« Fringe fries | Main | Webcam art »


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

Seke Rob

Oh dear, today, the 18th... yes the wedge has broken free: http://lance-modis.eosdis.nasa.gov/imagery/subsets/?subset=Arctic_r03c02.2012170.terra.1km.jpg

Tor Bejnar

Not deceptive, I say! I've been printing out DMI's Kane Basin TERRA or AQUA images on occasional clear days, and noted that yesterday's bulge (toward the Canadian side) extends beyond where the bridge edge was 3 or so weeks ago. Other part of the bridge have receded, but not near Greenland (except for close to the shore). (I hold the two pieces of paper together - with the landforms matching - and note the bridge edge differences.) I have not checked today's image carefully, but it looks like some ice has sloughed off - the bulge may be gone.

Espen Olsen

Ice in the Canadian Archipelago is disintegrating several places now, polynyas are popping up here and there and everywhere!
the ice is like thin chocolate on an ice cream!

Account Deleted

I dropped CT a message about the apparent discrepancy - it's not 1% or 5%, it's a 30+% discrepancy (something like 200k km^2) between what MODIS is showing and what their data is indicating. The http://www7320.nrlssc.navy.mil/hycomARC/skill_public.html
site has the area of the CA we are talking about at 100% concentration with 2m thick ice.
I think CT and the other groups do a great job of getting this data out and in the end it won't change how much ice we loss this year - but it is potentially giving ammunition to the denailists and unfortunately they are so adept at creating doubt - which allow politicians/etc. to do nothing about CC.

Account Deleted

I should have added: apart from having holiday junkets in various exotic locations to sign/argue about treaty/documents that they most often ignore.

Artful Dodger

Aaron, looks like a good model run until you see that they're using 1985-86 forcings for their 2026 predictions... :^(


While we are discussing ice - the impact in Siberia of early ice and snow melt is now obvious - fire.

Forest fires are burning across central Siberia - alot of them. Where there is fire - there is smoke. Lot's of new GHG CO2 and black soot.


"According to the environmental group Greenpeace, more land in Russia has burned this year than in 2010, a year that intense wildfires affected western Russia and produced rare pyrocumulus “fire clouds.”

See also: http://www.themoscowtimes.com/news/article/state-of-emergency-declared-due-to-fires-in-eastern-regions/460515.html

Account Deleted

Unfortunately it's not just Russia that is burning - there are lots of fires in my part of the world. Some of these hotspots appear to be in areas of peatswamp - so not only lots of new GHG CO2 and black soot, but also further damage to one of the long-term carbon sinks :(


Rob Dekker

I'm sorry if I misunderstood your point. But when you state "it's not 1% or 5%, it's a 30+% discrepancy (something like 200k km^2) between what MODIS is showing and what their data is indicating. "
which MODIS vs CT data did you refer to exactly ?

Also, Hycom/ARCc is not really meant to be used as an accurate ice concentration or thickness assessment tool. It's intended to present 'worst case' ice encounters for US Navy vessels. On many occasions has this tool shown to overestimate ice thickness by a factor of 2 or more.

Peter Ellis

Rob: CT thinks the ice area in the Canadian Archipelago has dropped by ~200k km^2 (at least 30% of the total) within the last week or so.

It hasn't. There are a few small polynyas, but the rest is at 100% concentration.

Somehow, the microwave sensor is getting it wrong, either because of melt ponding, thinning ice, or some other issue.

Account Deleted

Sorry Rob, I was working on something else when I was writing that comment.
Peter is right - I was referring to the CA sea ice area. CT is also showing 50%ish sea ice concentrations in the Viscount Melville Sound - while Modis is showing it fairly full of sea ice.

Rob Dekker

Thanks Peter.

Forgive me, but without pixel-counting the Lance-Modis pictures from the last week or so, I can't determine the area loss.

On the other hand, Masie does also report a 200 k loss over the past week :

Where is the data that contradicts CT/Masie ?

Peter Ellis

Um, you don't need pixel counting to see a THIRTY PERCENT change, just your own eyes!

Also, I think you're looking at the wrong column in the MASIE file: it in fact shows a minute increase over the last week, and no more than a 22.5k loss over the last month.

Rob Dekker

Sorry to be playing devil's advocate here. But as Colin points out, denialists are prone to use discrepancies in data as ammunition to create doubt. And the best defense against doubt is hard data.

Peter Ellis

Yes, and CT is the one that's way out of line.

Rob Dekker

Peter, I think you are right that I looked at the wrong column. It's late.
Seems indeed that Masie reports only a minor drop for the CA.
So I'm with you guys now. Sorry for the confusion.

Rob Dekker

There is still something odd, guys.
Masie shows 823810 some km^2 for the Canadian Archipelago, while CT peaked out at 600 k. So it seems obvious that the two measurements use a different definition of CA.

Still does not make the 200k drop in CT any more or less believable, but it does suggest that there is a mismatch in measurement standard.

Do we have a third measurement from anywhere ?

Rob Dekker

Final thought on this CA area :
SSMI measures microwave emissions indicative of water versus ice. That's how it determines 'ice concentration', which is the determining factor between area and extent. It's very well possible that CT's 200k drop in area estimate is created by melting of snow in the CA, which SSMI will now see as 'water'. On Modis, it would be hard to 'see' that water, unless you take a very close look at the albedo changes throughout the CA. Thus, not only pixel-counting but also taking gray-scaling into account.

What I'm saying is that there is a plausible explanation for the 200 k drop in CT which does not imply a system failure.

My assessment : Either CA ice is getting pretty rotten and is top-melting badly, although it's still compact (without many polynia) or CT is truely out of touch with reality.

Time will tell.


In the meantime CT reports another drop in SIA of 158K.

In the meantime CT reports another drop in SIA of 158K

Indeed. That makes an 11-day drop of 1.87 million km2, an average of better than 170k per day. Over the past two weeks, we've seen the following: the 1st, 3rd, & 4th largest ten-day drops in the CT record; the 1st, 3rd, & 4th largest nine-day drops; and the 1st, 2nd, & 4th largest eight-day drops.)

However...if CT is not accurately reflecting reality, we'd really like to know about it, and soon. As others have noted, denialists--being deeply frustrated by the fact that science is, you know, winning--love nothing more than rooting out the most inconsequential data discrepancies and magnifying them into "proof" of major global conspiratorial fraud. If CT is, as Peter Ellis said, "way out of line", that needs to be made right--or we need to disavow.

Steven Newbury

From what I can see on MODIS, (where it's not obscured by cloud) large polynyas are opening up in the central region of the CA in the last couple of days, along with continuing disintegration in the west as the Beaufort hole expands. Also looking at the 250m resolution clearly shows some areas, particularly near the polynyas are no longer solidly packed ice, certainly less that 100% concentration. That said, surface melt where the fast ice is still continuous is certainly adding quite considerably to the concentration number, but at the current rate and conditions I don't think it's going to matter too much.


Regarding sea ice melt as blue tinged or colored ice, Nasa Earth Observatory has commented that melting ice, or water covered ice does appear blue.

Image, Hudson Bay, June 5, 2012: http://eoimages.gsfc.nasa.gov/images/imagerecords/78000/78274/hudsonbay_amo_2012157_lrg.jpg

"Two months later, snow has melted and ice has retreated, leaving Akimiski Island and the Belcher Islands easier to spot. The ice in Hudson Bay has taken on a bluish tinge, likely due to waterlogged ice and surface melt on the ice."


Here is a closeup shot from the Healey in August 2009 of ponding on sea ice.


Hopefully this helps document the melt water issue that may be impacting CT SIA numbers.

L. Hamilton

The DMI index leaves more coastal area unclassified, so it is not tracking much detail in the CAA. And the DMI decline has stalled over the past 3 days. Even so, the mean daily decline this month sets a record:

Here's the comparable graph for CT:


I don't think it's going to matter too much

I agree. It'll even out soon enough, once those melt ponds drain or freeze over. It's like I say in this update: sure, melt ponds may exaggerate the decrease by a couple of tens or even hundreds of thousands of km2, but at the same time it says something about insolation and heat accumulation.

DMI extent has indeed been stalling, but the other graphs are a bit slower to respond. At the same time the weather forecasts keep showing that highs are going to take over on the Canadian side of the Arctic in a couple of days from now. It's a bit unclear, looks slightly Dipole-ish in 5-6 days, but if this comes about I think we will see more big drops.

BTW, that low did churn the pack up a bit (see here on Arctic.io and zoom out to see what I mean).


The area Neven highlighted in "Kind of Blue" is instructive here.

A few days after that post a lot of the Laptev reverted to white (local freeze up?), and the area anomaly halved (from ~-240K to ~-120K). A few days later again, it had largely turned blue again (warmer weather?). In the last few days, the Lena delta has completed thawing, and the area has drained(?), changing a lot of blue back to white, and the anomaly has reduced by ~75K sq km. So the condition of the ice certainly impacts on area readings, but is a temporary phenomenon, seen mainly in June.

Now while it is annoying to have incorrect data in absolute terms, having it consistent is what is relevant to year-on-year comparisons. This blue area appeared in 2010 and 2011 at around the same time.

If you look at the current CT CA graph you will see that from mid-June to mid-July last year area was pretty static. But that doesn't mean it wasn't changing: between day 170 and day 190, most of the "blue" went out of the CA (which should have caused the appearence of an increase in area), while at the same time ice was being punted out (causing an actual decrease in area). These factors roughly balanced out for a month, causing the apparent statis.

So if blue ice (whatever the cause) results in artifically low readings in June, but this problem is largely past by mid-July, and - critically - if this happens every year at around this time, I have to ask what the big deal is.

It may be that area data is 200K too low for the moment, but it was a similar amount too low in 2010 and 2011. If you want add, say 200 K to this years data, you have to make the same adjustments to other years you are comparing it too. So long as we are comparing like to like - and we are - the we are still ~500 K below the old record, whether you factor in this artifact in the data or not.

Seke Rob

The MASIE A integrated chart has been further expanded (Ctrl-F5 to refresh), almost a full webpage worth of data included... latest, sparklines [courtesy MS], that break and visualizes each region, for those who lost track in the spaghetti. If a sliding week total change, or 7 day average is of interest, or some other fixed period, fill in the form [a post somewhere] and it will be considered how to inject swiftly.

On the overall data, 2012 is per the 18th 112KKm2 lower than 2011. From Max to current low, there was 5.4 million km2 reduction, where last year max to the 18th had 4.6M reduction... quantity is not quality as this "most terrible of terrible winters" demonstrates. The laugh is on in denyloland... suddenly that ocean heat that never existed, Trenberth's travestial "missing heat", that Pielke claimed went all poof into space when the last La Nina series kicked in, magically returns. Consistency is not printed big in those quarters :P


Further to the above - it is not difficult to reverse engineer the CT CA graph to overlay the 1979-2009 average onto this years data (by adding the anomaly back in). When you do so, you can see the cumulative effect of this artifact.

The melt progresses fairly smoothly from 600K to 450 K at the end of June. It then pauses for about 17 days in July at this value, before recommencing its decline to 300 K (average minimum). This exercise highlights that this white-blue-white shift has occurred earlier in the last few years (which is interesting in itself), But it also highlights this is a regular occurrence in the CT data.

It would be hard to argue that this flat spot actually represents a sudden pause in melting (what could possibly cause that?). But it is very easy to argue that the effect that we see for the last three years - the balance between "whitening" (+) and actual loss of area (-) - is a regular seasonal effect, and it is most likely embedded in all years of data.

michael sweet

The current area issue with Cryosphere Today in the CAA is why the NSIDC uses extent. Extent is much less susceptible to melt pond issues. Frank is right: compare like with like and you will be OK. If you don't like the area data, use extent from NSIDC or IJIS. The area issues will shake out by the middle of August anyway.

Andreas Muenchow

The ice arch at the southern end of Nares Strait extending into Kane Basin is very solid and has been in place since at least Dec.-8, 2011. I thought it was about to collapse when I noticed it change its shape, but it turned out that new ice formed making it look as if it had moved when it did not. Processed and gridded Nares Strait MODIS visible and thermal imagery in a common frame and projections can be found at http://muenchow.cms.udel.edu/MODIS/index.html for almost every day since 2000-present. I expect this ice arch to hold another 3-4 weeks before it collapses. This will allow thick, old, multi-year ice to leave the Arctic via this gate.


The discussion about MODIS imagery not showing holes and cracks in the ice is a little like not seeing the soccer ball from space and claiming that there is obviously not a game being played. Each pixel is 250 x 250 meters. 62,500 sq meters. How many soccer fields are in each pixel? Yes water on the ice reports as a lower area number. Water on the ice has always reported as a lower area number. But MODIS does not have enough resolution to make claims that area numbers are wrong and end up shilling for the deniers. It is just another measure of sea ice and it is not wrong!


Dr Andreas Münchow, thanks a lot for that link to your website. It feels like somebody left a treasure map on my doorstep. Your blog is fantastic too!

Doug Bostrom

Just a quick caution that Apocolypse4Real's "Moscow Times" link above seems to be infected. Opening the article produces:

ClamAV: Exploit.JS.CVE-2006-1359

"Exp/TxtRng-A is an exploit for a vulnerability in Microsoft Internet Explorer (CVE-2006-1359)."

So if you're running an ancient copy of IE, best avoided.


Tor Bejnar asked me to post this on his behalf:

On the order of 1/2 to 5 km disintegrated off the face of the Kane Basin ice bridge between June 17 and 18 MODIS/TERRA passes (per my estimates from DMI website images). I find it interesting that the ice is disintegrating rather than breaking off in large floes that would show as obviously discrete units floating in the water. We’ll see big pieces when the bridge finally crashes. (Up close, I’ve been impressed with the smallest piece of floating ice, but I have different standards for satellite images!)

I now detect movement of identifiable ice features within Kane Basin between June 17 and 18: “downstream” movement near the bridge and near the Canadian side and “upstream” movement in the center of the Basin. Clouds obscure some parts of each image, though, so only some areas can be tested. I cannot wait for today’s images!

Espen Olsen

I have followed the discussion of what source of data is most reliable, my answer is simple: Modis Aqua/Terra images, after a few years of watching and studying those images and almost knowing all fjords, straits and glaciers in the polar area nothing beats these images, and to have a little deeper understanding these images are really fine and reliable as well: http://igloo.atmos.uiuc.edu/cgi-bin/test/print.sh?fm=06&fd=17&fy=2011&sm=06&sd=17&sy=2012

I only rely on all the other sources in case I am confused, and that happens often, though!


Apologies about the Moscow Times link.

I am running the latest IE and anti-exploit, etc. software. Looks like cyberwar includes the news websites. I won't be going there again. Additionally, take care in reading Chinese web-based news.

Ar Vb

Look on the current Modis-Aqua image how the icesheet is broken ~ 80°North, 165°East
These aren't meltponds.



Indeed, Ar Vb, that's what that low-pressure system did. The melt ponds are mostly on the fast ice in the Canadian Archipelago, Laptev Sea and East Siberian Sea, and I'm guessing on ice floes in the Beaufort Sea.



Thanks so much for the new (to me) Nares data.

On another topic Ward-Hunt can be made out, possibly with a small polynya forming at MODIS aqua 203c03. It will be interesting to see is either or both halves make it through the season.



The NSIDC has an extra June update explaining the recent nosedive.

Bob Wallace

The 2012 nosedive has occurred about a month earlier than the 2007 nosedive.

That's an additional month of lower albedo/higher open water absorption.

Rob Dekker

Yesterday Colin and Peter mentioned that Modis does not support the steep drop (30%) in SIA in the Canadian Archipelago that CT reported over the past week.
Peter even suggested that "CT is the one that's way out of line".

Masie suggest that ice extent did not drop much at all, so the drop in SIA that CT reports must come from ice concentration reduction.
CT area estimates 'ice-concentration' reduction. Here Colin and Peter are right that Modis does not show an obvious reduction in ice concentration.
However, as Frank and others pointed out, significant snow-melt and melting ponds of all sizes also shows up as 'area' reduction, and tends to show up om Modis as shades of 'blue'. Neven did a good blog on the blueish ice recently.

To see if snow-melt and water on the ice may be the cause of CT's area drop, I looked at the Modis pictures starting day 150 (5/29) :

Then, in the browser, I opened up every day since day 150, so that you can step back and forth through time, creating a high-resolution 'movie' of the Arctic.
I can recommend this for everyone : Amazing how you can see the Beaufort gyre spinning and crunching ice, huge chucks breaking up and desintegrating, and, indeed see the Canadian Achipelago turn from bright white on day 150, to increasingly blue-gray (certainly in the 50-70% shade area, suggesting water content of 30-50% in the top-layer), especially the last week or so.

So there we have it. There is no apparent mis-match between Modis and CT area numbers. The CA is simply top-melting badly, and desintegrating in-place, with chucks breaking off at the boundary.

Peter Ellis

Um, pretty much my first post on the subject suggested that the reason CT (or any other area measure including ROOS) is out of line is abundant melt ponding. It's what I said for the "blue ice" in the Laptev and East Siberian seas too - which was/is also erroneously seen as low concentration by CT. You can see it on the CAPIE measure - pretty sure that if Neven produces a new one, we'll be off the bottom of the scale.

The bottom line is that area measurements are confounded by melt ponding to a much greater degree than extent measurements. This year, for whatever reason, we seem to have more melt ponding. I think it's premature to claim that it's due to increased top-melt of the ice: remember that the initial melt ponds form from thawing of the overlying snow. A further factor is that ice may be more or less permeable to water, and hence allow more or less drainage of the melt ponds.

Of course, the fact that there is increased melt ponding means a reduction in albedo, which will if weather permits lead to more rapid top-melt. That's subtly different from saying that unusually large melt pond fraction is itself due to top melt.


MODIS looks particularly terrible today. Even in the "stronghold" north of the CA there are hundred-mile leads opening up. Here's the north coast of Ellesmere and Greenland with the cloud-covered Nares Strait in the center:


At some point one might expect that the pressing of the ice against the north coast of the CA and Greenland, instead of strengthening the ice through plastic deformation, would just crush it like a giant Daiquiri.


You can see it on the CAPIE measure - pretty sure that if Neven produces a new one, we'll be off the bottom of the scale.

Indeed. Here is the latest CAPIE graph:

Ratios for June 17th:

2005: 87.08
2006: 83.16
2007: 80.12
2008: 81.34
2009: 86.31
2010: 83.97
2011: 83.41
2012: 76.90

Veli Kallio

It is quite while when I last wrote onto this block. I have been tied with Rio+20 where our nations are moan that we warned the United Nations Genral Assembly back in 1992 that any sustained warming on the polar regions would lead to a rapid responses: retreat of seasonal snow cover area, damage to permafrost, loss of sea ice and destabilisation of Greenland Ice Sheet.

I just noted that one important factor has not been covered in your block and decided to drop a line to remind existense of this.

If you go to Cryosphere Today, you can compare 2012 to same dates of 2011, 2010, 2008, 2007 and 2006. Each one of these years seem to have produced a late spring combined terrestrial snow cover + sea ice area that is smaller than the preceding years. There are small anomalies locally, but the totality of the white cap on high latitudes has shrunk. In 2012 loss of snow in Yakutia must have been a major contributory factor to high temperatures and capture of solar light and heat.

As the winds move air from the land over sea and vice versa, the dark areas produce heat which is transported whether it is sea or land. This transportation in Yakutia has brought intense sunlight converted to long wave radiation and heat which is reabsorved by ice on East Siberian Sea, or transported as melt water into the Arctic Ocean. In Kara Sea we have seen reverse, the ocean lost its sea cover and the warm air from open sea had licked away the snow cover on its coasts.

We must not forget the consolidated impact of marine and terrestrial snow cover to each other due to the intense conversion of dark surfaces of light to long wave radiation and heat. As per earliest snow cover loss, you are right that it is July, it is because of the early start and the strongest sunlight season in Siberia.

Also Canada has seen strong early loss on its terrestrial snow cover, this helping to damage the Hudson Bay which is now going. I expect the Central Arctic to be hit seriously by the extra energy from snow free Siberia since may, that has pushed season a month ahead.

It also bears to be remembered that the soil temperatures are also rising, making snow ever more volatile, however abundant in may be. This is the reason why Ewing-Dunne Ice Ages cause theory does not work. As ocean warms, vapourises and rains more, the soils too warm up with ever higher bacterial activity to keep it thawed even in the mid winter.

Veli Albert Kallio, FRGS

Rapporteur, First Nations United Nations General Assembly Motion 101292 (RIO 1992)

Vice President Environmental Affairs, Sea Research Society, Charleston, South Carolina

Chairman, Frozen Isthmuses' Protection Campaign of the Arctic and North Atlantic Oceans, Helsinki, Finland

Rapporteur, 1992 Motion at RIO+20, Rio de Janeiro, Brazil

Aaron Lewis

Look at dew points across the CA. They are above 0C. Anywhere ice is exposed, water is condensing. For every gram of water vapor that condenses, 7.5 grams of ice melt resulting in 8.5 ml of melt pond water.

The other side of this is that the water vapor acts as a local green house gas to hold in heat trapped by the melt ponds.

Veli Kallio

We at FIPC developed the last year a novel sea ice algorithm which we hopefully can make viral this year. It generates sea ice survival forecasts for the rest of season of various types, this could point out if the sea ice loss becomes a runaway event in the season or how much short it is from becoming one. It is a linear progression, but we will refine it further by a variable and perhaps regionalised weightings.

Al Ramsey

The melt ponds on the Kimmirut web cam have turned into open water today. The tidal action will soon open up the whole sound. Davis Straight (to the south) is mostly ice free as well. This I suppose is an example of what is happening - will have all over the Canadian Arctic.

Account Deleted

No matter what is happening in the CA - all of the data, images show a depressing story. We are something like half way thru the melt season and we have lost a shit-load of ice (+/- a bit). I really hope we see a slow down soon - at least todays temperatures aren't as brutal as before

Account Deleted

Bill Chapman (Cryosphere Today) got back to me about the CA sea ice figures. As has been discussed: "the apparent discrepancy is the problem the passive microwave sensors have in determining sea ice concentration from melt ponds. This time of year, this issue is particularly problematic as the melt season is in full swing in the southern parts of the Arctic." and also ". Another issue that can cause underestimation of the concentration in passive microwave products is cloudiness, especially optically thick clouds". He also points out that the heavily melt-ponded ice is not long for this world. I appreciate that he took the time to reply.



Thanks for getting such a detailed reply and sharing it. We are still watching unprecedented melt, and the NSIDC report seems to support continued rapid melt.

Artful Dodger

Hi folks,

To understand how sea ice area is determined, eventually one must delve into topics like sea ice emissivity modelling and emission spectra. But first, it is helpful to have a basic understanding of the 5 stages of melting that sea ice undergoes:

  1. Puddle:
    An accumulation melt-water on ice, mainly due to melting snow, but in the more advanced stages also to the melting of ice. The initial stage consists of patches of melted snow.
  2. Flooded ice:
    Sea ice, usually first-year ice, flooded by a melt or river water layer.
  3. Thaw holes:
    Vertical holes in sea ice formed when surface puddles melt through to the underlying water.
  4. Dried ice:
    Sea ice that was earlier at the flooded ice stage, from the surface of which melt-water has disappeared after the formation of cracks and thaw holes. During the period of drying, the surface whitens.
  5. Rotten ice:
    Sea ice which has become honeycombed (laced) and which is in an advanced state of disintegration.

The topic then advances to defining how remote sensing is used to identify each of these stages.

Seke Rob

Fundamental difference to MYI in large numbers then, and the large floes of FYI today is that the puddles of stage 1 spread out much wider and interconnect, if the ice not broken up already. The North Pole Web cam has been set up to watch a large flat floe [see a pressure ridge in the far], so will offer a good opportunity [does the cam have a higher point too?]

Rob Dekker

As Apocalypse4Real already mentioned : Thank you for sending that inquiry to CT and for posting the detailed response.

Peter, it seems to me that the confusion (of if CT area matches what Modis shows, or not) comes down to what you define 'sea ice area' to be.

Currently, 'ice concentration' (and by extention 'ice area') is determined by microwave analysis of the amount of water versus ice under some resolution.

This means that water on the ice (and transparancy of the ice) DOES affect the 'ice concentration' and thus the 'ice area' numbers.

So, ice 'area' of a pixel with 50cm ice with water on top is NOT counted as the same 'area' of 4 meter ice with snow, even though their 'extent' will be the reported as the same.

As we say in software development terms : It's not a 'bug', it's a 'feature'.

But if you agree or disagree with the current measurement and definition of 'ice concentration' and definition of 'ice area' seems rather irrelevant. As FrankD pointed out above : "having it consistent is what is relevant to year-on-year comparisons.".

Rob Dekker

Quick post about bottom-melt :
Results from ITP53 bouy (located in the Beaufort) suggest that a lot of solar irradiance makes it through the ice and into to ocean below.

In the top 75 meters of ocean under the ice, enough heat has been building over the past month or so to melt about half a meter ice from beneath in the next few months :
This half a meter ice melt-to-come is on top (actually 'below' :o) of whatever atmospheric heat will still melt out from the top over the remainder of the melting season.

Seems that the top 75 meter ocean under the ice serves as a container of heat accumulated during the spring and early summer...

Rob Dekker

DMI extent is nose-diving again, and the low has not even dissipated yet :
Seems that my prediction of a 'stall' was incorrect.

Seke Rob

Wish you'd been right.

Atmos/CT latest with a small uptick:

2012.4575 -1.9107149 7.9473367 9.8580513
2012.4603 -1.9876204 7.7897234 9.7773438
2012.4630 -1.8462970 7.8461437 9.6924410

Peter Ellis

Modelling oddity: look at the ACNFS thickness model for the last month


That lump of dark green ~2.5-3m ice in the middle of the pack drops down to cyan in the last few days, indicating sudden melt of at least half a metre in the middle of the pack. Something to do with the big low/storm sat over the area?

Espen Olsen

Barrow web cam is now on-line again:


Peter Ellis

(Actually, looking at the gif more closely, ACNFS is forecasting that weird central melt to happen in the next 5 days)

L. Hamilton

It seems that DMI sometimes leads CT? DMI stalled for a few days, then today dropped a century. CT has now caught up, in calendar days, with the DMI stall period.

Seke Rob

Would I(We) not love to get the hands on the DMI daily data points. Just for the loose exercise overlayed the MASIE/IJIS 2012 curve on top of the DMI image... readily done in Excel without need of digitizing pixels. MASIE/JAXA/DMI Extent Merge

Trust me, no Viscountian tricks ;>)

Seke Rob

Off topic, but the right side news reel bar popped up at least for the second time the same article at CO2 Science [a denylo site], not sure why they are included in the news roll, lest there is no influence on that or for 'balanced' reporting. How it got to the top again... anybody's guess.

-- Rob

Peter Ellis wrote:

That lump of dark green ~2.5-3m ice in the middle of the pack drops down to cyan in the last few days, indicating sudden melt of at least half a metre in the middle of the pack.

I suppose the bottom line--well, one of them, anyway--is that regardless of whether winds conspire to prevent 2012 from surpassing 2007 or 2010 records in area and/or extent, volume is all but guaranteed to do so. And when you get down to it, volume is the single truest measure of the state of the ice, no?

Seke Rob wrote:

...the right side news reel bar popped up at least for the second time the same article at CO2 Science [a denylo site]

I saw that. That's the ALEC-connected "CO2 is healthy for plants and all living things" site, right? I imagine--though I really have no idea--that Neven's site simply consumes an RSS or XML feed built by a bot that scours the web for ice-related headlines. That kind of thing is unavoidable, I reckon...


That lump of dark green ~2.5-3m ice in the middle of the pack drops down to cyan in the last few days, indicating sudden melt of at least half a metre in the middle of the pack. Something to do with the big low/storm sat over the area?

Well-spotted, Peter! A lot of the cyan in the Siberian Seas also turns to black (1.5 m). It could be that the model is forecasting for that low to have an effect. It did show up in the LANCE-MODIS images yesterday.

I think I'll check those ASCAT radar images 4-5 days from now. This area that suddenly is said to disappear in the next couple of days, is the area that caused some uncertainty with regards to ice age (was it MYI or FYI). See the first part in the Winter Analysis post called 'ice age'. Diablobanquisa brought it up, so maybe he will have something to say about it.

This could be interesting, although I don't really trust the model from the NRL as it seems to overestimate thickness.

Espen Olsen

I bet the ice of the coast of Joekelbugt; North East Greenland, will disappear this year, it rarely happens, but this season I believe it will, the ice is now now catching that blue crispy fragile color and it will soon disintegrate! Please check the Modis images regularly it is pretty fascinating!

Kevin McKinney

Speaking of the sidebar RSS feed, there's currently an item about the Emperor penguin being endangered by sea ice loss. I was interested to note that in it is mentioned that one Emperor colony has already been wiped out by climate change. Googling back a bit, I found what appears to be the primary citation in the literature:


Interesting to note that there is a discrepancy; recent news reports claim a form population of 250 breeding pairs, whereas the actual study claims 150. Whichever, they seem to be gone now.

L. Hamilton

"the ice is now now catching that blue crispy fragile color and it will soon disintegrate!"

Espen, as if to punctuate your sentence, there's a bite missing from the N edge of the Joekelbugt ice on MODIS today. Looks like it was still attached yesterday.

Alberto Silva

Those forecasts:


do NOT show a clear dipole in the next few days.

Does anyone knows the source of that forecasts and how much trustworthy is that source?

Daniel Bailey

Congratulations, Alberto! You have made it to a climate science website from a climate nonsense website! Not all who dabble in the dark arts find their way back to the light!

Alberto Silva

Daniel Bayley:

In my link there is the description: "NCEP GFS 0.5x0.5 degree MSLP 8-day Forecast Animations".

It looks like an improved version of those maps:


The latter website, NOAA map room, is known to me by already some years.I used it to follow heat waves, the Arctic Dipole, Trade Winds,etc. In particular, when there weak highs instead of strong highs in the Eastern Pacific(i.e. a negative pressure anomaly in the Tropical/subtropical Pacific), it's likely the birth of a Kelvin wave (and if that continues, an El Niño, a persona non grata in my country, Peru.

I do not followed the link to the blog, and effectively, is a typical right-wing anti-science blog. I however trust NOAA.

The "policlimate" maps are from NOAA or are fake?

Account Deleted

CT ice area increased slightly today - with "gains" in the CA, Laptev, East Siberian, Beaufort and Chucki - I guess we are into stage 3/4 of the 5 stages of melting in those areas and the melt ponds are disappearing. We have maybe a bit more to "gain" in the CA.

Account Deleted

And Happy Summer Solistic by the way.



"I bet the ice of the coast of Joekelbugt; North East Greenland, will disappear this year."

There is soild contiunal melt coming for the Canadian high Arctic and North Greenland the next 10 days. Alert, NU, Canada is expecting increasing temps. It has already experienced several days since 15 June above 50F/10C. The next 2 days call for highs of 7-8C and rain showers. See:


Also, Eureka, Canada has had above freezing temps since 1 June. Tuesday it hit 13 C.

See: http://weatherspark.com/#!dashboard;ws=27709

Eureka's 14 day forecast calls for above freezing temps, with daytime highs above 10C. See:


Thule A.B., Greenland had above freezing temps on the 18-19th with light rain. See:


Thule's 14 day forecast continues with above freezing temps and mainly sunny days. See:



Dodger, June 17, at 19:25

Yup. About 365-370 W/M^2 on 3 days of out of ~300 starting June 5th over the period 2000 -2010 on the chart that Rob Dekker (and Neven?) posted here early on last year, during “spring training”, when century breaks were only still only approaching. Those 3 dates as well as another 43 days exceeded 350 wm^2 out the total. (33 of the 46 with more than 350 w/m^2 values were in the two years of 2004 and 2007. Those 2 years also have between them another 16 days missing due an early ice break-up.)

Adding them all together yields about 20% of the dates around the solstice that made it up to a 350 wm^2 or greater. The chart is hard to extract all the date points needed to construct an average insolation for a date (besides being much work). But, by my eye, it’s clear the average day is between 250 and 300 wm^2 for the 4 weeks around the solstice. There’s a wide variance for a date between years, as well as the bulk of the high level days occurring earlier in June, markedly less variance with no day over 300 wm^2 in any year, save the rather cloudless 2004 and 2007, after the solstice. Now under NOAA, the observatory is at the base of the Point Barrow Spit and is dominated by the ocean. As the ice and snow melt it gets increasingly cloudy. If the atmosphere were similar to northern Africa’s, surface insolation would be similar, though for quite different reasons.

The Sahara has some atmospheric attenuation factor of it own and not shared with the Arctic; it’s dusty.


Shows in Table 1 peak surface insolation values, in M/m^2, averaged over the period 1979 – 1992, with (peak dates) of:

Sahara-Whole Area: 377.07 (17 Jun)
West Sahara: 364.29 (16 Jun)
East Sahara: 380.37 (14 Jun)
Saudi Arabia: 391.87 (21 Jun)

Figure 2 gives a good picture of things. And, Figures 3(a), (b), (c) show the problem of extending one particular spatial or temporal point to a much larger area.


G of L, June 17, at 02:23

Have a link for the Brazil stuff?

I spent some time with the Canadian site that you link. It shows an area receiving 6.7 kWh/m^2 (670 w/m^2 to get consistent units). That just isn’t creditable. In the Sahara the peak average day is ~390 and Barrow, at the latitude of the southern tip of area you noted, ~1% of the days reach ~370. (See my comment in reply to the Dodger.) Those other locations have issues, but seeing claims of values 70-80% greater and the bananas vs coconuts balloon forms over my head.

That site is concerned with solar volcanic, and by varying the angle of a panel away from parallel to the surface, I could get a bit more than that the 670 value (actually I expanded the longitudinal area of greater than 670 w/m^2 and shifted it south). When I did that, the north end of Ellesmere Island, the part roughly above 80º N, also dropped down into the band labeled 560-670 w/m^2. That argues against the idea that insolation at the surface simply increases with latitude up to the Solstice and down again linearly after.

Alberto Silva

It's my impression seeing this:


or the Russian Summer is shattering the Siberian Ice like ground beef?


Thanks for those numbers, Whitebeard. I've added: (edit: almost as much as in the Sahara). Sounds even more impressive that way! ;-)


Neven, June 17, 00:32

The areas of the planet that are now receiving the greatest amount of radiation at the surface are the ones near the Tropic of Cancer where the Sun is close to perpendicular to the surface, and as I noted, are deserts with few clouds and quite low water vapor levels.

The phrase that I italicized when I quoted you is what’s still arguably wrong. I’ve no beef with the rest of that sentence; I think it stays below the language usage fence top as you almost invariably do. I wouldn't council any change in your blog, but would, instead, point it out to people as an exemplary example of how to run one.

The Arctic is now experiencing very high levels of insolation, as high as any other part of the globe’s surface, or near enough not to matter much. My original comment, a bit of tongue-in-cheek, was intended to provoke thought on factors, really never mentioned that I’ve noticed and, perhaps, not so obvious to those residing at the end of palm lined drives, well away from the poles. It’s more than just photo period!

On re-reading my comment, I see I’m guilty of overstatement in saying the surface insolation is relatively puny. Relatively doesn’t adequately qualify what I meant which is: punier than most probably think. On something of a roll and pleased with “puny” emerging from the rats nest of my mental process, I over reacted to what I think is people’s tendency to read a (using air quotes) “more that anything else” description, especially when following and amplifying a “it’s huge” type statement as “and it’s lots more”. The perception of what others actually think and understand is squishy, of course.

To the point of fact I raised. After a limited amount of searching for seasonal isolation data at the planet’s surface I'm mostly frustrated. The best source of info to illustrate what was getting at is here:


It includes the graph that Rlkittiwake @ June 17, 17:52 linked and is from a page sourced to U of British Columbia, Okanagan’s Dr. Michael Pidwirny as part of High School level text. The third graphic (fig 6i-3) integrates day length and geometric factors (fig 6i-1) for isolation. I don’t understand where the additional 30% in the last sentence of his 2nd paragraph comes from.

Again, the increased surface area (fig 6i-1) that a “unit” of radiation impinges on offsets a good portion of increased photo period. Notice that the difference between 60º and the Pole is about 25 w/m^2 or ~5 %, and at the 3Oth parallel the difference is slightly less than at 60º. During quite a bit of the increased photo period, at the height of insolation, the Sun is near the horizon. “The dawn [does not] come up like thunder outer China 'crost the Bay!” rather it's like Sandburg’s fog that “comes on little cat feet”. At the Circle it will be on the horizon at midnight but only ~45º above it at noon. Further north it will above the horizon longer, but always nearer to it. Then throw in clouds and water vapor from a marine environment.

Finally, at lest for now, as my fingers weaken and cramp from extended exercise, the ice covered Arctic is more that just the pole. The surface area between 70 and 80 degrees, where almost all of ice coverage is going by-by, has an area 3 times greater than that between 80º and the pole. Just as the weather and sea surface temps at various places go into an understanding of what’s occurring, so does insolation at different locations, longitudes, and intensities.


Nevin, June 21, 07:13

Ah, a little spicy and quite consistent with the data. Keeps the customers coming back for more.



I went looking for those insolation numbers, having read that the Arctic receives a lot of sunlight around summer solstice. I had found that pink and black graph that has been linked to here, but didn't realize that was for the top of the atmosphere.

Anyway, 'almost as much as the Sahara' is plenty enough for me. ;-)


I think I'm seeing effects of that low again: Arctic.io.

Seke Rob

This Arctic.io conjured up the [horrible] picture, ca 1955, of an Italian Icecream maker [now representing humanity], who got stuck with his tie into the big mixer and got spun in to get choked. Evil word was that it was no accident [the doing of a mobster, today represented by the Heartland].

Sorry, if too graphic :|

Seke Rob

IJIS/JAXA in the mean time seems to have dropped attempts to give a current day number and is now headlining:

The latest value : 9,793,750 km2 (June 20, 2012)

The original 19th value got twice updated to go from 9881563 to 9896563 (small adjustments), moving the < 10 million extent step to the record at 11 days or 91KKm2 reduction per day, the quickest since 2002, except it was done before 2002 went below 11 million square http://bit.ly/IJISMD. So happens this is the start of the JAXA fly time. NSIDC counted 9 days for the 11 > 10 million step in 2002 (off from more smoothed statistics).

Rob Dekker


I spent some time with the Canadian site that you link. It shows an area receiving 6.7 kWh/m^2 (670 w/m^2 to get consistent units).

That does not sound right.
Isn't 6.7 kWh/m^2/day something like 280 W/m^2 ?

Seke Rob

Breaking out NSIDC [daily data through 31.12.2007] and JAXA [Since July 2002], this is the picture for the 11 > 10 Million step: http://bit.ly/EXT10M

Seke Rob

As a forward look, last year made this chart: http://bit.ly/CTASTP (needs update) and http://bit.ly/EXT06M and pondered on the Area step change smooth getting below 6MKm square starting 1998 [the Super La Nina year]. Correlation not causation and few weeks away from reaching that 6 million area point again, will there be a more graphic step change or a smooth extension of prior decade?


So this is something of a curiosity. I was looking at the OSTIA SST charts, and I noticed this patch of positive SSTs to the north of NW Greenland, but this is down as 100% ice concentration on both the Uni Bremen and NSIDC charts.

As far as I can see all the other areas of positive SST coincide with ice-free conditions, so it's a bit of a puzzle as to how OSTIA manage to calculate a positive SST there.

Ghoti Of Lod

WhiteBeard, Rob Dekker:

Rob you are exactly right. The numbers I posted are daily totals not flux densitys. In case anyone still cares I stumbled onto the Brasilian numbers in this publication http://sonda.ccst.inpe.br/publicacoes/eventos/Rio05_FRMartins&SLAbreu.pdf

I've been frustrated by the fact I know irradiance is measured by weather services such as Environment Canada but they don't seem to publish the data anywhere on the web.

I'm sure I've seen (but of course now can't find) research out of Barrow AK correlating seasonal total solar input to the time of fast ice breakup.


Meanwhile, over in lalaland, Walt Meier pops up to cast a pearl or two, here:


...mainly in response to an earlier post from WUWT regular "adjustthefacts", which set out to dispute whether warmth melts ice.


Here is a different view from the UK/Godiva sea ice imagery. it is displaying an area of approx 40 percent concentration NW of Alert, CA in the Arctic Ocean.

You'll need to follow the previous instructions I posted. Also, if you want to display or save in Google Earth - use the lat-lon view of "sea ice fraction" in the NCOF - Global Arctic Ocean tab to get the Google Earth view. You can save each days view in Google Earth or image save to a jpeg.


Kevin McKinney

"...mainly in response to an earlier post from WUWT regular "adjustthefacts", which set out to dispute whether warmth melts ice."

Bet he tilts at windmills, too!

michael sweet

The Barrow and Kimmirut web cams show the fast ice has broken up in both locations. this is the second earliest date for breakup at Barrow.


Michael, I don't think the grounded ice (in the background) has broken up officially yet. The ice in the foreground is not the ice that has to break up. It's those floes a mile or so out from the coast that tilt and then get stuck in the sea bottom.

michael sweet


On the Barrow web site they say

"We define
break-up as the first detected movement of landfast ice shoreward of grounded ridges within the 20 m-isobath off NARL"
my emphasis
They do not count the first 100 meters of ice that has dirt on it from the road. It is hard to see the middle area from the web cam but it looks like most of the ice is moving. The outer area of the ice (past the grounded ridges broke off a week ago. The grounded ridges are the last to go.

Espen Olsen

Barrow and Kimmirut: I believe it is only tidal water at both location, but we are not far from a break up!


I wrote several blog posts about the Barrow Break-up last year. In Barrow Break-up 3 there is a comparison of webcam images for the 22nd of June.

It could break up soon, perhaps it even has already. Too bad the Barrow website hasn't started issuing forecasts yet.


Greenpeace and various slebs are starting a campaign about the Arctic:


Christoffer Ladstein

Truly times they're a changing...
Temps in Kimmirut rose as a rocket in the sky today, ended up at staggering 20,1 C, was'nt even close last June, happened first 23. July! So both the landfast ice (not much seem to remain!)and seaice, will "evaporate" at a staggering rate if this continues.

Also Polarstern is well positioned just west of Svalbard, wonder if they dare cross the Northpole this early in the season?! Perhaps last year was to easy a ride, and thereby seek out more resistance and "action" this year?

I am slightly puzzled about the lack of logic concerning the steady drop of volume of ice, meaning this just have to in the next turn lead to earlier meltout both in extent and area?! So even if we don't get optimal meltout weather, we nevertheless ought to break even to prior minimas, or just below...

Enjoy the summer solstice parties and bonfires tomorrow (or today!), just don't make it look like Siberia these days...

Peter Ellis

Must be melt ponds (or tidal water rising through thaw holes?) at Barrow - the radar's still showing the fast ice as present.

Note also that the webcam is in Barrow town, while the forecast (when it's running) refers to the northernmost tip of Point Barrow.

The comments to this entry are closed.