Steve McIntyre on Climate Audit brings our attention to an interesting sea ice extent forecasting “contest” conducted by the Study of Environmental ARctic CHange (SEARCH). With the end of the Arctic melt season likely just a few days away, it appears that the experts have a lack of forecasting skill for the subject they are experts in.
SEARCH writes:
We received 13 responses for the September Outlook based on July data (Figure 1). Estimates for September sea ice extent are in a narrow range (4.2 to 5.0 million square kilometers), as were the Outlooks based on May and June data. As the submitted uncertainty standard deviations are about 0.4 million square kilometers, most of these Outlook expected value estimates overlap. All sea ice extent estimates for September 2009 are much lower than the past climatological extent of 6.7 million square kilometers.
Here’s the SEARCH graph (Figure1 PDF available here) showing forecasts from several well known Arctic experts and organizations. I’ve added the most recent available data, the September 6th ice extent from IARC-JAXA of 5,345,156 square kilometers in magenta for a current reference.
While we can’t be certain what nature will reveal as the final number, it is likely that the end number will end up somewhere between 5.1 and 5.25 million square kilometers. What is most interesting is that it appears that all of the Arctic experts overestimated the amount of melt back in August, using July data as a forecast basis.
McIntyre made his own prediction two weeks before this report was published saying:
2009 is now slightly behind 2008. My prediction is that 2009 will end up over 500,000 sq km behind 2008.
His wording is a bit confusing, but what he means is that the final number will likely be about 5.15 million square kilometers.
As Steve McIntyre writes:
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.
Back in early March 2009, I asked WUWT readers what they thought the 2009 Arctic sea ice extent would be.
With 67% saying then that the 2009 extent would be greater than 2008, and with McIntyre’s forecast also, it appears that bloggers and laymen just might have have a better handle on sea ice extent than the majority of Arctic experts themselves.
The next few days will be very interesting.
Discover more from Watts Up With That?
Subscribe to get the latest posts sent to your email.
Arctic Sea Ice has to be related to Arctic temperature.
There is an excellent set of Arctic temperature records from right round the Arctic Ocean, many going back to 1900 or considerably earlier (the best in the world IMO), mainly UHI-free. Funny thing is, they show scarcely any change at all longterm though short-and-medium-term there are huge variations in some of the records. I’ve put the best of the late John Daly’s temperatures all together here. Especially for Joel who likes to claim the evidence against AGW is cherrypicked. Now reading thermometers is the most basic evidence you can get; moreover it is courtesy of NASA GISS and CRU.
Science and religion is not a very good combination.
[SNIP – sorry Flanagan, you don’t get to steer this thread in OT your desired direction, its about 2009 and July/Aug/Sep]
“Phil M (13:21:53) :
Is this the minimum extent for September or the average extent for September?”
This
http://www.damocles-eu.org/artman2/uploads/1/Sea_ice_outlook_2009_AWI_FastOpt_OASys_contribution.pdf
seems to be the best forecast. And the way I read it: They prognosticate the MEAN September ice extent. Which means that their number is higher than the extent on the day of the minimum. Which means that for those who prognosticated the absolute minimum, their estimate is even worse than thought.
And to remind us of summer’s past:
“Exclusive: Scientists warn that there may be no ice at North Pole this summer”
Steve Connor, ‘science’ Editor, The Independent, June 2008.
http://www.independent.co.uk/environment/climate-change/exclusive-scientists-warn-that-there-may-be-no-ice-at-north-pole-this-summer-855406.html
Good old kim says “Sure, I may be wrong”. Took the words right out of my mouth 🙂
Quite so, no ice at the North pole.
But as I understand it our satellites can’t actually see the North pole itself.
Sounds like a safe be to me, can’t be proved wrong unless there is somebody there to observe it.
I notice Dr. Serreze got into the act.
Kindest Regards
Prediction for 2008 by The National Snow and Ice Data Center; “Could the North Pole be ice free this melt season? Given that this region is currently covered with first-year ice, that seems quite possible.” http://nsidc.org/arcticseaicenews/2008/050508.html
Looks like they missed the boat. That’s why I don’t make forecasts, only aft-casts. So my prediction for minimum sea ice extent this summer is … ask me in a few weeks. And I bet I’ll be quite accurate too.
I still feel that area is a better indication of temperature than extent.
For example: say there is a given area of solid ice. A storm comes up, breaks some of it up and the winds disperse it some. Extent will increase, area will stay the same. So while extent increased, it didn’t get any colder and there wasn’t any more ice. Now say the winds change and compacts the ice. Now extent decreases again and area still stays the same. So you had extent increase and then decrease but there was no change in the overall amount of the ice.
Hang on. The minimum extent hasn’t been adsjusted yet to line up with the predictions
Lucy Skywalker(13:30:25) :
Arctic Sea Ice has to be related to Arctic temperature.
Couldn’t agree more. I checked John’s website for data from Eureka (where i worked in 79-80). I have collected data from Environment Canada’s website for it as it has a special place in my heart, and checked my data (means’) againsts John’s and concur. very little change in temps since 1948.
sunrise/set is now taking place all over Ellesmere Is. and temps are at or near the zero C mark and will start dropping fast. Final sunset at Eureka will be around Oct 26, however, the ice will form fast in the next week or so. It covered the fjord around mid sept, and by the spring was 8 feet thick.
The purpose of having a number of different experts make a prediction is to get a range it is unlikely you’ll exceed at the top or the bottom –that reality will be bracketed by those guesses. Of course, here it wasn’t.
I wonder what one would have said the “reliability” of the range of the original predictions was at the time? 95%?
If they didn’t take it down, my guess posted at RC a few weeks back was 5.1M, and as far back as last fall I was looking for roughly ballpark of 2005, but perhaps a bit below.
It is remarkable that in the Ensemble 1 team are people from the Alfred Wegener Institute. Remember: They measured Eisdeckendicken in Spring with their airplane Polar 5, and Radio Bremen published a short notice that they got thicker ice than expected.
I have not heard from them since. Why?
1) They have to wait until after “Kopenhagen” ?
2) They did not want their competitors to have the information ?
3) They work very slowly ?
4) I simply missed it ?
Funny, I emailed SEARCH about this yesteday. Email below. Of course given the rather heavy sarcasm. no reply received, or expected.
Dear Professor Schlosser.
May I congratulate you on a absolutely spot-in prediction.
“Although the majority of the responses indicate either persistent conditions or a slight increase over the 2008 sea ice extent, there appears to be about a 20% chance of reaching a new minimum in 2009. The September 2009 extent, as we track it for the rest of the summer, will depend on several factors, including the dynamics of the relatively high levels of thin, first year ice; temperature and wind conditions; and sea level pressure.”
What genius.
Kon Dealer
So much for “experts” and computer modeling! If I do an eyeball estimate of the average minimum of the “expert” ensemble, it’s about 4.6 km^2. That’s likely to end up more than 10% below the actual. Wouldn’t this render the “expert” results as insignificant? And remember, there forecasts were less than 2 months before the event!
Dr. Lubchenko of NOAA says the computer models for 50 years out will be more accurate than short term forecasts. What a hoot!
It is amazing how poorly experts are at evaluating their own biases and producing useful ranges for any uncertain quantity.
In a 1989 article in Am. J. Physics, Henrion and Fishoff ( v. 54, p. 791, 1989) examined the evolution of measurements of physical constants such as speed of light (back in the days when the meter was defined separately from the speed of light), electronic charge, etc. What they found is that the estimated uncertainty bounds were almost always too small–so small in fact that when the true value was measured again more accurately, the old measurement was so far away that it was completely improbable. In one case the “new” revised value was 42 standard deviations away! How improbable is that?
What Henrion and Fishoff proposed was that researchers tend to bias their results in the direction of what current consensus dictates, or they are unduly influenced by work they view as “definitive”, or whatever. They all get on a bandwagon. They are incapable of thinking of reasons why they should deviate from consensus or why their knowledge is less certain than it actually is, or why they might be biased in some way.
Certainly that is what has happened here. The experts have clustered all around values close to one another, hoping to be within a ball-park of the correct value (after all one of their colleagues must know what he/she is doing and have produced a good estimate, right?). If the estimates were not biased in some way, then the central value of expert estimates ought to be within one or two standard deviations of the actual value. Instead they are many standard deviations away in the “alarmist” direction.
This is a demonstrated trouble with global warming science over and over and over again, but few people posting outside these [stop that now. I don’t care which side of the debate you are on. ~ ctm] sites seem able to learn so.
As usual, “poorly” in the first paragraph above should be “poor”.
Mildwarmer 13:44:00
Heh, I was really going out on a limb 18 months ago. Andy Revkin didn’t believe me because of the sort of stuff he was getting from his experts like Mark Serreze, who made a spectacularly wrong guess just a few months later, I think in June of ’08. Please excuse a little triumphalism, and, by the way, have you ever seen me wrong? Double heh, heh.
===================================
You all need to remember, all these “ice-free arctic” pundits use “prediction-neutral” words to make their predictions. Words like “could”…”might”…”possibility”…
to wit:
I think there is a possibility that our planet might be covered by ice and snow in as few as 20 years from now. It could happen, based on the rate of arctic ice increase from 2007 until now. We are possibly witnessing the start of a global “summer death spiral” where many regions may no longer experience warm summer temps or be able to grow food crops.
It COULD happen…it MIGHT!!!
[I understand that was your intent, but I am taking a harder line as time goes on ~ charles the moderator]
Kevin The Skeptic.
I believe Kaufman et al. will make a reconstruction of Arctic ice extent based on sediment proxies, which will beyond doubt show that 2008 WAS ice free and 2009 had even less ice than that.
I admit that I voted for ice-free Arctic – just could not help myself 😮
Ron de Haan (13:31:21) :
“Science and religion is not a very good combination.”
Throw in a twist of bent politicians into that mix, and what have you got? – A steaming pile of polar bear poo.
The truth will out eventually.
“How have the scientists done on Arctic sea ice forecasts this year? – Maybe not so good.”
Maybe even not so well!
/Mr Lynn
I have begun work on my GW machine.
I believe as this global cooling progresses and the Arctic freezes over Algore will be hawking CO2 machines to warm the Earth. I will be in on the ground floor.
2007, the year the Northwest Passage opened up. And subsequently closed.
The previous time it got warm in the Arctic it drew the quests to find the NW passage and got them all jumping up & down. Tragedy struck as many as got stuck in the ice.
2009, the warning signs were up with the news of rescues, lucky to be alive.
Where are those stories?
Just one, our best buddies the Caitlin crew. Science 4 sale, inquire within.
It was just a little bit too obvious, due to unusally cold Northern winter & spring that foretold where the summer melt was headed. South.
The DMI for 2009 should be preserved in acryllic.