« Yamal to the rescue | Main | SEARCH 2013 Sea Ice Outlook: June report »

Comments

Feed You can follow this conversation by subscribing to the comment feed for this post.

jdallen_wa

Robert is a tad more aggressive in his estimation than I am, but I think he's definitely "In play" as far as what may happen.

I've been looking back a bit; there's nothing like the "hash" we're seeing across the arctic right now, in so far as the structure of the ice is concerned. There's all sorts of ambiguity - some stuff showing little melt, other locations showing lots. It speaks to there being a lot of local phenomena in play, in ways we don't yet understand, nor have the sensing to capture.

We might see a bounce, but as likely, if not more so, we may see the worst melt out yet. It is awful hard to tell, at this juncture.

Rlkittiwake

I'm with Robert as far as what the final extent/area will look like.

The only mechanism that I see working for ice preservation this year is the possibility of floes smashing into the CAA, creating summer ridging. A strong high forming over Greenland would serve to push the ice offshore, though.

It's all over but the screaming. And the approval of Keystone Pipeline.

Chris Biscan

GFS still bringing the basin wide melt.

Susan Anderson

I'm going to stick with 2.9 plus or minus 1.0.

Level 1, 4 years Arctic interest (about 10 on climate overall), particularly here, passionate but lacking technical ability. I have strong visual acuity, having taught drawing (mostly to scientists and beginners!) for many years.

As many others have noted, the multi year ice is mostly gone and what's left is fragile and broken. Somebody posted a revealing thickness graphic (Gates?) and I was started to see something estimating underice loss of 25 or more centimeters (!).

I would have said in May that I was too low, but now I'm thinking it might be too high.

Having swum in the north Atlantic, I know that water doesn't get warm until mid-July, and stays that way almost until October. There's at least a six-week lag time for warmth in general.

Weather weirding probably has a few more surprises, but this week of near 90s in Alaska will provide a jump start in that region (Yukon).

Susan Anderson

Also, significant increased moisture overall.

Rlkittiwake

(Oh, and I'm at level 1 and 2007 was my rookie season.)

Remko Kampen

Mignonette, simple misunderstanding, sorry for that. I´m obsessed with pancaking of the large ´toast´ part (R. Gates, 01:38 today) of the Arctic ice, meantime the zones with melting ponds I'm not reckoning with at all (anymore): they're toast anyway.

I thought we were discussing /A-Team | June 17, 2013 at 23:22/ which contains this question: "Let's take another poll: tell us what you think these two are showing, then remind us of your estimate."
I think R. Gates just answered that one.

I Ballantinegray1

Hi Mr Gates!
Since 07' I had used the 'High concentration' ice on Aug 1st ( C.T.) to have an idea of the final ice mass left come Sept.
At some point this correlation must fail as ice thickness in that mass of high concentration ice does not have it in it to survive the bottom melt end of the season (or exceptional export events?).
When I consider the fracturing we saw in this area ( your outline) of ice, and remember the loss of the shorefast ice along N.Greenland last year, I wonder if this year will be the one to fall lower than the 'High conc ice' on Aug 1st?

NeilT

@ jack

"I would like to see the top and bottom fifteen (15%) percent eliminated and the remaining 70% averaged"

Personally I'd say that was applying your own slant to crowd sourced data. Both the top and the bottom gave reasoned analysis of why they thought it would be that way.

The whole point of an average is to take the good with the bad, mix it in the middle and come out with a potential projection. Whether you agree with it or not.

If you keep on trying to take the average of the middle ground, in a shrinking Arctic, you will always be behind the curve. Because conservative estimates are always behind a curve like this.

Personally I'd say average them all. You may be surprised. You may even hit the nail right on the head.

NeilT

In my mind I had that yellow line a touch further out to the pole and bulging slightly out towards Canada and Alaska.

Well that was back in May anyway.

For those who doubt the sheer speed at which it can melt out.

Here is Barrow today

Link

Here is the 2011 prediction chart with the previous years breakouts

Link

Here is the 3 day animation which shows how the ice is moving and melting.

http://feeder.gina.alaska.edu/feeds/webcam-uaf-barrow-seaice-images/movies/current-3_day_animation.mp4

Are we still so sure that it's not going to melt because it was colder this year? I certainly am not.

Neven
Regarding whether results from this cycle should form an official prediction contributed to SEARCH -- I think that remains an open question. I am happy to write up and send off a July contribution on behalf of this blog, if the sentiment here seems to favor that; or not, if people prefer otherwise. It seems there are both pros and cons. What do you think?

I ran monthly polls last year, and people proposed I submit the results to the SEARCH SIO. I didn't want to do that because of 1) my own personal reticence, 2) the fact that I wasn't able to do anything interesting with the numbers, 3) I felt embarrassed seeing the WUWT poll (I believe the English term is 'vicarious shame') and didn't feel some online poll deserved a place among serious predictions.

This year it's different, because 1) my reticence was a mistake last year, and it wasn't my call to make, 2) you, Larry, do have the statistical skills to give the whole enterprise some more cachet, and 3) by letting people explain their choice for a prediction this has moved somewhat beyond a simple on-line poll.

So, as far as I'm concerned, you can send the results in as an official ASI Blog prediction. And thank you for making this effort. And thanks to everyone else for explaining your rationale.

me.yahoo.com/a/nSjChi4X3vr8X3DRw93GkY1.cerja.8nvWk-

Neil the current radar animation (3 days) is even more dramatic!
http://seaice.alaska.edu/gi/observatories/barrow_radar

Clare

3.4
I'm sticking with the same figure as for the last poll. Not based on any great scientific analysis of my own. Just looking at the state of things this year ie. slow start so I wondered should I up this but then reading a bit more & A-Team's image of the state of the ice says 'no'.
That image looks v like the xrays of my lungs when I was very crook a couple of years ago, seems like there could almost be an analogy here? Tho' the black areas on my xrays were the good parts, so I guess that doesn't work. But whatever, the ice looks 'sick' to me.

On the Fufufunk scale I'm a 1, with 8-9 years of 'obsession' about things ice-y, first Antarctic, but since 2010 its been almost all Arctic related.
Clare

FrankD

3.9 million sq km.

Extent is a horrible metric and one which I pay little attention to, so this is with low confidence. I forecast a dead-cat bounce six weeks ago on the Forum, and I see little reason to change yet.

Those who have projected low end numbers based on expected volume drops are not considering a key aspect of extent, IMO. Since since a healthy tight pack with good volume and a broken diffuse pack with low volume can produce the same extent, falling volume does not necessarily correlate well with falling extent. This year's pack seems thin but (until recently) well spread. That perhaps makes it vulnerable to massive reductions in area, but I believe extent is likely to hold up somewhat better.

As to Fufufunknknk's scale, I'd give myself about 10 years as a 1.3. That is, 5 years as a 1, 3 years as a 2, 2 years as a 1. The three years as a 2 was fun, but took up too much "real life" time to sustain.

Paddy

I feel we could do with a little more discussion of the extent of our individual uncertainties. Personally, I'd be very surprised if the extent this year was either less than 2.6 million sq km (a drop of more than 1 million sq km from 2012's record, and a very steep drop from where we stand today), or more than 5.1 million sq km (ie, more than any year since 2007, and a rise of more than 1.5 million sq km from 2012). But anything in between seems possible, especially with the great uncertainty and disagreement about what the implications of the cyclone's behaviour might actually be.

Tor Bejnar

My June guess is 3.6 M km^2 for average September Arctic sea ice extent as calculated by NSIDC.

This is a ‘huge’ increase from my May guess of 3.15. The main influences include
1) current poor state of the ice in the eastern Arctic Basin (endless floe edges to erode and melt from; lots of room for compression under an extended high pressure regime) and lots of ice in southern areas that is sure to melt out anyway,
2) Jim Pettit’s graphs showing what previous year’s loss from the current date would yield (no new record for area or extent is projected - I’m guessing a new record anyway - less MYI, more potential, due to fractures, etc., for MYI export through Fram Strait),
3) Mr. Pettit’s table showing Million km2 steps in CT Area, per day-of-year (2009 had the latest ‘cliff’ start in 9 years, and then lost 3M km2 in a record 26 days - a 3M drop in 24 days would catch 2013 up to 2012) [An aside: I’m missing updates of Seke Rob’s ASI Extent M km2 table; I hope and pray he’s OK.].
4) NSIDC’s graph of Daily Sea Ice Extent Time Series shows no ‘cliff’ yet (but I expect one to start any day - maybe I’m ignoring the evidence?).

2007 had record melt under spectacularly pro-melting & compacting weather. 2012 had record melt under mostly unspectacular weather. 2013 will require spectacularly pro-melting & compacting weather in July-Sept. to make new records, and it may have started (per Neven’s ASI Update 2013 #3); I’m guessing this will happen/continue. However, we could also end the melting season with an Arctic covered in dispersed slush creating a huge extent recovery. Therefore, the range of average NSICD September Arctic sea ice extent I see possible (and not be shocked) is 3.0 to 4.75 (somewhat in line with Paddy’s thinking).

In Fufufunknknk’s June 19 at 12:39 experience scale, I’m somewhere about a 3 (passionate) or 2 (occasional amateur statistical models), and started reading climate science 4 or 5 years ago, starting as a skeptical geologist (how could puny man ...), but quickly gained ‘respect’ for (dismay of) our ability to affect Gaia.

Safe (and low-C) travels, Neven. But have fun anyway.

Mdoliner43

Mathematicians, like me, are (by definition) lazy. Most of what goes on here seems like too much work to me. I am a 1 and will stay that way. Global warming because of increases in greenhouse gases is a scientific theory that says the radiation budget, the difference between insolation and radiation, is positive and growing larger as more greenhouse gases accumulate. This extra heat always moves from warmer to colder, and most of what is discussed on this site is observations of the mechanisms of this movement. And of course heat melts ice. The real question is not about extent or area, both poor proxies for the amount of ice, but volume. Even volume is only a proxy given the variability in mass of a volume of ice due to its condition. I note that volume has only increased in one year since 2001. This makes sense to me given that the radiation budget must be quite large by now and growing larger every year.
What could cause a volume increase? There have been reductions in insolation and increases in radiation, but these are the result of long standing conditions. Prior to the accumulation of human generated greenhouse gases the radiation budget must have been very nearly in balance (zero), the ice amount going up and down at random, these changes due to long standing conditions, must also be small, at least during the period of stable climate in the recent past. So I think the only possibility for volume increase is the result of the rate of movement of heat to the colder areas, mostly the arctic (in the NH). The only year of increase was 2008 after the huge drop in 2007. Therefore I guess that that huge drop was a result of heat moving into the arctic far faster than the usual rate. 2007 was, notoriously, a “perfect storm” for ice melt. However 2012 was not such a perfect storm. On the contrary, while it was happening many said it was a relatively bad or at least average year for melt (heat movement into the arctic). Therefore, I do not think the conditions are right for a volume rebound this year. What extent will be, given the slushy condition of the ice, is anybody's guess, but if crowd source averaging can get it right I guess the science, such as it is, is superfluous,

P-maker

@Mdoliner

Good thinking there. The brain is also a muscle, which needs to be exercised from time to time, and you are doing a pretty good job, despite your self-declared laziness. However, since the Arctic melting is a combination of two accumulating factors this year, I think you should be a bit more bold in your statements. The first part is the globally accumulated heat excess, which will be working on the ice from now on. The other accumulated effect is the fracturing of ice due to persistent Arctic cyclones. This churning has multiplied potential melting surfaces manifold. Combining these two factors into one, inevitably leads to a more dramatic drop in volume, than you may have presumed.

Hans Gunnstaddar

Good point Mdoliner re: only year since 2001 with an ice volume increase (2008). Not sure 2012 didn't at least end that melt with a perfect storm with the big cyclone though.

I wonder if the only thing keeping the genie in the bottle is the MYI clogging the Fram drain, and once deceased each season will be much like the one before, freeze, shatter, melt, flush, rinse, repeat). At that point each season ice-wise may become steady, with all attention and predictions centering on methane and CO2 release ascent up closely followed daily graphs in the flush season.

L. Hamilton

@Neven:
So, as far as I'm concerned, you can send the results in as an official ASI Blog prediction. And thank you for making this effort. And thanks to everyone else for explaining your rationale.

The SEARCH folks have also requested that I send our July prediction to SIO (along with something like Figure 4). If that's agreeable to the posters here, I will do it. And either way want to add my thanks to everyone for their predictions and rationales. If the predictions are data, this thread is rich metadata.

Rlkittiwake

Hans said: "I wonder if the only thing keeping the genie in the bottle is the MYI clogging the Fram drain..."

Is that still the case this year?

http://igloo.atmos.uiuc.edu/cgi-bin/test/print.sh?fm=06&fd=20&fy=2011&sm=06&sd=20&sy=2013

I know the CT concentration maps don't give an accurate impression of the quality of the ice, but even as recently as 2011, it looked like the MYI coming through the Fram was being extruded in a plastic manner more like toothpaste from a tube and notsomuch like iced tea being poured from a pitcher by a careless waitress at Denny's.

I've wondered for a while how long it would take the basin to clear through the Fram if all the ice became unconsolidated, but most of the numbers I can find assume plastic deformation across the entire pack.

That genie might be about to bust out hard.

Hans Gunnstaddar

Thanks for the link RLkittiwake, which compares 6/20/2011 to 6/20/2013. The link below compares those dates in 2012 with 2013. Sure enough as compared to both of those years 2013 looks like it could uncork at anytime.

http://igloo.atmos.uiuc.edu/cgi-bin/test/print.sh?fm=06&fd=20&fy=2012&sm=06&sd=20&sy=2013

jdallen_wa

Hans G said...
Thanks for the link RLkittiwake, which compares 6/20/2011 to 6/20/2013. The link below compares those dates in 2012 with 2013. Sure enough as compared to both of those years 2013 looks like it could uncork at anytime.

Ooof. I hadn't compared 2012/2013 same-day for a while. 2012 has more "voids", but far and away, the extent in the 2013 ice has been savaged, and overall is consistently lower concentration, if the maps are even close. Not terribly reassuring.

Paddy

@Mdoliner,

Even though it's been rare in the past, I wouldn't count out a possible volume increase this year, since when last reported, the PIOMAS ice volume was tracking "425 and 901 km3 above those of 2011 and 2012 respectively": http://neven1.typepad.com/blog/2013/06/piomas-june-2013.html.

David Vun Kannon

Somewhat off topic but based on Hans Gunnstaddar's comment above, I was thinking what would an essentially ice free Arctic look like? Taking 'essentially ice free' to mean 1 million sq km at the end of the melt season, I think that will be (approximately) a half ellipse of ice. The major axis extends from the western tip of Prince Patrick Island to Nordostrundingen, Greenland. The minor axis extends out into the CAB about 300 km.
I think this amount of ice could be resistant to summer melting in place, and export, so sea ice minimums might come to rest around this number for a while, perhaps a few years. (Freeze, melt, repeat)
After that, I think the main driver for change will be how much heat the Arctic retains over the winter, how many of the regions stop icing over completely.
A sea ice maximum of less than 12 MM sq km would really change the weather, IMHO. Scary.

Mignonette

Neven, Larry,

I realize now that some of my comments on June 19 are a bit too impulsive, partly due to my non-English mother tongue. Sorry about this. I had no intention at all of being arrogant, but only a genuine concern about scientific reticence. Still, that doesn't justify getting too impulsive.

I may indeed be underestimating the speed at which the Arctic catastrophe is happening. It's far from certain that the standard statistical models still make any sense in this unprecedented situation. But since the extent measure, more than all the other ones, could really go either way, I'll stick with my earlier estimate. Keep up the good work.

Mdoliner43

@ Paddy

I do not deny that a volume increase is possible, but since the radiation budget is positive and growing larger, it becomes less and less likely. It can only happen through a slowing of heat transfer to the ice while the gradient in the NH hot to cold grows ever larger. Remember, the earth is hotter every year. This gradient smooths out with the melt, for heat used in melt, rather than affect temperature, is then stored as latent heat or, in an expression I like to use to tweak physicists, heat that's not hot.

Kevin O'Neill

"...in an expression I like to use to tweak physicists, heat that's not hot."

Yep - for most of the year the largest source of heat in the arctic is sitting under the ice at 0 to -2C ... the ocean.

That's why clouds during most of the year are a positive feedback - they keep (or diminish) ocean heat from escaping to TOA.

Ned Ward

4.4 million km2, if it's not too late to join this game.

For each year 2002-2012, I calculated three regressions, to predict Sept. monthly extent based on (a) most recent IJIS daily extent, (b) most recent PIOMAS volume, and (c) a simple linear decline from year to year.

I then calculated an ensemble mean, weighting each model based on its std error. The ensemble prediction is 4.4 plus or minus some fairly large number that I forgot to write down. That would make 2013 the third lowest on record, after 2012 and 2007.

[Yes, given all the foolishness at WUWT lately, I just couldn't resist using an ensemble mean as the basis for my forecast....]

Jtstewart

It shouldn't be counted but I thought I should add this prediction of the sea ice from Japan. 5% less than the 2012 record of 3.41 million square kilometers. http://ajw.asahi.com/article/behind_news/social_affairs/AJ201306210059

L. Hamilton

SEARCH has just announced the deadline for July SIO submission, July 8. Let's say, July 5 on this site so my grad student Matthew Cutler and I have enough time to analyze all the responses, draft an ASI contribution and send that off to SEARCH. As done above, we'll use the median and IQR as summary statistics.

Yikes, 230 posts so far.

Erimaassa.blogspot.com

3,500000 Mkm2 , illustrating the difficulty of choosing categories. the longer ice stays dispersed the more surface area it has. expecting clouds over the arctic to reach temperatures high enough for rain after a fortnight (normal fog not ice fog)

epiphyte

3.2 Mkm2. Based on observing the slope of the decrease in area toward the end of June, + the minimum for the year, compared with the same observations for the previous year, going back to 1995 or so. The slope on this day hasn't been steeper than now for at least ten years. IMO it's more likely to steepen further than shallow for the next month.

I'd regard myself as a 1.5 (I'd have said 1.0 had I not been living in MN watching lake ice melt every spring for the past 20 years) I know the melt started slow - but IMO it's on a trend to catch up and surpass 2012.

Tommi Kyntola

I'll stick to my earlier 1.8. It's a guestimate based on the information both in comments and linked to in here. I believe the ice is in a worse condition than the graphs and models presently show, both extent and piomas. In that sense I dont think it has been a slow start in general and I see no reason why we'd finnish higher than last year barring unexpected weather for remainder of the melt.

Massimo

September mean extent: 3.75 M sq km
- average melting in June
- low ice concentration just south of North Pole, western/central Siberian sector

James Lovejoy

Guessing 4.25 Million K2 monthly minimum.

Mostly guessing that the cool spring is going to outweigh the poor state of the ice.

I won't be overly surprised to be wrong. In fact a new record low wouldn't surprise me.

IMO we're at a time where we're just discovering how valid formulas that worked earlier are now.

In any case, at best this is just a reprieve for the ice. It still looks like we're on a path for an essentially ice free arctic around 2020, if not earlier.

johnm33

.75
+/-1, 30 years,3 first interested in 76 when it didn't rain[daytime] for 5months 3 weeks and then rained for a week ,if memory serves, stopped flying in 83 'til 99 when i considered situation lost, have since joined the lunatics in their wanton consumption. Relied on my subconscious,[ what i thought when i woke up] after following every lead and considering every hypothesis/assertion, here and elsewhere, as time permitted. I'd be surprised to enter September above 3.8, the only variable too difficult to take any account of is the model NSIDC uses, so I expect to be wrong.

Rob Dekker

4.38 with a standard deviation of 227 k km^2.

I hope its not too late Larry, but I needed Ruther's snow cover numbers before making this prediction.

Argument for this prediction is based on June snow cover, ice extent and ice area, as a measure of how much energy the Arctic is absorbing.

Details in this post, and my latest comments therein :
http://neven1.typepad.com/blog/2013/07/problematic-predictions-2.html

Let me note that my gut feeling is that this prediction is still on the high side. The data suggests that 'holes' in the ice (extent-area) play a more significant role in determining the final Sept ice area or extent numbers, and this year these 'holes' seem to be smack in the middle of the pack.

Thus my 'prediction' may still be on the high end. I think there is a more than 50% chance we will break the 2007 record, but virtually NO chance of breaking the 2012 record.

Fufufunknknk

Just eyeballing the extent changes in the last two weeks, it seems like 2013 has made up about 20% of the difference with 2012. I think there is a decent chance but it comes down to weather as to whether 2013 passes 2012.

Werther

I just voted 3.25 – 3.5. I’m still within my range 3.28 – 4. But when pressed now, I'll go for the lower bound. I understand why blogfriends like FI Rob Dekker come up with 4.38. But the state of the ice in the part that matters is worse than this time ’12.
In a season, characterized by weather turbulence, I regard the hazard for a Nemesis cyclone to be high. Any wild jet stream branch could spawn one through July and August.
A bit like our tropical weather colleagues over at Masters Wunderground. You know all the signs flare red, although the Atlantic “looks quiet now”.

Faustusnotes.wordpress.com

Hi everyone, long time lurker first time poster. I thought I'd try a predictive model this year. My prediction is 4.69 million square kilometres, 95% CI: 4.06 – 5.32 million square kilometres.

Basis: a Prais-Winsten regression model using snow anomaly, temperature data and area and extent data for april, may, June and lagged value from the previous September. Details are here.

My model suggests a huge recovery. It barely seems plausible, but my model has shown itself capable of identifying the two previous crashes, with even some skill at five years out. So it could be right! Here's hoping ...

Gavin Cawley

My prediction of 4.1 was the median of the SEARCH predictions. I suspect the actual September minimum will be a little lower than this statistical prediction (which is based purely on the September mean in previous years), but it is likely to be higher than the September 2012 mean because of "regression to the mean". The 2012 event was well in the lower tail of sensible statistical predictions (that had previously worked well), and it is unlikely that the various physical processes will all be lined up in the same direction as they were last year.

Rob Dekker

Hi Faustusnotes,
I like your analysis, and I recognize some of the coefficients you found.

It seems that you chose the same variables that I did, with the exception of one :
- April and May surface temperature

That brought up a question for me : Why the 'April and May' surface temperatures ?
After all, main freezing occurs during Dec, Jan, Feb, and thus the surface temperatures during that period seem more physically important for the state of the ice starting a new melting season.

Could you please comment on the choice of that variable as a predictor of the next melting season ?

Faustusnotes.wordpress.com

Hi Rob Dekker, glad you like my analysis. It's a pretty classical modeling method, really, and subject to all the usual terrible constraints of such a method. I'm thinking of trying a more flexible ensemble method for August.

I used April and May because I couldn't get June and I just kind of thought that how hot it is now would be the most important temperature measure. I know nothing about sea ice, how it melts, the arctic or, for that matter, the northern hemisphere, so I just grabbed the first thing I had in mind. It occurred to me that I should use January extent instead of last year's september extent as a predictor, so maybe I should try winter temps.

This is part of my ensemble thinking. I have a crap-ton of variables: realistically, all the extent and area values for each month since the previous Sep, all the temp vars since Dec, snow cover since Dec, year, plus volume. I know of a technique for ranking models and then combinining them, so I'm thinking of doing a massive ensemble using mixtures of these covariates. Unfortunately I'll probably have to use R to do it (shudder) but now I've prepared the data in Stata it should be easy ... I'll take your suggestion under advisement for August.

May the best estimate win!

Rob Dekker

Faustusnotes,
I was not thinking of this as a competition. In fact, I was hoping we could all learn something.

The problem with statistical analysis is exactly this "I have a crap-ton of variables:".
With every variable you introduce, you increase the risk of overfitting.
As the common saying goes : with 4 variables you can fit an elephant, and with 5 you can let it wiggle it's tail.

So the trick is not to use MANY variable, but instead as few INDEPENDENT variables as you can.

When I read your analysis, I appeared to me that just like in my analysis, the focus in your analysis is on variables pertaining to the 'summer melt' season.

We have no variable that describes the 'winter growth' season.
And the very best variable there is describing how much ice will grow in winter are the winter temperatures over the ice regions. Specifically Dec, Jan and Feb (and maybe March), when the sub-ice ocean has cooled down enough so ice can start to seriously grow.
See for example the ice growth over time from this buoy data from the Beaufort :
http://imb.crrel.usace.army.mil/irid_data/2012H_thick.png
Or this one in the Laptev sea (moved across the NP area during winter) :
http://imb.crrel.usace.army.mil/2012J.htm

Since you have your system set up to included temperature data in your analysis, and I did not, you should be able to get a better correlation than I did, as long as you use and INDEPENDENT variable like winter temperatures.

Could you try that out, and see if you can get better correlation by using winter temp data instead of spring temp data (when neither melt nor ice growth occurs) ?

L. Hamilton

"In June through early July, participants in the Arctic Sea Ice blog posted 82 individual predictions for the mean NSIDC September Arctic sea ice extent. The median value of these 82 predictions was 3.2 million km2, with an interquartile range (approximately the middle 50% of predictions) from 2.7 to 3.9 million km2."

Sent to SEARCH for the July SIO, along with a rationale saying a bit more about the experiment. I'll write an analytical post here next week once we have the official SEARCH data for comparison. In the meantime here is the distribution of July predictions from this blog. They ranged from 0 to 5.6:

Stem-and-leaf plot for asi_blog
(Predicted September mean sea ice extent, million km^2)

asi_blog rounded to nearest multiple of .1
plot in units of .1

0* | 03
0. | 89
1* | 1
1. | 7888
2* | 0002233
2. | 555678888888889999
3* | 000001222333334
3. | 5566888888889
4* | 0011233333444
4. | 55557
5* | 0
5. | 6

Kevin O'Neill

Sorry I missed the deadline. Would not yet vary from initial prediction of 2.9 Mkm^2.

Faustusnotes.wordpress.com

Rob, sorry for taking time to reply. You're right, I'm only using summer variables at the moment, and extending the analysis to winter variables means taking a great deal more care in how I approach the modeling. I am going to generate a new prediction using an ensemble method, similar to that used in the Global Burden of Disease studies, using all the available data. I think I should include temperatures from Sep - June, and also the Arctic Oscillation index. I think an ensemble model will be more representative of the truth.

I think it's possible that winter data will be less relevant now than it might have been 10 years ago, because the prevalence of FYI means that late summer phenomena are the key drivers of ice change. We'll see!

Rob Dekker

Faustusnotes, thank you for your reply.

In summer, almost all of the ice that melts is FYI. After all the MYI became MYI because it is in an area that does not melt out during summer.

Also, the thickness of FYI is mostly determined by the winter freezing temperatures.

So, it seems to me that the winter temps are of utmost importance to the developments during the proceeding summer.

If this reasoning is correct, winter temps (above 70 deg North) should pop out as one of the dominant principal components in your analysis.

Please let me know your further findings. I'm real curious what you will find out.

Verify your Comment

Previewing your Comment

This is only a preview. Your comment has not yet been posted.

Working...
Your comment could not be posted. Error type:
Your comment has been posted. Post another comment

The letters and numbers you entered did not match the image. Please try again.

As a final step before posting your comment, enter the letters and numbers you see in the image below. This prevents automated programs from posting comments.

Having trouble reading this image? View an alternate.

Working...

Post a comment

Your Information

(Name is required. Email address will not be displayed with the comment.)