« An observer's prediction | Main | New poll is up »


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


Hi Neven,

OTOH, 2011 CT area is still significantly below 2007, and falling fast still.

Cross comparing the CT maps for the last available date - currently 28 July - there seems to be a much greater area of thin ice this year, which could yet melt out before September. By this date, most of the Central pack which melted in 2007 had already disappeared. What remained looked very soild, and survived.

The NE Passage and the Canadian Archipelago are also much more open this year.

Also worth noting that 2008 SIE seems to have decreased more through August and September through to the minimum than 2007 SIE.

Finally, I think that the main figure outstanding before we write off 2011's chances is the PIOMAS volume update. If that is still running at the same anomaly, lots more could happen yet.


I agree with you, idunno. But we'll know much more around August 10th.


CT area drop of 116,000 just in!

Chris Biscan

with that SLP still sitting there...that new blue area on CT concentration map will be compacted/melted into the main pack in a day or two with the steering currents that way.

Chris Biscan

even without and updated pips.

look at the new UB concentration map from July 31st east of Greenland.

Has that area ever gone ice free?

Chris Biscan

I am worried the main ice pack is going to get shifted to much towards greenland.

the ice north of Alaska is getting ripped up by the storm there.

that ice will compact/melt inwards as the entire pack keeps getting shoved towards Russia into warm SSTS.

this may not be as "steady" as some are saying.

Lord Soth

I figure we could go as low as 400K behind 2007 by Aug 10 and still make up the ground.

Remember the ice has already melted, it just needs to be compacted.

Paul Klemencic

I think the last ten days really has two stories. Yes, the weather has slowed the top melt from solar radiation and air (wind) convective heat transfer to the ice pack. The slowdown in ice extent has resulted from these factors and the spreading of the ice pack toward the Beaufort and East Siberian seas, which prevented even bigger extent losses (although the area losses in the pack in those areas were quite large).

The other story has been the resilience of the ice pack boundary bordering the Laptev, Kara, and Barents seas. In spite of ice pack movement over the warm waters in those areas, with accompanying bottom melt, the boundary of the ice pack didn't recede. I expected the ice pack would lose extent in these regions, but it clearly didn't lose much. But the edge regions of the pack in these regions have moved back and forth above some very warm water. Clearly, the ice pack must be sustaining some significant melt from underneath.

So if the weather systems return to a more neutral impact on ice melt, these regions of the pack should be susceptible to extent loss, although with some replenishment as new ice moves in. The pack out toward the Chukchi, Beaufort, and E. Siberian seas between 140 E to 130 W will eventually lose a lot of extent, as it usually does. The key to setting a new low, is melt into the pack above 80N from the direction of the Laptev and Kara seas. If the pack has been substantially weakened there, and can't continue to hold the boundary, as it has over the last ten days, then 2011 will lose a lot of ice extent over 80N. This result will be much different than 2007, even if the extent doesn't break the low.

What happens to the pack between the Greenwich meridian and 140E will tell the story of 2011.

Lord Soth

Century breaks for August are all over the board.

2008 had 5, 2007 had 4, 2010 had 1 and 2009 had none.

Due to the low CAPIE value, there is great potential for century breaks, if the weather pattern starts favoring compression.

However, there is no guarantee that the weather will not cooperate.

2011 may be remembered as the year of the near miss.

Kevin McKinney

Thanks for the context, all.

To anthropomorphize, 2011 spent a long time in 'slow but steady' mode, forging a lead that looked quite impressive; but sea ice (meaning, really, Arctic weather) is fickle, and now we're once again in a posture of uncertainty.

Well, it does make things interesting.

Russell McKane

Another thought in the melting pot.
The importance of daily rate of extent decline.
Let us assume that the majority of melt is at the edge (in terms of extent it can really only be this) So when we are looking at decline rates it is a percentage of change over area as deliniated by extent method. But it is happening at the edge.Now if we model this on a shinking circle. Draw an arch with the centre somewhere near the Ward Hunt Ice shelf extending to all points of arctic shore lines before the melt begins. Contracting in as we move toward september. So early in the melt using the circumference of a circle the line is long and the edge of melt is great so therefore the cumulative melt per day is a function of the length ot the line. Now as we progress toward the minimum this line reduces in length significantly having an obverse affect on the rate of daily decline. (less to melt) so to maintain the decline rate the melt must increase rapidly along the radius line. What does this tell us about rate of decline?
Any thoughts?

Seke Rob

Curious numbers: Change from July 21 to July 31 per...

JAXA: -475000
MASIE: -868000

Cooks using different recipes for sure.


As I earlier wrote, new 2011 records are possible, maybe not the extend figures, which I dont find that interesting compared to volume figures, because I believe more in the "overnite" disappearance of ice in the polar sea. The new record 2011 could be +/- 45% of the ice mid-september will be within the 80 parallel and the rest within the 75 parallel, I dont know what your guys think?


L. Hamilton

FWIW, both CT area and UB extent still have 2011 and 2007 very close (within +/- 20k), although IJIS shows them diverging.

That charismatic chunk of shore ice off NE Greenland broke apart late last summer but then largely re-froze in place. It looks stuck-together and fragile this year,


That chunk of ice, will soon be gone more or less, you can already see how thin the ice is below the glacier!



Does anyone know if this ice is permenant or melts every year, its in NE Greenland



Dormolin, does this answer your question



Sorry about that dorlomin, trying to post


Chris Biscan

the euro and gfs are both on board on a quick warmup and a flip to the Dipole Anomaly.

that low concentration ice is going to go fast.

after being burned the last few days on predictions....I still think by August 3rd or 4th we will see the extent drop quickly.


dorlomin: by a strange coincidence I just mentioned that area of landfast ice next to the NEW polyna in my latest blog:

Images of the NEW area are posted here regularly:

Arctic watchers may be interested in this site - especially the pdf of The Draft Summary iAOOS 2010 report ( top right of home page ):

Andrew Xnn


Are you asking about Flade Isblink or the
nearby sea ice?


Flade Isblink is a small icecap in northeastern Greenland. During the summer 2006 we will drill through the ice cap in order to test a new drill fluid, which may be used in a coming drill fluid. The ice core is expected to reveal interesting climate information from this scarcely investigated part of Greenland. It is not known whether the ice cap contains ice age ice, or whether it meltet away during the climatic optimum 8,000 to 4,000 years ago. The ice cap faces directly towards the polar ocean, and the core will hopefully give information on the climate history of the Polar Ocean and the sea ice


FWIW, both CT area and UB extent still have 2011 and 2007 very close (within +/- 20k), although IJIS shows them diverging.

I am also wondering whether all is right and well with IJIS. We had another 30k ish decrease for August 1 following a string of similar decreases for the previous 4 or 5 days. However, Masie is reporting an extent decrease of 190,510 for day 212 ( July 31), CT has an area decrease of 272,000 over the last 5 days ( up to July 31). Even the DMI graph shows a sharp downtick for Aug 1st.
Arrows on the PIPS graph are not that great, but they have been worse in 2010 and we did not get such a long stretch of 25- 30k decreases....

Looks odd to me?

Alan Wallcraft

I know all I do here is point to ACNFS, but its Archive of ice speed and drift includes a 5 day forecast, with NOGAPS atmospheric forcing, every day. The Aug 1st run isn't posted yet, so the last snapshot there now is 00Z (midnight UTC) on the 6th of August (from the July 31 run). The arrow tails are following streamlines.

This may not be telling you anything new over, say, the ECMWF pressure forecast. Except that it is showing sea ice directly.


The last trade on intrade's 2011 minimum IJIS ice extent higher than 2007 contract was at 93. This implies there is only a 7% chance of a record this year.

I think there is more chance of a record than that and to a certain extent the market agrees with the range now 76-84 implying 16% to 24% chance of a record. The contract price hadn't been above 74 until recently.

Even with the price moving back down again, is there too much reliance by market participants on the IJIS record and not enough consideration of area, volume etc.? Does that make this a good speculative opportunity to sell the contact?

(If unclear about implications of selling a contract at 76, it would mean risking $2.40 with a potential $7.60 profit.)

Chris Biscan

UB took a dive.

CMI(30% went even again..probably some compaction going on.

the 00z euro would end the string of low extent days in a big way.

Rob Dekker

Crandles, take it !
We all know that 2011 will have a few tricks up it's sleeve, and by no means is 93 a fair assessment of the probability.


>"Crandles, take it !"

At 93, sure, I would jump at it. But at 76 it is considerably less clear cut - For the moment I am putting in book order to sell at 88 in case anyone is mad enough to buy at that price rather than push the price down below 76. After all someone has been mad enough to buy at 84 and 93.


Thanks for the info, intersting to watch. Especially Andrew XNNs.


NSIDC monthly average data:

2007 7 Goddard N 8.13 5.03
2008 7 PRELIM N 9.06 5.70
2009 7 NRTSI-G N 8.82 5.71
2010 7 NRTSI-G N 8.39 5.21
2011 7 NRTSI-G N 7.92 4.99

So first time below 8 and first time below 5. Though they are only record lows by small margins.

L. Hamilton

CT area for 2011 is slightly below 2007, UB extent slightly above. These two time series still don't show much divergence between the two years.

Daily extent decrease probably won't stay this low for long,

Looks like I was wrong about this. IJIS has revised yesterday's extent decrease so much that only 18.5K remains. That's abysmal.

2011 is already 218K behind 2007, which hasn't even started its series of century breaks yet!

I really don't think we'll see a new record. First things have to switch again, then they have to stay that way for the engine to get going, but by the time it has, things are starting to refreeze again.

I still think 2011 can stay in front of 2010, and perhaps even 2008. But we can wave goodbye to 2007.

Just before posting this comment I checked ECMWF one last time and it doesn't look as if things are going to switch. We might see a lot of MYI transport through the CA though. :-)


Seke Rob asked me to move this comment of his here:

When there's all this confusion in the heat of putting hands on the latest greatest, I revert to NSIDC monthlies. Just out and combined for Area and Extent to visualize the slosshy content of extent, concentration monthly mean 66.92%, not a record, but Area did have one at a mean of 5.3 million km square [including the black hole they can't see the ice in of 0.31m]:

Arctic July 2011 since 1979

and for those looking at the whole world:

Global July 2011 since 1979

Global did have a new low July concentration of 72.3% and both Extent and Area set a new low month-of year record. The previous low concentration for Earth in modern times stood at 73.3% for July 2008.


Lord Soth

I wouldn't throw in the towel yet Neven.

I'm seeing more and more open water in the main pack, and all we need is some compression between now and the middle of September. The AO has gone to -1 and the lows are beginning to fill.

The Dipole is still not favorable however, but this may change.

Actually the dispersion of the ice will help in in the next few weeks, as it will expose greater surface area to melting.

Also, does anybody know, how big a polynya must get before it gets included in either the NSIDC or IJIS sea ice extent ? That polynya in NE Greenland is growing, and I wonder if it is being counted for extent purposes.

I did some quick visualization of the NSIDC graph (area below vs above the curves defining 2007 and 2011 extent), and 2011 is definitely a winner for lowest NSIDC extent for July.

Chris Biscan

I tried to copy and paste my post from another site.

it says we can not accept this data.


Chris Biscan

the last thing we need this year is favorable conditions.

Even without a major DA, the extent losses are about to pick up quite a bit.

Sea Ice Area is still 300,000K below 2010. It takes 2010 another 7 days to get to where 2011 is now. It also takes 2010 another 15 days to go under 4.0km2. Unless 2011 stops right now it will shatter that. It is nice that the extent has been slowed from some amazing divergence of the ice at this point in the season. But the ice pack itself has still taken a beating. Now the pattern is changing for more favorable melt/compaction with very warm temps moving into the arctic. And the SSTs are already very warm. Do we have historic SST data for this date in history?

Other Sea Ice Area Recent years on this date:
2009: +700Km2
2008: +500km2
2007: +100Km2
2006: +850Km2(we are only 100K from passing 2006 min Sea Ice Area That is a radical shift in arctic sea ice behavior. Maybe it wouldn't be so noticeable if we weren't also that far ahead of years that plummet passed 2006 and prior. It's on August 2nd and we are almost under 4.0km2. That just tells us how much open Ocean there is. And how much water is being warmed right now by a sun that is still potent this month.

Three major differences:

1996: We are 2,600,000 Million km2 lower than this date in 1996 in sea ice area..
1983: 2,850,000km2 lower
2001: 1,545,00km2 lower

Chris Biscan


Uh Oh

Paul Klemencic

Apparently the slight high pressure system has moved over Greenland far enough for the export of ice into the Fram Strait to start again. Comparing the Bremen images from yesterday with today: there has been a significant movement of ice through the Fram, as well as toward Svalbard and the Olga Strait. With the movement of the ice pack in that direction, you can also see the edge of the ice pack receding as warm waters from the Kara and Laptev seas eat into the pack. The pack has spread out toward the East Siberian sea over the last ten days, and today, the image shows a lot of polynas opening up in this area. Clearly the pack is moving differently over the last day compared to the last week or so.

As I have commented before, the really interesting Arctic ice story for 2011 will be what happens to the central Arctic Basin pack above 80N between the Greenwich meridian and 140E, the region adjoining the Kara and Laptev seas and ice exiting through the Fram and Olga straits. This area of the pack has been subjected to warm water for a long time (since late June and first week of July). With neutral or favorable wind and weather, the export of ice toward the Fram, coupled with warm water incursion from the Laptev and Kara, could slice and dice the central Arctic pack in this region. Normally, the pack would shift and fill in from the direction of the East Siberian sea, but this year, the ice pack there has been spread out and looks broken up. We may not see the normal filling in, so the region I am observing could see some unprecedented loss of extent. There is a slight possibility that open water could intrude into the pack above 85N in that region, something that hasn't happened in a very long time.

I tried to copy and paste my post from another site.

it says we can not accept this data.


We can only accept original content. ;-)

No, it's some quirk from Typepad. If you refresh your browser, data should be accepted.


CT daily area is 4.186 a drop of 85,000.

Seke Rob

Did not see one over at MASIE's that pulled all 16 aquatic region curves into one chart, so did one to get a single view of which areas are the main contributors in the extent decline... and those holding ground, somewhat. The Central Arctic is one not blinking, yet.

Did not start catching them until early July, and with only 29 days on-line, it's till now a short dataset. Any tip off where the past data from 2007 onwards can be had is much appreciated.

Paul Klemencic

Further to my previous comment:
Here's another way to look at sea ice extent comparing 2011 with 2007: the Cryosphere Today daily map comparison at the end of July for each year. Zoom in several times with your browser, and then assume all the green,yellow and orange areas (less than 50% concentration) in each image melt out by the minimum. Then look at the red areas (less than 75%) and assume 50-60% of those areas melt out. Clearly 2011 is still in a position to melt off to a lower minimum.

Secondly, look at how the 2007 ice pack was still firmly buttressed to Severnaya Zemlya and the central Siberian coast, whereas the 2011 pack is only slightly buttressed against Severnaya Zemlya. The pack in 2011 is much better positioned and weakened to see higher transport of ice through the Fram.

I still think seeing a record low minimum extent this year is a 50/50 proposition. But even if we don't, the pattern is completely different than 2007, where the pack melted back above 80N between 150E to 180E. Given the amount of ice below that latitude in that region, there is no way that will happen this year. But the other side of the central pack could see an extensive loss of extent this year.

I wouldn't throw in the towel yet Neven.

I'm not throwing in the towel. I'm taking a shower later on. ;-)

Like I say in this update: "I personally believe 2011 still has a lot of untapped potential, but it will have to assert itself in the coming weeks, as the window for extent decrease is slowly, but inexorably starting to close."

But I wasn't aware of ice moving towards Fram again. Thanks for the info, Paul!


Being watching ice come and go for years near the lake where I live, and to be honest I think it behaves very much like the polar sea.
Sometimes we have wind, other years a lot of rain or sun. So the extend discussions in my view is not that interesting, what is more interesting what do we have left (volume)!
My claim is the ice will be gone before you even thought about, none of us really have figures telling about the volume, to me it sounds ridiculous, because it must very easy to get this information from numerous expeditions and money spend on polar science!

I can only say ",
Houston we have a problem"!


Seke Rob,

I have masie data from 2010339 and have not seen older data being available though it might be somewhere.

I have a couple of short gaps - I am missing 2011004 to 2011008 inclusive. I don't think data was ever available for 2011088 to 2011092

Want me to put this in a googledocs spreadsheet?

Anyone got a longer or more complete record?

Paul Klemencic

OK, any help from people who understand weather systems. Right now an interesting pattern seems to be setting up. One high is sitting over Greenland, with a low over the Chuckchi near the Bering Strait, and another high sitting in Siberia south of the New Siberian islands. With this setup, I would guess we have some pretty good wind patterns bringing warm air into the Arctic from both eastern and central Canada, and from central Siberia, with the wind then moving toward far eastern Siberia. This pattern looks like one of the most effective patterns to pull hot air up into the Arctic, since it is importing warm air from both Canada and central Siberia.

Am I reading and interpreting the maps (see Neven's Daily Graph page) correctly?

Chris Biscan

yes Paul, the pattern is changing towards one that will aid in melting but also compact/melt lot of the low concentration ice.

we will start seeing much greater extent loses.

Bob Wallace

A rank amateur question, I'm afraid...

Looking at the Cryosphere Today comparison Paul links, this year has a lot more 50%-60% concentrated ice than did 2007. Does that not mean that there are lots of pieces of ice floating around with exposed water surrounding them?

If so, does that mean that there is a lot more 'ice edge' exposed to sun-warmed water?


IJIS (pre-revision) down 9.1k! 2011 has really hit the brakes!


IJIS (pre-revision) down 9.1k! 2011 has really hit the brakes!

Probably looking at an INCREASE in extent after revision....

According to IJIS, this has never happened in the period July 1- August 15. The only time it was close was on 8/8/2008 which recorded a decrease of just -625. But this was in between two largish decreases so it could have been the result of an adjustment.
Either IJIS got it wrong or something is very very strange up in the Arctic


"Something is very, very strange up in the arctic..."
I think the pack is near the moment of sudden melt out. If PIOMAS is right, combining volume, CT area and visual MODIS images leads me to that proposal.
Extent isn’t a reliable feature of the real state of the Arctic sea ice. It hasn’t been since 2010, when the complete picture was already worse than that in 2007.
I remember we have been discussing how thin the pack could get while still covering large extent. Well, what we see now is an illustration. Relatively warm surface waters usuallly cause an extra 2500-3000 km³ volume loss during august and september. Bob Wallace seems to be on a comparable track; the spread out, thin pack is indeed very vulnerable now. I still think 3.8 MK is possible, but the real thing may be less than 3000 km³ volume in september.
Once a special weather sit strikes, it is over. Maybe not this year, but 2012 could be historical.

Seke Rob

@crandles | August 02, 2011 at 23:45

Yes please, whatever you go of this MASIE series, either in original ASCII or Google spreadsheet. The 5 day gap will just be bridged by the software.


PS, The concentration, or what I've dubbed the "break-up index" is going past anything... 63.81% taking the matching date (Jul.31) of JAXA against latest of Atmos/CT. There's a Hans Klok act in the making ;>)

Seke Rob

On Atmos/CT, happened to take a snapshot of their Arctic image on July 31, 2010 and did same for July 31, 2011. Comparing that, the central Arctic did not look that dandy either in 2010. Spot the other differences.


DMI extent (30%) showing a big UPtick for August 3rd!


Masie data:


Chris Biscan

00z Euro and GFS suggest rapid melt coming the next 10 days.

very warm air moving in, with high heights, and the flow to flush ice out the fram straight.


re DMI uptic and ijis flattening, sorry not a very scientific comment but I have this picture of snow and ice in water. When it all gets very mushy it spreads out? and then suddenly goes completely.


We are now in the same area of speculation as last year. Sure, if the melting season would end two months later, a lot of that low concentration ice - probably all of it - would melt out. But in a couple of weeks almost everything will be dependent on weather patterns, as temperatures drop. Things could start refreezing again, before all the weak ice melts out completely.

So, of course, the big question is: are large portions of the ice thin enough to melt out in situ, without needing transport to lower latitudes?

At a certain point during a melting season the ice will not care one bit what the weather does and just disappear. But I'm not sure this melting season is it.

Of course, all that brown, mushy looking ice in the Beaufort and East Siberian Seas is very interesting in this respect. Some of that ice in the Beaufort is multiyear ice, but this year is markedly different from last year in that now we don't have The Arm.

With those highs moving in across the central Arctic, we might get a good look through LANCE-MODIS. Let's hope so.


When ice shelves collapse they certainly have a brief period when the extent increases significantly. However, I am not sure but kind of doubt that sea ice can do the same thing.

A ice shelf can be stressed to have lots of near vertical fractures such that width between fractures is less than height.

I just doubt you could get enough rivulets from melt water to be close enough together to have that effect. Also I would suggest ponding is likely to cover larger area than rivulets.

But I am certainly not an expert.


I suppose ridges could break off and cover more extent and area but be thinner. Such reduced thickness might not necessarily melt out (or it might melt out). However it would also imply area elsewhere would have to be plummeting to avoid showing similar slowdown in area.

This would have to be happening in a lot of places which makes it seem unlikely that this is the main effect happening all now rather than slowly.

Seke Rob

With [Top hat tip] for crandles' MASIE data share, here the previously posted spaghetti chart, but now back to just before the extent peak of March 8, 2011 [same as JAXA].

Interesting to observe which seas / bays start declining when... Hudson, a near fully enclosed area holding firm longer and then free-falling. Baffin even surpassing in extent acreage, underlining the sheer size of that corner as a contributor. Chukchi having a rather linear decline and the Baltic, though showing records [Read the news on that about stuck cruise ships], rapidly going to zero. Now to match the curve colors to the CT background image, so the geography challenged [not Tim Ball] know where to place the melt... ;>)

Lord Soth

I don't think we have to worry about air temperature for the month of August.

It's the water temperture that really counts, and its going to take a good while for this water to cool down.

The AO has gone approximately -1.4. So things are getting really whacky.

DMI is confirming what IJIS is showing, so I don't think its a math problem, either

We may very well see our first extent increase for the first week of August. We will know in a few hours.

I would really like to see the PIOMAS update, and will one of those expeditions in the high arctic, please make a slight detour and upright web cam #2. I really miss that webcam :)

Seke Rob

The NSIDC Monthly is out: http://nsidc.org/arcticseaicenews/index.html... Record Low!


Thanks, Seke Rob. Post will be up in a minute.

Lord Soth

NSIDC July report is out.



July is record low for average sea ice extent.

Oldest thickest ice is on the decline.

Ice in Beufort and Laptev sea has melted back to a tongue of older thicker ice (See Figure 4)

I will leave it to Neven for interpretation of this important fact.

Bob Wallace

If widely distributed ice is more vulnerable to edge melt is it also more vulnerable to warm water melt (aside from incoming solar energy)? Melting from underneath.

It would seem that there would be a quicker transfer of cold, just melted, water away from the ice and a replacement with warmer water.

It seems to me that spreading ice out essentially puts it in a larger 'bowl of warm water'. The time to transfer warmer water into contact with the ice would be shortened.

(If you can steer me to a site that explains all this stuff, I'd appreciate it. And quit interfering with your most interesting discussions. Thanks.)


with the latest figure of 4.121 we now have a daily loss for the last 5 days of 73,000 per day. so I think your thinking works Bob. Lot of in place melting going on in that expanding extent.
This is in no way over! Area numbers are going to go through the basement unless we get some darn cold weather.

Paul Klemencic

The title of this blog post "Two steps back, one step..." , well, whenever we see a step forward, seems to have been very prophetic. The extent loss has stalled completely.

Examining the Bremen maps for August 2nd versus August 1, the ice pack extent increased substantially in the Greenland sea with big import through the Fram strait, and along the edge of the pack from Svalbard to the Franz Josef islands; and the pack spread out in the direction of the East Siberian and Beaufort seas. These extent gains canceled out the receding pack boundary adjoining the Laptev and Kara seas, and in the direction of the Chukchi sea. The shift in the pack moving back toward the Fram and toward the far East Siberian coast as shown in the PIPS map, led to the extent gains as the ice spread out.

Regarding whether this spreading should increase actual volumetric melt rate... yes, but remember that with the warm surface waters, the most effective heat transfer occurs with wave action against the pack edges. An uneven pack edge with lots of wind and wave action hitting the pack will really melt and breakup the ice. At this point, the pack is weak, but weather conditions will drive the melt and extent loss.

Paul Klemencic

The title of this blog post "Two steps back, one step..." , well, whenever we see a step forward, seems to have been very prophetic. The extent loss has stalled completely.

Examining the Bremen maps for August 2nd versus August 1, the ice pack extent increased substantially in the Greenland sea with big import through the Fram strait, and along the edge of the pack from Svalbard to the Franz Josef islands; and the pack spread out in the direction of the East Siberian and Beaufort seas. These extent gains canceled out the receding pack boundary adjoining the Laptev and Kara seas, and in the direction of the Chukchi sea. The shift in the pack moving back toward the Fram and toward the far East Siberian coast as shown in the PIPS map, led to the extent gains as the ice spread out.

Regarding whether this spreading should increase actual volumetric melt rate... yes, but remember that with the warm surface waters, the most effective heat transfer occurs with wave action against the pack edges. An uneven pack edge with lots of wind and wave action hitting the pack will really melt and breakup the ice. At this point, the pack is weak, but weather conditions will drive the melt and extent loss.

Seke Rob

Someone asked to 'right' the NOAA2 webcam? Look at this image by NOAA1... something cubic is floating up, or tilting.


The buoy position map has not been updated since July 21, so not sure where the webcams are... probably drifting in Fram direction going by PIPS etc.

Paul Klemencic

Sorry for the duplicated post above; I refreshed the browser by mistake.

The Bremen map for August 3 is out, and the extent is going to drop a lot. The pack adjacent to the Beaufort and Chukchi receded signficantly; the East Siberian pack moved around, but lost extent and area; the pack edge adjoining the Laptev and Kara shows smaller pullbacks; and the ice pack edge closed in on Svalbard a bit, with even stronger movement toward and through the Fram strait, BUT the pack in the Greenland sea melted off even faster.

We could see 100k decline in extent today.


Hi all,

The CT area has now fallen below the record minimum for the whole of the 20th Century satellite era - as now illustrated by the nice new graph on the daily graphs link.

There seem to be some gorgeous turquoise algal blooms off the North of Norway.

There seems to be some melt going on in the glaciers flowing into the Ward Hunt Ice Shelf. I hope that there are at least some samples remaining of the ice here that is thousands of years old.

For what it's worth, I very much agree with several commenters here that extent could be about to plummet. The recent weather patterns have left a lot of well-spread slush puppy, surrounded by several patches of very warm surface water. If these have not interacted in some way by end September, expect the autumn refreeze in the marginal ice zones to be extremely slow, and Northern Hemisphere weather to go a bit berserk.

The recent weather patterns have left a lot of well-spread slush puppy

That's funny. I had just been looking for Slush Puppy images for a blog post somewhere next week.

According to ECMWF some strong highs are going to take over the centre spot in the Arctic in about 3-4 days from now. If that comes about we might get to have a nice view of the ice on the Kara and Laptev side. Maybe a bit of the East Siberian as well.

Slush puppy? Probably. But which flavour?

Paul Klemencic

The evening Bremen map came out for August 3, and there was a lot more movement in the half of the pack opposite the Fram strait. The map shows the continuation of many of the flows from this morning. It is amazing how fluid the ice pack is this year. Last year, I did the same thing, and I didn't see this kind of shifting around in just one day.

In any case there was some further spreading into the East Siberian sea, and some ice shifted into the melted areas there. The loss in extent will be somewhat less than appeared this morning, but still should be high.

Meanwhile, the high over Greenland, has now extended and shifted over the Greenland sea and toward the North Pole. I guess the Fram will close down a bit, or shift to stack ice against Svalbard or into the Olga strait.

Chris Biscan

-28K prelim.

Another weak extent drop.

but the ice pack is on thin ice.

the arctic is about to blow torch.

even with crappy winds it the extents will start to drop.


IJIS has already revised the decrease for August 3rd to -25.9 K

More significant than a single figure is the daily average for the extent decrease over the last two weeks ( July 21 to Aug 3). For this year the average was -40,446 sqkm.
Daily averages for the same period over the last 8 years were as follows:

2010: -71,596
2009: -81,640
2008: -83,437
2007: -82,496
2006: -59,497
2005: -59,375
2004: -71,696
2003: -76,484

The low extent decrease for this year is very unusual and it seems unprecendented this century ( would need to check the figures for 2000- 2002). If this weather pattern perists for another week or so, it is not impossible that we will see a minimum over 5.5 million for this year as extent losses after August 15 are generally much lower.

Chris Biscan

There is next to zero chance we get 5.5 mil or higher.

multi year ice is down again. Half the ice pack is broken up now.

I can see 4.8-5.0km2 if the pattern doesn't change much...but not 5.5km2 that would be insane considering we have 40+ melt days to go.


Interesting to see the two sides in this discussion. It seems to confuse me.

On the one hand we have the resident commenters who seem to agree, day after day -for weeks already- that large portions of ice 'could melt out any time now' because all the signals seem to be there.

On the other hand we have about two weeks of continuous very low extent decrease reports from IJIS, NSIDC etc.

It looks like the local ice is simply not aware of it's total state and assumes that -because it resides in the Arctic- it will not melt, no matter all the visuals and model outputs we humans can produce.

Is Neven 'on the ball' musing:
"Things could start refreezing again, before all the weak ice melts out completely."?

Either extent is providing little information about the ice anymore or our understanding of it's current state is severely lacking. Or my representation is skewed (probably). Who cures my confusion?

Seke Rob

Here's a little kitchen sink formula out of thin air. Take the first day < 6 million Area day of year off CT and add 32-35 days to get to the approximate date that JAXA would slide below 6 million in extent. As things go, JAXA will be passing the < 6 million km square no later than August 12 [wet finger in the windless air], no record.

As for Paul's great ice map reading, let's not forget that JAXA uses a 2 day average, which is a quick release, error prone number as they print on their chart page. For a century to be reported by them, you need 2 days that sum to > 200 kkm^2 decline before it will show on their chart to even have indication that there was a 100k change on a day. Anyway, my thinking remains on "lower than 2007". PIOMAS gave the extra hint needed. Their day 212 data since 2001 and anomaly shows:

2001 212 15.074 -324
2002 212 13.984 -1.414
2003 212 13.149 -2.249
2004 212 13.498 -1.900
2005 212 12.253 -3.145
2006 212 11.939 -3.459
2007 212 9.172 -6.226
2008 212 11.255 -4.143
2009 212 9.891 -5.507
2010 212 7.026 -8.372
2011 212 6.494 -8.904

2.678 cubic km less than 2007.... 532 cubic km less than last year same day.

Seke Rob

The picture to go with the glass bowl prediction: Area-Extent 1st Day Below 6 million km square

The near linear decline of Area in the last 12-13 years might be an indicator... are we past the point that no matter what, things will go, regardless any ENSO/AMO/PDO or whatever oscillation can be conjured up? Hight salt content melt-water from thinner FYI sinking to push up the warmer layers below... long as there is warmer water below, that's a near perpetuum mobile



Let’s make things simple.
A.There is about 2,8 MK left in the central pack (over 80 degrees north). The overall structure is like meshed iron, meaning large, elongated squares of more or less unified floes every surface between 15 and 1000 km², floating in broad, hundreds of km’s long leads filled with rubble. A lot of these floes consist of 2 up to 5 year old ice.
B.On the northeast side, stretching into Beaufort-, Chukchi- and East Sib Sea, lies about 2,5 MK of what Idunno so conveniently refers to as ‘slush puppy’. It has no structure left, is being ripped anywhere dependent on the wind and can easiliy fit any grid-count on a 15% or 30% area/extent norm on whatever surface the day-to-day satellite images offer.
C.Remains of about 1 MK, not being of much importance (they will melt out eventually), are to be found on the east side of Greenland and in the Archipelago.
Total A+B+C = 6,3 MK.
Thickness A: floes (40%) 2 m, leads 1m, median 1,35m x 2,8MK = 3800 m³
ThicknessB/C: 0,75 m x 3,5MK = 2700 m³
Total volume left: 6500 m³
Usual melt august/september: 2800 m³
Minimum 2011: 3700 m³
Translate that back to thickness for A and B (C will be gone):
1,15 m x 2,8MK = 3300 m³
0,5 m x 0,8MK = 400 m³
The already built-in melt capacity should be enough to do this, and volume loss will express itself, evidently the most in area. Whether we’ll see an extent record, depends much on the % norm in grid-counting. So when it’s going to be 5,5MK,Phil 263, 2,7 MK will be just 0,15 cm thick or 30% concentration.
Too simple?


So when it’s going to be 5,5MK,Phil 263, 2,7 MK will be just 0,15 cm thick or 30% concentration

Good point Werther, but how are you getting your estimate of thickness? I see thichness to the North of Ellesmere and Greenland at between 3 and 5 metres???
Estimates of thickness are based on models , not observations. Correct me if I am wrong but extent calculations are based on sattelite observations and are more reliable than area calculations (any way, that's what i have learnt from this blog and other blogs)

Lord Soth

Werther I believe you are right. With this year and somewhat with 2010, by late summer, rather than having a somewhat unified pack moving in unison, the pack is behaving more like slush on a grand scale.

I believe we have reach the point where extent is no longer a useful indicator, in predicting the state of the ice, for the August - September time frame.

Perhaps we should use the CAPIE ratios to do a correction to the IJIS extent, to better reflect the true melt, and weigh out the spreading, at least in late summer.

Seke Rob

phil263, think there is quite a bit of current year observational going into the PIOMAS model. Buoys, AWI flights, Catamaran sled Track [2 French guys], The Ellesmere-Baffin track, Navy subs, and probably more.

The icebreaker captain wrote last year: "The ice is rotten". Little doubt is left on that.

Lord Soth

Just to explain how CAPIE adjusted IJIS would work.

First we would take the average CAPIE for a given date based on the full 9 years of IJIS record, and then adjust the extent by the deviation from averaged CAPIE for the given date.

For example. Lets assume that the average CAPIE for the August 2 date is 67% and the actual CAPIE for August 2, 2011 is 63%. That is a deviation of 4% below normal. We would then reduce the extent for August 2 2011 by 4 percent which equates to 261K and the adjusted extent for August 2 would be 6278000.

Now do this for every date for the complete IJIS dataset, so we are comparing apples to apples.

I'm going on vacation for three weeks, and if nobody takes on the CAPIE adjusted IJIS, I will do it when I get back.


Should have a peek on the "other side of the mirror"...
Almost ready to start popping up some champagne: " sea ice extent making a sharp right turn" " .. the lowest melt since 1973"

BUT I would totally subscribe to the following comment made by a bystander:

“What does this mean? The short answer is, probably nothing. When we approach the minimum, and the ice pack becomes more fractured and scattered, it also becomes more susceptible to the vagaries of local and regional wind and weather.”

Well said – although I’d expect the usual “victory lap” from those that will try to seize on this as evidence that we’re not seeing longer term changes….

Seke Rob

Lord Soth, think sticking to the IJIS value and then slot in a couple of error bars may do the trick. Not doing a CAPIE like yet, just the monthlies off from NSIDC to get a state of the Arctic indicator... we all agree I think we're in uncharted territory. AR5 can't come soon enough to cause a few straight in the kissers.


How exact can things be? Phil, forgive me my beercard calculation fashions. But there’s not much need for an exact calculation (or a half billion euro’s worth satellite like Cryosat) to see where this is going.
Yes, The Arc model US Navy shows 4.5 – 5 m thickness north of Ellesmere, Archipelago and Greenland. I assume that that is no average. It reflects the depth of keels under the ice, and they can go down to 40-50 meters deep. Sure you can find those under the floes 15-1000 km² I mentioned. Produced by slabbing in the past. Besides, the minus 3 m-surface on Arc is less than 1 MK square surface. Base don PIOMAS and, FI , the Wegener flight measurements this spring, i’m pretty sure the average for the floes is 2m.
The leads might seem to be the place for extensive slabbing. But given the mobility and spread the last couple of years, they probably resemble a ‘coffeegrinder’. That’s why I assume average depth over there is even less; 1 m. Look at all the speckles of dark, open water in them!
You are right that it is almost impossible to do exact area counting/calculating. I, too, learned through Neven’s blog that FI melt ponds can easily be counted. But that is offset by the counting of ‘ rotten ice’ as being 100% volume.
That said, I agree with you that there are some contradictionary datasets. It’s clear that any method has it’s own limitations. I’d love to know what the specialists would have to say about that. They are bound to what Hansen calls ‘scientific reticence’.
But I guess they’re holding their breath like we do.


A new record will be difficult now, but a minimum extent over 5 million km2 will be even harder, I think. Of course, the longer extent decrease is at a complete standstill, the more probable it becomes. But something's going to give, even if the weather patterns stay adverse to extent decrease.

2010 flipped again around mid-August and the managed to gain a third spot with 1 month of extent decrease. If 2011 flips too around that time or before, it should be able to take over that third spot from 2010.

I'm hoping for some clear weather, guys, so we can see what's going on down there!

ECMWF still has highs taking over the central Arctic...


PIPS is also showing slightly larger arrows, pointing the right way:

I think we'll start seeing daily extent decrease surpassing 50K in the coming few days.


PIOMAS is a model but it does assimilate observations. I don't think we should rule out it being understated by 1000 km^3 (I think Werther has the units wrong, it is km^3 not m^3 only a factor of a billion ;o) so we know what you mean.) But I think that analysis is pretty sound - even if you change 6500 km^3 to 7500 km^3 it is only 15% more volume and the figures are not affected that much. Importantly, we still end up with less ice volume than ever before.

If the extent is greater than 2007 but area similar it probably means it is broken into smaller pieces and there is more surface area possibly allowing greater volume loss than 2007 if conditions were otherwise the same.

As multiyear ice is replaced by first year ice, the volume that can be melted in the season goes up and the amount of ice at maximum is decreasing (despite the negative thickness feedback) and approaching that meltable volume. Not this year, but within a few years.


CT has reported a -128K drop for August 2nd. Combined with the meagre -2.5K from IJIS we get a CAPIE percentage of 61.06%.


Hi Neven, I just checked the same. For what I can recall, there was just one time CAPIE got lower than that, mid-august 2008.
This continues to be a very, very interesting race...

Seke Rob

Talk about weird, probably 3 weeks ago someone observed that Hudson Bay would be melted out in a few days... per MASIE the extent went up from 88229 to 106363
to 108467 kmsq on Aug 2. It's clearly visible in the CT images. Is this spreading, refreezing or drift-in? This map of RTG/NCEP indicates it's hotter than hot in anomaly terms. In absolute the SST is clearly above zero Celsius

Lucia (The Blackboard)

The low extent decrease for this year is very unusual and it seems unprecendented this century ( would need to check the figures for 2000- 2002).
It seems to be the slowest melt rate for this time of year since 1972. I've been plotting the melt rate based on the 7 day smooth extent:

The vertical lines indicate dates pertaining to betting at my blog.

A new record will be difficult now, but a minimum extent over 5 million km2 will be even harder,
Using a fit based on 1972-now data to current 7-day average extent to minimum, I have ±95% confidence intervals of 4.18 million sq. km to 5.49 million sq. km. I'd initially done it with JAXA only data, and current melt rate didn't have predictive value above current extent. I haven't checked this with the 1972-now data. Also, I haven't included uncertainty in the fit parameters in my confidence intervals-- just scatter around the fit as if it's perfect.

Based on this info only, it looks like there is still a greater than 1/20 chance of breaking the minimum low but there is also a good chance of getting something above 5 million km2.

Lord Soth

Hudson bay is total ice free. South West Hudson Bay (the last ice to melt) went from open water (less than 1/10 ice) to ice free on August 1.

It's just sensor noize. You will see it in other areas occasionally, where there is never ice.


Seke Rob

Environment Canada says;

In early August, the ice melted completely in Hudson Bay.


Seke Rob

Gents if 108,000 kmsq is sensor noise for Hudson, what confidence is left there for anything in the extent ensemble to be right? MASIE reports it, CT sees it, Bremen does not: http://www.iup.uni-bremen.de:8084/amsr/arctic_AMSRE_nic.png... The extent curves of JAXA/NSIDC/BREMEN are well aligned on the 15% measure, so are they seeing Hudson as free or noised over?

Twemoran, think I'm overdue to look at more of the local reports: "Ice decay is 2 to 3 weeks ahead of normal over the whole region."

Lord Soth, accept there's sensor noise. Those large patches in the Atlantic way south of New Foundland could not be true... out of nowhere... more pinches of salt in the sea ice.


I disagree that it's sensor noise (which relieves me, as 100K sounds like a lot of sensor noise in an area as small as Hudson Bay).

Looking at the Bremen and NSIDC maps, I can see a fair amount of ice in the bay just to the west of Baffin Island (which is part of the CT Hudson Bay region).

Seke Rob

Bfraser, no issue with the ice near the Spicer/Prince Charles islands... it's uniformly present on all maps. If that is good for 108kkmsq and part of the ''Hudson Bay'' label, than we're fine.

Chris Biscan

CT dropped 113K again.

now at:


Eve though the ice extent has slowed a ton.

make no mistake it's still in shambles.

The comments to this entry are closed.