crosspatch (#3),
It’s been cold with many below freezing temps this week in the Canadian Arctic. Should be making ice.
http://www.weatheroffice.gc.ca/jet_stream/index_e.html
I think what we saw in 2007 was a lot of the dirty multi-year ice melting, and since then newer, cleaner ice has been gradually building back. It will be interesting to see if this coming autumn and winter will continue the trend of relatively rapid sea ice accumulation in spite of low ice concentrations at the end of the melt season.
None of these forecasters published their numbers without a ‘sanity check’ or PR check.
let’s face it if your model or heuristics or Stats gave you the same answer as the linear trend, you would check your
approach. Above that trend would also alarm you. And being above the 2008 would also give you pause.
In fact BEFORE they even did these studies you could guess that 80% would fall below the 2008 Min or be close to it.
Why? because of how 2007 was positioned. In fact, if you knew nothing at all about the state of the ice and were asked to Guess where the consensus of climate science would be.. you’d guess between the 2007 and 2008 minimum. Why?
it’s hard for them to believe anything else. Wish we had the code for how they guess.
Something strange happening at the IJIS website.
They have only just posted the Sept 4 data, 5,365,781sq Kms.
Their first Sept 3 data showed ice at 5,363,594 sq kms. With this extent the ice would have shown its first increase of the season by 2,187 sq kms.
On the 4th they upped the 3rds ice extent to 5,379,844 sq kms. With this it now it shows a decrease by 14,063 sq kms.
They are still 3 days behind. Strange.
oops the figures are just out. Sea ice decreased by 2343 and 14375
sq kms on the 5th and 6th. The melt is definitely slowing down. I predicted a low of 5,100,000 sq kms and the minimum on the 10th of Sept
Latest IJIS figures are:
09.03.2009 5379844
09.04.2009 5387969
09.05.2009 5363438
09.06.2009 5349063
So Sept 4 was the first day with an increase this year. It seems likely that the 2009 minimum will end up below 2005, but only just.
In sleepy Far North Queensland two guys were working their way through a slab of beer. Two lizards were slowly climbing up the wall to hunt moths at the light. There was a bet on. Mine on the left will get there before yours on the right.
Then the trend changed and the one on the right looked like winning. Out came a pistol and BANG – the one on the right was blown away.
“Hey! Waddaya do that for? There’s a bet on. That’s no way to win”
“Yep. Local rules.”
(I cannot comprehend several hundred posts on melting ice. Do you guys need to make up some local rules?)
Hopefully what happened this year can add to their knowledge pool to help in future years. Last year they underestimated ice thickness at the end of winter due to lack of snow insolation. Of course the weather does play a large part, but if they can more accurately pin down underlying factors then all for the better.
Regards
Andy
My revised minimum ice extent estimate – 5,277,000 sq kms
Local rules? There should be bets on the min extent and the date
I’m officially throwing in the towel wrt to my prediction that 09 would finish a tick above 08 (made July 29). The 150K melt that I had expected to occur late last week never materialised. In fact a close comparison between 09’s and 05’s minimum ice extent photos may be necessary to determine if 09 will indeed end up above 05. It’s that close.
Concerning the models above, normally you’d have expected them to be distributed about 5 – 5.3 million square meters. Instead they will ALL end up below what is likely to be the end result. This to me proves a that these modelers have relied too much on the bogus theory of steady man-made warming. They seriously need to examine why they all got it so wrong.
So dear modelers, some advice: Next year, leave out the widely assumed AGW. Get back to science, and stop producing sensational headlines.
What was 2005’s minimum extent??
Re: Tucker (#17),
2005 minimum = 5.315156
All of these forecasts, whether heuristic, modeling, or statistical use algorithms which are based on assumptions. One can use any jargon to describe them. but assumptions are still invloved in all of these forecasts. For all of these forecasts to be off of actual, and all on one side, means that the suthors of these forecasts need to look at how their basic assumptions impact their algorithms.
It would be reasonable to expect a random distribution with the mean value closer to actual than the highest value for sea ice minimum. Clearly, this is not the case with these forecasters.
My prediction for the Arctic – no matter what the minimum, new stories will focus on “it’s the age not the area that matters …. not enough multi year ice blah blah” with photo’s of polar bears on ice floes dragged from the archive.
Pondering multi year ice and noting that the seasonal variation from winter to summer is about 50% of area, does this give a simple proportion of each years remaining ice as 50% of the previous years minimum and how does this get modified for age? My simple thoughts start at 50% of remaining ice at the minimum as first year ice, 25% 2nd yr, 12.5% 3rd yr etc. which means that you wouldn’t expect to find much ice older than 4 or 5 years. If this was modified for increasing thickness due to accumulation with age, what sort of numbers/age does that give? Then how does the wind and water currents affect these number with flushing out of the polar region?
Not exactly Sea Ice, but related: I try to distill, and publish in my website, the best of the real science – seen as a self-taught layman – to help reclaim the science.
I’ve just done Circling the Arctic which collects onto one page all the Arctic best of the late John Daly’s magnificent collection of temperature records, many of which go back well into the 19th century – and of course, not a hockey stick in sight, in fact many show no overall temperature change at all; some show considerable fluctuation well in excess of overall “trends”.
I regard this basic evidence (using NASA GISS and CRU, no less) as of the first order of importance in appreciating the truth about the Arctic – in which a “hockey stick” of sudden-recent-never-before-seen temperature rise is unthinkable – just like the Emperor’s New Clothes were unthinkable to the little child.
Thanks to John Daly and, more recently, updaters like Paul Vaughan.
Re: Lucy Skywalker (#22),
Do bear in mind that John Daly was a self-proclaimed anti-AGW promoter. I complained to him once or twice that I thought we should have higher standards than the pro-AGW stalwarts, but he said there were tons of them and only a few of the skeptics. The point being that while I love all his work, he didn’t pretend that his data was unbiased. He picked stations which showed the opposite of the conventional wisdom, and thus the stations with the most warming won’t be found in his database. Just consider that before you make too much of his database.
Re: Dave Dardinger (#24),
Quite so. I haven’t analysed Daly’s work for potential bias and cannot provide a personal recommendation or endorsement. The focus of this site is on close examination of mainstream studies. If Daly’s views became mainstream, I’d spend time doing the same sort of analysis on his work.
Please don’t debate Daly here. This is a sea ice thread.
This fascination with Arctic sea ice was created by the oft repeated notion that the Arctic is the most sensitive region (the canary in the coal mine) to GW.
Trend lines continue to go down even after a bottom is reached. So if the Arctic is beginning to recover in a meaningful way, trotting out that axiom neither confirms or disproves anything.
One of the more enjoyable things about this site is it deals with observable and measurable phenomenon without hysteria. I thank everyone associated with the site for keeping it that way. How good does a web site have to be to hook you on reading about Arctic sea ice? BTW, that’s rhetoric.
Lucy If you want old temperature data, go to
http://rimfrost.no/
They have a lot of the oldest records around: Uppsala, Berlin, Paris, etc. from around the world. One of the more interesting one relating to Arctic temp is the record starting in 1873 from Upernavik NW Greenland, showing it had warmer spell in the 40’s then now.
Relevant:
http://www.newscientist.com/article/dn17742-worlds-climate-could-cool-first-warm-later.html?DCMP=OTC-rss&nsref=online-news
“Arctic ice loss in recent summers was partly a product of natural cycles rather than global warming” “There has been much less melting this year than in 2007 or 2008”
It does look like after a couple of very low years we are heading back towards the long term trend:
http://www.woodfortrees.org/plot/nsidc-seaice-n/mean:12/plot/nsidc-seaice-n/mean:12/trend
I’ve let myself get a little carried away here. It’s quickly getting cold up north and the actual likelihood of more melting due to air temperature is going to zero in a hurry. However, funny things happen to the ice and could yet in the next two weeks. The fat lady has not yet sung.
Re: Rob Spooner (#29),
Ths fat lady went to the all you can eat buffet.
Re: Shawn Whelan (#31), you lost me. Is the fat lady too stuffed to sing? Or is she putting off her big moment while she eats?
There seem to be only two years in the previous seven when a significant drop took place in the middle of September. They don’t seem to be related to the situation on September 1. 2008 was well down and turned early. 2005 was middling until about now, and then dived. The odds that we’ll see another late drop are low but not trivial.
Since the Arctic sea ice extent minimum depends mostly of the atmospheric circulation i.e. MPH trajectories, meteorological observation allows a thorough appraisal week by week of this evolution. That indeed should take the “surprise” out of the equation as per my post on Air Vent back in July: http://noconsensus.wordpress.com/2009/07/28/31-years-of-july-27ths/#comment-8439
Therefore, a meaningful prediction of Arctic sea ice minimum would mean proper modelling of atmospheric circulation phenomena, and in particular, understanding the reasons behind MPH trajectories changes and evolution. That would be climatology as defined by Marcel Leroux, not Madame Irma’s palm reading predictions.
Well at this rate crosspatch’s estimate of 8th September for minima may be best, unless there is another late dive like 2005 though, perhaps we are already on the cusp? Just going to stuggle past minima for 2005, perhaps. We needed flatulent cows up there this year and not polar bears for my guess to have been close at 4.8.
Regards
Andy
Re: AndyW35 (#33),
I made my prediction for area minimum, not minimum extent. Extent at this time if year is more a function of wind, in my opinion, than it is temperature. I believe area is a better indication of temperature.
Re: crosspatch (#47),
All these threads are based on the JAXA graph which is extent so the minima is connected to that rather than area; do you think ice minima day for extent will be signigicantly difference to minima for area and if so why? You can change you day if you wish ??
Regards
Andy
She is building up her strength for the big performance.
Turned awful cold in the Arctic.
Likely freezing is started.
Re: Shawn Whelan (#34),
Nice map, but what are those absurdly high temperatures throughout Siberia? Hard to believe that can be right.
Re: Willy Nilly (#44),
There is no error, temperatures are exceptionally high in Eastern Russia and Western Siberia at the moment. On the other hand it is colder than usual in Eastern Siberia. One should remember that Siberia has a very continental climate with hot summers and cold winters with very short spring or fall.
Incidentally the supposedly desertifying Sahel is flooded for the third year running, and there is rain in large parts of Sahara as well.
Weather, not climate.
Re: “many”
Can we move past the layman versus ‘science’ squabble? The questions of interest ought to be:
1) is the error exhibited by each model unusual based on expected variance?
2) why is the ensemble of models systematically low?
One apparent source of bias is the predominance of ‘runaway’ parameters–e.g., the use of multi-year sea ice area as a predictor–which reflect the tipping-point advocacy taking place and the effort to rationalize 2007 as indicative rather than as a weather phenomena.
Anyone else interpret Steve’s prediction as meaning 2009 will be more than 500,000 sq km of ice extent below 2008 levels?
Steve: I doubt it. Had that been what I meant, I would have placed it in a 2007 context. I was influenced in part from how cold it was in northern Ontario in June and July.
Re: MikeN (#36), Yes, I had trouble with the wording. I think the intention was that when 2009 is at its minimum, there will be 500,000 sq. km. more ice than in 2008.
Phil, I have no doubt that the methods of the groups are sophisticated. The problem is that their methods are wrong. Don’t be so pretentiously dismissive of people.
Not surprisingly, the forecasts seem to have been bested by the linear trend. How is that “based on emotion”? It seems to me it is based on the data.
“…That prediction didn’t look all that great a couple of weeks later, but right now it looks pretty much right on the money. As of today, 2009 is 470,000 sq km behind 2008 and the chances of 500,000 seem pretty realistic…”
Nonsense, Steve, you’re wildly out. You are working off base data. You should be following the concensus, and working off corrected data. Then you will see that the Arctic barely contains enough ice to float a polar bear.
The details of the corrections will be contained in my next missive, which you unfortunately cannot see because of commercial sensitivity. Or an agreement with someone who I can’t remember…Oh, no, I know what it is – I’ve lost them…
“2009 will end up over 500,000 sq km behind 2008.”
I first read that as you predicted something in between 2007 and 2008.
After seeing the charts, I guessed your quote back then was about ice melt.
The competition is not about sea ice extent on the day of the minimum, but mean extent during September. So their prognoses automatically are all higher than the very minimum they expect.
Team Ensemble 1 includes people from the Alfred Wegener Institute. In Spring they used an old DC 3 (Polar 5) to measure the thickness of the arctic ice; according to Radio Bremen they got up to twice as much as expected. They said they would use the information, and it appears they did so relatively successfully.
Re: Alexej Buergin (#57),
“The competition is not about sea ice extent on the day of the minimum…”
I don’t think that is true. According to SEARCH Overview:
“Re: Geoff Sherrington (#13),
“…two guys were working their way through a slab of beer…”
Geoff, I’ve had Foster’s and I’d swear it’s not that heavy or thick….
Re: rephelan (#59), A slab is 24 cans. Oz slang.
Re: pjm (#61),
Frozen or unfrozen? :/)
2009 is now 516,000 behind 2008. Have there been some revisions? 2009 seems to be gaining more overall, than the early daily reports indicated.
month day year ice dd diff
9 7 2002 5.663281 250 0.005937
9 7 2003 6.104375 250 -0.009375
9 6 2004 5.887344 250 -0.015781
9 7 2005 5.657031 250 -0.011407
9 7 2006 5.921875 250 0.003437
9 7 2007 4.436719 250 -0.010312
9 6 2008 4.808281 250 -0.017344
9 7 2009 5.324688 250 -0.020468
21,000 loss….we will see how it holds up tomorrow at 10:00 AM eastern time
It was fun to track this melt season. Many surprises for some and few for others. Up until mid-July I was able to graph the progress of this season’s melt until taking on a second job, which left me little time for special projects and only enough time for quick glances at the JAXA site.
When I saw these predictions posted in the entry above by some of our leading climate modelers I found it quite shocking that ALL of them over-estimated melt (barring some unforeseen September Arctic heat wave/compaction event).
I had to search through a few blogs to find my prediction for this year’s melt season. I left it in the comments section of a friend’s blog on wunderground back in May.
“27. sullivanweather 5:01 AM GMT on May 05, 2009
Hey, Jer!
I added a 2009 arctic seaice watch section on my blog.
Everyday I will have posted the extent and daily ice melt in text form. Plus displayed in graphical form will be melt rate compared to 2008/2007 AND difference in ice extent between 2009 and (2002-2008).
As of 5/3 2009 was ahead of all years in the JAXA period of record with 2008 in second place ~180,000km^2 lower.
2006 currently is in last, 840,000km^2 lower.
This summer’s melt season is looking interesting and may catch some of the ‘experts’ off guard. Ice is a bit thicker this spring across the North American side of the Arctic which may keep more ice north of Prudhoe Bay in the Beaufort than last summer. With that being said and deference paid to the unusually cold spring in the Canadian Arctic thus far (something that looks to continue for some time as well) the Northwest Passage should remain closed as well.
Melt still looks strong on the Russian side of the Arctic but I think when averaged together with the conditions expected on the North American side, melt should come in lower than both 2007 and 2008.
I’m expecting a minimum of ~5.3 million km^2 around 9/14 (using an average min date here). That may still be good enough for third lowest (maybe 4th after a fight to the finish with 2005) but it’s also about 1 million km^2 more ice than 2007, a decent build-up. “
Re: Tom Woods (#66),
sullivanweather’s comments were pretty spot on apart from those on the NW passage which opened up just to prove hengav wrong 😀
I’m hoping for a really low maxima figure this winter just for a bit of variety and see how that impacts next summers figure.
Regards
Andy
A bit over minus 20k for today’s loss. Still havent quite reached 2005’s minimum. I’m surprised Phil hasn’t chimed in with “the ice is melting” comment and “2009 has a bigger loss than any year since 2004”.
J/K Phil, I actually love your posts. This year has been a ride and I think we are figuring out that 2007 was more of an anomaly rather than a trend.
9/7/2009 5324688 -24375 -17500
9/6/2008 4808281 -17344 -40245
9/7/2007 4436719 -10312 -24330
9/7/2006 5921875 3437 -7544
9/7/2005 5657031 -11407 870.43
9/6/2004 5887344 -15781 -6406
9/7/2003 6104375 -9375 -25401
A pretty decent day for 2009, with the strongest lost (before correction). On the other hand, 2008 still has a few strong days coming soon. In 2005 the minimum was 5315156 on Sep 22.
Molon Labe Nr 64
The background page:
http://www.arcus.org/search/seaiceoutlook/background.php
(Notice 3. )
The probable winner (paper from July):
Click to access Sea_ice_outlook_2009_AWI_FastOpt_OASys_contribution.pdf
Their prognosis: “With a probability of 80% the mean September ice extent in 2009 will be in the range between 3.9 and 5.3 million km^2.”
(A prognosis looks better when you use 1.3 standard deviations.)
With the melt/compaction season coming to a close, now we will need a thread about the “growing” ice season to compare with previous years. Oh what fun that would be Steve! 🙂 I think my Sept. 11th ( As an American I almost hate typing that day) prediction will be too soon for the minima.
I’m not sure if we’re comparing apples to oranges or maybe galas to fujis. I looked at this page at NSIDC and found that they had 6.26 mk2 on August 17th. I looked at the JAXA data for the same date and got 6.11. So how bad the predictions are is going to depend on which source they’re using, so JAXA may even be making them look better at this point. I wish they’d post the numbers that make up their charts like JAXA does. It would make analying this much easier, but then maybe that’s why they don’t.
Re: BarryW (#73),
Those predictions still seem terrible whether its 100k off or not. So many of them had 2009 ending with less extent than 2008, which looked obvious not to happen even a month ago despite those predictions being issued more recent than that. You wonder what they were looking at when making those.
Re: NEwxIce (#79),
Well with respect instead of ‘wondering what they were looking at’ perhaps you should go to the link Steve provided and find out! Also a month ago the 2009 and 2008 values were ‘neck and neck’, their predictions weren’t ‘issued more recent than that’ they were dated the first few days of August and were based on July data often early July (see Phil. (#81) for example).
Re: Phil. (#82),
The fact it wasn’t ahead of 2008 at that point should have been a warning bell. I know most people here didn’t think the 2009 melt would be as drastic in August as 2008 was…which is why most here also thought that 2009 would finish well above 2008 for extent, simply because of how anomalous the late ’08 melt was. Any way you slice it, that’s a poor set of predictions when you look at the end points and the mean.
Re: NEwxIce (#84), Re: Johne S. Morton (#85),
Which shows you haven’t read the reports and don’t understand what is being attempted here!
This is the second year of this program, last year they overestimated as a group this year likely under but a lot has been learned.
Re: Phil. (#86),
So this is the Goldilocks school of Arctic ice prediction? Next year they’ll be just right!
If that doesn’t work, I suppose they can go back to their previous favorite fairytale: Little Red riding hood; and scare us with the Big Bad wolf.
Re: Dave Dardinger (#89),
Given your immature post and fascination with nursery rhymes I assume you’re still at elementary school, come back when you’ve learned something about science!
Re: Phil. (#82), If you look at the months of May and June, the last several years had fairly similar sea ice extents, with 2009 being in the middle to upper part of that range. There wasn’t much reason to suspect values as low as those predicted.
http://www.ijis.iarc.uaf.edu/en/home/seaice_extent.htm
Well, Barry you might be right! The Sep data found on the NSIDC site
ftp://sidads.colorado.edu/DATASETS/NOAA/G02135/Sep/N_09_area.txt
strangely corresponds to what is referred to as “minimum 2008 extent” in the above graph.
If the difference with JAXA remained similar, then the NSIDC extent (that should be compared with the above) would be around 5.17.
Anyone check out the south pole lately. The sea ice extent is headed straight up.
Re: realist (#75),
The area isn’t. It’s a little (0.35 Mm2) above average, but the rate of increase is well past its peak and slowing at a very average rate.
Calculating average loss to minimum by reference to the measured minimum day rather than calendar date and comparing the current smoothed rate to the smoothed average rate, the minimum is expected 7 days from now with an average loss to minimum of 115290 km2. However, there’s still a large range on the estimate of the loss to minimum.
The area curve is very flat right now. The 3.571 Mm2 area on 8/30/2009 is still the observed minimum.
With correction, 9/7 finished at 5,328,906 for a daily melt of 16,250.
Am I the only one struck by the size of this range? And even with such a huge target it looks like the actual figure will just barely fit.
Re: Soronel Haetir (#80),
That was the result of their calculation with a starting point of May 22, their results with a starting point of July 11 were between 4.52 and 5.52 million km^2. I don’t think that the spread is particularly wide considering the method they used.
From the Open Passage news letter we have:
Maybe we will hear from the rest. I was trying to find out if the Canadian Geographic ship is able to follow its planned trek through the northern part. It was to sail Sept 1-17.
Re: Gerald Machnee (#83),
Fiona (the one that called the CCG) is entering the Bering St, Ocean Watch is off Newfoundland, Baloum Gwen was on Bylot Island with the Silent Sound yesterday, Bagan is also entering the Bering St. Fleur Australe was in Nome about a week ago and are now in the Aleutians. The two marines ended their trip in Gjoa Havn due to head winds over the last couple of weeks.
So three through from the East and three through from the West.
Re: Gerald Machnee (#83), Sounds like they need to change the name of their organization…
So in about 6 months will we be told that the buildup of ice is lower than the last two years?
Phil,
I find it amazing that y’all still play with funny numbers.
Trends are only meaningful when the underlying CONDITIONS are consistent in their change causing the trends.
When you have changes in temperature, and especially currents and winds, that vary from those that established the trends, the trends become meaningless.
Re: kuhnkat (#92),
I find it amazing that you don’t avail yourself of the ‘reply & paste link’ so that I’d have a clue what you were talking about?
Those forecasts are a joke although to be fair the date when they were made is important.
Steve is dead right to point out the forecasts are lop sided, why?
I can post forecasts for 2010 peak and 2011 dip although I would prefer to wait a couple of months.
Here are two new arctic ice snippets, best use PDF.
First one shows the longer satellite data without the annual cycle, shows fair game for funny money projectionists.
The second one shows what I found when I wondered about that 1996 wiggle, what gives?
Drop UAH North Pole over the top and a nice similarity appears. Doesn’t explain what caused the temperature event, just that there was one. Not checked for SSW.
Could be ice disappeared and the effect was seen by the satellite, or it might have been temperature causing the ice to disappear. Or something else.
“Canada has some exposure to northern weather”
LOL!
20,000 loss today……but if recent history is any indicator that number will drop somewhat. be interesting to see what tomorrow holds. i am still holding out a sliver of hope that we end above 2005. needless to say, if this number stands we go below…..but with an up day 5 days ago and an almost certain correction tomorrow morning, we still have a chance, albeit very unlikely.
According to CPC it looks like 2009 had the lowest (or at least close to) Arctcic Oscillation value in 50 years.
http://www.cpc.noaa.gov/products/precip/CWlink/daily_ao_index/month_ao_index.shtml
So this would mean lowest average atmospheric pressures, and presumably much more cloud cover than normal?
What is odd is that the AO seemed to be most negative early in the melt season when we had fastest melt, and switched to positive recently as the melt slowed down significantly. Have I somehow got my interpretation of the AO backwards?
If I’ve got this the right way round this would suggest that the strongest AO cold event in 50 years has influenced the Arctic melt this season, and presumably we can forgive the models their failure to take into account this possibility. From my reading none of the models tried to predict the likely weather conditions for the 2009 melt season, but imposed weather patterns based on recent history.
Re: Michael Hauber (#97),
Judging from the pictures this year from buoy 2009A this year was much cloudier than the past two. There are some webcam animations here but I haven’t looked at them in a while.
“Steve McIntyre:
September 7th, 2009 at 9:27 pm
2009 is now 516,000 behind 2008. Have there been some revisions? 2009 seems to be gaining more overall, than the early daily reports indicated.”
I am keeping the daily differences from 2008 to 2009:
1 September 2009 459,063
2 September 2009 474,062
3 September 2009 452,813
4 September 2009 519,063
5 September 2009 537,813
6 September 2009 536,875
7 September 2009 589,062
8 September 2009 593,125
Its been going up ever since the 23rd of August.
They have just posted the 8th extent 5,308,594. The melt at this figure is 20,312, but everyday they have been posting an update increasing the extent and reducing the melt, since the 2nd.
Visually the ice seems more on the 8th than the 7th. See here. Use the drop down arrow below “Show one image”, choose “Select time step” and toggle between the 7th and 8th.
Re: Richard (#98),
Steve usually allows for the leap year which might account for the difference.
The NSIDC latest report is out which is as interesting as always, available on their front page so I won’t link to it. One thing that made me chuckle was the statement “While this year’s minimum ice extent will probably not reach the record low of 2007…”. I’d happily back them up on that none too risky claim! Some good comparitive diagrams for sea loss by region for the last few years.
My 13th as extent minima is still looking not so bad.
The Royal Marines stopped as Phil said, they are now off to Afghanistan for their proper day job. I have the utmost respect for that sort of person; I have to admit they are lot different to me who favours safety and comfort. Hope they can return next year to finish the route.
Regards
Andy
snip – piling on
Note that the latest extent is actually lower than the 2005 minimum – at least before correction.
Re: Flanagan (#103),
Yup, but it does look like we are going to be headed for an “unprecedented” increase in minimum extent over a two year period. Not that there’s anything wrong with that.
Re: jeez (#104),
No indeed, but you might want to look at 1990-1992 and 1995-1996.
Re: Phil. (#105),
Think percentage increase of ice extent, not the anomaly graph. If the melt stops around 5,250,000 we are looking at about a 23% increase in 2 years. I believe by all accounts that would be considered “unprecedented”.
snip – bickering. I also snipped the piling on comment to which this was responding.
Re: Phil. (#106),
We have to let the old ice build up again. At this point in time I would wager that the probability is greater that next year will be greater than this year and the recovery will continue provided the winds cooperate. In fact, if temperatures continue to decline, we may see Arctic ice minimum at average or possibly even above within 3 years or so.
And according to NCDC, the continental US temperatures in August were lower than last year. Over the latest 12-month period since 1999, we are in a 0.9C/decade cooling trend. That is a very significant cooling.
Go here, in the “Period” pull-down, scroll all the way to the bottom and select “latest 12-month period”. For “First year to display” select 1999. The trend is -0.9C/decade. Last month it was -0.86C/decade.
We are currently less than 0.5C above the 1901-2000 mean.
Since one man with a brain did better than 13 teams with all the information and every technical help aviable, we must conclude that this is still an art and not yet a science.
When practicing the art it helps to look at more than one internet-site; Nansen use different satellites than Jaxa and different smoothing, and this summer both were correcting each other.
Re: Alexej Buergin (#108),
I don’t regard my guess as a “prediction” (nor do I regard similar guesses by others.) That’s one reason why I don’t usually get into such games.
All such guesses do is indicate a sort of null distribution, not “skill” on the part of the guessers.
What we don’t know right now is the basis of what the climate centers did? Were they calculations from climate models? Were they calculations from time series? What is a “heuristic” estimate – as many of them are called? Sounds like a guess to me.
Maybe a reader could look at the underlying reports and provide some information on this. Maybe ask one of the most aggressive predicters how they got their result.
If their estimates result from models, then it would appear that there is some sort of systemic bias in how the industry prepares their models, as it is otherwise highly improbably that ALL of the estimates should have erred in the same direction.
Re: Steve McIntyre (#112),
Steve there’s no excuse for such ignorance since all the information you mention is available at the link you provided.
I have but apparently no-one here wants to know that, they appear to want to continue with their fantasy about what was done.
[Steve: What an unfair thing to say. The very words that you quote include a request for a reader to summarize information on the projection methodology to save me some time. Why would you say that I wasn’t interested in methodology?? Please.]
You’re always talking about openness in the process yet when there’s a research program being carried out in the open no-one is interested!
Well they didn’t did they, at least three got it right, if you want to find out which ones and how you should read the reports.
Steve: I’m covering a lot of topics here, not just sea ice. I have only so much time and energy. What’s wrong with asking someone to do summarize the reports.
Re: Phil. (#114),
The herd has congregated around its hopes; three are mooing at the ice ghost they see.
=====================================
Re: Steve McIntyre (#112),
Currently reading “the Undercover Economist” by Tim Harford. Quote on modelling of the cost of coffee including Starbucks examples:
“Each model is useful for different things, but a ‘model’ that tried to describe the design, the engineering, the ecology and the economics would be no simpler than real life itself and so would add nothing to our understanding.”
Knowing when to get into a project is 30% of the work. Knowing when to get out is 70%. Just like mineral exploration, eh?
p.s. Did you get an email from me yesterday on CRU? It came up as not transmitted, no such email address, but also reported OK as a sent item.
Re: Steve McIntyre (#112),
Maybe they overestimate the climate sensitivity to a doubling of CO2 concentration, again.
In terms of energy budget, would not the area under the curve for a year be a better metric that the minimum cusp on some day in September?
Re: Geoff Sherrington (#110), Geoff, an interesting twist would be to factor insolation into the calculation. A reduction of ice in high-insolation June does considerably more to heat the Arctic than ice lost in lower-insolation September
Richard…i agree! it looks like the ice increased from the 7th to the 8th. it will be interesting to see if this morning’s correction reflects that.
Re: markinaustin (#113),
“They have just posted the 8th extent 5,308,594. The melt at this figure is 20,312, but everyday they have been posting an update increasing the extent and reducing the melt, since the 2nd.
Visually the ice seems more on the 8th than the 7th.”
Richard…i agree! it looks like the ice increased from the 7th to the 8th. it will be interesting to see if this morning’s correction reflects that.
Sure enough they have posted an update and guess what! This time it has gone up by a whopping 21,875 sq kms! and the ice extent is now not less by 20,312 sq kms but it has INCREASED by 1,563 sq kms the second time it has done so. (The first on the 4th of Sept)
3 got it right?
Maybe you are reading a different report to me.
Re: Etienne (#115),
There’s more to reading a report than looking at a summary graph, I suggest you reread it more carefully this time.
3 got it right?
The only prognoses that are really challenging are those from June.
I am sure that they are very busy trying to find out what went wrong; but I believe they would not tell it to 2 categories of people:
1) Those who know a lot about statistics and models
2) Those who do not know enough about statistics and models
Phil 114: Pleas provide a link to to August reports of the teams that got it right
Re: Alexej Buergin (#117),
I’m busy in the lab it shouldn’t take you more than 5 mins to find them.
Current Extent is 5.30
Kaleschke/Halfman’s range goes as high as 5.35
Lindsay’s prediction range goes as high as 5.20
Stern et al goes to 5.09
Perhaps those are the 3 that are right?
Re: Agland (#119),
Morison et al
“It” hasn’t officially happened yet and won’t until NSIDC releases the September data some time in October. So saying someone got “it” right is a tad premature.
The 10AM correction was substantial and the change from 9/7 to 9/8 is now positive (1563 km2). That also means 2009 stays above the 2005 minimum for another day.
The corrected extent is out – showing an increase! Could it be we already reached the minimum?
Re: Flanagan (#122),
Yeah that was a large revision upward.
Phil., I do not follow this thread in great detail, but I would suggest that you could make your case clearer for readers like myself by including some more details from your references in your posts.
You seem to be saying that others do not read in sufficient detail and then do not present details of your own.
Re: Kenneth Fritsch (#125),
No Ken I’m busy in the lab I don’t have time, I’m not the only one here who can read a report surely?
Re: Alexej Buergin (#127),
See Alexej managed it!
Re: Kenneth Fritsch (#125),
Hear, hear. [Yes, I’m piling on.] Saves me from having to retype Ken’s words, which mirrored my thoughts this morning. You can’t handwave about “reports” without linking to them or be willing to summarize the most salient points. If Phil’s busy today, then perhaps tomorrow?
Maybe the details are simply what you get when you go to the top of this here an click on “detailed report is here”, then go to the bottom and click on the desired name. They describe what they changed since July.
Read the last part of page 3 of Dr. Walt Meier et al: If I understand it correctly, they did what we all do. They took the momentary value, looked what the ice did in preceding years, combined that and got their result. That’s all?
Most results are mean(+-)standard deviation; Ensemble 1 has an 80%-Intervall, others do not say. Well, what should be the criterion to say if the prognosis is correct? Just use 2 or 3 sigma and all are. But all mediums might be too low.
Flanagans Predictions:
Flanagan: August 11th, 2009 at 12:53 am .. 2009 was slow again, which I find personally surprising in view of all the low-concentration ice there is up there. Maybe there’s not enough wind?
No maybe you could blow on it a bit
Flanagan: August 12th, 2009 at 9:18 am That’is it now, we’re all doomed. The Arctic melted completely during the night.
Good god we are indeed
Flanagan: August 13th, 2009 at 12:51 am …So it seems 2009 actually just beat the JAXA daily melt record .. I think the most honest conclusion is that we’re back to the 2008 level..
Flanagan: August 13th, 2009 at 9:01 am Mmmm… What I had in mind was more like 4.7-4.8 million km2
Flanagan: August 19th, 2009 at 1:33 am So here we go….Imagine that by Aug. 27, the 7-day averaged daily rate dropped to -18000 in 2005! IMO, the real “fight” is with 2008.
No H?
Flanagan: August 24th, 2009 at 2:08 pm ..In the meantime the extent was revised and the loss is “only” -82656 km2.
Flanagan: August 31st, 2009 at 9:01 am Hi there,I prefer to refer to “consensus” in terms of models/scientific projections. ..
I bet you do!
“..arctic ice decays faster than antarctic ice, and northward ocean heat transport increases into the northern high latitudes), models have poor agreement on the amount of thinning of sea ice (Flato and Participating CMIP Modeling Groups, 2004; Arzel et al., 2006) and the overall climate change in the polar regions (IPCC, 2001; Holland and Bitz, 2003). Flato (2004) shows that the basic state of the sea ice and the reduction in thickness and/or extent have little to do with sea ice model physics among CMIP2 models…” IPCC Climate Projections
Back in 2005 there were 3 daily increases in extent before the minimum – the 2nd, 3rd and 14th of September. After the 14th sea-ice extent in 2005 still dipped by over another 200,000km^2. Not saying that will happen this season but some of the 5.1’s and 5.2’s out there still have a slim chance of happening if sea-ice compacts in the right area. Although I would personally say that any significant drops look unlikely now, especially considering the week 2 forecast being put out by some global weather models showing cold building very quickly after the 16th of September.
Now nearly 600,000 km behind 2008.
month day ice dd diff
2002 9 8 5.649688 251 -0.013593
2003 9 8 6.103750 251 -0.000625
2004 9 7 5.856563 251 -0.030781
2005 9 8 5.642656 251 -0.014375
2006 9 8 5.935781 251 0.013906
2007 9 8 4.413438 251 -0.023281
2008 9 7 4.739844 251 -0.068437
2009 9 8 5.330469 251 0.001563
Re: Steve McIntyre (#132),
I make it 615,000 Km^2 behind 2008 as per IJIS data.
Ice extent 8 September 2009 5,330,469 and on 8 September 2008 4,715,469
HEre is a link for the Rigor et al algorithm-based projection of a new minimum of 4.3 million sq km: http://www.arcus.org/search/seaiceoutlook/2009_outlook/august_report/downloads/pdf/panarctic/2_Rigor_etal_AugReport_JulyData.pdf
I guess that their model must have left something out.
Re: Steve McIntyre (#134)
It’s Funny, There Looks Like A Good Deal Of Rigor In That Discourse, But Appearances Are Deceptive. More Rigor Needed, Or Less Rigor Needed? What-do-the-coauthors-think?
Rich.
Re: Steve McIntyre (#134),
Obviously they got the thickness of the ice wrong at the beginning, did not hear about Polar 5 (or ignored it), and stayed with the model to the bitter end. That does not mean that the model is useless, only that we might not have enough information in spring to do calculations instead of guessing. In addition to that no one knows the summer WX in advance.
Re: Alexej Buergin (#139),
Even now after the second year in a row of decreasing Arctic ice the consensus of science still acts as though the ice has been increasing for the last two years.
This is like the housing market. You can’t base future prices on the trendline.
Re: Shawn Whelan (#139),
I think “eerga I” to what you wanted to say.
“Agland 119
Current Extent is 5.30
Kaleschke/Halfman’s range goes as high as 5.35
Lindsay’s prediction range goes as high as 5.20
Stern et al goes to 5.09
Perhaps those are the 3 that are right?”
So Kaleschke looks better than Lindsay but only because his SD is 0.43 and Lindsay’s 0.26. And Ensemble 1 goes up to 5.52 because they use 1.3 SD.
I believe they prognosticate the September mean, which will be higher than (behind?) the lowest point. Perhaps a prognosis should be right if the predicted mean ist above the minimum?
Ice from the 8th to the 9th. Again visually it seems to have increased, everywhere except off the East coast of Greenland and the NW coast of greenland. Overall I think it has increased a smidge. Methinks IJIS will again show a decrease followed by an update increasing the extent.
at this rate, we may actually be at the bottom…..but only time will say for sure. interesting to hear that a few years ago there were some upswings (2005) before losing another 200,000, but as stated that seems highly unlikely this year. would be funny indeed if this year stayed above 2005’s low! we are also setting up for some incredible accumulation if we start growing this early in the season and have the cold fall and winter that is being predicted and which seems likely considering the el nino pattern we are in (if that is the correct understanding of how el nino affects the arctic).
Aglan #119
“Current Extent is 5.30
Kaleschke/Halfman’s range goes as high as 5.35
Lindsay’s prediction range goes as high as 5.20
Stern et al goes to 5.09
Perhaps those are the 3 that are right?”
Kaleschke/Halfman are the closest “Our forecast remains at 4:92 +/- 0:43 Mio. km2.”
Still thats a 17% spread or basically 4.92 +/- 8.5%.
For ‘science’ thats not good enough. Maybe the 2 farmers could do better.
Predicted mean sea ice extent for September 2009 in million km^2 for predictions with largest extents:
Kaleschke & Halfmann: 4.92 +/- 0.43
Kauker: EnsI = 5.02 +/- 0.26; EnsII 4.42 +/- 0.38
Lindsay: 4.94 +/- 0.26 (June data = 3.99 +/- 0.30)
Re: Kenneth Fritsch (#142),
The +/- is for one standard deviation.
How would a minimum extent for the day of the minimum compare with the mean extent for September over the past several years? I may be wrong but is not the mean extent, as used by the modelers, a better and more reliable indicator than that for the day of?
Why not more discussion of the modelers predictions and what might have gone wrong/right with their methods? Would not that be more in the tradition of CA?
I see too many food fights over some less important details on this thread.
Re: Kenneth Fritsch (#515), A more reliable indicator than what? The minimum extent? Yes I would agree. And would not the average ice for the whole year be an even more reliable indicator than either the average for Sept or the minimum extent?
Case in point – one would get the impression that there is less ice in 2009 than 2005 from either the minimum or mean extent for Sept, when in actual fact the ice in 2009 is far more than that of 2005
Re: Richard (#518),
The Year to date average of JAXA extent data is currently higher than 2005-2008, but not 2003 and 2004. I don’t expect that to last because 2009 is running lower much lower than 2006 right now. It might stay even with 2005, though. Here’s the data:
year ytd avg full year avg
2003 10839129 11400980
2004 10657076 11149287
2005 10344909 10830545
2006 10220438 10698124
2007 9965810 10550617
2008 10460820 11002406
2009 11037168
Re: DeWitt Payne (#536), The Year to date average of JAXA extent data is currently higher than 2005-2008, but not 2003 and 2004.
How do you say that? The Year to date average of 2009 seem to be higher than any of the previous years according to the figures you have published. None of them are above 11 million
Re: Richard (#539),
Sorry, messed up the column labels and formatting. The first data column is the full year average and the second column is the year-to-date average. The 2009 value should be in the second column, not the first. I’ll try again:
year full year ytd
2003 10839129 11400980
2004 10657076 11149287
2005 10344909 10830545
2006 10220438 10698124
2007 9965810 10550617
2008 10460820 11002406
2009 no data 11037168
The tide has turned. The world is starting to cool and the Arctic ice will increase. The Earth has always done this.
Steve: Please stop such triumphal announcements. This site is about verification and data testing and not about such announcements.
PS Arctic Ice 5,312,813 Km^2. Melt 17,656. Wait for the update.
well….if we have another correction like this morning, we will have 2 days in a row with growth. of course, this one may not correct upwards…we shall see!
Yes it is just below 2005 so the correction will be important. Here’s a reminder of peoples guesses for minima day.
8: crosspatch
9: Jon P
10: Michael (+ DeWitt Payne)
11: Michael Jennings
12: BarryW
13: AndyW
14: MikeP
15: markinaustin
16: Neven
17: Earle Williams
18: Roy1915
There could still be a 2005esq late dive so it will be hard to tell for a while on who won.
Regards
Andy
There was some melting yesterday.
The latest value : 5,312,813 km2 (September 9, 2009)
(From IARC-JAXA)
yeah….but Urederra, every morning at 9:00 they have a final amount which becomes the permanent amount recorded for the day. yesterday the night value was even lower than 5,312,813 and then rebounded up to over 5,330,000. if the same thing happens today we may have our first 2 days in a row of up.
Re: markinaustin (#151), Thanks for the info. What time zone is the update?
Re: MikeP (#152),
JAXA posts the first estimate for the day at 11 PM Eastern. The final update is 10AM Eastern the next day.
ok…so it Just updated to 5,315,938 km2, so that means it is only SLIGHTLY above the 2005 low. as in a few hundred (600?). so for now we are still above 2005. one thing seems certain…..we are going to be very close to 2005. every day that passes is another day off the calendar and one day closer to the true minimum if we haven’t actually hit it yet.
Btw, the zone for the time stamp on the posts is Mountain. I assume that’s because the server is located in that time zone.
Well I was right in my prediction at Richard (#137).
But the increased update still shows a reduction of 14,531 over the 8th.
Re the 10th visual: there has been an appreciable decrease of ice in the North, with an increase in the NW, SE, and SW. Ice is forming south of Baffin Island and the North Hudson Bay.
My prediction. The 10th may show another decrease followed by a big increase the next day. The 10th may be the minimum.
Re: Richard (#156),Richard, how can
ice form when SSTs are 5-10C likewise air temps that is N Hudson Bay
and S of the “Running Terrier”, Baffin Island…?? Source: Environment Canada [EC has some “warming agenda” but…] Maybe down
to around zero, early morning some places though. What is your source??
Re: Staffan Lindstroem (#159), My “source are my eyes. See here. Maybe that data you are referring to is outdated?
Re: Staffan Lindstroem (#159),
Where are the SSTs 5-10C ?
Re: Staffan Lindstroem (#159),
Re: Richard (#156),Richard, how can ice form when SSTs are 5-10C likewise air temps that is N Hudson Bay and S of the “Running Terrier”, Baffin Island…??
Cant answer that question. Obviously ice cannot form if sea temperatures are 5-10 C. The corollary to this is: if ice has formed then the SST’s cannot be 5-10 C.
Common mistake of the AGW proponents. Believe the theory and question the evidence, when actually it should be the other way round.
What is your source??
My eyes. Have a look here
Obviously that ice has not blown there, if you toggle between the 9th and 10th. It has formed there in one day.
Re: Richard (#162),
What is your source??
My eyes. Have a look here
Obviously that ice has not blown there, if you toggle between the 9th and 10th. It has formed there in one day.
They’re called clouds! You won’t find them here
Re: Phil. (#163), No sir. That white thing on the ocean is ice, not clouds.
Re: Richard (#165),
Afraid not!
Re: Richard (#163), Re: Gerald Machnee (#161), Gerald and Richard…Pathetic…OK, I was wrong about one thing, SSTs in N Hudson Bay and W of Baffin Island, weren’t up to 10C maybe 5-6, cold upwelling after N and NW winds, when I turned my back for a while…tty and I are among the 100[ironic number] or so Swedes that DON’T buy CAGW… Richard, I don’t say you called me an AGW proponent, you
did not but quite close, no?? Then I suggest you talk to some people
at SMHI…I would say Sweden is obsessed by “climate change” I have
the idea of letting SMHI know that climate=change…FYI…MORE TO COME LATER…
Re: Staffan Lindstroem (#178), No no Staffan I didnt actually call you an AGW proponent. Welcome to the sceptic “gang”. I am among the dozen or so “kiwis” (New Zealanders) who dont buy AGW either. The religion is sweeping this place, taught in schools and churches, but weakening by the day.
We have to battle against people like Phil, who converts ice to clouds and Flanagan, who is a whiz at predicting after the event.
Re: Richard (#184), Richard, {and tty}
actually, Phil. MAY have been right about “clouds”, [low fog clouds,
possibly, very common up here?!] here…unless you
think “underfrozen” water hides at the seabed near SW Baffin Island,
coming up by the upwelling event the other day and freezes the surface water…after all, we’ve all heard the story of when tons of cod surfacing
around NF, frozen solid…?? Old ice can linger in Hudson Bay into
September,ie. 2004 [record event] So folks, everything boils down to
the BIG question: Are the poles cooling the rest of the planet more
than the rest of the planet warms the poles?? …and Richard, FYI,
I’ve always been a sceptic, some 54 years… I was even sceptic to
the ordinary delivery time of 9 months, so I came out 9 weeks early…[RIP Lars Gyllensten and thank you for improving incubators][more than 30-40 % oxygen caused retinal damage]…And the day I see
a persistent warming trend and the rest of humanity is screaming
“oh we have a catastrophic global cooling”, I’ll let you know…
Re: Staffan Lindstroem (#195), Phil. MAY have been right about “clouds”
He may have been but he was not.
AndyW35 (#187), Sorry Richard but that much ice does not melt in one day completely.
Guess what? The ice was there on the 10th gone on the 11th.Phil.:September 11th, 2009 at 10:50 am Richard,if you go to the DMI site and do the same toggle between yesterday’s and today’s SST map you’ll see that the clouds have gone…
And how much ice disappeared on the 11th? 41,094 Km^2. So “that much” ice does melt in one day. And thats why its gone.
PS I have been dead right in my predictions. The secret? I have been studying the pictures of those “clouds” and presuming them to be ice, because thats what the DMI website and picture is all about, (controversial I know – should be clouds), and have been right each time. DMI is a day ahead of IJIS
Re: Richard (#197),
snip
You (mis)interpreted cold patches on a SST map, an acknowledged problem with the IR instrumentation used, as ice. The same site in their map of ice concentration, using the more appropriate passive microwave method shows no ice in that location on that day! Neither does any other site, your suggestion that this ‘ice’ contributed to the loss of ice over the last two days is laughable since that the IJIS site doesn’t show any ice there so it couldn’t possibly contribute to the loss since it wasn’t there in the first place.
Re: Phil. (#199),
I did not say that this particular ice contributed to to loss on the IJIS ice. Merely that ice had reduced visually and that the ice had melted was a possibility, when I was told such a thing could not happen in a day.
Whereas I accept that there could be a possibility that the ice was shown as a mistake, you will have to show me proof that there was indeed mistake in this particular case in the the DMI map.
Re: Phil. (#199),
snip
And Richard, before DMI takes it away, take screen dumps of 1-DAY changes around mid August 2009 until now, and you’ll see they have
big problems…If you don’t realize then that you MAY be wrong, what can I say, you’re entitled to your opinion…let’s move on to
some other aspects of Arctic ice…Best regards Staffan
PS I have known them to update twice in a day, sometimes late in the day and sometimes an update for the day can come 2 days later. So nothing set in stone (after all we are dealing with floating ice).
this could be the bottom….or not…time will tell.
Re: markinaustin (#158),
Looking at Nansen (different satellite, different smoothing): They still have a downward trend. That indicates we might have to wait a few more days.
http://arctic-roos.org/observations/satellite-data/sea-ice/ice-area-and-extent-in-arctic
In talking with a co-worker (former USN sonar operator), I learned that the USN measures the ice thickness (and extent?) whenever they’re up under the ice.
I wonder if the Navy has made that data available to the icecap researchers – there should be data going back to the early sixties, if not before.
Re: Scott Lurndal (#161),
See here.
Re: Scott Lurndal (#162) and Re: Phil. (#165), The coverage is so sparse that I don’t see much coming of it. Plus, I know how the thickness is measured, and the resolution is not that great. Perhaps it will be useful for calibrating satellite sensors, but not really for much else.
a down shift of about 17,000 (estimate)….we shall see what tomorrow’s correction brings.
Re: markinaustin (#166), Melt is 14,688 from yesterday. I did say it would be less. I dont think the update will show more, maybe less. But I think tomorrow it will be more because of the ice (or “clouds” according to phil) forming SW of Baffin Island and ice expanding to the NW.
Re: Richard (#169),
Richard,if you go to the DMI site and do the same toggle between yesterday’s and today’s SST map you’ll see that the clouds have gone, also if you check their ‘Ice concentration’ map for yesterday you’ll see that they don’t show ice where you claimed it was.
Re: tty (#171),
Actually the wavelengths DMI use do have problems with clouds: “Satellite observations of sea surface temperature is currently retrieved from several satellites. The most accurate infrared satellites have an pixel size of 1 km and an accuracy of 0.3 degrees but they are limited by clouds.” They use passive microwave for ice measurements which is why the clouds didn’t show up there.
Re: crosspatch (#172),
The temperatures shown by DMI are for the high Arctic (above 80ºN) which is almost all ice covered throughout the summer so no positive change in summer temperatures is possible until a substantial part of that ice is not present in the summer. In which case there would be little argument about whether the Arctic was going to become ‘ice free’.
Re: Phil. (#182),
Richard,if you go to the DMI site and do the same toggle between yesterday’s and today’s SST map you’ll see that the clouds have gone
The ice has gone. ICE not clouds. It forms and melts you know. There are still some “clouds over the north pole” and surrounding areas, that looks surprisingly like ice. Seems there is a bit of clouds around.
Also if you look at the SST’s purple is below 0 and there is a bit of purple where the “clouds” were.
Re: Richard (#183),
Sorry Richard but that much ice does not melt in one day completely. You need to take multiple data sources to correlate. People who just take one and bank on it are often left out to dry. I mentioned the drifting sensor above, that was a classic case.
snip
Regards
Andy
Re: AndyW35 (#187), Sorry Richard but that much ice does not melt in one day completely.
Depends on the concentration. And it may not have completely melted. It maybe just at lower concentrations than recorded at the DMI site. It was thin ice, but ice nevertheless. That DMI site deals with ice, not clouds.
If you look at the “observed” Sea Ice Map of SSM/I here, you will see that on the 9th there is a bit of thin (hazy) ice at the precise area of Baffin Bay on the 9th that shows up on the DMI site on the 10th. This map shows ice at lower concentrations than the other maps.
Multiple data from multiple sites does not mean that some sites are not correct. Just that different sites record the ice differently. Loose pack ice for example may not be accounted for by most sites.
snip
Re: Richard (#189),
And you say “that much ice does not melt in one day” How do you account for an average daily ice melt of 28,250 sq Kms last week and 50,300 the week before that? When that ice disappeared we also had an ice reduction on that day, so is it surprising?
Re: Phil. (#182),
I am aware of that. This summer’s temperatures were below average for much of the season. They are currently about average … looks to me like something around -5C. Maybe if you compared this year’s current temperature to, say, 1991 when temperatures were well above normal, you might get what I am referring to.
Richard (#165) what you are seeing in Hudson Strait and northern Hudson Bay isn’t new ice. It is too warm for that. The only continental coast where new ice is apparently already forming is the northern Taimyr Peninsula (where temperatures are well below zero now).
However it isn’t clouds either, Phil (#167). Microwave sensors simply aren’t very good at distinguishing ice in coastal waters, particularly where there are lots of islands and skerries. They often show ice were there isn’t any, This is/was much worse with SSM/I than AMSR-E, and in my opinion is a large enough source of error to make it more or less impossible to compare ice data before and after 2003. I base this on the SSM/I data from the Baltic and the White Sea which shows large areas of ice in summer when these seas are invariably ice-free (indeed warm enough for bathing in the case of the Baltic). For some reason (probably differences between sensors) this is particularily bad 1989-2003, and if it affects other areas as badly as it does the White Sea (and I can’t see why it shouldn’t) summer ice during this period must be significantly exaggerated.
Re: tty (#171),
At Erice, Y-s Choi discussed the tremendous difference between old satellites and new satellites in definition of clouds. It would make sense that there might be problems with sea ice. Are there adequate overlaps in the SSM/I and AMSR records to demonstrate that there is no inhomogeneity in the methods?
Re: Steve McIntyre (#173),
I have noted a higher prevalence of cloud detail in the newer portion of the video’s I produced. It’s enough difference that at times there is no cloud cover in the old video and a lot of clouds in the new.
Re: tty (#171),
Microwave sensors aren’t relevant here since Richard was referring to SST images made using IR which does have problems with clouds.
Re: Phil. (#215), Microwave sensors aren’t relevant here since Richard was referring to SST images made using IR which does have problems with clouds.
From the DMI website: “Satellite observations of sea surface temperature is currently retrieved from several satellites. The most accurate infrared satellites have an pixel size of 1 km and an accuracy of 0.3 degrees but they are limited by clouds. DMI has developed a statistical method that uses the individual error characteristics to combine satellite observations from about 10 different instruments in an objective interpolation method. The satellite observations include polar orbiting infrared and microwave sensors as well as geostationary satellite observations and are obtained through the Ocean & Sea Ice SAF project and the GHRSST-pp project…”
Air temperatures in the Arctic appear to be very close to the 1958-2002 mean according to DMI though that is an overall temperature and things can vary greatly from one specific location to another. This year’s pattern has shown a continuation of the recovery from the 2007 wind event. 2009 saw more time above the -100km anomaly than any year since 2005. The deep minimum would be due, in my opinion, to the relative lack of old ice and I would expect to see further recovery next year, all things being equal to this year. There could possibly be a reversal of the trend of general decline seen since 2004 but we don’t have enough data yet. I will say that I think there is a greater than even chance of next year’s minimum being above this year’s barring any odd weather event such as we saw in 2007.
Steve, to avoid the inhomogeneity, the NSIDC uses only the SSMI data to build the “long term” trends.
By the other hand, overlaping:
http://www.agu.org/pubs/crossref/2008/2007JC004257.shtml
http://www.agu.org/pubs/crossref/2008/2007JC004255.shtml
Re #173 The same overlap question can be asked of the late 1970s transition from pre-satellite methods to satellites and of other methodology changes over the years. I continue to wonder if there was a methodology-related step change downward in sea ice circa 1979.
Arctic sea ice area has dropped 0.168 Mm2 in the last three days. IMO, ice area is a much better indicator than extent. You can have big gains in extent, but little or no change in the area if the average concentration drops. It was looking like 2009 would show close to a 0.5 Mm2 gain in area over 2008 only three days ago. It’s looking now like we’ll be lucky to see a gain of 0.3 Mm2. The Summer average area for 2009 should be above the 2002-2008 trend line so the slope should get somewhat smaller with the inclusion of 2009. But it’s still going to be well below -0.1 Mm2/year. If the average area continues to increase by the same increment as for the last two years, it will take many years for the rate to go to zero.
New total after adjustment is 5,295,313 km2 which pushes the loss to 20,000 for the date.
Re: Michael Jennings (#180),
New total after adjustment is 5,295,313 km2 which pushes the loss to 20,000 for the date.
17,500 to be precise. This is in line with my prediction above:
Re: Richard (#169),
Melt is 14,688 from yesterday. I did say it would be less. I dont think the update will show more, maybe less.
Re: Richard (#185),
Richard, you missed the adjustment for the 9th that went from 5312 to 5315 thus the 5295 for the 10th gives a 20,000 loss.
Re: Michael Jennings (#193), I did indeed. The melt was 20,625 for the 10th. It maybe a little more for the 11th.
Yes, now I prefer to wait for the daily update before posting – corrections have been large these times. So here they come
9/10/2009 5295313 -20625 -12075
9/9/2008 4707813 -7656 -30915
9/10/2007 4367188 -32343 -30401
9/10/2006 5926094 -10000 -4575
9/10/2005 5599844 -28437 -12790
9/9/2004 5835313 -3906 -6383
9/10/2003 6060469 -25937 -19776
9/10/2002 5655156 8281 -11540
I think we can safely say now that 2009 will be below the 2005 minimum (5315156). Since the probability that 2009 will be below 2008, looks like we’re heading for number 3! But how far from the others?
2007 just had its minimum in our daily series: 4707813. 2007 was at 4267656. So is 2009 going to have a 2004-like event and drop below 5.0 or stay well above the two preceding years?
Re: Flanagan (#181),
Need to change will to is in your first sentence Flanagan.
Whether the “ice” is clouds or other spurious data it is taken into account by the scientists. Don’t forget these instantaneous maps and plots are just for the public and the proper data that is used in science papers has a lot more work gone into it and corrections. NSIDC have stated that as being the case, even their averaged 5 day graph is not what they use in retrospection for scientific work. It’s a quick fix for junkies like us. I have no worry at all that spurious data is accidentally taken into account. Even when the sensor drifted and gave too high readings it was only a matter of importance for us rather than any scientific paper being impacted, Walt Meier has said as much on WUWT.
Regards
Andy
PS My prediction, on the 11th, the IJIS data will show a further reduction. The melt continues.
i still say we are very close…but my guess of the 15th is looking good, but my guess of 5.15 million is still looking quite low.
whoa! huge drop…..what is up with that? 41,000?
PS – Staffan Please let me know if you get the answers to the big questions. Right now all I have to go by are a few inconvenient truths defended by a guy waving a hockey stick.
Re: Richard (#198), Richard, excuse me, who’s the hockey-stick waving man??????????
Re: Staffan Lindstroem (#203), Richard, excuse me, who’s the hockey-stick waving man??????????
Staffan I cant remember the fellows name. But he is the right hand man, the enforcer, of this great prophet who started a new religion I believe. He handed out a whole lot of hockey sticks made from some sort of pine to deal with anyone who would deny the religion.
This guy said that you Vikings make up all kinds of stories in your sagas. Like Greenland being green, when everyone knows it was always covered with ice except now.
Re: Staffan Lindstroem (#204),
Staffan in the spirit of compromise I am willing to concede that the white patches MAY have been polar bears who had drowned while trying to get onto a small ice floe that may have drifted into that area. But I will stop short of clouds. There is only thus far I am willing to go.
Maybe you and your small band of sceptics could organise another viking expedition in one of your longboats to investigate if there is any ice over there? Take along some ice from your freezer to chuck overboard in case there isnt any. You are on my side after all.
Re: Richard (#214), Ah, Richard, you
mean Michael Mann, HSM No 1. … Your idea of opening my freezer
is not very good…because…there’s where the ice IS…I don’t want
to make another “snowball earth”…LOL…[IS is Swedish for “ice”]
And why do “we” love the apocalypse so much?? Oh, time is running,
gotta work…SL
Re: Richard (#214), Sorry, Richard…
your information about Swedes seems a little outdated…The longboats were so 2008…Hint: One of “my” morning paper subscribers
made a “documentary” about undertakers etc in Lodz, “Necrobusiness”,
so Mr Cohen/Borat/Brüno could make “Climobusiness” OT… ON TOPIC:
Did you take screendumps from the DMI site and only explanation could be patches of polar bears the size of Iceland???
Re: Staffan Lindstroem (#228), What!!?? you dont use longboats anymore? and go on raiding parties? By Thor that’s a pity.
“The size of Iceland” maybe a bit of an exaggeration. And if other people can exaggerate why cant I? And remember an exaggeration is an exaggeration, the extent doesn’t matter. If I can do it better than others so be it.
You haven’t congratulated me on catching someone out on a question of fact? Hint: “DMI has developed a statistical method that uses the individual error characteristics to combine satellite observations from about 10 different instruments in an objective interpolation method. The satellite observations include polar orbiting infrared and microwave sensors…”
Here are my predictions based purely on “clouds” and a secret formula I use:
1. There will be a BIG reduction in ice on the 13th. (Unlucky number).
2. The melt may continue for perhaps another week. There is a strong warm anomaly off the NW coast of Alaska, consistent with warm water of the El Nino coming in from the Bering Straits and causing a huge “Bay” in the arctic ice. The Kamchatka volcano has proved to be a damp squib. The waters seem to be cooling in the “Bay” though. The waters also seem to be cooling where some polar bears MAY have perished. We may see more polar bears there in the near future.
3. While the melt continues there will be longer and more detailed posts from Flanagan. He will provide sterling service on the extent of the melt and how close we are getting to the catastrophic 2008 figure, in case anyone misses it.
4. The cries of gloom and prophesies of doom will increase.
Re: Richard (#230),
Congratulations on reading the site I referred to, regarding facts care to point out where it says that the SST is compiled without using IR? In fact their algorithm was unable to correct the spurious IR signal in this case, the microwave data used for ice concentration showed no ice in that location as I’ve pointed out before.
Hi,
given the size of the drop yesterday, I will wait till the update because there might be some huge correction. See you later (well, in 6 hours)!
Richard, I’m sorry but you are wrong. An amount of “ice” that size, roughly the size of Iceland, does not suddenly appear one day then disappear the next, especially when the temps are too high for it to appear in the first place. Temps in the north of Hudson bay and south of Baffin Island have been constantly a few C above zero during the day and about 0C or -1C at night. Therefore there is no mechanism for the creation and then the destruction of this “ice” in a 24 hour time span. It was simply an artifact on a map that you misread.
You can tell this was spurious because if 100 000 KM of ice appeared in one day, as that map suggested to you, the JAXA graph would have spiked highly positively too, but it didn’t. It is still decreasing as the ice is is nibbled/compacted around the periphery.
Quite a large drop by the way today !
Regards
Andy
Seems we could be seeing a late surge (barring a massive correction) for 2009 that, if it continues, will end up taking it closer to 2008 after all. Whether due to compaction, melt, or some other reason, the next few days will tell the tale. I fear my Sept. 11th prediction will be too early for the minima
Re: Michael Jennings (#205), and also markinaustin:
The 15th is certainly looking more likely than the 11th now – we need to see some small days mixed in with some positives for minimum. And 5.15M is no longer looking out of the question, it just needs 40, 30, 30, 20, 10, 10 or so.
There may be a sting in this tale [sic] yet!
BTW did anyone see today’s Independent (UK) front page alarmism on the NE Passage being open, including a rigged picture of the Arctic with no clouds and the northern NWP totally ice free?
Rich.
Very large correction puts the new figure at 5,278,594 km2 for a loss ending up at 17,000 for the 11th. Interesting that the correction was more than the adjusted total loss was.
Here we go. Large revision upwards.
9/11/2009 5278594 -16719 -15625
9/10/2008 4729688 +21875 -28191
9/11/2007 4343438 -23750 -26383
9/11/2006 5896719 -29375 -5513
9/11/2005 5581875 -17969 -14375
9/10/2004 5825000 -10313 -8861
So we’re still getting away from 2005, but more slowly than first expected. It might however be a sign of stronger days to come.
Does anyone have a link to the 2008 NSIDC forecast, analogous to the one shown at the start of this article? I have found a few pieces as text, but the whole shebang as a graph has eluded me.
Re: Rob Spooner (#209),
The graph is slightly different format, but here it is:
Their July outlook was much better last year, as it averaged 4.77 million sq km out of the 16 predictions, and a median just below 5.0 million sq km. A smidge too high but not bad.
Their early season outlook though was too low though averaging 4.25 million sq km and median of 4.2 million sq km.
Re: NEwxIce (#212), there are two possible reasons for a consistent miss by all models, ao we’ll have to wait till next year’s result to form any opinions. One is that the modelers all followed a bias in the same direction, the other is that the target made a surprising move. It’s hard to look at 2009 and see anything surprising, but I could be missing something. It will be interesting to hear the post mortems this winter about why the models consistently estimated a lower minimum extent.
ok…i am ready for another up day!
There is another explanation.
The ice is increasing as the Earth enters a cooling phase.
This is most likely the cause.
In the ’40’s when the PDO turned cool the ice increased and it is doing the same again.
If ice is so damm important, why but why is the SH ice ALWAYS ignored? Its been above anomaly for nearly 2 years now!
previous year not visible but check records yourself
Re: Vg (#216), Vg, 1. You mean
“above normal”…gws… 2. If you’re to believe CT’s diagram, “nearly one year” obviously? very rapid spring melt until
mid October 2008…!? But as many bloggers have remarked CT has issues,
too…
The El Niño is expected to strengthen and last through the Northern Hemisphere winter 2009-2010, according to NOAA. The SST’s are also very high just north of the Berring Straights. This would account for the ice melt pattern NNE of it.
It may also effect the Min extent day and winter ice extent.
Oh oh – It appears that a major eruption is underway at the Shiveluch volcano in the Kamchatka peninsula. Another factor to consider in the arctic ice scenario? (spanner in the works?).
Pesky nature trying to pit its puny might against the almighty humans, who we all know control the global climate. In New Zealand we have have opted to fine tune the Global Thermostat by cap n trade. So AGW proponents never fear. The ice maybe low today – but in a few years, thanks to our measures, it will be back to freezing normal again.
The lost yesterday was again 20k+, but I’ll wait for the traditionally upwards correction. Preliminary extent is 5255313
Re: Flanagan (#221),
Yep, the size of the adjustment does tend to increase as we get closer to the minima in most years
Does anyone know if the wind pattern will carry the ash of the Kanchatka volcanic activity toward the Arctic? I would assume where the ash goes, temperature dips will follow.
stephan an.. above anomaly…….I dont consider any of this significant in the slightest…. in climate terms its all normal
Re: Vg (#225), Vg, of course and if all ice in the Arctic would disappear in late summer and NOTA BENE I’m not ironic,
it’s not our fault…And if it had been our “fault” I have to inform you we are not unnatural or better contranatural as all these “alarm-agendists” bore us with…
New adjusted total for the 12th is 5,259,375 km2 which nudges the loss JUST under 20,000
Looks like the minima will not be 13th as I supposed, got a feeling now it as a few more days left in it yet before it makes the turn. Shawn’s theory on the early freeze seems to be looking suspect at this time. Richards UFO’s (Unidentified Frozen Objects) have not turned up again either
Regards
Andy
Re: AndyW35 (#229), I like UFO’s – less controversial. We shall call them that in the future.
The size of Iceland is not that much of an exaggeration if at all, see here from an eyeballing.
spurious
Not only is it large scale but it is miles away from where the main action is taking pace which is at higher latitudes. Still no answer from Richard how this large amount of “ice” can suddenly appear one day and disappear the next when temps stayed the same.
Regards
Andy
Re: AndyW35 (#232), The size of Iceland is not that much of an exaggeration if at all, see here from an eyeballing.
Hi Andy,
I have dutifully eyeballed the spurious image and processed it with an algorithm which connects my retina to the computer in my brain. I have further processed the pixels with the point of a sharpened pencil and have come to the following inescapable conclusion:
It is “very likely” (> 90% probability of occurrence), that the combined size of the white patches is 40% +/- 25% of the size of Iceland. Hence it is “Virtually certain” (> 99% probability of occurrence) that “the size of Iceland” is an exaggeration. The words “not that much” are hereby expunged as being unquantifiable and hence unscientific.
Not only is it large scale but it is miles away from where the main action is taking pace which is at higher latitudes.
That argument is spurious. In a battle while the main action takes place in the frontal assault of the contestants, the more mobile light cavalry is known to attack at the flanks and rear of the enemy, then melt away at the counterattack.
If you notice that map of the SST’s there is considerable purple (denoting SST’s below 0), in the Hudson Channel, the Foxe Basin and Committee Bay, well south of where the main action is taking place. It remains possible that ice can form in waters where the temperature is below 0.
If your eyes are keen you will also observe that much of the white patches are surrounded by purple, which gives rise to the POSSIBILTY that the white patches, even though may not be ice, are also areas where the SST’s are below 0, much like its white higher latitude counterparts.
Still no answer from Richard how this large amount of “ice” can suddenly appear one day and disappear the next when temps stayed the same.
“..when temps stayed the same” is an unsubstantiated allegation. Large amounts of ice do suddenly appear and disappear in the course of one day around this time. On the 7th of September 2008, 68,437 sq kms disappeared and on the 10th and 11th of Sept 2008 43,750 sq kms appeared. Before you say that this is not comparable to 40% of the size of Iceland, let me point out that these are the net amounts. Much larger amounts could have appeared and disappeared, and probably did, at different areas of the Arctic.
The map of the 13th shows a white spot in the Hudson bay, surrounded by purple. Though not quite the size of Iceland, it is possibly the size of London.
Having said all that, to be more honest, (than the IPCC, 2004), I am willing to concede that it is possible that all that white bit MAY not have been ice, while retaining the POSSIBILITY that some of it MAY have been.
Regards
Richard
Re: Richard (#237),
Richard, you have a good turn of phrase there with the cavalry etc and I take my hat off to your posting style, but I still have a problem with the content. You now say that the SST’s are cold enough for the ice to appear, purple, but the SST map I look at shows it warmer than this in that region.
Unisys
Even if they are cold enough the ice has gone again! So it appeared and disappeared with very little change in conditions, both air and sea.
Anyhow, getting back to the minima day it looks like my guess is likely to be wrong and it will be closer to the end of the month.
Golly, I hope I don’t get snipped, I am in the prime of life 😀
Regards
Andy
Re: AndyW35 (#253), Andy for heavens sake. If you are in the prime of your life you should have no difficulty spotting the purple. That picture you referred me to is like showing me evidence from a pin hole camera, taken from a mile off, to refute the evidence of a close-up from the latest SLR. See here again. If you cant spot the purple get your wife or gf to point it out for you. There is some white again just north of Hudson Bay surrounded by purple. The white spot in Hudson Bay has gone, but now there is plenty of purple there. I would bet there will be white there within the next two days.
So many of you knowledgeable people here told me they were clouds, that DMI has got it wrong, and certain people who are absolutely sure of themselves have got it right, that I started to believe you. But now I am having more and more doubts.
And I’m betting that tomorrow will show an increase.
Re: Richard (#254),
Hi Richard,
I used Unisys data set which shows it warmer, but taking your source purple signifies still around zero so no freezing would occur. It would have to be around -2C for it to start freezing and then only very slowly.
Even if it was cold enough to freeze you are still not commenting on where the ice has gone too? It appeared and then disappeared in one day with hardly any change in temperatures at all.
There is no mechanism to create it one day and no mechanism to destroy it the next. So common sense says that it never existed, as the AMRSE maps agree with, as do the temperature plots.
Regards
Andy
Re: AndyW (#289),
Hi Andy,
..taking your source purple signifies still around zero so no freezing would occur.
Wrong. Purple does not signify around 0 if signifies 0 and below. If it were -5 or -10 or -30, for that matter it would still be purple. But in this case I agree the temperature was probably just a little below 0.
Do not discount the DMI product so lightly. They are smart people not a bunch of idiots. If you have a look again. Ice is forming just north of Hudson Bay and from the SST’s I would say in Hudson bay soon.
There is no mechanism to create it one day and no mechanism to destroy it the next.
If that be so what is the mechanism whereby thousands of sq kms of ice get formed and “destroyed” every day?
Regards
Richard
Re: Richard (#299),
“Do not discount the DMI product so lightly. They are smart people not a bunch of idiots”
Very true, I never said they were, just that the reader did not understand what they were showing 🙂 I have emailed them and this is what they said
“The appearing and disappearing of ice from one day to another is, in this case, most probably an atmospheric effect that makes our ice-concentration-algorithm generate false ice concentrations. Large amounts of water in the atmosphere can cause such artifacts.”
So they are agreeing with Phil.
Regards
Andy
Re: AndyW35 (#303),
Andy in the face of that I will have to concede that “in this case”, the ice was “most probably” an “an atmospheric effect” which led to the generation of false ice concentrations.
So they are agreeing with Phil.
I hold no brief for Phil but he deserves to be quoted accurately. Phil is ticklish about such things, having pointed out the difference between “will” and “may well” and “expected to”.
There is a difference between “they’re (called) clouds”, (operative word are), to “most probably”, “in this case”, “an atmospheric effect” which led to the generation of false ice concentrations.
I therefore concede that in this case the ice concentrations shown were probably false.
However the reason for this is “atmospheric effects” effecting their algorithm and not that there is “no mechanism to create it (ice) one day and no mechanism to destroy it the next”. Take some ice out of your freezer leave it for an hr and put it back and you will discover the mechanism. Though I would be loath to use the words create and destroy.
In the arctic ice freezing and melting of ice occurs on largish scales every day. Possibly due to falling temperatures and the mixing of warm waters from the Pacific and Atlantic.
Regards
Richard
My congratulations to all for the fair and spirited comments.
Mike
It appears from this graph that the freeze should be starting for all areas north of 80N. I estimate that the area above this latitude is already 5C below that which sea water freezes.
http://ocean.dmi.dk/arctic/meant80n.uk.php
Re: Tucker (#235),
Tucker, apart from a small area at the periphery the area north of 80ºN is already ice so I don’t think that temperature is too relevant.
Re: Phil. (#238),
I think it would not be unreasonable to assume that if it is -6C or -7C above 80N, that the freeze line might extend to someplace below 80N. It isn’t like there is some temperature barrier right at 80N that prevents cold air from moving any further South. It might not be unreasonable to also assume that the colder it gets above 80N, the further South the freeze line might be.
The latest picture at buoy 2009A was -2.5C on 8 Sept. It was getting quite near 80N.
Re: crosspatch (#239),
Buoy 2009A is not transmitting data but it’s at about 83.5ºN, nevertheless seaice is still melting at the periphery and the surrounding waters are still warm. There will be sharp difference between the temperature over ice (often an inversion) and that over open water.
Re: Phil. (#240),
Buoy 2009A is not transmitting atmospheric temperature data but the pictures returned do report the internal temperature of the camera which was -2.5C on September 8. That temperature usually runs a little higher than the atmospheric temperature.
Re: Phil. (#240),
Also, see here
Re: crosspatch (#239), Crosspatch the way I look at it the Arctic ice is forming because of the cold and melting because of the warm waters from the Pacific and the Gulf Stream. The battle is between the freeze and the melt. Right now the melt has the upper hand, more melt less freeze, but soon methinks the tide will turn in favour of the freeze.
wow tucker! if that graph is correct, this year is already much colder than the past 2 years which stayed mostly above the average by quite a bit. this year is already below the average. i think based upon that graph it is reasonable to expect the minimum will occur within the next 5 days…..heck, my guess of september 15th may have been correct…..but it was made LATE in the game i realize
Try here: http://psc.apl.washington.edu/northpole/index.html
The daily extent lost was the same as yesterday, so waiting for the update before publishing it.
STEVE: snip – please don’t get into this sort of tiff about whether one reader accused another of something. Blog policies prohibit accusations of bad faith against other posters. If this has been breached anywhere, please ask for the comment to be removed, rather than filling up the thread with bickering.
Steve: snip – if you want to debate WUWT, please do it at Anthony’s blog.
snip – for reasons above
snip
. Let’s focus in the ice. The extent is 5249844, meaning a loss of -9531. Things seem to be slowing down. The 13th might well have been the minimum.
CT Arctic ice area has increased for three days now. It’s possible the minimum was 3.425 Mm2 compared to 3.004 in 2008. The uptick has been fairly large, though, and I expect, based on previous years’ performance, a correction and a test of the minimum sometime in the next ten days.
snip – for reasons above
snip – for reasons above
For the second year in a row the Arctic ice has increased. You will not hear that from the MSM.
[snip]
RomanM: In his inline statement in comment 244, Steve asked for a modicum of restraint on bickering and sniping. How’s about a little cooperation?
so, was this the biggest refreeze ever (or at least in the last 125000 years) ?
Steve,
A suggestion, if you will. [Snip …]
Cheers,
Rich.
P.S. This will be the second of my postings you snip, and I’ll be aiming for it to be the last.
RomanM: Good advice – done]
snip – editorial policies at this blog are against attempts to debate AGW from first principles in one paragraph on every thread. Otherwise every thread becomes the same.
Can anyone tell me if there is a site with Antarctic sea ice extent data/graph, similar to that of the JAXA Arctic Sea Ice Extent comparing recent years.
PS the Arctic extent,etc debate has hooked me over recent weeks, however I do wonder at the lack of comment on the Antarctic extent as from the NSIDC site it seems to be running well above average for the whole year, at least if you accept that 79-2000 is a valid average to compare to.
Re: Scott (#259), Scott I suggest that Flanagan has got it wrong. The minimum in the antarctic does not happen till March and it is trending up.
The absence of interest at the moment is precisely because there is practically no upward trend in the Antarctic during this last month. Had there been a downward trend, statistically significant or not, you could bet that all attention would be focused on it.
Note how attention will be refocussed away from the Arctic in a couple of months, as it gathers ice, and on the small Antarctic peninsular and its melting glaciers.
You may have noticed the news is a little lopsided. Snow in England this June must have been a little unusual and newsworthy in this allegedly dangerously warming world, but search for it in the Met Office news or on BBC weather and you wont find it. Instead you will find in the NEWS – Barbeque summer predicted, and London deaths to increase in 20 years time due to heatwaves caused by Global Warming.
Re: Scott (#259),
You can find the information for area on the Cryosphere Today site.
Over the last couple of years the max and min have hovered around the average, the key difference is that the refreeze starts earlier giving a growing positive anomaly for about 6 weeks, thereafter it grows at the normal rate until about July when the anomaly drops to about the normal max. It may be that this is a ‘new’ pattern which means that 79-00 isn’t a good average, however it will take more that a couple of years to establish this.
Re: Scott (#259),
There’s Spreen and Kaleschke’s Uni-Hamburg’s sea ice page with links to area and extent pages for both poles (FTP link). The data comes from the Aqua satellite so the dates are the same as for JAXA, but the numbers are different because they use a different algorithm for converting the microwave brightness temperatures to ice concentrations.
Scott: actually, there’s no trend in the Antarctic (at least no statistically significant one). This explains the absence of interest.
Re: Flanagan (#260),
Right, particularly around the minimum which bounces around with no obvious direction.
snip – OT
snip – Al Gore has nothing to do with this thread. Please do not debate general issues on topical threads. Otherwise every thread becomes the same.
Re: Richard (#265),
I hold no brief for Al Gore but he deserves to be quoted accurately, a recording can be found here.
Re: Phil. (#267),
I went to your link, and Al Gore says exactly that. “The entire North Pole ice cap will be gone in five years”. (Actually, more like “enti-i-i-i-re”).
He then reiterates it is “expected” to be gone in five years.
How is Richard’s quote inaccurate or misleading?
Re: John M (#268),
‘may well’ is not the same as ‘will’.
Arctic Melt that is
Please do not debate Al Gore on this thread. Otherwise all threads become the same. There’s lots to discuss about the NSIDC models without dragging Al Gore in.
snip – please stop this editorializing.
wow…we had a 20,000 INCREASE in ice tonight. i guess we will see what happens in the correction, but that’s a fairly severe increase. has the low happened?
Re: markinaustin (#271), wow…we had a 20,000 INCREASE in ice tonight.
See my prediction here: Richard (#254), based on eyeballing a DMI product dismissed as defective by a knowall.
We’re pretty close to the bottom if we haven’t reached it already. In most of the JAXA years, there’s more ice 10 days from now than today.
month day ice dd diff
2002 9 14 5.759077 257 0.028338
2003 9 14 6.146563 257 -0.033750
2004 9 13 5.854375 257 0.047656
2005 9 14 5.522344 257 0.015781
2006 9 14 5.781719 257 -0.032344
2007 9 14 4.291250 257 -0.032500
2008 9 13 4.742344 257 -0.002812
2009 9 14 5.269531 257 0.019687
yeah…my guess is that we have seen the bottom. even if we have another down day it is difficult to imagine it being for almost 20,000. i guess we could if we did 2 in a row, but with each passing day that becomes less likely, not to mention our down days were averaging less than 20,000. in fact, we have only had one day of more than 20,000 loss since the 3rd of September 11 days ago. i am going to be dogmatic and say that regardless of another melt day or not, we have already seen the bottom on the 13th and i am HAPPY to report that it was 2 days earlier than my prediction and 100,000 square kilometers ABOVE my prediction. (that is if my dogmatic pronouncement is correct).
here are some other interesting facts:
with today’s large jump upwards of 20,000 square kilometers we MAY already be past the minimum. however, another week of ups will confirm that. IF we have already reached the minimum, this year showed a 542,000 square kilometer increase over 2008’s low of 4,707,813 square kilometers, which showed a 453282 square kilometer over 2007’s historical low of 4254531. All in all, we are looking at 995,000 square kilometers of regrowth within 2 years. next year will be very pivotal.
Now, 2 years is hardly a trend worth mentioning although this one is admittedly fairly dramatic. But if we see a similar growth next year, then we will have a year with more ice than most of the past 8 years and a bit of a better argument could be made towards actual regrowth. all that can reasonably be said at this point is that the melting arctic has at least taken a 2 year respite from it’s dramatic downturn. we are currently only 65,312 below 2005’s low…..so any serious recovery next year would put us dramatically above years 2005-2009, but this year is still 782,187 below the minimum of 2003’s minimum of 6,032,031 so we still have a ways to go!
Re: markinaustin (#275), Makes it hopeful it wont be gone in 4 more years.
Dont be too sure we have reached the min yet but I would say we are pretty close.
Re: markinaustin (#275),
I am also interested to see what next year brings. Current temperatures above 80N have now gone below normal.
Re: markinaustin (#275), that’s a mighty big dependent clause for such a tiny two-letter word, “if”
Re: RealityBytes (#278),
i agree…and i am risking some crow. but regardless, even if it dips a bit more, we are very close to the end and the nuts and bolts of my post still holds true. (not that i am backing down, i really do believe we are at the bottom).
Re: markinaustin (#275),
Indeed.
Using Uni-Hamburg data, 2004 minimum area was 4.698 Mm2, 2007 3.259 Mm2 or 30% reduction, similar to the loss in extent. But area is not coming back as fast as extent. The 2008 minimum area was 3.521 Mm2 or an 8% increase compared to a nearly 11% increase in extent from 2007 to 2008. As of 8/31/2009 the area was 3.934 Mm2 with an additional loss of 0.15 Mm2 expected, calculated from CT data. That’s only a 16% gain over 2007. Even if by some miracle that was the minimum, it would still be only a 20% increase over 2007 compared to more than 23% increase for extent if 9/13 was indeed the minimum. The percentage gain and loss calculations are uncertain because I’m not sure how Spreen and Kaleschke or Cryosphere Today for that matter treat the unobserved hole at the pole for area. It’s possible that they mask out an even larger hole to make the data more comparable to data from older satellites.
err….eating some crow was what i meant to say.
The extent, I am told, has gone up but we have had no data from our regular publisher. The silence is deafening. I rely on his data and feel positively uninformed and disappointed.
Richard: is it me you’re mentioning? In case you didn’t notice, we don’t live in the same region of the world. It’s now 10 am in here, so please give me some time in the morning. BTW, please don’t come saying you “predicted” a minimum on sep. 13. Remember you said on this very thread that
“I predicted a low of 5,100,000 sq kms and the minimum on the 10th of Sept” (Sep 7)
“My prediction, on the 11th, the IJIS data will show a further reduction. The melt continues.” (Sep 11)
“PS I have been dead right in my predictions” (Sep 11)
“1. There will be a BIG reduction in ice on the 13th. (Unlucky number).
2. The melt may continue for perhaps another week.” (Sep 13) Rem: the loss was 9000 that day
“And I’m betting that tomorrow will show an increase.” (Sep 14)
So meybe you got it right finally, after changing “predictions” every 2 days. My guess was largely worse (4.8) so don’t feel like I’m trying to show I’m better here.
Concerning ice: There was an increase but, as usual, given we had several large corrections these days I will wait for the update before posting it. Happy everyone?
Re: Flanagan (#284), .. Remember you said on this very thread that “I predicted a low of 5,100,000 sq kms and the minimum on the 10th of Sept” (Sep 7)
I hold no brief for myself, but deserve to be quoted correctly. Yes I predicted a low of 5,100,000 sq kms and the minimum on the 10th of Sept. But that was in August sometime somewhere on WUWT. I subsequently upped that to 5,277,000, with the min on the 11th, which may possibly be pretty good going.
“My prediction, on the 11th, the IJIS data will show a further reduction. The melt continues.” (Sep 11)
And the reduction was 16719 sq kms. So right on the button
“PS I have been dead right in my predictions” (Sep 11)
True
“1. There will be a BIG reduction in ice on the 13th. (Unlucky number).
I have subsequently re-examined the data and determined: There was a big reduction in the ice “bay” to the north, but this was compensated by an increase to the south and west.
2. The melt may continue for perhaps another week.” (Sep 13)
Note the word perhaps. Compare the difference between “perhaps” and “will”, to “may well” and “is expected to” and “will”. Would you agree that “perhaps” is less definitive than “may well” and “is expected to”?
“And I’m betting that tomorrow will show an increase.” (Sep 14)
Which was correct.
Thank you for illustrating my predictions and showing how correct they were.
Looks like minimum about equals 2005? http://wattsupwiththat.com/2009/09/15/arctic-sea-ice-melt-appears-to-have-turned-the-corner-for-2009/
2005 minimum occurred 1-2 weeks later?
The recovery may be spectacular this year and predict next year will be above anomaly/normal minimum
Flanagans new attitude is very welcome haha
well…i think the fact that we are in an el nino could have interesting ramifications. on the one hand, it may mean more precipiation up there (not sure of this so really asking people to help clarify)…..but on the other hand, it may cause for more prime melting conditions next summer IF it remains that long. on the other hand, if it only remains during the winter months and then fades, it may be the perfect scenario to have MORE ice than we have had in quite a while and then to remain longer next summer. should be a very interesting 12 months.
Re: markinaustin (#286),
Mark it is already fading. It will likely be gone by December and then we will go into a short neutral ENSO phase followed by a fairly strong La Nina early in 2011. The models have constantly had to correct downward for this El Nino in their predictions (which have been uniformly poor) and that will continue.
snip – OT
the update ADDED 7,000 square kilometers so now yesterday’s final was UP 27,000. i think it is now fairly safe to say that we have seen the minimum and it was on the 13th. but to be SURE we should wait another 3 or 4 days.
Well I will be happy with 13th as it was my guess, but 2005 shows what can happen.
snip – OT
Regards
Andy
Here we go
9/14/2009 5276563 26719 -7477
9/13/2008 4742344 -2812 -9419
9/14/2007 4291250 -32500 -20781
9/14/2006 5781719 -32344 -20022
9/14/2005 5522344 15781 -19241
So a 27k increase took place yesterday. That’s something! We still should wait one week though before calling 13 a minimum. That would place 2009 at about 500k above 2008, almost 1 million above 2007, 500k below 2006 and 100k approx below 2005.
It would also place 2009 at about 1.5 million below the 79-00 average and 300k below the long-term decreasing linear trend, towards which the sea ice extent seems to go back. Note that it will nevertheless make this trend even more negative.
Re: Flanagan (#292),
I hate to say it, but I thought that misuse of linear trends on non-linear systems went out in the mid to late 1800’s. It’s a bit disconcerting to see a resurgence. I’d suspect that even a version of Marcel’s monkey (a weighted average of the last 3 occurrences – giving the most weight to the most recent) would prove a better predictor. Even better would be to try and work from the factors that influence minimum area, including the oscillatory ones.
Re: Flanagan (#292),
And yet the minimum will be easily higher than all expert predictions made in June…from here: http://www.arcus.org/
Not significant? Let’s face it, the experts were way too low.
Hi,
I don’t think the linear trend should be seen here as a predictor, but rather as a reference for the dynamical behavior. If you want to qualify a system as “non-linear” you first have to check what the linear behavior really is. Actually, all models show a nonlinear response (some sort of a sigmoidal curve). So it only proves that the decrease is more-than-linear.
Re: Flanagan (#294),
You also have to be careful to specify what period you’re talking about. 2009 is clearly above the linear trend from 2001-2007, for example. If you’re going to reference “dynamical behavior”, then you need some sense of dynamics and what periods they operate over. Arctic ice has shown a relationship to the AMO, so a plausible investigative dynamic reference would be to look at the Arctic’s behavior over some multiple of AMO cycles to see what kind of long term variations might exist. Of course we don’t have that much quality data, but that shouldn’t justify pretending that what’s been observed by satellites has long term meaning. If next year the minimum extent goes above the 1979-2010 trend line will this either mean anything or give any insight into the behavior of the arctic? Will you change your POV if it does? If so, why?
Shorter term, the extent minimum decreased in a locally significant way from 2005-2007. Discussing this over a similar time scale makes some sense. Maybe our grandchildren will be better able to say what causes such events and say more about how unique such things are.
Re: Flanagan (#294),
It is my experience that in pretty much all fields, people tend to weight more highly information that validates their own hypothesis and discount information that is counter to it. I think it is a normal human response that one must actually work to overcome. If one’s career, income, social status, desired social identity group, etc. are also tied to a particular hypothesis, there is even greater tendency to weight validating information highly. As climate science has come into the spotlight mostly based on this AGW/CO2 hypothesis, maybe the last person you want doing objective analysis is a climate scientist and that is where I believe this site has its greatest value. An objective, statistical look taken by someone who does not stand to gain materially from one hypothesis or another is valuable.
While we could say that the overall trend on annual minimum ice levels is declining, is the overall trend of mean annual ice coverage declining? And what about trends in annual maximum ice coverage? And by “overall” it depends on where your start point is.
If you are in the stock market, which is also a daily number, you are well aware that linear trends don’t really predict anything. They are a helpful bit of information but they are lagging indicators.
Looking at CT graphics to see where the ice might be increasing, I would say that the northern route of the NW passage, which was looking not far from opening, is now closing up. And even the southern route now looks problematic.
Also, increasing ice to the NW of Greenland, which I thought might have been pushed there by winds from the Pole, seems to be great enough to be new ice forming.
But the NE passage is looking good just now.
I don’t claim to be an expert in assessing these things, so I’m awaiting instruction 🙂
Rich.
We might have reached the minimum. What is the criterion to judge the experts, so that all are treated alike?
I propose to give every group the same HALF A MILLION square kilometers of wiggle room; but that would probably mean that nobody got it right (remember: they estimate the September mean, not the minimum).
Any better propositions?
Re: Forecasts – Methinks the Climate Audit bloggers have done reasonably well, (though the median may have been brought down a trifle with Flanagan’s forecasts).
We did better than the climate modellers and better than some other bloggers. The bloggers on Grumbine Science for example had a median of 4.5 million km^2
I am not aware what the forecasts on RealClimate were but I would bet they were more pessimistic. Here are the opinions of the Great Gavin himself “…In this year’s June outlook, there is significantly more clustering around the median, and a smaller spread (3.2 to 5.0 M km2) than last year…While it doesn’t look at all likely, the best outcome would be for all the estimates to be too low.”
Would “it doesn’t look at all likely” count as a forecast?
@302 The poll you reference at my site was open to all readers. As far as I know, I’m the only sea ice person who entered a guess. It also closed 30 June, with most of the responses having been in by mid-month. Congratulating yourself for predicting September perhaps better than a group of amateurs, after giving yourself until 3 weeks before the month instead of 10 seems rather low standards.
@Original: The reports are released 2-3 weeks after the start of the month (June’s, for instance, being on the 24th), but the predictions have to be in by the end of the preceding month. My guess and rationale was in the June report, sent in on the 29th of May. It looks like I’ll have underestimated the cover, but by less than 1 standard deviation.
What was interesting to me was the complete lack of takers for a ‘recovered’ sea ice bet (7.4 million km^2 for the September average), in spite of the number of folks who had been saying that the ice has recovered.
Re: Robert Grumbine (#304),
You’re taking issue with a comparison that was never made.
As noted above, I set little store by any such guesses, including my own.
I made a guess in early August and did not compare that guess to similar guesses made earlier in the season, but to model guesses published a couple of weeks later – model guesses that were uniformly low. As noted elsewhere, I do not interpret the success of my guess (and other guesses) relative to the models as evidence of insight on my part, but as evidence of systemic bias on the part of the models.
Re: Robert Grumbine (#304),
You speak of amateur bets, yet ignore the fact that NSIDC expert ice prediction panel had a mean WAY below the actual (4.5 mean to an actual around 5.2). And none of them were high enough. But no recovery, no change in expectations, right?
Dr. Grumbine, thank you for the clarification.
Perahps you can divide computer models into three categories:
1. Useful models, found mostly in limited and specific applications.
2. Useless models, whose variability exceeds that of the real world.
3. Worse Than Useless models, which are biased according to the modellers’ non-scientific preconceptions (or funding needs).
If the Arctic models had predicted, say, a range of 4.0-6.0 m sqkm, they could have achieved a “Useless” rating. However, as it stands they can only be rated in Category 3.
I will go out on a limb and say that barring a repeat of the 2007 wind event, I believe there is greater probability of more ice at minimum next year than there is of less ice. So put me down as minimum 2010 > minimum 2009.
I base this mostly on a belief that we have a continued recovery in the inventory of “old” ice which better withstands summer melt, on arctic air temperatures being near the historic (1958-2002) average so far, and on nothing scary happening with sea temperatures in general.
Sounds a reasonable position crosspatch. Any thoughts on the maxima before that? We have have 2 years where the maxima on our JAXA graph has been relatively high (with the proviso it has only been a few years of course), do you think it will be around 2008/2009 figures again?
Regards
Andy
Re: AndyW (#309),
For maximum, I don’t see any compelling argument either way. If temperatures stay in the pattern that they are in (near normal), starting from a place of greater ice than last year, I might expect maximum 2010 > maximum 2009 but we also have a weak-ish el nino so while I would lean toward more ice this coming season, that is tempered by the uncertainty of how Pacific temperatures will pan out. It will all depend on the weather. I can’t commit on maximum though I lean toward higher than last year.
Another day of increase. The 2005 minimum was 5.315 on day 264. Not much different than where we are today in 2009 if we’ve bottomed this year.
month day ice dd diff
2002 9 15 5.787415 258 0.028338
2003 9 15 6.049844 258 -0.096719
2004 9 14 5.901094 258 0.046719
2005 9 15 5.492500 258 -0.029844
2006 9 15 5.794063 258 0.012344
2007 9 15 4.267813 258 -0.023437
2008 9 14 4.747188 258 0.004844
2009 9 15 5.287500 258 0.010937
yeah…it will be interesting to watch the refreeze…..starting this much higher than the past 2 seasons will be interesting if this is indeed a very cold fall and winter like it appears to be setting up for…
i am interested to see the morning’s final tally on today. if the trend remains it could be dramatically more. that graph is starting to show a fairly extreme uptick.
Re: markinaustin (#314),
It’s nice to see what is hopefully the 2009 minimum, but the uptick looks too steep / too early to continue without a “correction”.
Winter 2009-10 sea-ice max:
3/18/2010 max date
14,750,000km^2 JAXA extent
http://i255.photobucket.com/albums/hh148/Sullivanweather/win09-10.png?t=1253083352
Yes, I will wait for the update as usual. Concerning trends in maxima/yearly average you can still go to the cryosphere today
which plots seasonal averages, as well as the annual trend. Obviously summer is the most rapidly decreasing one, but none of them shows an increase or anything flat.
Re: Flanagan (#317),
Flanagan, I don’t think anyone is ignorant of or disputing those longterm trends. Instead, we are following the short term trends in Arctic sea ice, which so far are defying expert predictions. 2007 featured an “unprecedented” loss of multi-year ice, and therefore many feared the Arctic would have serious trouble recovering at all. This is reflected repeatedly in the NSIDC discussions/outlooks they have issue each season – the overall thin-ness of the ice was expected to be a major factor in the minimum, and therefore the consensus was that both 2008 and 2009 would have difficulty getting much above 2007’s min, if at all.
Clearly, there were other factors involved that have compensated for the lack of multi-year ice. The Arctic is now virtually guaranteed to have more multi-year ice going into next summer than it had this year. In my opinion, the experts and climatologists should reconsider some of their assumptions, as Nature has shown a surprising resiliency so far.
Re: Flanagan (#317), Flanagan, maybe I’ve missed
something obvious…but until around 1950 trends are almost positive…except
for summer JJA…How have they reconstructed the period before 1950??
Re: Staffan Lindstroem (#322),
Hello, we went through that graph last year at this time. The author said to use it with caution as it is based on estimates. I cannot repeat all the discussion so go back about a year and check.
Re: Gerald Machnee (#327),
Thanks for the tip about 1900s sea ice extent.
And since Steve don’t want us to discuss the Amundsen bit again, here is the link to last year’s Sea Ice Stretch, just in case anybody is interested <a href=”javascript:edInsertContent(edCanvas,’Re:%20Shawn%20Whelan%20(#448),%20′);updateLivePreview();”>in the story.
Re: Flanagan (#317),
What satellite did they use back in 1900?
Frankly, that graph looks Mannian. Amundsen managed to navigate through the Northwest passage back in 1903-1906 when, according to the graph, the summer ice extent was around 10.5 million square kilometers, and now, when the ice extent is around 5.5 million square Km ships have to be helped by nuclear icebreakers in order to navigate through the same passage. Polar bears must be happy because it smells fishy.
Anyway, any bets on which day (if any) 2009 JAXA ice coverage line will cross 2005 line?
Re: Flanagan (#317),
Flanagan,
I wonder if CT will ever update that graph… It has been awhile…
Mike
Re: Flanagan (#317),
Those graphs are essentially pure invention, particularly the part before 1950. There simply are no data for most parts of the Arctic.
Remember that in 1900, when the graph starts, the northern Parry Archipelago and Severnaya Zemlya hadn’t even been visited by humans ever, and northeastern Greenland had never been visited by Europeans, so how could you get annual, all-year ice data from there?
The only areas where *any* winter data would have been available at that time would have been part of Alaska, southern and western Greenland, parts of Hudson bay, the southern Barents Sea and perhaps Svalbard (but not every year).
For Svalbard some year-round information is probably available from c. 1915. For the Siberian coast annual summer data is probably available from about 1930 and the same may be true for the southern branch of the northwest passage and eastern Greenland. Winter data would be much scarcer, and restricted to the vicinity of the few settlements (no satellites and no aircraft up north in those days, at least not in winter).
For the other, more inaccessible, parts of the Arctic there is almost nothing until the 1950’s.
Do not be too sure of the minimum. On the 11th and 12th of Sept 2008 there were increases of 21,875 each followed by a loss of 32,969 over the next 6 days. As far forward as the 25th of Sept 2008 there was a loss, albeit a small one of 5,000 sq kms.
The 16th will show another increase. May have reached minimum afterall
OK, I am checking the news and now I am not sure if it was the same passage or a different one. Apparently there is a Northwest passage and a Northeast passage, there are a few ships trying to go across one or the other but many press releases fail to actually tell us which passage are they crossing.
Anyway, it seems odd that Amundsen managed to travel across the northwest passage when the ice extent minimum at that time was 10.5 million square Km.
Re: Urederra (#326),
Amundsen also crossed the NE passage in the early 1900’s.
From Time Magazine 1937. This is nothing new.
“*Across the Pole is the Northeast Passage to China along the top of Norway & Russia. Sebastian Cabot initiated its search in 1553. Henry Hudson twice attempted a passage but it was not until 1879 that the route was navigated. Now Russia currently operates 160 freighters on summer schedules in the Northeast Passage’s more open but colder waters.”
http://www.time.com/time/magazine/article/0,9171,770864-2,00.html
Re: Shawn Whelan (#328), What’s the record for years in a row passing through the NE passage?
Re: Neven (#329),
Icebreakers go through there every year.
Re: Shawn Whelan (#332),
And are they breaking ice, O Master of irony?
Re: Neven (#329),
At the current time the figure for consecutive years of passage through the Northeast Passage is something like 75-80. Soviet/Russian ships have been passing through regularly at least since 1932, though often with icebreaker assistance. Of course there has been difficult ice years, and even occasional disasters like in 1937 when several ships were beset by ice in East Siberian waters and a whole shipload of prisoners headed for Kolyma starved to death.
Re: Urederra (#326),
Bear in mind that it took him 3 seasons to complete his journey, also the southern arm of the NW Passage which he traversed is the most southerly part of the Arctic ocean.
Re: Phil. (#335),
Amundsen was in the Arctic to do scientific experiments. That is why the trip took three years. We have been over this several times.
Re: Shawn Whelan (#336),
Did any freighters use the passage for commercial activity this year?
Re: crosspatch (#339),
Thank you for offering a forum. I made my “prediction” of ice extent back in March/April (?) of 5.1 mKm^2. It was posted on Lucia’s site and was my first stab at such a thing. I’m mostly interested in the factors that one would need to introduce to provide a reasonable prediction, rather than “being closest”. I’m not in favor of purely statistical projections since I think that this is just what many projections are and I think that that approach is fundamentally flawed. Such projections really only project a continuation of the past. It’s no better and perhaps a lot worse than projecting Stock market valuations or baseball averages. Both of these latter have a lot more information available, yet projection based solely on statistics has major issues even for these.
In my simpleminded model, I tried to include the heat transport into the Arctic by currents from looking at the AMO and PDO. I tried to include my own guesstimate of ice thickness (I didn’t buy the hand-waving argument that first year ice will always be the same thickness – I thought that the first year ice existing closer to the pole would be thicker than “normal” for first year ice.). I tried to guess long term weather patterns (again looking at the ocean, since the ocean provides the atmosphere’s memory). What I came up with was 5.1-5.4, but I chickened out and only posted the lower limit on Lucia’s site. Everybody else was much lower and I admit that this intimidated me a little. Does anybody else think that my list of “inputs” is reasonable? Does anybody else have favorite parameters to look for that they want to share?
Re: crosspatch (#339),
Thank you for offering a forum. I made my “prediction” of ice extent back in March/April (?) of 5.1 mKm^2. It was posted on Lucia’s site and was my first stab at such a thing. I’m mostly interested in the factors that one would need to introduce to provide a reasonable prediction, rather than “being closest”. I’m not in favor of purely statistical projections since I think that this is just what many projections are and I think that that approach is fundamentally flawed. Such projections really only project a continuation of the past. It’s no better and perhaps a lot worse than projecting Stock market valuations or baseball averages. Both of these latter have a lot more information available, yet projection based solely on statistics has major issues even for these.
In my simpleminded model, I tried to include the heat transport into the Arctic by currents from looking at the AMO and PDO. I tried to include my own guesstimate of ice thickness (I didn’t buy the hand-waving argument that first year ice will always be the same thickness – I thought that the first year ice existing closer to the pole would be thicker than “normal” for first year ice.). I tried to guess long term weather patterns (again looking at the ocean, since the ocean provides the atmosphere’s memory). What I came up with was 5.1-5.4, but I chickened out and only posted the lower limit on Lucia’s site. Everybody else was much lower and I admit that this intimidated me a little. Does anybody else think that my list of “inputs” is reasonable? Does anybody else have favorite parameters to look for that they want to share?
oops: I meant to reply to this and hit the wrong link: Re: Robert Grumbine (#338),
Re: crosspatch (#340),
Did any freighters use the passage for commercial activity this year?
Many, they do every year.
Re: Shawn Whelan (#361),
Is there any record of traffic in the NW passage this year? Reading the various blogs of private pleasure craft taking that route, I didn’t see any remarks of meeting commercial shipping, that doesn’t mean there wasn’t any, just that nobody mentioned seeing any.
I do know that the Canadian Coast Guard regularly patrols the passage with ice breakers but the only commercial shipping I hear about tend to be something of the oddity rather than the rule.
As far as I can find only one commercial vessel got through the Northwest passage this year and it was a passenger cruise. Not exactly a vital economic route. Two ships sailed the Northeast passage.
Re: crosspatch (#363),
To say “Two ships sailed the Northeast passage” is a bit misleading. The MSM has grossly misunderstood and misreported the voyage of the two Beluga ships. Their purpose was not to sail the Northeast passage but rather to deliver equipment to a settlement on the mouth of the Ob River which empties into the Artic Ocean. They were in the Northeast passage because their destination was on the northern coast of Russia.
See http://eureferendum.blogspot.com/2009/09/turd-eaters.html
and
http://eureferendum.blogspot.com/2009/09/pictures-tell-story.html
Re: Shawn Whelan (#361),
Regular commercial ships do not typically operate in the ice. Ice is very hard on the hull, rudder(s) and screw(s), and modern hull designs with a bulbous bow are not designed for ice breaking. There are classes of ships called Ice-Strengthened-Ships that are specifically certified by Classification Societies for operation in ice. Some of these ships go in reverse when they are going through heavy ice. They are equipped with reversible Z-drives and have a reinforced stern that is designed with the correct shape to break ice when going in reverse. These ships are not very common. I think it’s misleading to say that “many commercial freighters” use the passage.
Re: Shawn Whelan (#336),
That is not why the trip took three years.
The Gjoa made it into Gjoahavn on the evening of Sept 12th 1903 after suffering damage from running aground in a storm. Amundsen decided that was where he would spend the winter and carry out his experiments, he reports “We had been working very hard during the last few weeks and needed a rest. As regards myself I confess that I wanted breathing time.”
They were iced up by October 1st.
Having completed their scientific work they left Gjoahavn on August 13th 1905. He describes the next 14 days as “the most dangerous, intricate and nerve-wracking navigation of the trip”.
By September 2nd 1905 they were forced by ice to over-winter at King Point, Amundsen described it as an ‘unusually severe winter’.
After the winter the Gjoa was finally able to leave Herschel island on August 10th 1906 and head for Nome which was reached August 31st.
So three seasons travelling and one on research (by Amundsen’s description it’s unlikely that they could have left Gjoahavn that summer in any case). During that summer during exploratory journeys along the shore of King William Land they were forced to leave two dories on the shore and walk/ski back to Gjoahavn.
Re: Phil. (#334), Rehash of Amundsen’s journeys is forbidden on this site, so I will just quickly say your narrative is somewhat misleading. He spent two of those three seasons establishing weather stations and learning from Inuit how to travel in polar conditions — which enabled him to win the race to the South Pole.
Re: An Inquirer (#350),
In which case I conclude that you think that Amunsden was misleading when he wrote his account of his travels in 1908? Everything in my narrative is exactly as he describes it, three seasons travelling!
Update is out: Arctic added 24,531 sq km of ice yesterday, and is now just a hair over 5.3mm.
I don’t know if it’s relevant or not, but I keep track of the five day average change, and today is the first day of the season that the number went positive (1,156 sq km).
Can someone please re-post the predictions (guesses) for the minimum day and extent?
The Northwest passage was crossed both ways in the 1940s by a small wooden boat.
Read about it here and here.
The Canadian Geographic Society was supposed to have a cruise ship start at Cambridge Bay on Sept 1 and go west and northwest then east to Greenland by Sept 16. I have not been able to find the location of this ship to see if it completed the trip as planned.
http://www.adventurecanada.com/rcgs
Re: Gerald Machnee (#337), Sorry, Meant to reply to the topic.
Hi Gerald,
Clayton from Adventure Canada here. We did, indeed leave Cambridge Bay and got out to Banks Island, but then turned around and did the southern part of the Northwest Passage rather than passing through the Prince of Whale Strait, as it was full of ice. On the previous trip, we made it through Bellot Strait about two days after the ice broke up there, and had 19 polar bears sighted who had, most likely, just finished hunting on the ice. We encountered 9/10 ice on our approach to Gjøa Haven, but had very little difficulty navigating through.
Cheers!
Clayton
Re: Clayton Anderson (#355), Thank you for that. So would you say that there was more ice in those parts this time than last year?
Re: Richard (#357), Yes, we found that this year thee was more ice than the past few. We had planned the journey based on the past 5 seasons of ice coverage and thought we had a better than even chance of getting through Prince of Wales strait if we timed it right.
Re: Clayton Anderson (#391),
Judging by the ice maps there was a window of a couple of days about a week ago but it would have been tight!
Re: Clayton Anderson (#391), Thank you Clayton. I recall reading other blogs where they said the maps showed 30% ice, (going to Gjoa Haven?), but when they got there it was more like 90%.
I guess you cant always trust even satellites. I would love to attempt the passage one day – but there are so many things I want to do.
Re: Richard (#401), do all the other stuff for 20 years or so and then the passage should be much easier. 😉
Re: Neven (#402), Seriously, Neven…
don’t count on it! For me, the window might be closing…[Life extension, someone…??]
Re: Staffan Lindstroem (#403), PS. I
hope I’m dead wrong, though…DS.
Re: Neven (#402), do all the other stuff for 20 years or so and then the passage should be much easier.
Quite touching your faith in a 20 year prediction, based on a consensus, when a much easier 10 week prediction, computed from a consensus, has miserably failed. You have to be a die hard believer.
Re: Richard (#415),
That position is absurd. The time scales of prediction are completely different. Weather versus climate. Is it possible to avoid such silliness on this thread?
Re: Richard (#415),
More touching is your precious suggestion that one head is better than two. I think Steve M has stated many times that he has no problem with a consensus approach to climate science, as long as it is transparent and open to criticism.
Re: bender (#420), ..I think Steve M has stated many times that he has no problem with a consensus approach to climate science, as long as it is transparent and open to criticism.
And in this case the criticism is that it failed in its prediction.
Re: Richard (#424),
That consensus can fail does not imply it is not the best method to formulate predictions.
Re: bender (#427),
Are you implying that consensus is the best method to formulate predictions?
Re: Richard (#415), Are you sure you are not Shawn Whelan, the glorious destroyer of the invisible consensus-enemy? 😀
Re: Richard (#357), Re: bender (#366), Our Mammal specialist from DFO felt the bears were well fed and in good shape. There were two sets of mothers and cubs as well. It was a good number of sightings for us, we usually don’t see that many the whole trip and this was in one day.
We found that this year thee was more ice than the past few. We had planned the journey based on the past 5 seasons of ice coverage and thought we had a better than even chance of getting through Prince of Wales strait if we timed it right.
@319 Jared:
I was responding to someone talking about the poll at my web site. Most people there did indeed greatly underestimate the September average (at least they did barring major surprises in the next couple of weeks). But they are also strictly or almost strictly amateurs on the topic, so I wouldn’t be terribly concerned about the error. Nor would I be impressed that people making their predictions at 3 weeks lead did better than the folks who stopped by my place and entered figures at 10 weeks lead.
In complaining about NSIDC, you’re looking at the wrong group. NSIDC provided some of the estimates, and has in each month. But the bulk of them are from elsewhere, including me in the June outlook. And the summaries are collected and published by James Overland and Hajo Eicken, at the ARCUS site, as part of the SEARCH. Neither Overland nor Eicken are at NSIDC.
If you go back and look at the June report, you’ll find a median of a little over 4.7 million km^2, with standard error 0.5 million km^2. (Mine having been 4.92). That’s going to verify within about 1 standard error, which is not bad for this kind of prediction at that lead time. The median, and even more so my, prediction is also far closer to what has come about than a prediction of ‘near normal’ — the prediction of those who claim that the sea ice has recovered. ‘normal’ is off by about 3 standard deviations.
If, instead, you want to go with ‘recovering’, rather than ‘recovered’, you’re welcome to stop by my place next May/June and leave your prediction based on, say, extrapolating forward from the 2007-2008-2009 series. Or whatever other method you prefer. (You all, that is, not limited to Jared.)
Re: Robert Grumbine (#338), If you go back and look at the June report, you’ll find a median of a little over 4.7 million km^2, with standard error 0.5 million km^2…That’s going to verify within about 1 standard error..
I think the mean is more relevant here than the median.
From the June report estimates: 3.200 4.200 4.300 4.600 4.600 4.700 4.700 4.900 4.900 4.900 4.900 4.920 4.920 5.000 5.000
Actually yours was given in the report as 4.9, as was K and H’s but I will grant you both 4.92
The Mean of these comes to 4.649. The Standard deviation of these estimates comes to 0.469. Thus Mean + 1 Standard Deviation = 5.118 so if the minimum is 5.24 then the actual does not fall within 1 standard deviation of your estimates.
Moreover if you look at forecasts based purely on modeling these are: 4.200 4.300 4.600 4.900
The Mean of these comes to 4.5 and Standard Deviation 0.316. Thus Mean + std dev comes to 4.816, hopelessly below the minimum.
I agree we are a bunch of amateurs but not you. You are a bunch of experts. We expect you to do better, not just “not bad for this kind of prediction at that lead time”.
Re: Richard (#353),
Actually Bob’s June prediction was 4.92 million km^2, with a standard error estimate of 0.47 and H&K’s was 4.92 ± 0.43.
Re: See – owe to Rich (#358),
I think Bob’s entitled to claim ‘not bad’ for his June prediction based on May data.
Re: Phil. (#359), Re: Richard (#353), Actually yours was given in the report as 4.9, as was K and H’s..
“Actually Bob’s June prediction was 4.92 million km^2, with a standard error estimate of 0.47 and H&K’s was 4.92 ± 0.43.”
Re: Robert Grumbine (#338), If you go back and look at the June report, you’ll find a median of a little over 4.7 million km^2, with standard error 0.5 million km^2. (Mine having been 4.92)…
I am not doubting that Robert Grumbine computed the figures as you say but he specifically said to look at the June report, which I did here. In this it says that both the estimates were 4.9.
Anyway I have used the correct figures and the fact remains that, so far, the minimum remains outside one standard deviation of the mean estimates, and it remains more so if we take the estimates based on pure modelling.
This leaves one to wonder how accurate climate modelling is, and whether it is akin to ancient astrology, as some Japanese scientists have alleged.
Re: Robert Grumbine (#338),
Sorry, but that’s bad statistics, which is usually hard to get away with on this site. The distribution in the June graph is not normal, the estimated standard deviation is vastly overestimated because of the outlier with the tiny prediction of 3.2, and if your standard error was correct on a normal distribution then 16% of the guessers, i.e. 2.4, should have predicted above 4.7+0.5 = 5.2, instead of all of them being 5.0 or below.
Please retract your “which is not bad” assertion.
Rich.
Does anyone know why DMI shows the extent to be so much lower than JAXA?
Also, the graph of 2009 relative to 2005 seems to be quite different as well. On DMI, 2009 is relatively quite a bit lower than 2005 than corresponding graphs are on JAXA.
Re: Daryl M (#341),
Someone asked the same over at WUWT. The answer seems to be that one (DMI) measures the extent of at least 30% ice and the other (JAXA) measures the extent of at least 15% ice. So the one with the greater reported extent is the one measuring down to 15% area.
A small addition to Steve’s post: the extent is now 5301094, so the increase was even more than first published. Another sign we reached the minimum? Only time will tell…
Listen
I posted this graph because some people were asking for trends concerning maxima and annual average. If you want, put your hand on the screen to hide everything before 1980 and then take a look at the trends.
I’m not responsible for this graph.
Re: Flanagan (#344),
I would prefer putting my hand over everything before 2003. As I have pointed out before (#171) even a cursory check of the data shows that summer ice area was significantly overestimated before then.
Incidentally the limited data that does exist suggest that there was as little ice as now some summers in the 1930’s. For example in 1932 when the Sibiryakov expedition sailed around Severnaya Zemlya (which has not been possible in either 2007, 2008 or 2009) or in 1934 when Sadko sailed north of Nordaustlandet in early August (not possible this year) and reached 82 deg 42 min north in open water west of Severnaya Zemlya on September 13(!), which would perhaps have been just possible to repeat this year. Incidentally Sadko then had to retreat south since the sea was starting to freeze, suggesting that the melting season was just about as long then as now.
You can read about the French sailor’s, Phillipe Poupon, trip through the northwest passage. He said it was not easy with the ice around. The only problem is the sub title which suggests ice is disappearing.
http://www.winnipegfreepress.com/canada/icy-northwest-passage-gave-sailor-the-chills-59214522.html
Hi Gerald,
Clayton from Adventure Canada here. We did, indeed leave Cambridge Bay and got out to Banks Island, but then turned around and did the southern part of the Northwest Passage rather than passing through the Prince of Whale Strait, as it was full of ice. On the previous trip, we made it through Bellot Strait about two days after the ice broke up there, and had 19 polar bears sighted who had, most likely, just finished hunting on the ice. We encountered 9/10 ice on our approach to Gjøa Haven, but had very little difficulty navigating through.
Cheers!
Clayton
Re: Clayton Anderson (#354),
Thanks Clayton. It looked quite solid on the northern route. Good to see the bears doing great.
Re: Gerald Machnee (#360),
He didn’t say the bears were “doing great”. He said he saw 19. To tell how “great” they were doing you’d have to measure body fat content. To tell how well the young are surviving you’d have to look at interannual rates of population change over time, not a single number from a one-time sample. Please.
Re: bender (#366),
OK, OK, I was having fun and in a hurry to get outside and and work in the overly warm September. I could not think of a better line. I really wanted to know if they were stranded on an ice floe. In 14 months at Resolute, I only saw about 3 polar bears, so 19 in a short trip is “great”.
Cheers,
Gerald
Re: Clayton Anderson (#355),
As far as I know Henry Larsen in the St. Roch in 1944 is the only non icebreaker to ever traverse the Northern route of the NW Passage.
Several others have run the length of the Northern Route of the NW Passage. Nothing new about low ice levels in the Arctic.
On the Arctic’s third expedition in 1910-11, Bernier took the vessel North to patrol the Davis Strait, Baffin Bay, Lancaster Sound, Barrow Strait, Viscount Melville Sound and McClure Strait. Open water in McClure Strait tempted Bernier to attempt the Northwest Passage, but because this would have exceeded his orders, he resisted.
http://www.ucalgary.ca/arcticexpedition/icebreakers/cgs-arctic
Steve: Shawn, as observed in #350, I’ve asked people not to keep re-hashing these old voyages over and over again.
PS IJIS have published an update. The increase is now 24,531sq kms.
And my prediction was that the ice will increase again on the 16th, so lets see
I am looking at JAXA on another screen at this moment. It shows 5,264,375 km2. That would be a drop of about 35,000 since yesterday. Rather large but not impossible.
But the chart itself does not show a tick downward. Admittedly that little red line is hard to read at the tip. So perhaps the posted number is wrong. The morning correction should be interesting.
http://www.ijis.iarc.uaf.edu/en/home/seaice_extent.htm
Big melt today 36,719 sq kms
Quite a big loss day, will have to wait to see what the correction is, but it almost eclipsed 13th as lowest extent day this year.
I have to say that I’m not too keen on historical passages because they had worse equipment but on the otherhand perhaps were “tougher” or more determined as it was a “first”. Descriptions are very subjective as well, “there was a lot of ice” for instance, what does that quantify?
Of interest
Cryostat2
And some details here on the resolution
details
250m is not bad at all.
Regards
Andy
yeah…but with the past 2 days corrections, i expect that to be greatly lessened in the morning. will be interesting to see. still, it is now the 17th of September so there can’t be too much melt left i would think and this is still above the minimum by 17,000……yet another day of this (if this holds) would bring us to a new minimum so i will watch the morning closely!
“This leaves one to wonder how accurate climate modelling is, and whether it is akin to ancient astrology, as some Japanese scientists have alleged.”
Ladies and gentlemen, I think we have a new world record jumping to conclusions here. 😉
Methinks they have got it wrong. The Sea Ice looks visually more if you toggle between the 15th and 16th here
@353 et seq.
The standard deviations are those of nature, as estimated by the groups making estimates. Several of us estimated the variability to be about 0.5 million km^2. That’s just the natural variability of the system. Being within 1 standard error is not a bad start.
In deciding that a predictive model is bad, one should compare it to something — unless you’ve got some a priori certainty about how good the model should be. You’re saying that the predictions are all bad. Ok. What is your a priori assertion about how good the predictions should have been, and why is it that (what method leads you to that requirement level)?
The usual starting point is to beat climatology, which all of us did except for the 3.2 estimate — which was labelled a wild lower bound by the group making it. Climatology is 7.4 million km^2, which’ll be more than 2 million km^2 off, vs. my own less than 0.5 million.
General note:
You shouldn’t get too excited about single day differences from a single platform (AMSR-E only, for instance) map. JAXA, for instance, is updating yesterday’s grids with today’s observations — where they’re available. Given the vagaries of both the satellite swaths and of weather contamination, there is jitter having nothing to do with whether the real full Arctic has more or less ice than yesterday.
Re: Robert Grumbine (#375),
You fairly casually throw out the word climatology. Do you think that you really have a good handle on what that is for the Arctic? What is your definition? My concern is that you are just picking a relatively short recent period of 20 yrs and labeling it climatology because this sounds a lot more impressive!
I think that there are severe questions about the quality of data prior to 1979 and severe questions about how typical the 1979-2000 period was compared to longer term Arctic conditions. Both of these would make any estimate of Arctic “climatology” problematic.
Re: Robert Grumbine (#375), I am an amateur so you will have to excuse questions for which, to you might seem, the answers are obvious.
“In deciding that a predictive model is bad” [or good?], “one should compare it to something …. The usual starting point is to beat climatology, ..”
“Climatology”? I thought you WERE climatology. All climate experts making an estimate. You could hardly be expected to beat yourselves.
Of course some of you, have done better than others, but if we go the consensus route, by taking the mean of each of your best estimates, then the actual, if it is indeed 5.25, lies outside 1 standard error of the mean.
If we take “climatology” to mean estimates made by climate scientists, then these would possibly be those based purely on modeling. These fare worse. The mean comes to 4.5, the standard error to 0.316 and thus the minimum lies beyond even twice the standard error. Would you still classify this as good?
It is a little unnerving because the estimates were made for a comparatively short period of time, and yet momentous decisions are being made on the basis of climate models which are supposed to predict accurately over much larger periods of time.
Then you go on to say “Climatology is 7.4 million km^2,…”, (which I presume is the prediction made by “climatology”). The only clue I have to this figure is from your blog, where you give this exact figure, as I understand it, as the mean of the last 30 years?
My questions to you are –
1. If you are claiming that the mean minimum of the last 30 years is the figure that is predicted by “climatology”, why should this be so? How does “climatology” dictate that this is the figure that the minimum ice should climb back to?
2. A question of fact – If you are saying that 7.4 million km^2 is the average minimum over 30 years, how do you get this figure?
I downloaded the data for Sept ice to 2008. I am pretty sure that these are not the minimums but the average figures for the month. And the average of these comes to 6.67 million km^2, a figure that is also given in your post above, but these I think are not the minimums but the average of the averages for the month.
I note that you would have lost 2 “Quatloos” in the bets you offered. One because the extent is now probably more than 4.92 in your evens bet and one because it is more than 4.97 in your 20:1 bet.
One way, I presumed, how one could judge a climate model would be to compare the prediction with the actual outcome. I believe there is an ultimate truth in nature which is oblivious to our desires and preconceptions. We make presumptions and calculations and based on these calculations, predictions. Sometimes the predictions prove accurate, which could reflect reality or be coincidence. But if the predictions poorly represent the outcome then it could be presumed that the model does not accurately represent reality.
In this case the models were brilliant, the reasoning impeccable, the maths dazzling, but alas the outcomes wrong. Ultimately I think this is what the model has to be judged by.
Re: Richard (#436), no-one seems to have answered him on “climatology”. My understanding (which is very limited) of this term is that it refers to a linear trend from climatological data (viz. the sea ice data in this case). I think it’s a silly term to use for that, but sometimes these things stick and no-one is brave enough to challenge the consensus and say “hey that’s a silly use of that word”.
I’m happy as always (well, nearly always) to be corrected.
Rich.
Not desperate your prediction might be wrong Richard ? 😀
I’ve got you at about 50% right up to now.
Regards
Andy
Comparing by switching back and forth between September 15 and 16, I don’t see how they could revise downwards.
http://www.ijis.iarc.uaf.edu/cgi-bin/seaice-monitor.cgi
Re: Pierre Gosselin (#378), The logical course of action is to wait, but for those who want to squint through another set of images, here is Cryosphere: http://igloo.atmos.uiuc.edu/cgi-bin/test/print.sh?fm=09&fd=15&fy=2009&sm=09&sd=16&sy=2009
i agree Pierre….but Robert says that could be due to other things. we shall see if our inclination is correct (although the underlying reasons may still be wrong?) this morning at 9 Central (30 minutes from now!).
and it also seems like the ice grew in those pics……strange. i would now not be surprised to see yesterday’s loss completely wiped out and an addition of ice replace it
Re: markinaustin (#381),
Remember it’s extent, if a region ‘fills in’ it won’t make any difference to the extent.
ok, so the new update is in and it went from negative 34,000 to negative 8,000
IJIS have posted an update and the ice reduction is now 10,000. They have increased the ice by 26,719 sq kms from their last posting.
I know certain people are going to crow that I predicted an increase, whereas their data shows a decrease of 10,000. To them I will say I stand by my prediction and that the ice has actually increased from the 15th. IJIS are not infallible. And if you believe that the ice has decreased by exactly 10,000 sq kms over the previous day then you will believe anything, and you probably do.
Re: Richard (#384),
Not much point waiting for the data since you know what it is going to be all along! Note that DMI also shows it going down. What is your problem with 10,000 it’s a valid number (hint 10,001 isn’t).
Re: Phil. (#386),
10,000 would be a valid number if it would mean 1*104. But knowing the Japanese I suspect they mean 1.0000*104; and that is not a valid number. I wish they would publish their numbers in Gm2 instead of Mm2, which would be exact enough.
Re: Alexej Buergin (#388),
“super” did not work; all the 4s and all the 2s should be high up.
Re: Alexej Buergin (#388),
It’s a valid number because it corresponds to an integer number of pixels.
Re: Alexej Buergin (#388),
Phil instructed me on this. JAXA areas are in km2 not Mm2. The prefix is assumed to be included when raising to a power. For example, a square of 10 km on a side has an area of 10 km2. A cube 10 km on a side has a volume of 1,000 km3. So what you want is for JAXA to be like CT and use Mm2 not km2.
Re: DeWitt Payne (#405),
So when you calculate with prefixes the mathematical laws are not valid:
ab^2 is a*(b^2) but if a=k=1000 then kb^2 is (k^2)*(b^2). Amusing, and I thought I knew the SI.
What I meant is: 1 (km)^2 = 1000000 m^2 = 1 M (m^2) and since those number are in the best case exact to about 1000 (km)^2 = 1 G (m^2) I would simply write the last 3 digits as 000.
Re: Alexej Buergin (#409),
I recall the conversation DeWitt refers to a while ago, as you illustrate above the prefix is part of the unit (i.e. a and b in your example are inseparable) so the unit we see so often ‘million km^2’ is 1,000,000×1,000×1,000 m^2, i.e. 1 Mm^2 (the standard SI nomenclature). Several posters challenged it, as far as I know DeWitt was the only one to adopt the SI style.
Re: Phil. (#410),
it is on pages 121 and 122 of
Click to access si_brochure_8_en.pdf
which is the bible of the SI
Re: Alexej Buergin (#411),
Yeah, I think that’s what I showed DeWitt, I was surprised that so many didn’t know.
Re: Phil. (#386),
DMI looks to me like the 30% extent was flat, not down, for yesterday. I opened the graph in a picture editor and blew it up 400% and it is flat but don’t have the exact number plotted. Also, Arctic air temperatures seem to be going further below normal.
Richard: is it some kind of a joke I don’t understand? When the data agrees with your “prediction” they’re valid, and when they do not then there’s a problem with JAXA?
Now, surely you’re joking.
What is the problem with 10,000 ? you don’t like round numbers?
Re: Flanagan (#390),
The point is that you only get a number as round as 10,000 once in 10,000 tries. In this case we’re dealing with daily readings or twice daily. So we should see such a number once every 13 years or so. Now maybe this was the magic time, but it’s much more likely it’s a “picked” number rather than a calculated one. In fact if I wanted to throw out a non-calculated number, I’d make it a very even number like this so everyone would know it was an approximation rather than a calculated one. Moreover if a calculated number ever came out like 10,000 I’d go to the special trouble of making a remark like “Yes, folks this is the actual number today, not something I just made up.”
Re: Dave Dardinger (#394),
10,000 is 64 pixels as I recall, so the probabilities are much higher than that, more like 1/100 if you did the stats on the daily change I’d think.
Re: Dave Dardinger (#394),
Dave, the figures don’t go down to 1km2, the quantisation is (6.25×6.25) so giving blocks of 39.0625km2. So in this case it gives 256 which means it is allowed. This is why you also get the same figure repeated more often than you would if they went down to 1km2 accuracy.
There’s no fudging here.
Richard, I gave you the benefit of the doubt before but now you are saying JAXA are in error just because it does not agree with your guesses. You seem to want to constantly be “right” even when obviously you are not. Your track record on guessing daily changes is poor and your assertions about what is true and how other people are wrong does not tie in well with reality.
Regards
Andy
Excuse me Dave, but you seem to have problems with probabilities. Where the hell does this idea come from that one has one chance out of 10,000 to get this number? 10,000 has exactly the same probability to appear as 10,000-1 pixel or 10,000 + 1 pixel. Maybe we should ask Steve? Help, Steve.
Re: Flanagan (#395),
No , you have a problem with reading what I said. Which was:
The operative term is “as round as” I.e. 10,000 is round to four decimal places. Sure 10,000 has (about) the same chance of occurring as 11111 or even 14159 but neither of the others are rounded.
OTOH, I stand corrected Re: Phil. (#396), and others. Though Phil and Andy ought to get together and decide exactly what number is correct. (I expect Andy is). Anyway, it’s clearly possible to have rounded looking numbers far oftener than I’d assumed.
Re: Dave Dardinger (#406),
Andy could well be right I remembered a 12.5×12.5 as the square used by the algorithm whereas the sensor was more like 6×4, but it could easily be half that. Just for fun I took the last couple of years worth of data, calculated the day-on-day change and divided by 156.25 to determine how many 12.5×12.5 pixels there were (all whole numbers no fractions). A quick eyeball check showed you might expect 10,000 (64 pixels) to turn up about 1/500.
Re: Dave Dardinger (#406),
“OTOH, I stand corrected Re: Phil. (#396), and others. Though Phil and Andy ought to get together and decide exactly what number is correct. (I expect Andy is).”
I wouldn’t count on it! It might well be 12.5 and I have halved it.Phil described it first and linked to the source, so it may be my memory. The general point about it being “quanitised” based on the method is valid though.
Richard, apologies if I sounded a bit bolshy yesterday, someone knocked my car wing mirror off so I was not in the best of moods. When you questioned the integrity of the JAXA data I suddenly thought if that is believed why the hell are we here? 😀
Regards
Andy
No time to explain the reasons in detail, will do so later this evening or tomorrow but:
1. If the 17th shows a big increase followed by a decrease this will cast further doubt on the 10,000 figure
2. If the 17th shows a big increase followed by another increase this will also cast further doubt on the figure. It will be an anomaly of the same probability as “clouds”
3. There is a way I could be proven “unequivocally” right. Flanagan you will never guess why so dont bother. If it happens you can eat your hat and everyone else too. You are very unquestioning and gullible, which has its uses ([RomanM -clip – out of bounds]).
Never guess how sorry
and when I say followed by I mean the updates not the subsequent day
And obviously if it had been a “manipulated” figure like some suggest, maybe they would have decided for somthing else than a round number which , to some, might look artifical no?
2002 9 17 5.844091 260 0.028338
2003 9 17 6.033281 260 -0.008594
2004 9 16 5.882813 260 -0.029375
2005 9 17 5.422344 260 -0.025312
2006 9 17 5.828281 260 0.022187
2007 9 17 4.268750 260 0.001094
2008 9 16 4.726250 260 -0.005625
2009 9 17 5.285313 260 -0.005781
How much talk is there now amongst alarmists about 2007 being a “tipping point” for the Arctic?
IJIS published data for the 17th – 8,781 loss so far.
Re: Richard (#416),
Richard, please see Steve’s post #413 which are the correct figures for the date. You should be more careful in your pronouncements because it does not aid your cause.
Re: Michael Jennings (#422), I beg your pardon I had incorrectly copied down 5,282,313, instead of 5,285,313. The loss is indeed 5,781 as given by Steve and not 8,781. Shall copy and paste in future.
Pardon the speeling above 😀
Wow! The update just loaded and the new extent is 5,326,094 km2, a positive correction of about 40,000 sq km.
If I’m doing my math right, the correction means that extent increased 35,000 sq km yesterday, which would seem to end the 2009 melt season.
* Ask a Question
* Do Background Research
* Construct a Hypothesis
* Test Your Hypothesis by Doing an Experiment
* Analyze Your Data and Draw a Conclusion
* Communicate Your Results
If your data does not match your hypothesis do NOT run for cover by claiming consensus. The best way to formulate predictions is to use the data and test your prediction against the data.
Re: realist (#429),
Semantics. When you have umpteen results from umpteen groups running umpteen models and none are 100% correct but all have certain degrees of correctness “the best way to formulate predictions” is to triangulate amongst the ensemble of results. Which is a mathematical definition of what it means to do “consensus” forecasting. Your peculiar – and ironically unrealistic – definition is irrelevant to the multi-party case introduced by Richard. Please, let’s not bicker over semantics, hmmm?
Revised 2009 is now above the 2005 minimum (which occurs in about 3 days). It’s up over 75,000 sq km over the past 4 days, 3of 4 being increases. Looks like the season is over.
month day ice dd diff
2002 9 17 5.844091 260 0.028338
2003 9 17 6.033281 260 -0.008594
2004 9 16 5.882813 260 -0.029375
2005 9 17 5.422344 260 -0.025312
2006 9 17 5.828281 260 0.022187
2007 9 17 4.268750 260 0.001094
2008 9 16 4.726250 260 -0.005625
2009 9 17 5.326094 260 0.035000
Under DeWitt Payne/BarryW/Bob Tisdale’s tropical ocean heat transfer/arctic sea ice hypothesis, a post-ENSO trend reversal (temporary) may in fact have started in 2007. Check unthreaded.
Realist #428
You forgot ‘Prediction”
If prediction is incorrect then restate hypothesis and redo experiments.
I pointed out in an earlier post that the guy who came closest had a 17% spread. 4.92 +/- 0.35. Imagine if NASA missed the moon by 8.5% or a satellite launch was 8.5% out of orbit. The people paying the money would not be pleased.
Its the prediction part that makes or breaks science. If you cannot predict then its not science and don’t pretend it is.
From NSIDC.
They feel it is appropriate to predict future warming by averaging all the climate models together. It seems to follow they would think it appropriate to average ice extent across all models.
That gives an average of 4.6 square kilometers.
D’oh!
4.6 million square kilometers.
Reduction yesterday 10,000 Km^2 Increase today after correction 35,000 sq kms. Does that not give you a just a teeney weeney bit of doubt that these maybe just artifacts. The broad data yes is correct – ice is increasing. But maybe, just maybe it didn’t really decrease on the 16th?
PS I would like to clarify: How does “climatology” dictate that this is the figure that the minimum ice should climb back to in 2009?
It may be a bit early to say that the NH sea ice minimum extent is behind us. A few days of melting or compaction could still occur, and I will personally wait a couple more days before calling the ‘extent loss’ season over.
I must concede though that my prediction of a minimum on September 17 has been proven wrong. No quatloos or brownies for me.
As I ponder the 2009 NH sea ice extent minimum, I am left with a few lingering questions. Firstly, is the notion that the arctic sea ice is in a ‘death spiral’ well and truly refuted? Secondly, what is the forecast for 2010 multi-year ice? And finally, is there a new phantom menace lurking in the crags and bergs of arctic ice? Ice volume was the boogey-man for 2009. What will it be for 2010?
Discuss amongst yourselves… 8)
Another small increase. 2009 is now above 2005 day 264.
month day ice dd diff
2002 9 18 5.872429 261 0.028338
2003 9 18 6.032031 261 -0.001250
2004 9 17 5.821250 261 -0.061563
2005 9 18 5.385156 261 -0.037188
2006 9 18 5.874063 261 0.045782
2007 9 18 4.281406 261 0.012656
2008 9 17 4.718594 261 -0.007656
2009 9 18 5.332188 261 0.006094
small increase but i suspect it will be larger when the correction comes in…
Re: markinaustin (#441), I wouldnt be too sure about that in this case. Visually the ice seems to have gone down in some areas, up in others. Not sure what the net effect will be. Could be down.
But one thing seems plain. The SST’s are dropping. We should see a lot more ice in the coming days.
Struggling to see how this fits in with NOAA’s declaration that the world’s ocean surface temperature was the warmest for any August on record? August warmest ever and ice minimum decreases. Strange
I wouldn’t say we are out of the woods, there could still be some decreases in store but I find it hard to believe we would face enough decrease to get below the previous minimum. Arctic air temperatures are now well below normal. Any significant change at this point is going to be wind related. The sun is offering nothing in the way of heat at this point. There are no sea temperature worries. It looks like it’s all uphill from here.
Small enlightening of this serious thread
h/t to radiant, solarcycle24.com
Re: Juraj V. (#446),
Thank you for posting that Juraj V., very interesting to see the parallels of thinking between then and now. It is becoming more and more apparent that 2007 was an extreme outlier in Arctic Sea Ice. I think it likely that the next several years will see flucuations in the totals enough to make both sides claim it reinforces their view. Like Climatology itself, Sea Ice is a subject we probably know less about than scientists think we do.
Re: Juraj V. (#446), Hmmm, that 40 percent decrease in volume in the 40’s and 50’s – unprecedented!!
Richard Black at the BBC has called it with Pause in Arctic’s melting trend.
He has the wonderful quote
though I personally doubt whether 2010 could be the final word. If 2010 were extreme, one way or the other, would that introduce more or less uncertainty about the future than if it were in the recent range?
so after the morning correction the amount of increase is 22,000 (i am rounding)…..
i am guessing we are not going to go down any more. at this current rate of freezing and the temperature drops it makes me think that there won’t be any more down days until the spring.
but i do realize that we may have one or two more down days….just seems unlikely at this point.
Funny how the morning numbers have recently been adjusted nearly always to more ice. Looks like 2009 passes 2005 today.
month day ice dd diff
2002 9 18 5.872429 261 0.028338
2003 9 18 6.032031 261 -0.001250
2004 9 17 5.821250 261 -0.061563
2005 9 18 5.385156 261 -0.037188
2006 9 18 5.874063 261 0.045782
2007 9 18 4.281406 261 0.012656
2008 9 17 4.718594 261 -0.007656
2009 9 18 5.348438 261 0.022344
It looks like the bottom is going to be 5.249844 on the 13th unless something happens to cause a huge compaction in the ice. Missed it by one!
It looks like 2009 is tracking above the average rate of increase. The slope of the extent increase could be interesting this season but too early to tell.
Re: BarryW (#454),
I missed it by 0 😉
Rest of my guesses were rubbish though 😀
Ok, so assuming we are now into the reverse ablation season will it rebound like last year or be normal? That’s the next interesting checkpoint as we work our way back to the maxima in 2010.
In the summer, I added something to my spreadsheet to smooth out the daily ripples, which calculated the difference from seven days earlier. That’s now gone negative and shows the refreeze over seven days. Now at about 70K, it’s the highest seven-day performance for any date prior to September 21 since 2002, except for a single point in 2004.
Not an especially meaningful factoid, but fun to contemplate. Maybe cold temps will form sea ice earlier and much farther south than usual this year. Maybe in Copenhagen, say in December. Oh, please, puleeeeze.
I have been thinking further on the question of Sea Ice Modelling, and perhaps I was a bit unfair to Robert Grumbine earlier. The CA participants, acting as the farmers leaning over the fence, have certainly done better in predicting Arctic minimum than the professionals this year. But that is probably due to an optimism about climate and a “feel” for time series.
If you were a professional modeller, what would you do? You need to have good data and a valid relationship to the output of the model. Should we make a list of variables here, and ask Robert how they match against the ones used?
* Sea temperatures at 70N
* Land mass temperatures at 60N
* Amount of multi-year ice
* Thickness of ice
* Sunspot average of previous 6 months
* Aerosol estimate from volcanoes
How to combine these must be problematic, and presumably one wants to compare to previous years’ values. I have no idea whether, as a scientist rather than a farmer, I could do better than the 15 groups of experts. But you have to say that, even if the standard deviation from a smoothed minimum extent is 0.5M(km)^2, the 15 groups collectively missed a trick this year – or the Arctic weather was freakishly unfavourable to melting and dispersal.
I write all that in regard to the early predictions (say based on May’s data). I find it hard to be as lenient on the estimates given July’s data. 2008 had a much greater than average drop in August, and the modellers had to be betting on a rough repeat of that event in 2009 for their predictions to come true. Nature “batted last” and won. What scientific data did the modellers have, which put them on the wrong side of reality? Did they actually look carefully at the July data?
Rich.
Re: See – owe to Rich (#461),
Sorry to harp on this again, but instead of pontificating and asking these hypothetical questions why don’t you actually read the report (which has been linked here several times)? We had the same nonsense on here about a week ago For example: Phil. (#82), et seq.
Re: Phil. (#462), I’m sooo sorry Phil, you’re right, I should RTFR. Well I have now – at least the “Full Report” and one of the submissions.
Cor, what a laugh, none of them use all my variables above. Perhaps the most amusing is Nguyen, who used a model which would have predicted 5.3 (spot on, man!) but subtracted 0.9 because his model had overestimated by 0.9 in earlier months. Eat your heart out…
I have to give some credit to Lindsay, the highest predictor at 4.94 – he increased his estimate a huge amount (from 3.99) compared with the previous month. At least he seemed to tell which way the wind was blowing and ice melting.
Phil, does this satisfy your standards a little better?
I certainly would welcome a debate on what the important variables are (Lindsay was using July ice thickness >1.0m and its correlation with September minimum – that’s just one variable used).
Rich.
They blame their bad predictions on the weather.
http://www.arcus.org/search/seaiceoutlook/2009_outlook/update_september.php
http://nsidc.org/arcticseaicenews/
And the Alfred Wegener Institute says that despite 2009 higher ice extent, it does not constitute a recovery.
http://www.awi.de/de/
Sorry, only in German.
Here they (Dr Gerdes and colleagues, claim they were almost right on the money!
Translated from the German:
http://www.awi.de/de/aktuelles_und_presse/pressemitteilungen/detail/item/minimum_2009/?cHash=273ba67544
I don’t know which bar on the above graph Gerdes. And it was strange to see almost everybody revise sea ice extent downwards in August. So I don’t understand why they are boasting about their models. Maybe to misleadingly impress government funders.
Re: Pierre Gosselin (#464),
Gerdes and his German team are “Schlaumeier” (“clever dicks”). First they cheat with the spread, using 80% probability (about 1.3 SD instead of 1), and then they publish 2 estimates, one high and one low. So in any case they can claim to be right, since the wrong one will be quickly forgotten.
(“Clever Meier” is not a criticism of an unsuccessfull American team, but just a German saying)
“4.5 bis 5.5”
bis = to
While all the attention is on the Arctic, the Antarctic is at about the sixth highest level of the last 30+ years.
Re: Xyrus (#466),
If Richard is pontificating I feel that you are bloviating in this entire comment. (-: This is a nerde statistics site, so please so the stats of long term climate conditions verses short term weather conditions in regard to artic sea ice.
Interestingly, at the south pole, the Antarctic still seems to be adding sea ice.
you kust forgot to include the two most important factors,
– oceann currents and
– wind patterns,
because much of the ice doesn’t melt in the basin, but is transported away and melts off Greenland.
the models didn’t include this, so their
– prediction was false again and
– their analysis of former events like the 2007 minimum was false as well.
Re: Manfred (#468),
Yes they do, if you can’t be bothered to read the report please don’t come on here talking about ‘the models’.
I find it very strange and alarming that our super reporter has gone absolutely silent on the ice. There is no need to be depressed just because the ice has stopped melting. Surely there must be something negative about the ice news that he could remind us about.
Pierre Gosselin and Realist:
6th highest maximum anarctic extent this year, 4th lowest maximum last year… There’s absolutely no statistically significant trend down there. Contrary to the North Pole.
Re: Flanagan (#479),
i thought the word “global” in “global warming” supposed to mean “in the whole planet” rather than “only in the North Pole”.
Re: Flanagan (#479),
1. Seems strange to be concerned about minima in Arctic and maxima in Antarctic. Other times of the year have ecological signficance as well.
2. The Antarctica maxima last year was affected by storms on the outer limits of the ice range.
3. I would be hesitant about using the term statistically signifant in samples less than 30, especially when oscillations are involved.
I guess this has been shown to be a wild flight of fancy?
http://news.nationalgeographic.com/news/2008/06/080620-north-pole.html
That was the “consensus” view, as of a little over a year ago. Now that these predictions have clearly been shown to be erroneous, and assuming that the trend (20-30 year) is now toward increasing NH ice, and next year has similar gains as the last two, where does that put us in terms of the ’79-? average? At half a million per year increase, where does that put us 5 years from now? How about 20? Should I sell my Iqaluit time share condo now?
Why isn’t National Geographic warning us of the impending ice age? After all, the consensus for a melt was apparently based on only 2-3 years of loss…
Just wondering.
Re: INGSOC (#483),
You need to understand climate parsing.
If it had happened, it would have been an amazing projection showing the brilliance climate scientists. Since it didn’t happen, well, he did say “may”.
Re: John M (#486),
My point is that to do the kind of analyses that you mention you need information that we don’t have. In the absence of this, I get offended at throwing around linear trends over an arbitrary time frame as if it has meaning. Elsewhere I suggested the idea of breakpoint analysis, not because I’m enamored with this, but because it gives a little insight into the robustness of claims based on trends. Claims such as your “below the trend line”. If a little more sophisticated analysis produces opposite results, then IMHO it calls in question the whole ball of wax.
Flanagan, there is a trend in Antarctic sea ice, although it is small. The NSIDC monthly slope is half a degree positive per decade from the 1979-2000 mean. Hope I said that corrrectly.
Aha…I seem to be first with the update! 5,358,594 km2, or another adjustment from down to up. I think that this should put us over 2005. I believe that most arctic researchers use the Sept. average rather than the daily minimum. In this case, it’s entirely possible that the 2009 summer monthly minimum will be above 2005 and hence the 4th lowest. Will this be presented on the various sites?
Re: MikeP (#489), Mike is 5,358,594 an update or the first posting? Middle of the night here, will check in the morning
Re: Richard (#491), Update.
Re: MikeP (#492), I see – what was the original amount then?
Wouldn’t be surprised if tomorrows figures (the 20th) shows a loss. If how the melt takes place is by the ice being transported south to be melted by the Gulf stream, as I have read somewhere above, that’s what seems to have happened between the 19th and 20th.
There is a dent in the ice and a clear warm water tongue west of Spitsbergen and also judging from the way the big bay is being moved around still some warm waters pouring in from the Bering Straits.
Re: Richard (#495), The original was something like 5337031 (from memory).
Re: MikeP (#481),
It’s only possible if ice extent increases at a rate greater than 100,000 km2/day for the next ten days. In other words, it’s highly unlikely. Here’s the data:
year Sept. mtd Sept av.
2002 5749230 5888005
2003 6110539 6126469
2004 5855359 5958901
2005 5558789 5530094
2006 5891805 5913271
2007 4393156 4380521
2008 4779977 4836974
2009 5332117
If I did my sums correctly, extent would have to average 5,930,000 km2 for the next ten days for the September average to exceed 2005. The NOAA average has been about 130,000 km2 less than the JAXA average the last two years. If that holds true for this year and the 2009 September average reaches 5,400,000 km2, then expect the NOAA average to be 5.27 Mm2 (tired of typing all those zeros).
Re: DeWitt Payne (#513), Agreed there is a much simpler and clearer way to see that. Draw the graph. From the 21st of August to the 19th of Sept the 2009 extent has been below that of 2005.
But if you ask the question: Which year had more ice 2005 or 2009? – the answer is 2009 – much more.
Flanagan: You are wrong. I did not bring up Antarctica. I was responding to your statements on Antarctica. And again, you are wrong. I am not saying that the Antarctica has a statistically significant trend.
snip
At any rate, this is off topic. More on topic it appears that the minimum has passed. We’ve had several consecutive days of positive accumulation. I know some on here have been speculating about a record refreeze, but there’s quite a bit of warm water up there still.
Steve– Yup.
The BBC news besides being biased is also inaccurate. They claim that the minimum extent was 5.1 million sq kms on the 12th of Sept., in their news item Pause in Arctic’s melting trend
Re: Richard (#502), No, the BBC correctly quoted their source, NSIDC.
Re: Nick Stokes (#506), Re: Richard (#502), No, the BBC correctly quoted their source, NSIDC.
1. No the BBC STATED “At its smallest extent this summer, on 12 September, the ice covered 5.10 million sq km (1.97 million sq miles).” No source was quoted for this statement.
They then talk about an analysis which “.. is compiled from satellite readings at the US National Snow and Ice Data Center (NSIDC) in Boulder, Colorado.”
2. Is the “source” correct in saying that the minimum extent is 5.10 million sq kms on the 12th?
snip – this is hugely OT. Plus it’s the sort of generalized discussion that I try to discourage editorially. I haven’t gone back and snipped some similar earlier comments only because of time.
5,358,594 square kilometers for the first measurement of september 19th. A tad more than yesterday.
Re: Flanagan (#485),
If measuring global temperatures is not easy and there are founded doubts about the quality and accuracy of the different temperature records (you know it because you read WUWT), more difficult is to know the actual heat content. And I stop here because is way too off-topic.
snip -OT
I noticed that the modelers predicting minimum ice extent were using the average extent for the month of September and not the minimum at the day of the minimum. Would not that be a better metric to use?
Re: Kenneth Fritsch (#512),
If by models you mean the predictions at the begining of this thread, then the actual average would be larger than the minimum value so there predictions are even worse than they first appeared?
Re: Richard (#510),
JAXA reports a minimum of just under 5.25 million km-sq on the 13th of September in its CSV data file at their website:
9 12 2009 5259375
9 13 2009 5249844
9 14 2009 5276563
That’s a sizeable 150000 km-sq difference. Who’s right?
Re: RomanM (#513),
Both and neither. Different algorithms ect.
Re: BarryW (#516), Re: RomanM (#513),Both and neither. Different algorithms ect.
Ok here is the link I think to the NASA data. Can someone download the data and paste it here for us to check? And also tell me how to unravel those bin files so that I could check it for myself?
Steve: I have some functions at CA/scripts/seaice but they are not user-friendly.
“22 men, 1 ball, and in the end the Germans win”; how do they do it ?
Kauker et al (including Gerdes), corrected to 1 standard deviation (probability ca 2/3):
Ensemble 1: Between 4.64 and 5.40
Ensemble 2: Between 4.05 and 4.81
Since these overlap, they cover the interval from 4.05 to 5.40
Can’t miss, can they?
Passed 2005.
month day ice dd diff
2002 9 20 5.929105 263 0.028338
2003 9 20 6.122656 263 0.033750
2004 9 19 5.808750 263 -0.000938
2005 9 20 5.345156 263 -0.023282
2006 9 20 5.846875 263 -0.045625
2007 9 20 4.310313 263 0.014063
2008 9 19 4.745000 263 0.008594
2009 9 20 5.357813 263 -0.000781
Re: Steve McIntyre (#502),
Not passed 2005 at 30% extent, but arctic temperatures are still well below normal. I would expect things to ramp up considerably if things continue in the same pattern as they are currently. Daytime highs are now in the 20’sF in Qaanaak, Greenland. Show is flying at Prudhoe Bay.
Re: crosspatch (#504), Daytime highs are now in the 20’sF in Qaanaak, Greenland. I think that should be 40 F not 20
Re: Richard (#504),
Not according to the forecast I am reading:
Tuesday
Chance of Snow. Scattered Clouds. High: 21 °F .
Wednesday
Scattered Clouds. High: 24 °F
Thursday
Clear. High: 22 °F
Friday
Scattered Clouds. High: 28 °F
Saturday
Scattered Clouds. High: 32 °F
Re: crosspatch (#506), Well I think you are right. In fact its even colder:
Sep 22 Sep 23 Sep 24 Sep 25 Sep 26 Sep 27 Sep 28
Low temperature: 10 °F 9 °F 9 °F 13 °F 9 °F 13 °F 10 °F
High temperature: 18 °F 14 °F 17 °F 21 °F 19 °F 19 °F 17 °F
Monday Tuesday Wednesday Thursday Friday Saturday Sunday
-12° C | -21° C, -19° C | -26° C, -21° C | -25° C, -17° C | -24° C, -19° C | -21° C, -18° C | -19° C, -16° C | -17° C
I think weather city is unreliable.
I plotted a graph of the Sea ice differences between 2009 from 2008 and 2005.
The total ice in 2009 has been way above 2005 all through the year. In fact it has been below the 2005 levels only, for any substantial amount of time, from the 21st of August till yesterday the 19th of Sept, which was the period the minimum lay in.
Another way of looking at it is the total ice (area below the curve) has been way above the 2005 level and it is a little above last years level (so far). But it is climbing rapidly away from last years level with a few months of freeze to go.
From the JAXA graph it currently has a nice “normal” shape compared to most of the years, unlike 2007 and 2008.
Antarctica has a large area this year compared to last looking at Cryosphere.
Regards
Andy
471 Phil.: September 19th, 2009 at 2:33 pm
I meant an inclusion of multidecadal oscillation with a magnitude to allow approx. 40% of volume loss as observed at the end of the last warming cycle approx. 50-60 years ago.
Apears to be that 50 years ago, scientists had better knowledge than today about the cyclical behaviour of climate and the temperatures 2000 years ago.
big correction upwards as is normal lately…..new amount:
5,383,594 km2
so now we are fairly far ahead of 2005 and the gap is growing. 38,000 and counting….
Well another large positive adjustment, from 5,357,xxx to 5,383,594 km2. It’d be nice to know where the extent is being added. Looking at NSIDC and Cryosphere I see a little on the Atlantic side, but otherwise can’t really tell.
Re: MikeP (#511),
According to the local charts at Cryosphere Today, area is increasing in the Arctic Basin and the Greenland Sea. Everything else is still fairly flat.
As pointed out above, the NOAA September average is what’s being predicted. 2007 and 2008 had fairly flat minima so the average was close to the minimum. 2009 so far has not spent as much time near the minimum and could indeed be very close to 2005 at the end of the month. I’ll do the September to date averages and post them later.
Some say that the melt-disaster-skeptics wait until the last minute to make their predictions. I got involved in 2009 too late to qualify as an early predictor, but here’s my short analysis for 2010. Other things being equal, 2010 should be like 2009, so I’d be tempted to say the minimum next September will be 5.25. There is a long-term downward trend but a big two-year gain. I’d call it a wash.
However, September 2009 is refreezing much faster than the past several years. Assuming first-frozen is last-melted, September is quite important. Only new ice that forms early has much chance next summer. The ice that forms from November on is pretty much doomed. An early start should mean that this year’s first frozen will be thicker by the end of winter and less likely to melt at the end of summer.
So I’m going to add 200,000 and say 5.45 for 2010.
It’s statistically significant that every model guessed low for 2009. What’s not clear is what is being signified. It could be a bias towards AGW, or it could be bad luck. I don’t see what would have been the stroke of bad luck in 2009 but let’s concede that explanation. However, if the ensemble of expensive computer models again misses low uniformly in 2010, one will start to wonder.
Re: Rob Spooner (#514),
“An early start should mean that this year’s first frozen will be thicker by the end of winter and less likely to melt at the end of summer”
Perhaps not true! If ice extent is large at the start of the freeze season then snowfall will insulate this expanse in Autumn and so it will not be as thick next year. That was actually a reason given for the ice to be thicker than expected going into 2008. Not much ice so no snow accumulated in the Fall and then when the really cold clear skies developed there was nothing to stop the ice getting thicker than expected over a wide area. I do agree it is counter intuitive until you think about it.
Regards
Andy
Regards
Andy
Re: AndyW35 (#517), an interesting observation, possibly true, and testable. Although not by me. Someone younger and hardier can venture out into the Arctic in the dead of winter to see what happens to the thickening process under different scenarios.
Still, it points to the value of my Null Model. The Null Model is time dependent, makes no pretense of science, and simply makes the most obvious assertions based on the gross values of generally accepted facts.
Scientific models should do better. For the 2009 minimum extent, they did not. Although there was no published Null Model forecast on June 1, it would have suggested a significant rise above 2008 because, like 2007, 2008 was unusually low, and on June 1, 2009 was in the middle of the recent pack.
So the Null Model forecast for 2010 won’t be altered as a result of this new information, because by design, it is mindlessly intuitive. My point is that science is supposed to do better. 2010 is a second chance. When will the scientific models make their estimates for 2010, I wonder?
Richard @ Post #518:
Mimimum extent is an important metric, I assume, because if the extent were to become vanishingly low at minimum we could have an accelerating feedback effect on climate and produce a condition that makes future recovery more difficult.
My point was that using the minimum at the day of the minimum might be too confining and perhaps more subject to a short period variation than using the month’s mean where the minimum occurs.
If I could see these minimums side by side, I might change my mind. Why do the modelers use the September mean?
If you are having a contest, I guess for the sake of maintaining short term interest that the day of mimimum becomes important.
Re: Kenneth Fritsch (#519), Yes and total ice is also an important metric, I assume, because it gives the average albedo effect over the whole year and not just the minimum that seems to be concentrated upon. It could have the positive feedback in the other direction that no one seems to consider.
Re: Richard (#520),
Understood. Does the maximum extent vary as much as the minimum? Or, better, does the mean for the maximum month vary more than the minimum for the minimum month? What relative importance do the modelers put on the minimum and maximum with regards to albedo and other feedbacks?
In Fauria et al. (2009) the authors used the maximum ice extent for the Western Noric Seas as their metric in a reconstruction. I assumed the maximum was used because for that area that that value varied more than the minimum – but what do I know as I am a rookie with regards to Sea Ice.
Re: Richard (#520),
Not exactly, if you consider the midpoint between the pole and the Arctic circle, that doesn’t see any sunlight for 112 days during which time the sea reaches it’s maximum extent. So that metric doesn’t represent what you think.
Re: Phil. (#540),
What we need, assuming that changing albedo is critical for feedback, is a metric that is weighted for ice extent (ice area?) by not only the day light hours but the angle of the sun light and also day light cloud cover. How do the models handle these factors? Could the minimum extent end up giving a good approximation for the changing effect of all these factors?
Or do we use minimum extent because of a conjectured minimum from which there is no return on a path leading to ice free Septembers?
I posted this over at WUWT as well.
We’re in for an interesting period over the next couple of months. This is the period where the seasonal jump in the Arctic refreeze significantly exceeds the start of loss in the Antarctic. Over the past two years, this has represented a large gain in the global ice anomaly in the region of 2.5 to 3.0 million km^2. :
With arctic temps below the last couple of years…:
http://ocean.dmi.dk/arctic/meant80n.uk.php
could we see a similar bounce? If so, it would approach or exceed the highest anomaly over the 1979-2008 period. Wouldn’t that be a paradoxical prelude to Copenhagen?
Re: Arn Riewe (#521),
I don’t think the ice rebound is very relevant to be honest, ok the rebound was high in 2007 and 2008 but that’s because there wasn’t much ice there. Same for years where the maximum was high, then a lot of ice melted quickly, but either way by June or Nov the yearly traces are all very well bunched.
This year, in relation to markinaustin’s post #545, I see no big increase yet, it still seems to be “turning the corner” like a lot of the years on the graph. I would guess it will follow closely to 2005 on the way up.
Regards
Andy
Re: AndyW (#546),
Your right, there’s very little scientific relevance in the whole ice extent issue (bragging rights only in my view). But the AGW doomsday proclamations of a couple of years ago make it very PR relevant. My point is that it’s extremely likely that the anomaly will turn positive with the potential for a 30 year high in the next couple of months. That would be worth shouting from the rooftops, but don’t expect to see it in the MSM.
I don’t agree that the higher minimum will be a constraint on an increase during the autumn. It’s got at least 5 million km^2 of open Arctic Basin to grow into. It would be unusual if it doesn’t grow by more than 3-4 million km^2 by the end of October.
Re: Arn Riewe (#549), constraint probably isn’t the word but there’s clearly an impact from the minimum. Looking at the graph of the last several years, there is a wide variety of minima, but by late fall, they’re converging. The minimum is not a platform from which a constant refreeze begins.
A question for the experts. Geographically, is the sequence of areas where the ocean refreezes roughly in reverse order from the thaw? In other words, is there first a refreeze of the areas that thawed in early September, then late August, and so forth backwards to the maximum extent? Not exactly, of course, but roughly?
Re: Rob Spooner (#550),
Yes, more or less. Cryosphere Today has separate area and anomaly graphs for different regions of the Arctic. Right now the central Arctic Basin and the Greenland Sea are increasing rapidly while the other areas are not changing much. You can mouse over the map, but there are a few glitches in the names. If you click on the area, the correct graph comes up. The Sea of Okhotsk is the area to the right of the Bering Sea on the Pacific side of Russia. It really should be a different color.
What is interesting to me is the anomaly graph at CT. The continued rise in the anomaly graph says that not only is the ice rising, it is rising at a faster rate than “average”. Notice that 2008 the absolute was rising but the anomaly stayed flat until mid-October which would indicate the ice was increasing at an “average” pace. This year seems to be recovering well so far.
Total ice in 2009 upto 20th Sept = 4.4 million Km^2 more than 2008
Total ice in 2009 upto 20th Sept = 54.3 million Km^2 more than 2005
Re: Richard (#523),
I am sorry, Richard, I am not sure that I am following you. Are you summing the daily ice extent, year to date?
Re: An Inquirer (#527), Yes or you could take the “anomalies” for each day, same thing. It will give you how much more extent there was for the period.
This is bound to have some effect on the albedo for the year
Yet the minimum in 2005 is more than the minimum of 2009 and that is supposed be far more important?
PS Total ice in 2009 upto 20th Sept = 128 million Km^2 more than 2007
It might be because the maximum is much easier to reconstruct historically than the minimum. There are trustworthy information in Icelandic sources about the annual state of the winter ice in the waters around Iceland back to the late Middle Ages.
Data on the summer ice edge are almost nonexistant before the seventeenth century and very sporadic until about 1900.
Re: tty (#528),
Thanks, tty, that explanation sounds reasonable to me.
Whats happened to the arctic temperatures today? They seem to have shot straight up
Re: Richard (#532),
Same thing that happened in NW Canada last week? Record temperatures throughout the region.
Record Warm Sea Surface Temperatures
Re: bender (#534),
I think it’s called El Nino and we haven’t had a big one since 1997-1998. It’s a little disingenuous, though, to make a big deal about how July 2009 was higher than July 1998 when SST’s were falling from much higher levels in previous months in July 1998. The big question will be whether the tropical SST is going to be higher at the peak of this El Nino. We won’t know that for months.
Re: bender (#534), But these are air temperatures. The only thing I can think of is increased solar magnetic activity. Like perhaps a severe geomagnetic storm brought on by it.
PS The trend of those SST’s are down not up.
Richard on the left click any of the years dmi site they ALL shoot up and down all the time that’s life
10,000 down for tonight’s update….
Re: markinaustin (#543), or 9,688 to be exact. But this is the first posting the update still to come. The ice is inching up. Nothing dramatic so far.
well….i would say it has been doing more than inching up…..considering how early we are in the season it has been performing quite splendidly don’t you think? i mean it has turned the corner and is climbing fairly quickly…..so far the “down” days have always been corrected up massively.
interesting take Andy…so what is relevant? if the low and the high isn’t relevant (and i am not sure you are saying that so feel free to clarify)….what should we be looking at?
Re: markinaustin (#547),
Sorry to confuse, I meant the rate of regrowth or melt, not the absolute values for maxima and minima.
Regards
Andy
New figure for 9/21 is 5,401,875 km2
Re: Michael Jennings (#548),
So adjusted figure for 9/20 was 5,383,594 so we are up 18,281 for 9/21?
Re: crosspatch (#551),
Yes, we went from over 9000 down to over 18000 up a difference of 28,000
Re: crosspatch (#551), Right
Re: DeWitt Payne (#552), Somewhere on this blog it was informed that the principal method for melting of the Arctic ice was transportation of the ice by wind and currents to the south where warmer waters would melt them.
What seems also very apparent, looking at the SST’s, is warm waters from the Atlantic and Pacific directly entering the Arctic waters and warming them, which seems to be happening west of Spitsbergen, Baffin Bay and past the Bering Straits. Although the Greenland sea ice maybe increasing this is probably also an area of high sea ice loss.
Re: Richard (#554),
I believe that’s the loss mechanism for multi-year ice. It takes on the order of a year to go from the Bering Sea side to the Fram strait. See the chart for the NP-36 experimental station for example. So seasonal ice mostly freezes and melts in place. Sunlight is a big factor. Average daily insolation at 75N and above is over 500 W/m2 near the summer solstice. Warm water from the Gulf Stream and the Japan current is also important. Air temperature is probably least important.
Richard:
There is always a branch of the Gulf Stream coming up the west side of Spitzbergen. This is quite normal. What is unusual for the last few years is that there is apparently more warm water than normal going up the west coast of Novaya Zemlya and north through the Bering Sound.
The uptick should really increase over the next couple of days. I’m expecting a couple 50,000-75,000km^2 gains starting around the 25th and continuing through the end of the month which should put us around 5,900,000 km^2 by the 31st. SSM/I is showing a fast concentration increase north of the Beaufort and conditions for ice growth in the Beaufort region is high. Northerly flow of cold, below normal temperatures are expected over the next 4-5 days helping to form new ice and push old ice into areas of open water before high pressure settles overhead to increase concentrations within the field.
On the Russian side of the Arctic ice growth will be slower with winds from the south expected during much of the next week.
Re: Tom Woods (#556),
While I agree with the substance of your comments, I believe the growth will be slower than you indicate. I see no more than 100,000km growth by this time next week. It may even be less than 50,000km total. To be sure, we both see growth in coverage, and in the same regions, but some erosion on the Russian side is already occurring and will continue to do so.
I have noticed that the average Arctic air temperature has taken quite a spike upwards over the past few days. Looking at past years, such spikes seem to be common so no reason for worry but the rapid increase in ice could slow down a bit.
Re: crosspatch (#558),
first of all, most areas above 80N are already ice, so I don’t believe it will have much, if any, effect on ice. Secondly, air temps are a minor player in ice formation. Wind and current have a higher impact. Third, even at these elevated temps, ice can easily form. Fourth, a defect in the algorithm is the most likely cause of the spike.
Re: Tucker (#559),
basis?
Re: bender (#561),
Bender,
I was referring to the possibility as discussed in posts 1563-1570 in the attached link.
http://www.easternuswx.com/bb/index.php?showtopic=168404&st=1560
Now that I think further about it, could it be caused by the release of heat due to ice formation that typically picks up this time of year??
Re: Tucker (#576),
I do not believe that the freezing would release enough heat to make a significant difference in the temperature. I would suggest the weather pattern was more significant.
Re: Tucker (#576),
“Now that I think further about it, could it be caused by the release of heat due to ice formation that typically picks up this time of year?”
Yes, I remember from last year that spiking temperatures counter-intuitively meant that there was a rapid refreeze. Freezing water releases heat in the atmosphere. Don’t ask me how, but that’s what I’ve learned last year and it explains those (increasing) spikes on the right side of the DMI temp-graph.
Let me show you how stupid I really am: Isn’t this what is meant by Polar amplification?
Re: bender (#561),
From the JAXA web page:
Re: Tucker (#559),
Please tell me how winds and currents form ice at this time of year with a greater impact than air temperature has. Winds and currents can play a significant role in breaking up, dispersing, and melting ice but I am not seeing how sea currents cause ice to form. I am under the impression that it is the cold air temperature that freezes the water turning it to ice and as the temperature declines, the ice thickens. When air temperatures warm, the ice thins.
But in this case I am operating on the logic that if the air is warmer at 80N, there is a high probability that it is warmer still South of there. I also didn’t say I thought we would see lower formation, but I did mess up and use the word “could” which I should have avoided. My sentiment was that I would not be particularly surprised if we get a couple of slower days.
Re: crosspatch (#562), and #559.
At this time of year radiational cooling causes the cooling air temperatures and freezing. There is a net loss with the lack of incoming solar radiation. Wind may slow formation of ice.
Re: crosspatch (#562),
I think what he should have said is that air temperatures there are already below zero and are unlikely to rise above that, so the net change from day to day is going to be related more to wind and currents than to air temperature fluctuations. Which is still questionable. If -3C, -5C, or even -10C are unseasonably warm for a give location and time then ice formation will be unseasonably slow. To determine whether it is so slow that ice formation is exceeded by ice melt requires a finer analysis than what we’re capable of in a text blog full of non-experts. An unprecedeted month of an unprecedented -3C would help make the point. Not sure that a week-long moderately warm anomaly will demonstrate much.
Re: crosspatch (#562),
Crosspatch,
I worded my post poorly. You are exactly correct. I meant to say that wind and currents play a large role in the retention (or lack thereof) of ice already developed. They play little if any role in ice formation. As mentioned in other posts, sst’s play the major role in formation.
What I was really getting at was ice buildup is occurring on the Alaska/Canada side due to SST’s, and ice is eroding on the Russian side due to wind/currents. That’s why I think we make minor gains up to 100,000km in the next week.
Re: Tucker (#574),
What I was really getting at was ice buildup is occurring on the Alaska/Canada side due to SST’s, and ice is eroding on the Russian side due to wind/currents. That’s why I think we make minor gains up to 100,000km in the next week.
I guess I should have bet quatloos on this prediction. Only a 7,000km gain total in the last three days. I don’t see much for today either. Might even be a small loss. Oh well, there’s always powerball for my retirement..
Re: crosspatch (#558), crosspatch the uptick is in the air temperatures which, as has been pointed out, are still well below freezing. The sea water temperatures are the ones that will cause the freezing and there seems to be plenty around at freezing point or below. The ice is at the moment inching up – I judge that by the curve which is pretty flat at the moment. (The 30% concentration iceseems to be going up steeper though than the 15% extent given by IJIS)
Re: DeWitt Payne (#560), Sunlight is a big factor. Average daily insolation at 75N and above is over 500 W/m2 near the summer solstice.
That didnt seem correct to me, about insolation, so I calculated it and you are right. At 75N it is about 500 W/m^2 and at 90 about 528 at aphelion, which is the 5th of July
Re: Richard (#565),
“as has been pointed out, are still well below freezing”
I am aware of that. That didn’t have anything to do with what I was saying. I didn’t say any ice would be melting. I said that if air temperatures have risen generally in the Arctic, we could see a reduction in the rate of freeze. Or, rather, if we see such a reduction, it wouldn’t be surprising. The focus was on rate of freeze, not any melting.
“The sea water temperatures are the ones that will cause the freezing and there seems to be plenty around at freezing point or below.”
Right, but it isn’t the sea water chilling the air, it is the air chilling the water. The surface of the water cools to below freezing in response to air temperature well below freezing. If the air temperatures rise, the water temperature will too, in a general sense. Say I have ice at a given thickness at -30C. Now I increase the temperature to -15C. The ice will begin to thin even though the air temperature is still below freezing.
What I was actually thinking of, though, is a notion that if air temperatures are at an average of -15C above 80N, it would not be surprising to me if the average below 80N is higher than that. And if average temperatures above 80N rise to -10C, it would also be unsurprising to see the average below 80N to also rise a bit. I don’t mean to imply that it absolutely would, just that it wouldn’t be surprising if it did.
I wasn’t making a prediction. Simply pointing out that the rapid freeze might return to more “normal” rates. Not that it would stop or reverse.
And Re: “Sunlight is a big factor.” at this time of year, I would consider solar isolation to be playing little if any role in melting ice at high latitudes.
Re: crosspatch (#567), Right, but it isn’t the sea water chilling the air, it is the air chilling the water. The surface of the water cools to below freezing in response to air temperature well below freezing.
I may be wrong but I doubt that very much. The specific heat of air is way smaller than water. The chilling takes place because of lack of heat. The temperatures of the air and water drop in the arctic because there is very little insolation – heating by the sun, as the year progresses. By Nov 5 for example it becomes 0, 75N and above.
The air moves around mixes with other warm and cold air and this possibly explains the big rises and falls.
And Re: “Sunlight is a big factor.” at this time of year, I would consider solar isolation to be playing little if any role in melting ice at high latitudes.
That depends on the time of the year. In summer the sun is over the horizon throughout the day and the average insolation is higher at the North pole than at 75 N. I think it is the lack of solar insolation that is the factor in the freezing of the Arctic later in the year.
Re: Richard (#568),
“I may be wrong but I doubt that very much. ”
I may be as well but for example, look at Greenland. Westerlies off of Greenland freeze the water off the East coast while at the same latitude on the West coast the ocean is clear until much later. The air coming off land mass is very cold. If it was simply due to lack of sun, the ice would be the same each year. Imagine you have a foot of ice. The ice touching the water on the underside is right at the freezing point. The ice at the top is, say -15C. Now lets drop the air temperature 5 degrees. The ice will thicken. Not because the water temperature changed, but because the air temperature changed causing the ice to get colder. If I have a lake in Minnesota and the air temperature is 5C, it doesn’t freeze. If the air temp is -5 it does. Same amount of sun in both cases. Drop the temperature to -10C, the ice thickens. Raise it to -5, it thins.
The ocean exchanges water. It isn’t going to get much below freezing. If it was lack of sunlight that caused ice, the abyssal plains would be well below freezing.
Imagine that there was no land mass around the Arctic. Imagine the entire Northern Hemisphere was water. There would be no ice cap. You would end up with water sinking and the creation of a polar gyre. Any ice that might form would be blown away by the winds. You MIGHT get a rotating island of ice right on top of the gyre if it was extremely cold but as soon as a strong wind came up or as soon as the circumpolar jet broke down in spring it would blow away. There would certainly be no year-round polar ice cap.
Re: crosspatch (#569), crosspatch I think there is ice off east Greenland because of the east Greenland current that brings cold water and ice from the arctic ocean along the east Greenland coast. It is not because of the cold air off Greenland. If you look at the geography of Greenland the Arctic ocean is all but cut off on the west coast compared to the east coast.
Someone more knowledgeable than us could clarify. But thats how it looks to me.
Re: crosspatch (#569),
Was not the conjectured snowball earth formed when a super one-continent (Rodinia) earth existed? The current Arctic and surrounding area was believed to be all water.
http://en.wikipedia.org/wiki/Rodinia
I believe that explaining snowball earth, which remains controversial, would have been less difficult if the Rodinia were located pole ward instead of the tropics where it was indicated to have formed. I believe I have heard conjectures on the History Channel that the poles froze because of changes in warm water transport to those regions. Snowball earth has detractors because of the difficulty of finding a mechanism for unfreezing the snowball. The History Channel, as I recall, showed volcanic activity, caused by the break up of Rodinia, doing the unfreezing.
http://www.wordiq.com/definition/Snowball_Earth
Re: Richard (#568),
More to the point it’s the radiational cooling of ~300W/m^2 (at around freezing) which occurs continuously in the absence of sunlight. Once the ice forms the water below the ice can’t lose heat to the surface very rapidly because of the slow conduction through the ice and the surface of the ice continues to cool very rapidly due to radiation, the ice thickness the reaches the ‘thermodynamic limit’.
Re: Phil. (#575), Yep thats what I meant. Radiation heat loss and no gain due to no sunlight. Not sure what you mean by “the thermodynamic limit”. The conduction of heat will be governed by Fourier’s law and doing some quick calculations the thickness will increase, in the absence of incoming radiation, with the square root of time. In other words given enough time the water will freeze solid from top to bottom. Which makes sense, otherwise your freezer water never would. That the Arctic ice doesnt I should imagine is due to the fact that it is not given enough time, and the movement of the water also comes into play.
Re: crosspatch (#567), I said that if air temperatures have risen generally in the Arctic, we could see a reduction in the rate of freeze. Or, rather, if we see such a reduction, it wouldn’t be surprising. The focus was on rate of freeze, not any melting.
Well your hypothesis has been proven wrong with the evidence. While the air temperature shot up, the ice extent increase has increased from 18,281 sq kms to 30,938 sq kms. Thats a 69% increase.
Re: Richard (#582),
I wasn’t “hypothesizing”. I wasn’t predicting. Just trying to put some caution out there. Freeze going faster than “average” and temperatures well below “average”. Now temperatures are well above average and still rising. Again, under those conditions I wouldn’t be surprised if the freeze rate slowed down somewhat. That is not to say “I predict the freeze rate will slow down”. It is saying that if it does slow down a bit, that would fit with current conditions in the area.
The warm air that is causing the average to go up could be very local in only a specific area. I don’t have access to data with enough resolution to tell that.
Re: Richard (#582),
If you assume that the water at the base of the ice is -2ºC then the thicker the ice the colder must be the ice surface temperature. There’s a limit to how cold the surface can get so under those circumstances the ice stops getting thicker, this is referred to by those who work in the area as the ‘thermodynamic Limit’.
Re: Phil. (#589),
Does this imply that ice after this point is by accumulated precipitation and would rough seas breaking up ice allow it to thicken by new ice being created then compacted? If I understand what you said it would seem that any thickening beyond the limit you mention must be by forces that have nothing to do with sea temps.
Any data on where the daily or monthly freeze line is over the number of years where discussing?
Re: BarryW (#590),
Exactly, it could pile up etc. but just by cooling it will slow right down to a limit.
In the Arctic it’s about 3m and takes a few years to reach.
Re: Phil. (#593), This must be due to special conditions in the arctic where the waters move and mix around a lot. I just read that the Antarctic sea ice can be upto a 1,000 metres thick.
Re: Richard (#595), A cite would be interesting Richard. Are you sure they were refering to sea ice?
Re: Richard (#595),
No the thermodynamic equilibrium thickness in the Antarctic is thinner (~1m), it nearly all melts every year, also has more snow which is a better insulator. Ice sheets attached to the land are a different issue.
Re: Phil. (#589),
There’s a limit to how cold the surface can get
What is that limit? If theoretically the surface of the ice kept radiating energy at around 315 W/m^2, given by the Stefan-Boltzmann equation, then the surface should eventually fall to close to absolute 0, in the absence of any other influences.
And the way I would look at it would be – the surface would get cold, the bottom of the ice would get colder due to conduction and the process would continue so long as there would be a temperature differential between the surface and the bottom. The more the differential the more the conduction.
I can understand that the rate of freeze would slow down with the thickness but why should it stop altogether? I’m sure it does but dont understand exactly why.
If one took a vacuum thermos flask with the sides insulated filled it with water and placed it in a freezer, with the top open, would it not freeze down to the bottom?
Re: Richard (#591),
A couple of points: firstly you wouldn’t go down to absolute zero because the atmosphere would be radiating down (the greenhouse effect). Last year (April) the NP weather station measured about 175W/m^2 of downward longwave when surface temperatures were -27ºC. Secondly as the surface temperature drops the radiational heat transfer drops by S-B ( @250K, ~220W/m^2; @230K, ~160W/m^2).
Nope the conductional cooling rate depends on ΔT/Δx. The thermal conductivity of ice at -30ºC is 2.5W/m.K so with a ΔT of 35K and a surface loss rate of 175W/m^2 would be balanced at a thickness of 0.5m.
Not a good analogy, you’d need a ~100m deep Dewar, in which case it wouldn’t freeze all the way down (the liquid shouldn’t be in a Dewar it should be in contact with an effectively infinite volume of water at about 2ºC)
Re: Phil. (#598),
What, you mean Richard’s one-paragraph refutations aren’t rock solid?
And what’s this “greenhouse” thing? The topic here is sea ice *recovery*.
Re: bender (#600),
Hi bender 😉
Re: Phil. (#598), I see. Well thanks for that.
Re: Kevin Johnstone (#597), Ice shelves. They are sea ice in the sense they are floating on the sea. But perhaps formed from snow accumulation.
Re: Richard (#603),
IceSheets Vast shields of thick continental ice that have formed through the accumulation of snow over millions of years. Ice sheets are frequently domed shaped and gradually sloped.
Ice Shelves Large continentally based ice sheets which have flowed to the coast where they then float in the ocean.
Don’t you read the stuff you link?
Re: Richard (#604),
No, not on this planet, not even true of Europa.
Re: Phil. (#605),
Ha ha 🙂
The quality of the thread is improving steadily.
Re: Phil. (#605), I was thinking of Europa. Several kms thick and possibly only liquid because of the tidal forces caused by Jupiter.
Re: crosspatch (#588), Well it seems to have slowed down ore reversed today. Loss of 23,125 km^2 so far
Re: Phil. (#605),
That shelf-ice definition isn’t quite comprehensive. The ice-shelf on the north coast of Ellesmere land (which has mostly disappeared during the last century) has not flowed off land-based glaciers. It built up from sea-ice that remained in place for a very long period and successively grew thicker, mostly from snowfall that did not melt fully in summer.
This type of “paleocrystic” ice could once also be found in some northern Greenland fjords, and there is (or was) even a special word for it in Greenland Inuit.
Re: tty (#618),
It was the one that Richard cited, in the Antarctic which he was referring to it’s the predominant source, basically it’s not due to in situ cooling.
Re: Richard (#603), Thanks for the link Richard. The cited definition for sea ice would seem to exclude the ice shelves.
Re: Phil. (#598), PS would you agree that if there were no atmosphere, no greenhouse effect, then the ice thickness would continue to increase?
Re 568
Hanafin JA, Minnett PJ. 2005 gives a hint of the problems of remotely sensing sea surface temperatures but lends some confidence that they may have nailed the parameters over limited viewing angles. They mention the difficulties of sensing in ‘non-calm conditions’, which I would think means most of the time. Natural organic films also alter emissivity, but the reference I found is old, so I don’t know if satellite sensors allow for that weak effect.
This stuff — like any aspect of climate science — is difficult, and doing it under pressure, knowing the world is watching, must be nerve-wracking. Perhaps the occasional shirty response to questioning is understandable.
JF
“As the only major southward flowing current in the Greenland Sea, the EGC transports recirculating Atlantic Water, Arctic Ocean water masses, and >90% of the ice exported from the Arctic Ocean (Woodgate et al. 1999, Rudels et al. 1999).”
I had mentioned above that I thought the Arctic must be losing ice from here. Richard (#554),
PS Re: crosspatch (#567), If it was lack of sunlight that caused ice, the abyssal plains would be well below freezing.
The temperatures of the oceans > 1,000 m deep are between 4 to 0 C. But they dont go below freezing probably because of exchange of heat from the higher layers and the pressure.
Lets hope that the “black borders” which appear on clicking NH ice sheets at cryosphere today are not the “average” used to calculate current ice extent. If so all data would be invalid the areas invariable completely encroach on land and would give the impression that they are purposely made larger to exagerate lower ice extents today. Just checking….
http://arctic.atmos.uiuc.edu/cryosphere/
Also note that link to article questioning “current global ice normal” has been removed.
The total (pre adjustment) of ice for 09/22 is 5,406,563 km2 for a gain of 5,000 from the 21st.
Updated results are in.
Today’s gain comes in at +30,962km^2
Total ice extent is 5,432,813km^2.
Tucker,
The average extent gain from 9/22-30 in the JAXA series for the years 2002-08 is 366,851km^2.
Here’s the individual years:
2002 – 458,282km^2
2003 – 300,625km^2
2004 – 514,521km^2
2005 – 436,875km^2
2006 – 130,313km^2
2007 – 316,250km^2
2008 – 411,094km^2
Assuming the gain in ice through the end of the month this season is similar to other years in the JAXA series the total ice extent at month’s end would be 5,800,000km^2. However, I’m basing my forecast of faster ice growth on the fact that SST’s in the Beaufort are low enough for ice formation, temperatures are forecast to be in the -5 to -10°C range over the next week and there’s plenty of ice edge to expand from (orientation of the pack ice). Plus open water is far to the north in this region of the Arctic, in an area losing more insolation than areas say, on the Russian side, where the ice edge is further south.
Ice formation will also be occurring in the channels of the Canadian Archipelago over the next week.
To judge from the Canadian Ice Service map (http://ice-glaces.ec.gc.ca/Ice_Can/CMMBCTCA.gif) the Northwest Passage sailing season is over for 2009. The northern deepwater passage never did open and the southern branch is now blocked by heavy ice at the northern end of Peel Sound.
go for it richard! Totally agree.
Bickering. crosspatch, like any normal person, meant to preface with the ubiquitous “all things being equal”, which, of course, was not the case, and never is in the real world. Stop arguing at cross purposes. If you can lay down evidence showing the relative effects of air temp vs. SST, let’s see it. Oh, yeah, it’s called a “computer model”, which does not communicate well in a one paragraph strawman refutation. They have jobs for guys like you … at IPCC and Sierra Club.
Re: bender (#585), I do admit to crosspatch that according to Fouriers equation the difference of temperatures between the ice and air is directly proportional to the conduction of heat through the ice, so would have a bearing on the ice thickness. But this may not have a bearing on the ice extent as the sea water would freeze when its temperature falls below its freezing point.
My god these threads get huge.
I’ve done some plots of Arctic sea ice from the gridded data again. This tiem the algorithm is improved and I’ve done some unusual ratio plots between extent and area.
Re: Jeff Id (#592), Interesting.
In the interesting but not particularly useful information department, the (positive) seven day simple average for 2009 is at least twice as large as any in 2005-2008
2009 18,817
2008 3,795
2007 1,272
2006 8,192
2005 -25,335
well….it went down 23,000 today, correction should be interesting.
yesterday ended up at 5,417,031 km2. A loss of ~15,000km
Re: Tucker (#609),
Not an unexpected result really. The south winds over Russia have been eroding/compacting the ice pack there for three days now, and the Scandinavian side isn’t much better. Until that lessens, expect more of these up and down results. By this weekend, I would expect the winds and erosion to abate and allow for a more robust gain. It just goes to show how important seasonal wind anomolies are in the sea ice extent arena. It can trump sst’s and insolation in short time scales, and possibly even seasonally. Many have suggested 2007 was caused by wind anomolies, and I concur. In fact, one may hypothesize that PDO and AMO signal combinations can and do lead to small multi-decadel jet stream fluctuations and associated decadel prevalent winds pattern changes. This in turn is what causes arctic ice extent patterns in the historical records going back to the 1800’s and longer. At least that’s my theory.
Re: Tucker (#610),
Sorry, the theory has already been elucidated, and even proven. Polar jet stream dynamics are modulated by the circumpolar vortex whose behavior (strength and position) is described by the QBO. None of these modes is independent. Post-hoc eigenanalysis just makes it look that way.
Re: bender (#612),
Thanks Bender. I thought I recalled that work, but when I went searching for it, I couldn’t find it. My computer died a violent viral death recently, and many of these papers/abstracts are lost to me. Anyway, since I couldn’t find it, I called it a hypothesis instead. I appreciate the notation.
Northern Hemisphere circumpolar vortex trends and climate change implications
Oliver W. Frauenfeld, Robert E. Davis
Department of Environmental Sciences, University of Virginia, Charlottesville, Virginia, USA
Trends in the Northern Hemisphere circumpolar vortex at 700, 500, and 300 hPa are examined to assess the relationship between circulation variability and air temperature. A vortex climatology is developed for the period 1949–2000. At each pressure level, three geopotential height contours are used to quantify the size and position of the vortex at 5° longitude resolution within and both north and south of the primary hemispheric baroclinic zone. This combination of spatial specificity and the long temporal record makes this the most comprehensive vortex climatology to date. The overall and seasonal vortex time series for the Northern Hemisphere are created for northern, middle, and southern contours at each of the three levels in the atmosphere. From the beginning of the record until 1970, the vortex exhibits a statistically significant expansion, but the vortex has been contracting significantly since then at all levels. The pre-1970 expansion and subsequent contraction is strongest in the lower latitudes and weakest in the higher latitudes. The trends are also stronger in the upper troposphere than in the lower troposphere. Spatial examination of the vortex indicates that the pre-1970 expansion and post-1970 contraction were driven primarily by expansion/contraction over Asia, Europe, and North America with little change over the Northern Hemisphere oceans. Although significant climate change debate focuses on the discrepancy between positive trends in surface air temperature and little or no trends in Microwave Sounding Unit (MSU) satellite temperatures, contraction of the circumpolar vortex at every level of the atmosphere implies that the atmosphere is warming at depth since 1970. Comparisons with the MSU temperature history indicate that the Northern Hemisphere circulation as a whole, as represented by the circumpolar vortex, accounts for almost two thirds of the interannual variability in midlatitude MSU temperature, indicating that vortex size and position are coupled strongly to atmospheric temperature and could be a good indicator of climate change. On a latitude-by-latitude and level-by-level basis, the lower latitudes are associated most strongly with MSU temperature in the midtroposphere while the middle and higher latitudes are more closely associated with MSU temperature in the upper troposphere. The vortex trends are also similar to observed surface warming trends.
Temperature-related trends in the vertical position of the summer upper tropospheric surface of maximum wind over the Northern Hemisphere
Courtenay Strong, Robert E. Davis
Abstract
The surface of maximum wind (SMW) is used to examine spatial and temporal variability in the vertical position of jet streams and fast upper tropospheric wind maxima over the Northern Hemisphere (NH) in the NCEP/NCAR Reanalysis for summers 1958-2004. At a given observation time in a gridded data set, the SMW is defined as the surface passing through the fastest analyzed wind above each grid node, with a vertical search domain restricted to the upper troposphere and any tropospheric jet streams extending into the lower stratosphere. The 47-year mean summer SMW generally resides below the tropopause, undulates in the tropics, and descends poleward in middle and high latitudes. Trends in the pressure of the summer SMW are primarily positive, exceed 30 hPa/decade at some locations, and are found over most of the tropics and subtropics for the period 1958-2004. The thermal wind relationship is used to establish links between the SMW pressure trends and temperature gradient changes in the upper troposphere. The changing temperature gradients are consistent with nonuniform tropospheric warming and are correlated with sea surface temperature (SST) variability related to El Niño and the Pacific Decadal Oscillation.
Relation of size and displacement of the 300 mbar north circumpolar vortex to QBO, El Nino, and sunspot number, 1963–2000
James K. Angell
Abstract
The size of the 300 mbar north circumpolar vortex, and its eastern, western, date line, and Greenwich hemisphere components, is estimated for the period 1963–2000 by planimetering the area poleward of 300 mbar contours in the main belt of westerlies on the mean-monthly polar stereographic analyses of the Institute of Meteorology of the Free University of Berlin. On the basis of the superposed epoch method, there is little evidence of a relation between vortex size and phase of the QBO, but significant at the 90% level or better is the tendency for the vortex to be less displaced into the eastern hemisphere in the east-wind phase of the QBO, for the vortex to be expanded near the time of Nino 3 sea surface temperature maximum (El Nino) but contracted 3–4 seasons after, and for the vortex to be displaced farther into the date line hemisphere when there is an El Nino. There is also an impressive tendency for the winter vortex to be less displaced into the eastern hemisphere at the time of El Nino. The tendency for the vortex to be contracted near sunspot maximum, and expanded near sunspot minimum, is significant at only about the 80% level because of the small sample size.
Isn’t planet earth fascinating? What a place!
To further illustrate bender’s point that we’re still a ways from a recovery or even a return to the trend I offer this graph. It’s the September average Arctic ice extent from 1979-2009. The trend line is calculated from 1979-2006 data. The axis of the JAXA data is adjusted to overlap the NSIDC data for 2002-2006. The 2009 September average probably won’t change much in the last few days of the month. Note that 2009 is still well below the trend line. A trend line calculated from 1979 to 2009 will have a more negative slope than a trend line calculated from 1979 to 2008. The September average will have to go above the trend line and stay there before one can conclude that the Arctic ice is recovering from the long term down trend.
Re: DeWitt Payne (#621),
Can you do us a favor and run the comparable analysis without 2009? I want to know by exactly how much the trend stats STRENGTHEN (as I predicted before last year was even finished) when you add the 2009 data point.
Re: bender (#622),
The slope was originally only through 2006. I discovered you can extend the trend line by including empty values for following years. Anyway, when you add 2007, the slope drops from -55157 to -66711 km2/year, add 2008 and there is a further decrease to -72458 km2/year, add 2009 (month to date) and the slope decreases to -73381 km2/year. The 2009 point is still below the recalculated trend line, but just barely. Almost any increase in 2010 will cause the trend to become less negative when included in the calculation.
Re: DeWitt Payne (#659), Can you explain why a linear trend would be expected to have any physical meaning over this time frame? I believe that you know enough statistics to know better …
Re: DeWitt Payne (#621),
Try trending a couple of other things:
1. Annual mean ice extent Jan 1 – Dec 31
2. Annual maximum ice extent
Re: crosspatch (#624),
If 2009 stays even with or ahead of 2008, then the trend of the annual average will become a little less negative. The negative trend deviation for the annual average started with 2005, though, so it will still be a while before the trend could return to or exceed the 1979-2004 trend. I don’t think the maximum extent tells you all that much.
I will bet you that whatever happened to drop extent so low in 2007 (south winds, etc) also happened in 1990. A couple of fairly deep anomalies around an otherwise smooth declining trend. i.e. NO RECOVERY. JUST NOISE.
Thanks as well for posting those abstracts Bender. I have located the full versions, and will devour them tonight!
Re: INGSOC (#625),
link??
snip – pontificating
Sorry…
http://www.agu.org/pubs/crossref/2003/2002JD002958.shtml
http://cat.inist.fr/?aModele=afficheN&cpsidt=18303901
http://www.agu.org/pubs/crossref/2001/2001JD000473.shtml
In the order posted.
Hope they work.
27,031 Loss. Methinks there is a loss of ice off the east Greenland coast. You can see that ice flowing southwards. I also think there will be a largish addition in the update.
Re: Richard (#629),
So the loss is not due to melting?
Re: Gerald Machnee (#630), Hardly likely with temperatures of the air well below 0 and those of the sea dropping.
Re: Richard (#631),
Depends on how you define “well below”. Sea water freezes at somewhere around -2C depending on salinity. Eyeballing the graph, it looks like the “average” temperature above 80N is something close to -5. Much (all?) above 80N is probably already ice. The question is “what is the average temperature between 70N and 80N” and I just don’t know. A reasonable expectation would be that it would more likely be above -5 than below -5 if one would think that temperatures generally rise as one moves South.
So I don’t have anything that tells me that temperatures are far enough below 0C at the current freeze line to freeze a lot of sea water and I do have an indication that tells me that chances are better than 50-50 that temperatures at the freeze line might be warmer now than they were a few days ago which could cause thin skim ice to melt.
I admit all of this is probability as I figure it from incomplete data available to me. But it looks like the temperature spike has peaked and could be headed back down now. Such spikes appear frequently in the record and are not abnormal or any indication that we are going to boil ourselves alive. It’s weather.
Re: crosspatch (#634), crosspatch this is how I reason it out – the air temps today are what they were about 10 days ago. 10 days ago the sea had started freezing, so thats likely to continue. During this time insolation has decreased in the area. There is probably a net radiative loss because of this, in which case the temps at the surface of the newly formed ice will be going down. The SST’s show more area which are cooler today than 10 days ago, with temps close to 0 or below. So I very much doubt if there is any appreciable melt in the Arctic basin. There is rapid southward movement of the ice along East Greenland, this would cause ice loss as it moves south. I’m betting on the balance the ice will start to increase about now or in a couple of days. But as you say its the weather and the weather is notoriously unpredictable. There has been snow here on the Rimutaka Hill outside of Wellington and the Manawatu Gorge, very unusual for spring. There maybe unusual wind movements which might change the scenario.
I still see no sign of the predicted early freeze season as predicted in some quarters. Antarctica area is a lot higher than last year at this point, will be interesting to watch over the next month.
Regards
Andy
Re: AndyW (#632), Pretty flat so far but large areas of the sea are now showing pretty cool so the freeze could increase over the next week.
Check this and this for a relationship. If the relationship exists, and if the AO forecast is correct, then Arctic temperature should decline sharply soon.
Re: David Smith (#636), Yes thanks for that. And for those like me who dont / didnt know what Artic Oscillation is – no need to get bewildered by the big names climate scientists use – its the high pressure over the Arctic vs the low at 45N and vice versa. When its positive like now the Arctic has low pressure and the warm air rushes in from the south warming things up. And when its negative the air current reverses and “Midwestern United States and western Europe” gets cold, which is going to happen pretty soon if that link is to be believed. Since western and midwestern Canada lies in path of Midwestern United States, it too, though inconsequential, will get cold by the 10th of October.
Re: Manfred (#651), yes I agree “pause”, “recovery”, “Reversal” – avoid. Use increase instead – totally undeniable.
Loss 24th 8,750 km^2 after update.
Re: Richard (#654), oops that should be eastern and mideastern, or midwestern, (middle) Canada. Get mixed up with NH geography
From the above graph (produced using R’s seasonal analysis stl function) the anomaly is negative once again i.e. larger melt relative to long term trend.
Hardly “recovery”. However, the anomaly is not exceptionally large relative to trend, as it was in 2007 and 2008.
If arctic sea ice is in long-term linear decline, then it seems fair to say that 2009 looks like an unremarkable year?
Re: joew (#637),
When you draw straight lines you will show decline for many more years. If you break it up and use the last 3 years you see an upward trend. Yes, I know 3 years is not significant—
Re: Gerald Machnee (#639),
So then why the constant claptrap here about a “reversal”?
Re: bender (#640),
Yes, I knew I was going to get a response same as last year. However, as long as linear trends are drawn with 1979 as a starting point, I will keep commenting. Can we back up from 1979 to totally ice covered? Similarly I do not accept that 2008 and 2009 are “temporary cooling” in a progression to an ice free situation. History started before 1979. What we need to do more work on what causes the variation in ice cover as some have commented on in the last few days.
Re # 642: I agree. I did not say it was significant. If 2007 should be the lowest extent in a number of years, it will be worth noting along with why it was so.
Re: Gerald Machnee (#667),
Can we back up from 1979? Not really. 1979 is the start of the satellite data era. Numbers before then are not a whole lot better than guesses. I’d very much like to see high quality data on Arctic ice extent and area from 1910 to 1950, but it doesn’t exist. We use what we have.
Re: Gerald Machnee (#639),
And when you draw curved lines you run the risk of overfitting. You can’t suck and blow on these matters of curve-fitting.
Re: Gerald Machnee (#639),
Yes, 3 years of an upward trend is not “significant” — but noteworthy when we are dealing with known oscillations. It would be wise to stay away from the term “significant” because of statistical connotations, but the fascination with a linear trend seems odd to me given what we know of climate and weather. For the next 30 days, the daily mean temperature in Chicago will reinforce a linear warming trend which we have observed for the last 270 days. However, that does not mean that the linear warming trend is a reliable indicator of what is going on.
Re: joew (#637), We could also avoid using words like “If arctic sea ice is in long-term LINEAR decline, then..”. We can see if that If is correct after another 30 years.
From Jaxa website:
The latest value : 5,390,000 km2 (September 24, 2009)
It seems winds are packaging the ice.
Also, nice round number. I guess significant figures are correct for once. I never liked those so many non zero cyphers on the ice extent figures, it gives a false feeling of accuracy.
New total is out 5,408,281 km2
so down 9,000 yesterday after correction.
An Inquirer makes an excellent point. People pretending that 3 years of fairly radical recovery aren’t at all important are sort of kidding themselves. we all know that the decline could continue this year, but to act as if 1,000,000 square kilometers of recovery in 2 seasons means nothing is a bit odd. next year will be a very important year if the recovery brings us over some of the past 8 years besides 2008 and 2007.
Another way to look at the past 3 years is what would people be saying if the 2009 minimum was lower than 2008, or worse, lower than 2007. Everyone knows it would be headline news and the AGW drums would be beating. So while the increase of 2008 over 2007 and 2009 over 2009 doesn’t constitute a “recovery” of Arctic ice, it is not insignificant. If 2010 continues the “trend”, it will be another step towards Arctic ice recovering, and another datapoint that calls AGW predictions of an ice-free Arctic into question.
I don’t think recovery is the proper word to use as of this time. Maybe after 4-5 more years of increase we could categorize it as such. The term which might be more accurate is a “pause in an underlying declining trend” until we have more history to go on.
Re: Michael Jennings (#646),
This years minimum is still below the trend line even if it’s part of a recovery. If it continues to rise above the trend line as you suggest then I agree that recovery is a realistic term to use. On the other hand alarmists shouldn’t be using the last few years to imply that the trend is accellerating.
Re: BarryW (#647), Re: Michael Jennings (#646), Your “declining trend” is based on an artificial time frame. I do not consider this to be an accurate description at all. If you are discussing climatological time frames, you have 1 data point. I defy you to create a trend with one point. If you are talking shorter time periods, then you are assuming that there is some kind of linear system response with noise about the trend. This is false on a number of grounds. This “model” ignores autocorrelation and persistence. It ignores the fact that there was a distinct change in the system behavior around 2001 (more or less).
By tying your linear trend to 1979, you are artificially decreasing the slope during the recent decline in ice extend minimums and artificially placing this year’s minimum below the “trend”. If you compare this year’s value to the slope from 2001, then this year is “above the trend line” and the word recovery is justified under any definition of the word. How do you justify creating a linear trend over a period that has no physical basis or justification and over a period where the pattern in the data is clearly NOT linear???
Re: Michael Jennings (#646),
I am sorry, Michael Jennings, but I do not believe that it is a leap of unwarranted faith to say we have had a “pause in an underlying declining trend.” That is presuming that there is an ongoing declining trend, and that might be an questionable presumption. I do not know in what direction the minimum will go next year; I would not be surprised by either direction. I expect that oscillations would tend favor an increase, but that tendency could easily be overwhelmed by Asian soot and weather events involving clouds and winds.
Michael Jennings: September 25th, 2009 at 11:07 am
“pause in an underlying declining trend”
is wrong in two ways:
firstly, it hides the information, that ice extend increased significantly for over 2 years now (“pause” implies that it did not change at all ).
secondly, pause implies, that after a while, the (infamous) linear trend will continue, while actually nobody knows, if it does or a new trend just started.
In summary, that wording would be just deceptive.
Re: Manfred (#649),
Manfred, the graphs show the declining trend has been undeniable from 1979 to present and that is what the pause is referring to, the trend. It does not mean that there was no change from the last 2 years, merely that the steady decline over the time period has paused (and hopefully stopped) because it is after all still below the trend line for the period even with the large increases since 2007. After several years of a positive trend (above the anomaly), it would be fair to call it a recovery.
I keep seeing the charts showing “the average”, and comparing the last few years to that. Has anyone done a “spagetti chart” showing all the years, to see if there are any points that are close to the current year?
So far, all the comparisons I see are tied to about 2000-present.
Or, on the other hand, has anyone redone the averages to see where the new avg lays?
instead using the biased words pause or recovery, it may be just called an “increase”, or if you like, the biggest increase measured in human history 😉
This probably off topic, but regarding the Arctic sea ice, I’m trying
to figure out if something weird happened to either the ice the or
satellite between 1986 and 1987.
I started playing with
ratio=extent/area
for the months of March and September between the years 1979 and 2008
using the time series data from NSIDC.
And I noticed there’s a huge downward step in the ratio
(relative to the variance) between 1986 and 1987.
In fact, it appears to partition the ratio into 2 disjointed sets
– one set from 1979 to 1986 and one set from 1987 to 2008.
And the drop in the ratio for 2007 is not even note worthy in
comparison to the drop between 1986 and 1987.
Any ideas what might be going on?
— Cinaed
Re: Cinaed Simson (#657), I downloaded the GSFC extent data this morning 1978 – 2007 from the “NASA team” and noticed that the data for Dec 87 was missing. Would that account for it?
Re: Cinaed Simson (#657),
If I recall correctly that was when they changed satellites and the polar hole became smaller.
Pre adjustment figure for 09/25 is 5,409,688km2 for a gain of a paltry 1400 (which I expect to be increased substantially by the adjusted total)
Adjusted total for 09/25 is 5,439,688 km2 for a total increase of 31,000 from the 24th.
Re: Michael Jennings (#665), 31,407 increase. Is the DMI site down?
Re: Richard (#666),
Yeah, looks like DMI is broken. There is a web server there but it can’t seem to find even the top level index page. Looks like a sysadmin has some work to do.
Re: crosspatch (#668), It is seems very unusual for it to be down for so long. Do you think that DMI have deliberately taken off their ocean data site, perhaps because they were unhappy with displaying free content and data?
Re: Richard (#669),
I am not going to ascribe motive. What has been the most likely cause of things like this in my experience are things like the document root being a remote mount from another system that has failed or a software upgrade that has resulted in the user that the web server runs as not having read permission of the document root directory.
Re: crosspatch (#673), It is quite frustrating though not to be able to see their data and see how well extent increase correlates with air temps. Besides the fact that it provided an independent source of sea ice data to AMSR-E and NASA.
I would be particularly interested in seeing the correlation with the Arctic Oscillation with air temps and sea ice. If I read those graphs correctly the AO is much higher than that forecast. It is due to fall soon though. I wonder if it will.
Re: Richard (#674),
I would expect there to be any significant correlation between extent and air temperature only during the initial freeze and late in the thaw. In the dead of winter, temperatures are so low that air temperature influences thickness not presence of the ice except at the margins of extent. What would be interesting to have is a 70-80N mean in addition to the 80+N mean. That would give a better sense of what is going on at the margins if the ice extent.
Re: crosspatch (#679), Well the DMI site is back up. You were right about not assigning motive, specially a suspicious one, to the site being down. Looks like it was a technical problem as it still does not seem to be functioning at 100%.
The extent is creeping up – still quite flat. But the Arctic Oscillation seems to be headed down. The air temps at the DMI site dont seem to be updated, though probably headed down and the SST’s dont work at all.
All points to the extent picking up pace and a colder October for the NH
Re: crosspatch (#679),
One thing to keep in mind is that the work schedules of various places don’t always correspond with my browsing schedule 🙂 So if something breaks in Europe over the weekend, it might be some time on Monday before someone gets a chance to have a look at it.
Re: Richard (#674), Richard,
You might be interested in the work of the late John Daly. He was concerned that adjustments to temperature readings and the algorithms to produce an average temperature were misleading. Therefore, he assembled information on actual readings of temperatures, including the Arctic. For example, see:
http://www.john-daly.com/stations/stations.htm#Greenland,%20Iceland,%20northern%20Norway,%20and%20the%20Arctic%20Ocean
and
http://www.john-daly.com/stations/stations.htm#Canadian%20Arctic
As far as I know, his temperature work was not continued after his death.
5,452,031, pre-correction 9-27-09
2009 is virtually mirroring 2005 on the AMSR-E extent graph.
Adjusted total for 09/26 is 5,480,781 km2 which gives the date a gain of ~41,000
Cryosphere has backed their Antarctic ice curve to mid-August, I can feel an adjustment coming….
Re: tty (#676),
There’s an adjustment coming at Cryosphere Today, but it’s to the baseline. They are planning on using 1979-2008 instead of 1979-2000. That will change the value of the anomaly, but shouldn’t change the area. I have been collecting data from CT since July, 2008 so if they do change, I’ll know.
Re: DeWitt Payne (#678),
Are they going to change the Artic baseline as well?
Re: UK John (#684),
As far as I know they will change the baseline for both Antarctic and Arctic. It will make less difference for the Antarctic. In a July communication Bill Chapman thought it would happen any day. I haven’t heard whether they have shelved the project or not.
Total for 09/27 (pre-adjustment) is 5,498,438 km2
Has the cryosphere adjustment occurred?
The Arctic anomaly is heading back towards 2% under at the moment, but did they make the change to the baseline as suggested above?
The AMRSE extent indicates the Arctic has not really got into gear yet on the climb to the maxima.
Regards
Andy
Re: AndyW (#685),
I assume you mean 2Mm^2? On the website it still says that the anomaly is wrt 1979-2000.
Preliminary total for 09/29 5,553,594 km2
Adjusted total for 09/29 is 5,580,625 km2 for a ~37,000 gain.
Looking at the 30% ice extent at DMI, it is tracking just a little below last year. Arctic air temperatures are pretty close to what they were this time last year. Around this time last year is when ice formation really started heading up in rate which was a little earlier than previous years. Then next couple of weeks will be interesting to watch.
so that’s a total gain of 56,000 for sept 29th right?
“it is tracking just a little below last year. ”
Meant tracking just a little below 2005.
The preliminary figure for 09/30 is 5,613,750 km2 for a gain of 23,000, before the expected large adjustment upward. This could end up being the largest single day gain so far this season.
yeah…judging by the slope of all years on the record we should see steep rises for the next few weeks.
so no update today?
Re: markinaustin (#695), No there should be a big increase in the update and a big increase tomorrow. The DMI site is at last up and running. Sea ice is increasing, air temperatures plummeting and the NOAA site shows the Artic Oscillation is headed downwards too. This will mean the winds will blow from the arctic southwards instead of the other way round.
Everything points to a cold October for the NH.
December is a long way away. But my betting is it will be cold in the NH, specially if the AO is negative, which will bring the cold arctic air to bear upon North America and Copenhagen.
As Richard Lindzen pointed out, incidents of cold and warmth are unimportant for the science, but are important for public perception. And it would help the public to perceive if the people at the conference froze their b..ts off.
5,613,750 km2 (September 30, 2009)
Slow start to the freeze season. Average ice gain last 14 days:
2009: 24,989
2008: 44,040
2007: 28,828
2006: 15,792
2005: 27,210
2004: 48,929
2003: 27,009
Currently there is a deep low pressure (about 960hp) system North of Europe and a strong high (around 1040hp) over Greenland. The Asian continent is currently mostly warmer than average and this warmth is being drawn into the Arctic in the Barents/Kara sea sector. The ice edge there has receded, whilst ice grows on most of the Pacific side of the Arctic.
Models show the warm plume peaking in the next few days with +ve 850 hp temps and very strong winds reaching well into the Arctic ice pack, although in a fairly narrow area. After this the pressure patterns slacken right off with a weak high pressure system providing light winds and clear skies for much of the Arctic, which should be perfect conditions for ice growth.
Re: Michael Hauber (#698), It is a slow start. Is that 2009 figure correct? I get 24,899.6 or 24,900.
The freeze seems to be picking up though. The last 2 days over 50,000 and I think the 1st maybe close to 90,000
Where did you get those pressure figures/ details from?
Re: Michael Hauber (#698),
Looks like a fairly strong flow out of the Fram recently and also into the Beaufort no doubt contributing to the increase in extent.
Re: Phil. (#710),
Yes, and the increase in total area is also contributing to the total extent. Cryosphere Today Arctic area is up almost 0.5 Mm2 from its minimum while JAXA extent has increased slightly more than 0.5 Mm2 from its minimum.
Well I get 24899.6 for 30 Sep, the 24,989 is for Oct 1 which is preliminary of course. For weather patterns I look at http://wxmaps.org/pix/hemi.fcst.html and http://www.weatheroffice.gc.ca/analysis/index_e.html
With no adjustment for 09/30, it is hard to tell what the increase was for yesterday but the new figure is 5,675,938 km2. This gives us a 2 day increase of ~62,000 which averages to 31,000 p/day.
Michael…i don’t think that’s right because they adjusted september 30th to 5639688. which means the 2 day increase is 95,625 BEFORE this morning’s adjustment. that is a 2 day average of 47,812. So unless you are talking about 2 different days than I am, I think you made a mistake.
Re: markinaustin (#702), I think we are talking at cross purposes here. The 2 day increase 29th and 30th has been 115,313 km^2 and the increase for 1st Oct before adjustment has been 72,343.
I did think the 1st would have a big increase. (From visually estimating from the DMI site). Probably over 100,000 after adjustment.
I think the discrepancy of yours and Michael Jennings figures is due to what you are putting down as the dates figures. I think you are using your own respective dates, wherever you are based, whereas I am using the website’s dates, which maybe different from your local times.
These are the final figures:
5,580,625 29 September 2009
5,639,688 30 September 2009
5,712,031 1 October 2009 (preliminary)
Re: Michael Hauber (#700), Thank you for those links. Very interesting. I will study them at leisure when I wake up (again)
Re: markinaustin (#702),
My bad mark, I did not see the adjusted total as of late yesterday afternoon (I was using the 5,613,750 figure) and missed it entirely. The adjusted total for today is 5,712,031 km2. That gives us a total gain for the 2 day period of 131,406 or 65,000 p/day. (quick, someone check my math :))
65,859 to be exact! ; )
JAXA averages:
Summer (July, August, September)
year average area km2
2002 7082683
2003 7260786
2004 7238991
2005 6695114
2006 6801629
2007 5797672
2008 6412946
2009 6611929
Summer 2009 was only 200,000 km2 higher than 2008 and still below 2005. The OLS trend for 2002-2009 is -139904 km2/year compared to the 2002-2008 trend of -191886 km2/year. Compare that to the 1979-2008 trend for NSIDC summer extent of -67595 km2/year.
For September
year average area km2
2002 5888005
2003 6126469
2004 5958901
2005 5530094
2006 5913271
2007 4380521
2008 4836974
2009 5382787
2002-2009 -170666 km2/year
2002-2008 -238951
1979-2006 -55177
NOAA has published their September average extent and area:
extent: 5.36 Mm2
area: 3.42 Mm2
Compare that with the chart above.
Re: DeWitt Payne (#707),
Which chart?
Re: DeWitt Payne (#707),
NOAA Sept Average extent 2009 – 5.36, AMSR-E 5.38 (5,382,540)
I don’t see anything yet that would prevent 2010 from surpassing 2005 and approaching 2006.
A thought experiment on the volume of sea ice in the arctic versus the volume of ice in Greenland. The Greenland ice sheet is roughly 1,700,000 km2 and has a thickness up to 3 kilometers with a total volume estimated to be between 2.6 and 5.1 million km3. The modern maximum seasonal arctic ice volume is around 14,000,000 km2 with a thickness of say 3 meters (thats 0.003 km) and a total maximum volume of 42,000 km3. If we looked at the volume of maximum Arctic sea ice as a percentage of the Greenland glacier it works out to being less than 2%. If you include Greenland as part of the Arctic ice extent then sea ice becomes a rounding error. This melt season only helped to highlight the fact that extent/concentration should not be construed as a global thermometer. Sea ice is a measure of weather not climate, ongoingly. My sources were from Wikipedia…
Re: hengav (#712),
Volume is only part of it. Sea ice has a bigger effect on albedo because it is spread out more for a given volume.
Re: Daryl M (#713),
Volume is volume. Albedo is albedo. What effect does albedo have in the winter when there is virtually no sunlight? If it is below zero, you make ice. The whole naval-gazing examination of dust/soot/snow/ALBEDO at either pole is nothing but belly button lint. Our climate changes start at the equator, mix with our oceans, and will turn up as weather at our poles. If you are looking for the place to signal large scale changes in climate, look to the middle, not to the poles.
Re: hengav (#714),
The sink of a heat engine is just as important as the source.
Re: DeWitt Payne (#716),
Well said.
At any rate, this is off topic. More on topic it appears that the minimum has passed. We’ve had several consecutive days of positive accumulation.
Re: Lory (#717),
Sorry, are you the “new” moderator? My point is specifically on topic. The topic is Sea Ice – Sept 2009 (and it is October now). First off you are correct, the minima has come and gone. Once that occurs, and for the past 3 minima, it has been tradition for those who have been interested to pack up the calculators and begin to reflect on the year that was. It is the raison d’être of the thread. Interspersed with the predictions we learn of measurement techniques and definitions of extent versus concentration. We study ice charts, ice flows, remote sensors, ice breakers, arctic expeditions, animations, and weather patterns. I have been part of that discussion throughout, so as part of the tradition, I would like to say thank you to all those that provide the daily concentration updates. You have provided very accurate assessments of what the daily and adjusted numbers look like. I would also like to shout out to all the regulars that make this one of the longest lasting threads at CA. The civility of the thread is what keeps me coming back.
But I rather doubt if we will see a Sea Ice – October 2009. It is time for me to move my curiosity to other areas. I was concerned when 2007 minima occurred. I think we have rebounded nicely. I have learned so much more about the arctic than just the sea ice extent. It is healthy, cold, and unpredictable.
Cheers
Brad
The adjusted total for 10/02 is 5,748,281 km2 for a rather small 36,000 increase. The Canadian forecast model for the Arctic shows much above normal temps for the next 8-14 days.
hmmmm 8,000 loss today. i am sure that will be corrected but won’t be that big of a day regardless.
They don’t. All models show periods of cooling then warming. There is no ‘steady’.
Re: slownewsday (#720),
I disagree with that statement. All projections that have been produced for general (as in not academic) consumption show a rather dramatic monotonic warming forecast. They show no cooling, but nothing but warming and a lot of associated hype using terms such as “runaway warming” and “runaway melting” and warnings that we “have only 10 more years to save the planet” that they have been repeating now for longer than 10 years.
Can you cite a link that shows “ALL” models showing warming and cooling? I would be most interested in seeing that.
Re: crosspatch (#723),
I believe that you highlight a key word: “ALL.” It is very easy to talk past each other on this issue. GW Pessimists have claimed that their models are not inconsistent with periods of cooling, but their explanations seem to be rather after-the-fact notice of “global” cooling. The proponents apparently continue to adapt their analysis & warnings to what is being observed rather than being consistent to model outputs. Nevertheless, proponents this year have in a variety of venues argue that their models are not rendered unreliable by periods of cooling. See http://www.sciencedaily.com/releases/2009/05/090518172442.htm#
Re: An Inquirer (#724),
Quite funny that you completely ignore the question and then point me to an article about a study out of Berkeley that, in the face of a changing reality, attempts to validate the notion that all climate change is due to human activity, even cooling. It is absolutely hilarious.
Now, can you please point me to a climate model that shows periods of cooling and warming? I haven’t seen one. All the climate models I have seen go in one direction … and that direction is opposite of the observed temperatures over the past 10 years.
And you don’t have to point me to “all” of them, just one that has been cited by the IPCC will do.
And this might be news to you but climate has *always* varied since long before humans have existed on the planet. And just in this interglacial, a very short span of time in the history of this planet, it has been both much colder and much warmer than it is now. The warmest period was between 7000 and 5000 years ago when temperatures were estimated to be at least 2C higher than today and sea levels were about two meters higher than now. The coldest period was the Younger Dryas about 12,000 years ago. The second coldest period was the period we have just started to recover from in the mid 19th century called the Little Ice Age.
We have still not recovered to temperatures seen in warm periods over the past couple of thousand years. The interglacial previous to this one was even warmer than this one has been. There was no human industrial activity happening at that time unless Homo Erectus had an SUV plant hidden someplace we haven’t discovered yet.
This interglacial is nearing an end. If this one ends like most others have ended, look for wild swings in climate as things become less stable. As we head into it, conditions have in past events swung wildly back and forth between glacial and interglacial conditions over very short periods of time.
And there is nothing that Berkeley, the United Nations, the United States government, you, or I can do about it.
I understand your confusion – usually only the ensemble averages are plotted, not the simulations themselves. See this for an example:
Click to access GRL2009_ClimateWarming.pdf
The key point IMHO is that the models are not rendered valid by this either. In fact model proponents have consistently failed to show that their models can be reliably used. Too many pro-AGW arguments are of the form “prove me conclusively wrong or consider me right”. This is supposed to be their “due” as the “consensus” opinion. Even worse, they seem to consider themselves as the sole arbiters of what conclusively wrong means. That attitude is what bothers me most.
Adjusted total adds slightly to the days decrease at 5,740,000 km2
For those expecting an above average freeze rate, so far you’re not getting it. The smoothed rate anomaly is -14,000 km2/day compared to the 2003-2008 average and the extent anomaly is -233,000 km2. The extent anomaly has dropped 140,000 km2 in the last two days. The daily change should be on the order of +75,000 km2/day right now. Unless things pick up drastically we could well see 2009 drop below 2008 in just a few days.
Re: DeWitt Payne (#729),
Amazing what the wind can do (compaction) when it’s blowing off Siberia, eh?
Re: DeWitt Payne (#729),
There are a couple of ways that the difference between area and extent can change during the cooling period, if you’ve got say 80% coverage the area can increase by filling in the gaps, this is happening in the Beaufort sea when I last looked, I would expect this to be fairly rapid growth in coverage. Also if you’ve got winds blowing onto the ice the ice floes could be blown together thereby reducing the extent (but not the area), for the water to freeze then would require growth over open water which I’d expect to be slower, this appears to be what’s happening on the Eurasian side.
Re: DeWitt Payne (#729),
“For those expecting an above average freeze rate, so far you’re not getting it.”
Arctic air temperatures are considerably above average right now. As in more than 10C higher than is “normal” for this date according to DMI. The freeze will pick back up once temperatures cool.
Re: crosspatch (#735),
Can’t resist posting the applicable HHG2tG quote:
Another low increase today (uncorrected) means 2009 is only 100k lower than 2009 at this point. All years on JAXA show a large increase in daily values around now apart from 2009, it will be interesting to see how far before 2009 goes before following the same pattern. I doubt it will be too far away from the bunch come November.
Regards
Andy
Adjusted total for 10/04 is 5,780,625 km2 for a date total increase of ~40,000
A little bit OT, if not completely. Did you know a tropical storm formed… just next to Spain? Looks like Atlantic temperatures there have been quite high…
We all hope here it will not develop into a hurricane. That would be my first European one though…
Re: Flanagan (#734),
From Jeff Masters at Weatherunderground:
Not really that warm where it formed. (Last link is dynamic.)
Probably best to move this to the Hurricane thread if you want to keep discussing.
Anthony Watts has done a post on the JAXA figures, see here
watts
Regards
Andy
Yesterday saw (pre adjustment) a large 44,000 increase to 5,824,688 km2. We will see which way the adjustment goes although it is likely to add another 20,000 or more to the increase.
5,844,219 after the adjustment. Michael’s guess of a 20,000 addition was spot on.
Preliminary total for 10/06 is 5,883,281 km2.
Final total for 10/06 is 5,907,188 km2. It appears the “freeze” is picking up steam now.
Re: Michael Jennings (#742),
Appearances can be deceiving. 2009 continues to lose ground to 2008 as well as the 2003-2008 average. The smoothed rate anomaly is -13,000 km2/day. The extent anomaly is -226,000 km2. That’s a loss of 7,000 km2 from the day before. 2008 anomaly for the same day of year increased from -427,000 to -402,000 km2 and the smoothed rate anomaly was +16,000 km2/day. Even 2007 had a more positive rate anomaly at -6,000 km2/day. In the next four days, the 2008 anomaly for the same day of year decreases by 300,000 km2, and the extent increases by 600,000 km2. It’s very likely that 2009 will be behind 2008 by then. It’s not beyond the realm of possibility that the rate anomaly will perform like 2007 and decrease even further in the next couple of weeks.
Quick view of temps seemed to show a splitting of temps at the moment with the Canadian side being colder than the Russian side, though it was only a quick glance, can anyone confirm? Extent increase will not be that quick until the Russian side really starts getting cold I would have thought. The gradient of the increase is connected to the minima value assuming that around start of November it will join the “crowd” at about 9×10^6 Km2. If it is a lot lower than this crowd then that will be interesting.
Regards
Andy
Area in the East Siberian, Beaufort Seas and the Canadian Archipelago is increasing. The Laptev, Kara and Barents Seas are still nearly ice free. The Arctic Basin isn’t doing much either right now. The ice free part of the Arctic Basin is on the Russian/European side.
If I’m reading the MODIS images correctly, there’s a lot of cloud cover over the Arctic right now. Radiative cooling of the surface is drastically reduced when clouds are present.
Re: DeWitt Payne (#745),
There is a large polar air mass spilling South into Canada and the US high plains. Helena, Montana is looking at single digits F tomorrow. If Chicago gets the forecast snow this weekend, it will be the earliest ever recorded should enough of it stick around to be measured.
Seems like the coldest polar air is moving where there is more land and less water.
Today’s extent data:
date ———area —difference EWMA rate
10/7/2002 7310000 61094 80473
10/7/2003 6816094 4375 47927
10/6/2004 6925156 151406 49237
10/7/2005 6222813 75782 45140
10/7/2006 6401094 18594 31452
10/7/2007 5044063 90469 44852
10/6/2008 5849219 117813 66052
10/7/2009 5950156 42968 32753
At this rate, 2009 will fall behind 2008 tomorrow when 2008 adds 150,000 km2.
Is it possible to start an October thread as this one is getting a tad top-heavy.
Agree with Crosspatch and DeWitt, Low over Kara/Barents and High over Beaufort is meaning warm(er) air is flowing from the East Siberian sea region and is also compacting the ice on that side. So most ice formation is occuring currently on the Canadian side which means the Beaufort sea and the largly land area to the east of that. Hence a very low initial gain for today and 2008 has gone way past …
In connected news, although I find WUWT very interesting I do find it very selective on what it reports. Last year there were about 2 or 3 posts on how great the “bounce” was after minima, but this year with the least bounce since the start of JAXA nothing! There is a piece on the media not reporting Antartic melt being low though .. hmm, pot, kettle, black…. 🙂
Regards
Andy
Re: AndyW (#749),
That is merely because of the overwhelming bias the media has in jumping on ANY report which claims warming and total silence on any cooling claims. Surely you can see the difference? If Anthony doesn’t report it, who will Andy?
True they do, but then again there is no reason to be same, but the other way. All should be balanced.
Regards
Andy
Re: AndyW (#751),
You will see reports on WUWT that show the monthly temperatures from all recorded sources for the preceding month (rise or fall) while you won’t see anything but the increases in the media, that’s the difference. I reiterate, Anthony has to report it or it will go unreported when it does not fit the consensus. Now back to the ice, yesterday saw a pathetic increase (pre-adjustment) of 7,000 to 5,957,813 km2.
AndyW,
I know this is off topic…
But I had to say something in Watt’s defense…
He is not a journalist, he is not a news organization… He has a blog… This is at the top of the page:
“Watts Up With That?
Commentary on puzzling things in life, nature, science, weather, climate change, technology, and recent news by Anthony Watts”
His blog is a commentary of things that he finds puzzling…
I’m sorry if you think that he has some sort of obligation to comment on whatever you think he should comment on…
But he doesn’t…
You’re simply incorrect.
Thanks,
Mike
Adjusted total 5,985,156 km2 for an overall increase of ~35,000
WUWT only exists because it covers the science that the asleep-at-the-wheel media doesn’t cover. The site is contrarian by nature. That’s it’s role. There are plenty of places to read the consensus views.
date extent(km2) difference rate
10/8/09 5985156 35000 32978
10/7/08 5999063 149844 74431
As expected, 2009 is now behind 2009 for the same day of year.
…behind 2008….
Amazing what you see just after clicking on submit.
WUWT is nice, but personally I prefer DenialDepot. Just read their latest entry on Yamalgate for some excellent Blog Science.
Re: Neven (#758),
Please don’t tell me you think that guy is funny. That is the most moronic sophomoric attempt at satire I have ever seen.
Re: Michael Jennings (#759), What struck me as funny is that many of the comments at DenialDepot could be copypasted as a comment at WUWT and not be noticed. I’ve tried that a few times, with success.
You must admit that there is a certain section on the “skeptic” side of the climate debate that is very easily parodied. DenialDepot does that quite well, I think.
Re: Neven (#760),
“What struck me as funny is that many of the comments at DenialDepot could be copypasted as a comment at WUWT and not be noticed. I’ve tried that a few times, with success.”
Why would you want to do that unless you weren’t trying to help find out the truth about these very important matters? Deliberately misleading people because it strikes you as funny is both juvenile and contemptible.
Re: kimberley_cornish (#761), “Why would you want to do that unless you weren’t trying to help find out the truth about these very important matters?”
What makes you think that we lay people can help find the truth about these very important matters? At best we only help muddle the issue and unwittingly push other people’s agendas.
“Deliberately misleading people because it strikes you as funny is both juvenile and contemptible.”
Maybe I should have said ‘peculiar’ instead of ‘funny’. The point of my ‘experiment’ was to show that outrageously satirical comments on DenialDepot don’t get noticed between a lot of the comments on WUWT. The articles at WUWT have improved somewhat over time, but the comment section should be a big turn-off for anyone, no matter on what side of the climate debate you are. I can’t imagine a skeptic would want to be associated with it.
When I just started reading WUWT I have asked Anthony Watts why he let all those comments pass (no matter how insulting, misleading or unintelligent), because obviously it reflects on the nature of his site. He didn’t seem to care.
Sometimes I think the only thing Anthony Watts cares about is how much attention and hits per month he gets, and probably translates this into the amount he’s helping to find out the truth about very important matters. But unwittingly or not, all he most probably does is push someone else’s agenda. And unwittingly or not, if global warming is anthropogenic and continues unabated some people are in a bit of an ethical predicament.
Re: Neven (#762),
I understand your point and agree some do go over the top occasionally here but would you like to look at some of the comments so called “scientists” have said on the other side (pro AGW) of the debate? These are SCIENTISTS, not random posters mind you, who claim (via the news release media) we are killing off the Polar bears when in fact their population is expanding, AGW causes both an increase and and decrease in hurricanes, AGW will cause all the ice in the Arctic to melt by 2015, 2030, 2050, etc, floods and drought in California are caused almost exclusively by AGW, all the skeptics are funded by the big oil companies (hey, where is my cut?), the models show both warming and cooling (depending on how they massage the data) during the next century, Antarctica is losing ice (I think they even may have given up on that one), coastal cities and islands will be inundated by rising sea waters etc etc etc. On top of all that Neven, people on your side are trying to ostracize and censor any scientist who dares to question their methodology and model projections even though their predictions do not match the empirical data. Now I realize this probably will bother you no end but I am going to end my post with this: God bless you.
Re: Neven (#762), And your closing comment “And unwittingly or not, if global warming is anthropogenic and continues unabated some people are in a bit of an ethical predicament.” can be easily reversed without losing validity, “And unwittingly or not, if global warming is not anthropogenic and stops or reverses some people are in a bit of an ethical predicament.” Which is more likely?
Re: Neven (#762),
It will have to be proven to be caused by CO2.
Anthony does not censor as in RC.
My concern was that you admitted to pasting misleading rubbish on the WUWT website without any of the concomitant shame that you ought to have felt for such actions. Whether or not lay people can find out the truth about these matters, it is clear that your own actions didn’t help them one bit.
The point of your ‘experiment’ (and I use your own inverted comma expression) having been made, did you then post that your previous saltings of the threads were falsehoods? Did you at any time ask Anthony Watts (and you stated that you did communicate with him) if you might conduct the experiment on his blog? You write “… if global warming is anthropogenic and continues unabated some people are in a bit of an ethical predicament.” Does this justify your propagation of falsehoods? And, of course, if global warming is not anthropogenic, can’t you see that the ethical predicament is one not for other people, but for yourself?
I’ll be very happy to be in that ethical predicament. 🙂
Big day for 2009 yesterday, the first 100,000 + day this season at 6,093,594 km2 for an increase of ~108,000
I knew I shouldn’t have commented on WUWT … putting everyone at risk from being censored for being off topic. 😀
It does prove to me, however, that this sequence of blog posts that Steve has let kindly run is in a class of it’s own in some respects. it’s very data centric whilst allowing a few flags being planted in the sand. The data always comes first though, no spin.
Yep, first 100k plus day, the worm has turned. Next stop 1st Nov and see how it compares. One thing for certain, the early freeze stated by Shawn has not happened. If Shawn is still around any comments on this?
Regards
Andy
Although 2009 has had a 100+ day and will likely have several more, I would not expect 2009 to catch up to 2008 until November 14. If it does catch up sooner than that, I would be willing to consider that Artic conditions have entered a new regime.
well we just added 122,000….so this may happen more quickly than you think Inquirer….
2009 now seems to be back in step with 2008, it’s as if 2009 is the twin sister of 2008 but just forgot the summer happened. 2008 has a slow start to the melt season also of course.
I can’t get the figures from the JAXA file at the moment, just the graph.
Regards
Andy
the data is here: http://www.ijis.iarc.uaf.edu/seaice/extent/plot.csv
I know, it jsut didn’t give me any figures. Corrected now though and the initial extent increase is less than 2000, will have to see what it is updated to.
Regards
Andy
[Please continue this thread at Sea Ice – October.]
As I observed last fall, after someone here took issue with it, I’ve got my poll up for what people think of September’s ice extent. You’re welcome to add yours to the anonymous poll. Or to comment with explanation at Sea ice estimations.
Just be forewarned that I’m about to be off-net for a month, so comments might take a long time to show up. But your predictions will be safe, and time of prediction safely held, and will appear when I do come back to the net.
2 Trackbacks
[…] Arctic sea ice forecasts this year? – Maybe not so good. 7 09 2009 Steve McIntyre on Climate Audit brings our attention to an interesting sea ice extent forecasting “contest” conducted […]
[…] McIntyre on Climate Audit brings our attention to an interesting sea ice extent forecasting “contest” conducted by the […]