« 2011 Slush Puppie Animation | Main | Poll Results 2 »


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

Chris Biscan

-98K Prelim.

Holy Smokes that is with a -60K prelim yesterday with a revision down to 58.5K

in other words the total was like 130-140km2

I would expect a much smaller one tomorrow but no matter what it will be 75Km2 or higher unless its tiny and that is not likely.

UB and DMI 30% will plummet tonight.

Bob Wallace

Prelim 5,178,281. Down 96,563.


For the record, today I predict an IJIS extent delta of 40,000 (that is, the corrected number for August 22 will be 5,236,719).
Posted by: Nick Barnes | August 22, 2011 at 22:52

That's going to require quite the correction, given that the initial extent is : 5,178,281 (delta of 98,438 km^2)

I like precise predictions, though - nothing worse than a "prediction" with gigantic "error bars".

BTW, we've now passed the minimums for 2005 (5,315,156 km^2)
and 2009 (5,249,844 km^2)

Kevin O'Neill


11 days and a 561250 km^2 difference. 2011 needs to average 51k and change over the next 11 days to pull into a dead heat with 2007.

I say it's game on :)

Chris Biscan

This is a new arctic system with so much thin ice.

We need to start realizing that.

These models and thickness charts need some serious adjustment and Cryosat. Not sure what to say there.

Paul Klemencic

Well, that was an interesting day... I am slightly surprised the extent loss wasn't even worse, with the final Bremen map. I have been doing day to day blink comparisons since 2008, and I never saw the ice pack edge pull back over 100 km in one area in twelve hours. Just one quadrangle between 150W and 165 W, between 75N and 80N must have lost over 50k sq km.

The storm looks like it will still hit that same area hard tonight, so tomorrow could be another interesting day. If anyone knows where this storm goes next, please share.

The whole pack is actually shifting toward the E. Siberian, Laptev, and Kara regions, so the pack is moving over some warm water on that side of the pack. If this trend keeps up for a couple more days, there will be several hundred thousand sq km in those regions underlain by seawater perhaps as high as 3-5 deg C.

The next three days in 2007 lost only slightly more than 100k, so if 2011 loses 66k a day, in three days the extent will gain 100k on 2007.


98K. Wow, I didn't expect such a high number straight away.

I have updated the Slush puppie animation, and I'm going to do a separate post on those Uni Bremen sea ice concentration maps. Fascinating stuff.

If anyone knows where this storm goes next, please share.

Paul, the ECMWF forecast maps aren't very clear, but I believe it looks as though this cyclone is moving straight over the ice pack towards the coast of Siberia.

Chris Biscan

00z GFS holds serve with a HP forming and replacing this SLP basically sitting right over the center of the ice pack with two major SLPs, one in Russia the other over the Barrents.

this is about the worse possible thing for this time of year for the arctic.

the GFS has this for 10 days....about 5-6 of them it's at it's peak with a flow from E. Siberian Sea all the way out the Fram.


Timothy Chase

Chris, I hope you don't mind if I ask you about some T L A s

00z -- midnight Greenwich Mean Time?
GFS -- NOAA's Global Forecast System (a model)
SLP -- Sea Level Pressure?

" major SLPs " ?

Just trying to make a little more sense of what is going on

T L A -- Three Letter Acronym

Chris Biscan

Looking at current conditions an the 24 hour forecast.

We are going to see that quadrant between 80 and 85N and 180E get pummeled tomorrow.

It may not breakdown as fast as the one today did but it will make sure the extent and area plummet.

Also expect a massive Area loss the next 3 days..probably under 3.0km2 by Friday at the latest.

The Fram will be open the next 10 days at least.

Ice is already pumping out.

winds in the Fram from NE Greenland to SE are 15-20Kts for the next 24 hours witha trail back 300 miles or so into the Arctic basin of 8-12KT winds

around 24-30 hours those winds are cut off and that area goes relatively calm.

however between NE Greenland and Svalbard winds increase to 20-35kts sustained with 12-16kts everywhere else. This is ridiculous and will flush out that weak ice we see in there now. That ice got fried the last week:

At the end of day 3 the winds there slow down a bit while a Wind off Russia by the E. Siberian Sea cranks up around a major SLP in Russia while a HP is organizing in the central arctic.

The winds curve outward into the barrents...so the pack will be moving south but still not connected to the flow out the fram but doesn't matter since that ice there is hitting warm water.

By day 4 winds from the Siberian Sea to the Barrents set up north to south around 20-30kts for a thousand mile strip.

Another area of 20-30kt winds will form around a SLP NW of the UK these both basically move slowly around the HP in the central arctic from day 4-10 on the models.

This continues to get stronger and more connected by day 7 with a flow from the Beaufort around an HP all the way to the Fram.

it is really incredible.

on top of that. the 850s will be coldest right over the conveyor belt from the Siberian Sea to the Barrents. Which is over 4-10C SSTs that will be churning up warm water and mixing so much there will be no chance for refreeze.

the rest of the arctic is above normal with 6-10C positive 850 anomalies in the the CA Islands which will help melt out any remaining ice, even this time of year with warm SSTs around and no cold.

To get even more insane. by day 6-7 it looks like 2-4C 850s flood the entire Beaufort with 4-6C 850s along the coast and in the Chukchi. at the same time 2-4+C 850s get wrapped into the SLP over Russia and pushed into the Siberian Sea. The rest of the arctic has a couple cold pockets but most of it is -3 to 0c 850s which means no refreeze.

This is supported by both models and has been shown now for for 5 days.

If this does pan out the next 10 days as models show There is no doubt we will be somewhere between 4.0 and 4.5km2 in ten days. this would be epic and flat out scary.

If that happens, given the warm water up there. A 3.5-4.0 min extent is not out of the question. This is by no means a prediction. but an observation made on what we have seen.

We also have up to date reports the ice is around 0.9 meters thick from the East side of the pack to the NP along Polarsterns route. This means that ice will move very fast in a DPA like this. Faster then we have ever seen with winds that are being forecasted.

Chris Biscan

00z is currently 8pm Eastern time in the USA and 7pm during the winter when we lose an hour of daylight.

GFS is the Global Forecasting System a computer weather model that is ran by the NCEP a division of NOAA.

yes SLP is right.


on that map you can see a 993 SLP over the Baffin Bay. A 989 NW of Norway and an Elongated SLP over the coast of Russia.

Unlike more circular SLP's the Elongated HP and SLP are even more brutal more like fronts or troughs of air movement instead of a cyclone.

This will just conform winds over a two to three thousand mile long stretch. This will keep the ice basically moveing Clockwise in the Arctic while shrinking inward from compaction/Melt. since the winds are so powerful SE of the Fram Straight and backed so far north the ice will literally be shunted out of the arctic...the only thing that would be worse would be the SLPs being a big further west and a stronger HP.

but this is brutal in itself. This will quickly push the entire Russian side of the pack towards the North Atlantic. ice on that side will fray and melt rapidly as it pushed into the warm waters...

also ice along the Eastern edge is already beat to hell. I can't believe it's this bad already. check this:


The Eastern part was still filled with decent concentration.


2011 is not, it's almost comical at this point.

that ice in the East Siberian Sea is going to get flattened like a bug. Literally vanish in a week. IF IF this pans out.

Chris Biscan

my friends on Americanwx who are on the other side of the debate have completely vanished the last 2 days. It is dead over there beyond myself and the Moderator attm.

It is clear why no one is showing up. I don't care I am not a spiteful mean person, this is a serious bad situation that could be the wake up call we need if this year see's the big plummet like 2007 but down to 3.5-3.9km2 and a record area loss.

If that does happen, I wonder what the "opposite" side will say about it.

After today many of us were in shock and awe, they have to be as well. I wonder if this will open some eyes to how dire this probably is.


Very low SLP in the Beaufort Sea right now, as buoy 2011K shows:

Chris Biscan


UB took a nose dive. Looks like 125-150Km2 there. Which means Jaxa will likely lose 100km2+ tomorrow. Jaxa might be under 5,000,000 by Wednesday night guys. Unreal turn of events if that happens.

So far in August we have lost 65km2 per day. It is all but certain it will be above 60km2 per day by the end of the month.

We have lost 70,000km2 per day since the 3rd.

Chris Biscan

neven, does that graph show the ice thickness at installation and now?

or is that the current?


Chris, if I'm not mistaken, those graphs haven't been updated since around July 20th. I've sent a mail to ask when the provisional data will be updated. Buoys 2010 C, D, E and F are of interest.


BTW, the post on the UB flash melting is up.

Paul Klemencic

Excellent description: I am stunned and in shock and awe, and any fun I was getting out of this about being able to predict a late season melt, and the big demise of the 'slush puppy' seas, is gone. What happened last night is scary.

I believe werther made the comment about "What will an ice-free Kara and Barents do to the weather in Europe early this winter?". Well, now we should be asking, "What is the effect of adding half the Arctic Basin free of ice, to the ice-free E. Siberian, Laptev, Kara, and Barents seas, and what impact will that have on the NH weather?"

I looked at the regional ice extent maps, and the Beaufort, Chukchi and E. Siberian could see dramatic falls if this storm does similar damage as it tosses and shakes up those regions during its trip across the Arctic. The maps don't have the last several days of damage shown yet.

The death spiral in the Arctic continues tomorrow.... Lahko noc' everyone (have a light and restful night) because tomorrow looks tez'ko (difficult).



Not yet giving up the ghost.


And if IJIS revises downwards, we'll have a century break. That would be number 17 (2007 had 20 in total).

Chris Biscan



Look at this guys:

Yesterdays Image taken last night, I am not sure if it is updated during the day:


you ready for it????

You sure?


That is from the last hour or two. Still have another 12-15 hours to go.

Here is the current winds:


Here they are tomorrow evening:


Expect the Ice pack to take another 20-30 mile ride off the Land up there. How far will it dent into the Arctic Basin..considering the direction of the winds are pushing into the thinner ice...who knows.

Chris Biscan

I agree Paul.

Those SST charts made me sick to my stomach.


Damn, there is a second page......

Sorry for repeating the IJIS data.

Nick Barnes

What a lot of egg on my face. I'll keep going, though. Predicted IJIS extent for August 23rd: 5,140,000.


I'm sharing some of that egg, Nick. I didn't believe such an extent decrease was in the books. But hey, chances are we get a century break today. You know how I love those.


Timothy Chase’s last post describes something I have been thinking of before. I regularly visit this blog but still find it difficult to follow many of the discussions. Neven’s posts are very well written and easy to follow, but the comment section often contains difficult acronyms which makes the many great comments difficult to understand for newer readers.

I suggest you add a link at the top of the page where first time visitors can go and get some basic stuff explained, Neven.

It would be great to get some common terms and acronyms explained (Arctic Oscillation, Buys Ballots law, Dipole anomaly, etc).
Perhaps a map of the Arctic, examples of the most common ways that sea ice in the Arctic decreases and some words about the different daily graphs would be useful as well? I think that would make more first time visitors stay.

Chris Biscan


they are getting close to the ice. but they have a bit to go before they reach any sort of solid ice with water around it.


I'll see what I can do about a glossary. Mind you, NSIDC has an excellent resource with a lot of explanations. Maybe I should just link to that.

And you can always ask here if you need to know something.


Glossary is now in the navigation bar. And don't forget there's a search function in the left hand bar as well.

Kevin McKinney

I'm thinking about crushed ice in a glass of water. Gradually, it's melting, cooling your drink.

Then you put a spoon in and stir.

Cyclone = spoon?

Anyway, sure didn't expect a near century on today's IJIS prelim. Nick missed that by a mile, unless this is the biggest revision ever.


That's great Neven and it is indeed an excellent resource!


Glossary search didn't find 850s

surely too low for a pressure.


CT area 3.18 only down 47.5k. Does the big melt come tomorrow? Or is 47.5k quite large if the 98k extent is 16% to 14% which is only 16k?

Waves hiding 50% coverage sounds more likely to me, which means 47.5k could be the big event but would mean no 80% down to 70% coverage changes or similar.

Chris Biscan

that number is not for yesterday but the day before.

I learned that earlier this year.

the number for yesterday will be tomorrow and likely be much larger since Jaxa nose dived in area even faster.


What's the wind speed in meters/second where this storm is, anyone?

Kevin O'Neill

>> Glossary search didn't find 850s

850 = 850 hPa
hPa = Hectopascal, equivalent to a millibar

A meteorological temperature measurement used in modeling. 850 hPa level is roughly at 1.5 km, usually above the atmospheric boundary layer. That means there is no diurnal temperature variation, and the underlying surface (such as a cool sea) doesn't affect it's temperature. That is why 850 hpa temperature is used to distinguish air masses and thus to locate cold and warm fronts.

Because the models have had several problem in surface parameterisation, 850 hPa temperature forecasts have been more accurate than those for lower levels.



Thanks, Kevin! I knew that 850 mb or hPa designated height, but nothing about the finer details.


My mail has been answered (real quick). The provisional data for the buoys will probably be updated this weekend, so we'll have some more ice thickness information.


Thanks Kevin.

Ian Allen

Kevin McKinney, you're spot on with your stirring. this goes for the air also, there is a radical difference between laminar and turbulent flows. the former render 850 temps irrelevant - if laminar flow prevails (typical slack high pressure) then no upper to lower mixing ensues and little non-solar melting.
However, all changes uncomputably when you get the lazy version of tornados sucking away the chilled surface air and letting more warmth melt the ice, and start the waves which need big basins of clear water (or loose slush) to really ramp up.

L. Hamilton

Whatever happened, it's registered across all the indexes. I updated the CT North area graphic, here are the numbers:

Year min(areaN)

1979 5.30673
1980 5.50771
1981 4.95649
1982 5.13906
1983 5.38693
1984 4.69589
1985 4.99285
1986 5.38184
1987 5.28899
1988 5.14489
1989 4.81592
1990 4.62893
1991 4.46038
1992 5.02678
1993 4.47295
1994 4.8161
1995 4.4103
1996 5.23818
1997 4.89971
1998 4.2624
1999 4.2045
2000 4.16877
2001 4.53362
2002 4.03471
2003 4.14166
2004 4.28297
2005 4.0918
2006 4.01692
2007 2.91944
2008 3.00356
2009 3.4246
2010 3.07213
2011 3.18004

Ned Ward

Chris Biscan writes: this is a serious bad situation that could be the wake up call we need if this year see's the big plummet like 2007 but down to 3.5-3.9km2 and a record area loss.

That's a big "if". Right now, based on previous years' extent data, the expected minimum would be 4.6 million km2, with a 95% confidence interval of [4.1, 5.0]. So anything below 4.1 is quite unlikely.

Ian Allen

How can you put your trust in those 95% confidence intervals these days? We seem to live permanently outside of 2 Sigma these days.


Hi all,

This storm is probably fatal for any polar bears who were in the area.

H/t to the goons trying to prosecute Dr Charles Monnett for drawing attention to his work on this issue in such a timely manner.

Paul Klemencic

idunno, I was thinking the same thing... Time for Interior Department Secretary Salazar to get him off leave and send him to the Arctic to a do a post-storm polar bear count. Get Obama to mobilize some military aircraft, if necessary, and get out there and count the carcasses!


I am stuggeling to understand many things today. When I look att a graph of wind speeds in the Arctic I see maximum winds of 15 to 24 knots over the ice-sheet today. Acoording to this page that means moderate breeze to strong breeze. Did the storm only last one day or am I missing something?

Ned Ward

How can you put your trust in those 95% confidence intervals these days? We seem to live permanently outside of 2 Sigma these days.

Those 95% confidence intervals are based on how much the ice extent declines from Date X to the minimum.

In order to achieve a minimum under 4.1 million km2 this year, we would need to see a larger decrease than in any of the past nine years (from this date to the minimum).

Now, maybe that will happen. On the other hand, over the past month the decrease has been the second-smallest, not the largest. So it would be a remarkable turnaround if such a thing happened.

I started running this empirical prediction model back in early July. I kept the first day's predictions, and then also made a second version that updates on a daily basis using the latest data.

As of today, we're about 178,000 km2 above what the original model had predicted for this date. So from the perspective of a model I created early last month, we're not "permanently outside of 2 sigma"; we're well inside 1 sigma and on the more-ice side, not the less-ice side.

It's obviously true that there is a long-term decline in Arctic sea ice extent, and it's likely that this year's minimum will rank 2nd or 3rd out of the past decade (though 1st or 4th are still possible). It seems likely that within the next decade, both the 2007 record and the 4 million km2 mark will be broken.

When you want to see a new record set, it's easy to grasp at any indication that the ice is about to vanish. But years like 2007 (for ice) or 2005 (hurricanes) or 1998 (temperature) are by their nature exceptional. Most storms aren't the perfect storm, and most exceptional years are followed by a bunch of non-exceptional years.

My prediction for the IJIS-JAXA minimum this year? 4.6 million km2, plus or minus 0.2 million km2. I don't expect we will come near the 2007 record this year.

Kevin McKinney

Ian Allen,

Thank you for the feedback on my simple intuitive picture. Nice to know that it's not completely naive--something of a risk when doing non-quantified extrapolation across many orders of magnitude--!

Ned Ward

neven wrote: But hey, chances are we get a century break today. You know how I love those.

Well, congratulations, you got your century break:

Difference: -102813

Ian Allen

Ned, the whole confidence interval thing is based on the central limit bell curve theorem, static distribution, random varialbls, etc. and we don't have a static situation, so 95% confidence is just a dream. would you bet 20-1 against?. I'd take the other side of that bet in a flash!

Kevin O'Neill

The satellite sensors (such as AMSR-E) and associated algorithms for computing sea ice concentration have measurement uncertainty errors that approach 10% for this time of year. The areas most likely to be affected are normally limited to the ice edges - but this year, with the large slush puppy seas, that's a very large area.

Any estimate of uncertainty that doesn't take into account the basic errors of the sensors and associated algorithms is significantly incomplete by not including the largest single source of error.

The AMSR-E Validation Plan can be found here:

Paul Klemencic

Ned Ward, your reasoning using statistical trend extrapolation walks right into a major trap when the underlying system changes. When the system undergoes a major change, prior data collected on the previous system isn't as useful or accurate as you believe in forecasting the response (output) of the new system. This is a common failure mode for statisticians who don't understand the system. To quote Dr. W. Edwards Deming, who pioneered the use of statistics for improving manufacturing systems, "There is no substitute for knowledge!" and he emphasized that one should study and understand, and model, the system generating the output. He was scornful of statisticians who simply extrapolated previous system output, and didn't observe system changes right in front of their eyes.

Over the last month, we have been discussing and expecting some kind of late season impact on a seriously weakened ice pack, if and when the weather changed. We believe the late season melt could be quite different than in the past, because of a host of system changes including:
- unusual mobility of the ice pack to changes in wind, possibly partly due to relatively early loss of all major ice buttresses against land and island masses in Siberia (all effectively broken by August 12th)
- unprecedented occurrence of 30-80% ice concentrations in some widespread and large extent areas, particularly below the 80N parallel in the Beaufort, Chukchi, and E. Siberian regions
- much higher sea temperatures in the seas like the Laptev, E. Siberian, and Chukchi
- unprecedented early loss of ice pack along the Laptev, Kara regions

We have been postulating and discussing the possibility that the ice pack system has changed enough that different and usual responses could result. Among these changes, we believed we might see:
- unprecedented late season extent losses in the slush puppy seas of the Beaufort, Chukchi, and E. Siberian regions
- large swings and movements of the ice pack, pushing the pack into warm seawater, leading to unusual late season loss of ice pack along the Laptev, Kara regions, and near Svalbard
- and potentially a late season "Arctic Big Flush" of mobile ice into the Fram Strait and Greenland Sea

In our view, the only thing that preventing the ice pack system from changing its behavior significantly this year, was that the weather fronts in August haven't been positioned to drive these events, except for a few days two weeks into the month. So these events that would be caused by a different ice pack system response remained "hypothetical".

In the last several days, the first of these events seems to be unfolding; but the size and extent of the event is much faster than we expected. And the response is unprecedented compared to previous observations of late season ice pack melt. Furthermore, the weather pattern seems to indicate that all of our hypothetical events could unfold over the next ten days or so.

Clearly the Arctic ice pack system is responding differently this year; the system has changed. Given that, a statistical trend extrapolation could fail to predict what will happen over the next ten days, and to the end of the melt season.


You also get different figures depending whether you assume same drops as previous years on average or predict the drops with area or (extent-area) or extent ...

L. Hamilton

"Clearly the Arctic ice pack system is responding differently this year; the system has changed. Given that, a statistical trend extrapolation could fail to predict what will happen over the next ten days, and to the end of the melt season."

Paul, that's an excellent summary. For the past couple of months, many posts here have detailed qualitative changes in the ice cover that were not well reflected by traditional extent or area indexes, but that could allow them to behave much differently than in the past. Perhaps we're seeing the start of that now.

Ned Ward

Ian, it's entirely possible that the distribution of the decline from mid-August to the September minimum is not stationary over the 2002-2011 period. In fact, there's some evidence that the mean decline over this time period is increasing, by ca. 0.05 million km2/year.

I haven't tried to incorporate that non-stationarity in my model, because I thought it would substantially increase the complexity for relatively little improvement in the predictions.

But some preliminary messing around suggests that modifying the model to include an increased rate of late-summer ice loss would bring down the expected 2011 minimum from 4.6 to 4.4 or even 4.3 million km2, depending on how one chose to handle that. The 95% confidence intervals around that expected value would be correspondingly lower.

Would that more complicated model be justified? Hard to say. The sample size is pretty small.

Ned Ward

Don't have time to write much, but ...

I greatly distrust a lot of the "qualitative" evidence that gets tossed about here. We humans have a natural tendency to focus on things that confirm what we want to see, and ignore things that conflict with what we want to see.

Seke Rob

MASIE update, all values in kmsq exact: Chukchi -43,707 and Beaufort -24,910 moved big, out, the Central Arctic took some in +10,479 as did the East Siberian +7,717 even a little into Barents +2,168, but the net of all the melt & compaction totaled to -50,148, half the century that JAXA finished on.

Did someone actually put a suspected straw in and over-sampled, given all the exited observations from various map sources? Some other places there will be emergency polls and escape clause posts ;>)


Kevin O'Neill

For those who haven't already visited the Beaufort Gyre Exploration Project website, I highly recommend it. http://www.whoi.edu/page.do?pid=66296

For some reason I've anthropomorphised the sea ice; I view it as a living organism trying to defend itself. The BGEP pages give a more scientific explanation; likening the gyre to a flywheel. http://www.whoi.edu/page.do?pid=66596

The question is, does the ice have a response to these winds and temps?

Chris Biscan

I don't get to see what I want to see. If that was the case the arctic would be at late 70s levels.

What we have seen is the thickness in many areas change so radically that the the ice/system has changed.

Those stats you use do not in any way account for this in determining extent. Because they see in 2D when the Ice is in 3D.

And that precious thickness/volume overwhelms using extent as a measure to predict future extent.

One side followed the data/facts that have rolled in.

The other chases down faulty computer models, Then Cryosat2 came out and we all went WTF.

Now it looks like Cryosat2 was not even close probably issues with calibration and and snow.

But we have known for years as the thickness decreases it reaches a tipping point in any given area of low thickness.

for instance:

The last few days we have seen what happens when the ice only rests in the top layer of the water that is moving the fastest.

We saw Ice in the Laptev and Kara recede 75-100km over 5-6 days under a smaller shelf of winds that were 12-18kts for 3-4 days two weeks ago. I have seen countless people on multiple forums and blogs be a bit miffed at this.

Well when the ice 1.5 to 2.5 meters or more in these areas the last 10 years it would of moved half as much if that. without a long string of Gale winds. The ice there is now 0.5 to 1.5 meters at best probably 1.0 meters out there or less which let it move much faster.

that area also developed a bit area of low concentration hole once it was stopped being pushed because 4-8c temps were shoved into the ice..and under, mixed out quickly but every other day or so the ice on the edges frays and concentrations drop from 90-100% to 50-65% This would not be possible in the old arctic. Another thing many were surprised to see.


They have already passed into where the ice should be from yesterdays UB maps. You can see in front of them the ice pack thickness increase dramatically.

Which means they are closing in on it around 79N.

The ice has literally vanished in three days.

if you look through the images you see some with ice no more than a foot thick that isn't even white. It's that nasty stuff we see on Modis that is like paper thin ice in a puddle.

we saw the Fram Straight in 4-5 days loss it's ice to a blow torch and fall into chunks and this thin veil of ice.

We have seen the ice in the NW passage disintegrate. Ice that was 3-4 meters thick in places melt out this summer. Well having 4-10C water temps is kind of a big deal after open water, let alone warm water was non existent up there 20-30 years ago.

It's snowballing on it self. The Feedback is in overdrive and getting stronger. The guise of extent has masked the under lying issue of the ice thickness dying.

No I do not want to see this. And unfortunately your stats do not account for this. So if that DPA comes for the next 10-15 days it will be put to the test.

You distrust our qualitative evidence. Like real time reports of ice thickness, SSTs, and real time Sat images that we have vigorously combed through. I have read many blogs and forums and I would be shocked if many of the folks frequenting there even look at Modis. The ice has taken on a different color the last few years. Especially 2010 and 2011. All of a sudden it started to look more blue. Well that happens when we can see the water under it.

We have had report after report of thin ice. Now another one from Polarstern saying ice that the Navy has 1.5-2+ meters is 0.9 on average up to the NorthPole, some places are 1.5 meters off on models.

Cryosat2 said the ice there was 3.5 to 4.5 meters or more in spots. If that is true we have lost 30km3(in millions) no way that happened.

We have dotted our T's and crossed our I's. I have no idea how the weather will pan out, which will determine where this will end up.

but if it does pan out like the models say, your stats will leave you wondering why another 2007 took place within 5 years. Our Qualitative data will be there to show this was coming.

Nick Barnes

FWIW, I'm with Ned, more-or-less (although I'd note that 2003, 2005, and 2009 are a statistical tie with 2011 for the same loss rate over the previous month, and only 2004 and 2008 were very much faster). And anyone suggesting we've never seen ice conditions like this has a short memory. Consider the terrifying state of the ice this time last year, all around the eastern longitudes, all the way up to the pole.


As for volume: yes, volume is a key number, but we really don't have a good handle on it.


On delusion…
Ned Ward, most of us, Neven’s followers, are like me, people with an interest in nature and physics. Not trained in science, not into climatology, meteorology nor cryology. Still, we use the formidable means that are publicly accessible through the internet. To watch, as good as we can, what’s going on. The purpose is to learn and to be able to communicate the worrysome future.
Could you be more specific when you say ‘I greatly distrust the qualitative evidence’ you find tossed around in our blog? So we can be sceptic on them together.
From time to time I get a comparable sense, when I read stuff about ‘blowtorches’, ‘ toast’ or whatever tasty superlative commenters spill to underline their concern. But please don’t generalize in dismissing everything the commenters come up with...


To watch the storm in real time go to


Set the images to 24, hit play, and be amazed!

A brave lad in a very small sailboat is undoubtedly having a very difficult time at the western end of the NWP


Ian Allen

Ned, I agree with Paul K, Chris and others.
Don't forget we are watching a phase change.
The arctic used to be mostly locked, it is now growing seas where storms may roam. If and when it all goes, there's a salinity-supported thermal inversion waiting to be destroyed also.
I don't believe all this about little bits of ice remaining, there'll be nowhere to hide.


As I as wrote earlier these days seems to be "The point of no return", we will soon see the Polar sea ice evaporate overnite, it will not be this year, but very soon!

Regards Espen


I like the note of caution that Ned represents, and wish to give a few statements in support of his comments:

First, we should welcome all (reasonable) models here. And the model "The ice will fall as far from here as it has in prior years" is certainly a good model to consider.

Second, if there is a gigantic drop in sea ice extent, it is valuable to be able to say that this was very unlikely, based on historical data.

Third, his statement of "if this year continues like previous years" is very similar to statements above "if the weather forcast holds".

Fourth, if the gigantic drop doesn't happen, it is good if we don't look like a bunch of intolerant alarmists, who could then be disregarded because our predictions don't even come true.

And finally, I wouldn't want anyone to feel unwelcome, as long as they aren't being disruptive. And I don't believe I've ever seen anyone disruptive on this blog -- except possibly people who were over-enthusiastically anti-denialist.

L. Hamilton

Regarding statistical vs. other ways of knowing, I view the competition as mutually beneficial. We learn from the discrepancies of different data/analyses, as well as their agreements.

Statistical predictions, like other models, are formalized kinds of "if ... then" statements.

Ian Allen

The stats would need complicating though and any model would not be linear, so confidence intervals are hard to assign. I've seen a quadratically disappearing volume prediction on this blog for various months which seemed to fit very well and to belie most other models.

Chris Biscan

12z GFS is out and has the DPA still. The ice pack is really going to move around the next 10 days.

1. The Fram straight is open for business. You can see this today on the prelim for UB. This flow will max out by hour 30. And diminish by hour 60. This is a mini event for that edge of the ice pack. ice will get pushed out and melted some over this time. The waters up against the pack here are very warm, so there will be lots of melting on this end of the ice.

2. The winds in the Beaufort Have turned out the SE, then ESE which is diverging the ice pack. But this is not good because the next 7 days will see 5-10C 850 temp anomalies with a blend of the Euro and GFS. Not only will there be no ice production. Outside of melting there will be nothing to cool the Warm SSTs there. In fact if the Euro is right the water in the CA Islands and Beaufort will warm the more closer to the shore. Which means this diverging ice will expand the extent possibly for a few days there but quickly melt out to ice flows or all the way gone with a warm wind coming over it. you can see the ice flare out on the prelim today but also lose a ton of concentration.

3. The East Siberian Sea is about to lose all of it's ice. A week ago it looked like it would be a possible spot for refreeze. 850s will be colder there on the order of -2 to -6 pending the day, but the winds and SSTs will compact/melt the ice out. by day 7 most of the two arms of ice will be gone. This will offset any extent gains in the Beaufort from divergence and then some.

4. The Kara and Laptev will see winds compact the ice towards the east south east for a few days then more east which is towards the North atlantic. While there will be ice moving towards the barrents sea the problem is the water is warm there and will melt it even under -4 to -8 850s and the mean flow will be to head towards the Greenland sea and north atlantic. Even if the backing winds do not come at a perfect direction, the ice by the greenland sea and barrents will be heading towards the fram and that will leave a hole. You can see that hole blow up again over the Kara or Laptev. This will get filled with ice from Siberia.

5. the Ice int the NW Territories will likely be nothing but ice flows by day 6-7. This will be flat out incredible.

Chris Biscan

I am not intolerant to anyone.

I just want to see Weather/Climate Data.

yes Analogs are helpful. However this is also real life. And in real life things change. We blend the two. We know the ice won't melt out completely because the sun is setting up there.

He said his model shows the ice is about 175,000km2 lower than it predicted a month ago.

That is great. But does the model account for diverging winds over the ice while the arctic baked?

the model doesn't account for the fact that we are 175km2 lower then it predicted and for the most part winds have been unfavorable for melt.

Does the Model account for the warm water? While the model predicted the ice extent within 175Km2 a month out. I don't think it's aware of the historically warm water sitting around the arctic over millions of miles.

Does the model account for winds? While the model predicted the ice extent within 175km2 a month out. Did it know the winds would be less than stellar for sea ice loss during that month?

Now the winds are changing to favor ice compaction/flushing. temps are to warm for the warm ssts of the Russian waters to freeze over. The temps on the Canadian side are 6-10C above normal with literally no ice left to cool them.

I am not saying the arctic will see a new record low extent.

but even if it doesn't we have buoys, ships, and modis telling us the ice thickness is at it's worse. Many areas are almost gone.

So yes we know how volume is doing, its doing horrible. Decimated. We are at the bitter ends.

Does this mean it will tip over now? No.

What it means is that on August 22nd 2011 the state of the Cryosphere is in shambles.

And the delicate balance on Earth is 4000-5000km3 from being hurt, hurt bad.


You're an optimist, Chris. I guess it's 3950 already. Now please let attention be shared. You're not melting, I assume...


Guys, great comments. I'm going to bed now, so don't overdo it! We still have approximately three weeks to go.

Paul Klemencic

Bremen map review:

I did my twice-daily (prelim + final) Bremen map blink comparison, and the edge of the ice in Beaufort and Chukchi did spread back in some, likely resulting in a slight increase in extent.... (Caution: prelim map; changes are common in this area in the final map.) Also some ice 'reappeared' in the lower latitude Beaufort, albeit at low concentrations.

But a lot of ice disappeared for good yesterday, particularly in the quadrangle 150-165W, between 75N and 80N. The USS Healy is in that quadrangle today (great timing), and the pictures linked to the 'Flash Melt' thread are terrific. In those pictures, the ice pack in the area of 78 deg 35' along the 163W, is almost gone. The Bremen map shows 75% to 90+% ice concentration there. Clearly the pack is still receding quickly.

On the other side of the pack, there was some substantial flow toward the Fram for the first time in about a week, and the pack continued to shift toward the Kara and Laptev seas. The 'hole' in the ice pack in the 105E to 135E quadrangles contracted some yesterday, but opened back up today and spread into an adjoining quad.

Overall this won't be a big extent loss day, unless further loss in the Chukchi region shown by the Healy photos shows up in the final map; or if the storm damage in the E. Siberian is pretty bad. In the preliminary, the most important part of the E. Siberian isn't visible.



I'm confused about your comments on the Arctic Dipole Anomaly.

I thought that this phenomenon, which was a prominent weather feature back in 2007, was like what is described in this Wikipedia link; http://en.wikipedia.org/wiki/Arctic_dipole_anomaly

ie. high pressure north of Alaska and low pressure off of eastern Siberia.


A 'classic' dipole anomaly occurred earlier this summer.(The shift to a weather pattern dominated by low pressure north of Alaska is why I changed my sea ice extent guess; which a forum member asked about a while back).

That said, I am inclined to agree that what we are observing now with the Arctic Sea Ice is a fundamental change that has not been seen before. Thus the explanations of the 2007 melt are useful to understand what happened back then, but the subsequent reduction in sea ice volume has changed the rules of the game.

Chris Biscan


looks like it to me.



I was looking at GFS 12z


The next 5 days here don't seem to show any conspicuous weather pattern.

Chris Biscan

H5 is like 28K feet in the air.

Explain to me how that effects the Ice?



you wrote "H5 is like 28K feet in the air.

Explain to me how that effects the Ice?"

You lost me bro.


The link you posted showed 500MB Atmospheric Heights and Absolute Vorticity.

500mb is roughly 28,800 ft above the surface.

This map in terms of the arctic can be useful to show us where
Areas of High Pressure and Low Pressure will set up. It also is an indicator of weather and temperature to some extent. High pressure at that level usually indicates weather will be mote prone to calm and clear. And warmer air. Low pressure is the opppsite.

However as we just saw we had 528-540 heights over the Beafort. However the surface slp bombed out and went poleward. This pulled warm air into the arcticwhile under the H5 low pressure.

We soley care about SLP and 2m winds or 10m winds if 2m is not available. Also we can use 850mb(5000) feet temps to decifer the surface temps most of the time. Theyre mean flow is ussualyin sync.

We also have historical data to know how cold 850s


Need to be to create ice. Recently they have failed under -12c this time of year. Where as a decade ago -6 to -8C would start to make it happen. Now dont get the wrong idea 850mbtemps are an indirect measure.


Extent down 58,593 on the 23rd (IJIS prelim)

L. Hamilton

JAXA prelim 5,115,313, down about 58k.


major kickback from yesterday but another big loss.

A lot of that ice in the beaufort will go byebye the next 2-3 days as warm winds push 3-5c 850s over it and warm waters into it and the ice into them.

The fraying helped skew the impact. That ice is junk


IJIS area just took a big drop and is now close 2007 levels.


Steve Bloom

Ned, I think expect is a more appropriate word than want.

Chris Biscan


That is winds 3 days out..those winds slide west through day 4 and 5 to about the NP, slightly East of that.

and strengthen a bit.

I would expect a smaller extent loss tomorrow because the wind in the East Siberian Sea will not turn favorable for compaction until hour 12-18. So there won't be to much time for it.

However after that, its game over. The ice out there may not be as bad off as the Beaufort but the winds will be there for 7 days at least. The Beaufort will have Southerly winds pushing the ice north but that is into warmer waters and only for 2-3 days before they get weaker or calm.

the models still show huge 850 temps anomalies in the Beaufort/Chukchi. this is not far reaching to the ice but will keep the torch up there to continue to supply the shallower water there with warm water.

The GFS is missing a SLP East of the Greenland sea for the complete dipole Anomaly, the Euro has a stronger feature there.

Either way there will be winds From the Bearing Straight to the Greenland Sea booming across the arctic. There will be massive compaction basically along and towards the Russian side of 180E/0W and flushing down.

I would expect 1-2 days of average extent losses then we are off to the races. Of course there will be push and pull but with winds blowing over the weakest areas of the ice. This could only be worse with a SLP East of Greenland during it. If that forms the flow will be out the Fram at a sick pace.

Ned Ward

A few other things to consider:

1. Remember that the ice area and concentration maps are less accurate than extent, for reasons that have been discussed before. One consequence of this is that from day to day, large areas of ice appear to disappear and reappear. If one is expecting that the sea ice is poised to collapse any day now, it's easy to see an apparently dramatic drop in concentration or area from one day to the next, and assume that it's the beginning of the end.

2. I would similarly caution people against placing too much confidence in models of volume. Volume is very difficult to measure. Eventually, we'll have pretty good Arctic-wide volume measurements from Cryosat (there are probably a bunch of presentations on this being prepared for AGU), and maybe longer-term changes in volume can be inferred by comparing Cryosat observations to the much sparser and less accurate IceSAT GLAS data from 2003 onwards. But in the meantime I personally would be cautious about making arguments based on models of volume.

3. The simplest and most conservative assumption is that things are proceeding about as expected based on long-term trends with weather "noise" superimposed. So far, this year's extent data are pretty much exactly what would be expected based on previous years. People are free to make the claim that the system is somehow radically different now, and that dramatic changes are just around the corner. But I haven't seen any indication of that yet in the extent data. We are still right on track for 2011 to come in 2nd or 3rd out of the past ten years, with a smaller chance of coming in 4th or 1st. There is no indication yet that sea ice extent is going to suddenly fall, say, a million km2 below last year's minimum.

4. Remember that we are supposed to be the skeptics here, in the genuine meaning of that word. The commenters at WUWT and Goddard's place are mostly not skeptics, they're true believers who are more interested in finding data or lines of reasoning to support their pre-existing beliefs than in understanding reality. I cringe every time I read a comment here about the ice being "blowtorched" or "hammered" in the midst of a week when the decline in sea ice extent is basically normal for this time of year.

That said, I enjoy visiting this blog and I'm glad to see so many people taking such an intense interest in the fate of the Arctic sea ice.

Paul Van Egmond

@Ned Ward - Thanks for providing some perspective on the current state of affairs in the Arctic.

Steve Bloom

People do get impatient watching ice melt, perhaps unsurprisingly.


Or paint drying?

I've said it before. Don't count chickens.


Hi Ned

I appreciate your opinion about the environmental difference between science and "normal" people just interested in a subject like the Arctic Sea, I believe a blog like Nevens is important because its not fund oriented and thereby unfiltered, and in the case of the arctic I am not so sure the science got it all right, and in some cases under heavy pressure for not letting sensitive figures or knowledge come out!

Regards Espen


Hi Ned

Thanks for your comments. I agree with you 100 % !



I gotta be honest with you. You built a mice strawman. I am not sure what a believer is. But I do know what Data and real time observation is. I do know we have thousands of sea ice thickness mesurements from buoys and ships every year now. We have real time arctic temperature profiles telling us how much the water has warmed under neath the ice from buoys since the 1980s. We have accurate verified surface temps from satellites verified against real time Data with small margins of error.
We have a high res satellite called modis that we can verify the amsre data.

I can turn the wheels in my brain on and logically put this data together and there is only one conclusion. And it's that most of the long term old ice is gone and summers are now down to scant thin ice over most of the arctic.


Ice doesn't seem to appear or disappear dramatically day to day on sea ice concentration maps.


First of all, while volume measurements are less accurate, it isn't exactly throwing darts at a wall.

But lets look only at extent and area- there is now a 30 year trend that shows the arctic will be seasonally ice free in 20-30 years.

Since just a few years ago some scientists were predicting that the ice would last 100+ years, this is a pretty extreme change.

While predicting the exact year is impossible, I think it is quite reasonable to publicize the fact that the arctic is melting way faster than predicted in earlier climate models.

At this point, all the statistics are on the side of the "alarmists".
But let's suppose that you now want to reject the statistical argument since it clearly doesn't show what certain groups would want it to.
Do you have any scientific reasons why the 30 year trend is likely to reverse itself in the near future?

Paul Klemencic

Ned, I think you are suggesting we are making a Type 1 error; which basically means that we are looking at the system data, and seeing a 'special cause' due to a change in the system, whereas the system is behaving as expected due to random variation. By looking for a special cause, we are missing the real story, and wasting our time. You feel we should be able to predict the loss in extent from late July to the end of the season, using a statistical normal distribution based on data from recent years. And you claims the data over the last 30 days support your view, and until two days ago, you were correct.

On the other hand, we suggest, that if we broaden our scope and look at data other than ice extent (ice volume, ice concentration distribution, ice area, and effect of weather events), this data point to a system change. The system is changing because of decreased ice volume, more low concentration ice in the extent, warm seas surrounding the pack, and increased ice pack mobility. And this system change could lead to unprecedented late season ice loss. Ignoring this possibility is making a Type 2 error; that is they attribute the system response to random variation (common causes, such year to year weather variations), when in a fact a significant system change has occurred (special cause). A special cause can cause unpredictable outcomes.

The statistical trend guys (including Tamino at Open Mind, Lucia at The Blackboard, and others like you) have the 2011 minimum extent estimated at about 4.6 million sq km. Two sigma 95% confidence level (assuming a normal distribution, which we have no reason to expect) lie at around 4.1 and 5.1 million.

I suggest by looking at the special causes, that breaking 4.6 million is over 90% certain, and the odds of taking out the 4.25 million low extent record set in 2007 is greater than 50%, and there is a significant change (over 20%) that we end up with less than 4.1 million (if the right weather combo hits, which looks to be at least a one-in-five chance). I think there isn't any solid reasoning to expect a normal distribution of expected outcomes, versus a skewed distribution.

BTW, the cost of making a Type 1 error is usually significant, but 'manageable'. The cost of making a Type 2 error can 'kill you'.

Nick Barnes

I suggest by looking at the special causes, that breaking 4.6 million is over 90% certain, and the odds of taking out the 4.25 million low extent record set in 2007 is greater than 50%, and there is a significant change (over 20%) that we end up with less than 4.1 million (if the right weather combo hits, which looks to be at least a one-in-five chance).
Paul, I'll bet you 50 euros, at even odds, that the IJIS extent minimum will be above 4.3 million square kilometres.

Nick Barnes

Do you have any scientific reasons why the 30 year trend is likely to reverse itself in the near future?
D, assuming your comments are aimed at Ned, he is very clear that he bases his opinions on the trends. His criticism is of those who make speculative leaps beyond the trends.

Paul Klemencic

Ned, the even up bet you propose is too close the odds I suggest to give me a reason to bet (the outcome would be random). The place to bet is where the different POV lead to a substantial odds difference;

So how about I give you 2:1 odds on the IJIS breaking below 4.60 million? If you believe Ned, you'll take the bet, you have a 50% chance (in your view), but make 2:1. I will take the bet because I think its 90% probable that I win.

So how about that? Put your money up?

Paul Klemencic

Sorry Nick, I got Ned on the brain. So want to take the bet?

The comments to this entry are closed.