« Everything you want to know about the jet stream | Main | If this is real... »


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

Even though AMSR2 still seems a bit jumpy, IJIS remains an old favourite.

As far as I know, the IJIS graphs (for 2012 and 2013) are still based on the Windsat satellite.
( therefore those graphs were still updated on 11-14th of May, when the AMSR2 maps were not )

R. Gates

Nicely done Neven, and I think your overall analysis is spot on. We are right at the edge where the Hudson, Kara, and Chukchi should begin fairly steep melt rates and bring 2013 back into close proximity to recent years. It is the thickness of the ice that give it staying power in the strong melt month of July and August-- your garden could "pop" then.


Masterful summary, very helpful. I share your view that "the Arctic sea ice pack could soon go POP under the right conditions."

Here is what a quarter-million sq km of the central Beaufort Sea looks like on 25 May 13 using a Modis Terra visible scene with 1 km pixels. The original is mostly shades of white on white but in fact retains sufficient contrast for the state of cracking to be drawn out with image enhancement.

Earlier years from this same satellite (2009-2012) can also show very extensive cracking and even more open water along the shore for this same date.

Amundsen Bay has broken up only slightly ahead of 2012. I will put up an animation as soon as Modis finishes the last swath this afternoon.

 photo BeaufortModisCracking_zps8efe0586.jpg


Great stuff, A-Team. I'm looking forward to your additions and excellent visual stuff as the melting season progresses.

I also forgot to mention that snow is melting very fast in Siberia, and it's just a matter of time before it'll start to affect air temperatures for real. Here's the Rutgers Daily Departure for May 25, 2013 (Day 145):


Question/request for A-Team: I would like to see if you would be willing to email me a high-resolution copy of the Narcissus image (looking at his reflection in the cracked ice). I think that image is such an incredible metaphor for what is occurring on this planet. Everyone who sees it is absolutely amazed. I have a friend with a high quality printer that can print on various substrates & I would like to create an image for the wall of my office. I have not been able to figure out how to send a private message or email to another typepad member & so I am risking clogging up the blog with this request (with Neven's permission). My email is brrader then that symbol that means at and it is hosted by yahoo. Thank you.


Wipneus, I believe I inadvertently deleted your comment while releasing it from the spam filter. It was something about Windsat.



Thanks Neven. Solid work as usual.CT has apparently been out of commission in recent days, so there are no numerical updates on what is happening to ice area in the various seas, but some things I've noticed in the meantime from NASA's Earthview include:

1.) The Chukchi is retreating apace. A new polynya has opened up in the northern Chukchi that is evident as of May 24th. Meanwhile, temperatures are in the upper 60s/lower 70s Fahrenheit in much of Alaska, and the upper 40s towards the coastline. So, there goes the rest of the snow in the PNW. The deepening of a polynya in Beaufort is also apparent.

2.) The Siberian side has slowed down considerably, but based on concentration maps, it's a slushy of ice floes and polynyas, especially in Kara and Laptev, so the melt that is to come in that region is waiting in the wings for the right weather conditions. As long as a sharp refreeze doesn't manifest (which becomes less likely as we press forward to the June solstice), they don't stand much chance anyway.

3.) Hudson Bay and Baffin Bay are continuing to hold somewhat, but they are visibly retreating at this point. Nothing too interesting going on here yet, IMO.

All of which are the expected steps that precede the core Arctic basin melt. The question, though, lies in seeing to it that the ice is really as fragile as we imagine it is. I believe it is, but whether we break 2012 or not is somewhat of a wash right now. But the central Arctic is looking more like a mesh of ice cracks waiting to totally splinter into individual floes than a cohesive and strong unit. It's late May, but it's much too early to say the ice is going to fare better than last year. Just another day in watching the Arctic circle the drain.


Well said and well observed, GreenOctopus.


If the minimum this year isn't below last year's, you can expect the Morano-types to say the Arctic ice is "in recovery." Count on it.

Dan Ellis-Jones

Thanks Neven, great summary.

I've been looking at the following forecast...

Which seems to suggest a general warming trend over the next week/week and a half, especially along the East Siberian coast, with a few days of 15C or more, and a strong southerly pushing across the East Siberian and Laptev seas. It also suggests a decent amount of warming through Bering Strait and Chukchi sea.

If the forecast is correct, then we might start to see a ramp up in the melt season.

As an incidental point, the temps over the lower latitudes of Canada/Alaska and Siberia seem pretty high, and so the snow cover should be having a tough time, and with temps of 20C+ there will be some permafrost melting I would suggest. Would be good to take a look at methane anomalies for these areas in 2 weeks' time. Also, northern Scandinavia seems really hot for this time of year.

Worth noting though - this forecast does seem to enjoy making the more distant forecast days warmer than they actually end up.


Great work Neven!
Your work is very important for this generation, and for future generations to understand what we are currently experiencing.
Keep up the excellent work.

I have been watching the weather, and the ice, for many years.
What, I think, we are currently experiencing, is a major climatological shift. Although the extent of the arctic sea ice this year, is off to a slower start, and is more gradual (less erratic) as compared with last year, I think that things will dramatically change, in the coming weeks. I think that we will witness an 'ice extent cliff'. I feel that there are several factors which will contribute to this. Firstly, with the anomalous air inversions happening since last year, wave after wave of warm tropical air have been forcing their way into the arctic, with the subsequent forcing of 'cold air' (remember that there is no such thing as cold, just an absence of heat) to southerly latitudes. Secondly, the ice is very thin, fractured, and spread out. Thirdly, the area of ice which is still 'pack ice', is dramatically reduced from previous years, and is composed of far less resilient multi-year ice.
This should result in far more 'ice transport' out of arctic waters this year, and a resulting positive feedback loop effect for adsorption of solar radiation, and hence faster melting of the fractured ice pack.

It is important to put these things into context, however. We know that the Planet Earth, was once much warmer than it is today, in previous Geological Epochs. More recently, the 'Little Ice Age' had only ended about 1850 (Maunder Minimum). It is also important to recognize where we are on the 'Sunspot Cycle' (Milankovitch cycles).
There is always a lag-time effect for all of these things.

Can we adapt to the changing conditions, on a very overpopulated planet?

I like to check the Canadian Ice Service data frequently. I think that the MODIS Composite images are telling.


Artful Dodger

Troll alert: "Spacezorro1"

If you are the same skeptic that Neven has banned twice already today, you are wasting your time. You will receive no feeding here, troll.

Better go back to tweeting on SouthPark, while following no one, and having no followers, on Twitter.


We know that the Planet Earth, was once much warmer than it is today, in previous Geological Epochs. More recently, the 'Little Ice Age' had only ended about 1850 (Maunder Minimum). It is also important to recognize where we are on the 'Sunspot Cycle' (Milankovitch cycles).

Can we adapt to the changing conditions, on a very overpopulated planet?
Posted by: Spacezorro1 | May 27, 2013 at 06:33

The Maunder Minimum ended about 1715 - there was another, less severe minimum of sunspots, the Dalton Minimum, that coincided with the end of the "Little Ice Age", about 1830 - with the "LIA" ending about 1850.

Milankovitch cycles refer to the interplay of eccentricity, axial tilt, and precession of the Earth as it orbits the Sun - this leads to orbital forcings which are fairly easy to calculate. As far as I know, these are independent of short term Sun variations in sunspots, etc. - I don't think they can predict, or even explain, all those slight irradiance variations of our G2V star.

Can we adapt to the changing conditions, on a very overpopulated planet?
Certainly small tribes of humans are very resilient - a few million humans could survive another Ice Age, or 9° F average global warming. But Global Civilization ? Which evolved during a very stable Holocene climate, built on top of global agriculture that expects a fairly stable climate?

I don't think so.

Changing hydrology will be the downfall - droughts and floods. Disappearance of glacier fed rivers in Asia. Agriculture will crash, billions will starve, wars will break out, and nuclear wars will be brutal. 90% fatalities brutal.

Can small tribes of humans survive on a hot, radioactive planet ? Maybe.

What about the methane hydrates? Depends on the inertia of CO2 in the pipeline before the Wars. Things could go from Ugly, to Very Ugly.



I hope Dr. Maslowski figures out the Regional Arctic System Model (RASM) before the summer Arctic Sea Ice is gone this decade - here's a short article from Feb 2013:


Could you offer some thoughts on the potential reasons behind the record sea ice melt of 2012?

In my opinion, increasing heat content in the subsurface western Arctic Ocean, together with the snow-ice/albedo effect, advection of warm summer Pacific and Atlantic water, and stronger air-sea coupling due to thinner or no sea ice is one of the main reasons why the summer sea ice cover has been declining in the Arctic. This extra energy and its storage in the upper ocean can help explain the longterm negative sea ice trend and especially its acceleration since the late 1990s. The entrainment of this heat into the surface mixed layer (where it can affect the growth or melt of sea ice or be released to the atmosphere) is controlled by small scale processes, such as eddies, upwelling, coastal currents, mixed layer depth and vertical stratification.

For a long time, Dr. Maslowski has argued that these hi-resolution, small-scale processes that GCM's miss because of their large grid sizes, are the reason they predicted the disappearance of summer Arctic sea ice towards the end of the 21st century. Regional, hi-res models can capture important small scale processes, and lead to better predictions.

2016 ± 3 years.

Artful Dodger

Anu, please DNFTT.


Artful Dodger -
I've been away for a couple years - has Neven's productive little blog been discovered by argumentative mouthbreathers? I suppose it was bound to happen.

But it's hard to tell a curious teenager from a Christopher Walter Monckton, 3rd Viscount Monckton of Brenchley...

Artful Dodger

Hi Anu,

Yeah, just click on the 'Username' to see the comment history of the User in doubt.

This particular troll is a first time poster, from a previously inactive Twitter account, and check out the fake name:


Sound like a 'superman1' clone to you? Have you seen what that troll has done to CP recently? Just best to drop this, and him, now.


Espen Olsen

Re posted from Arctic Sea Ice Forum


11,816,563 km2 (May 26, 2013) dropping 6,250 km2. This is 229,579 km2 above the 2000s average. And 164,500 km2 less than the 1990s average.


I realize there's been a recent de-emphasis on day-to-day SIA/SIE values, and for good reason. But, still, these stats caught my attention:

--IJIS sea ice extent decreased by 911k km2 from 1 May through 26 May. That's the smallest extent drop for that period in at least the past 11 years, and it is, in fact, smaller than the average 5/1-5/26 drop for the decades of the 1980s, 1990s, and 2000s. (The average over the past ten years--2003 to 2102--has been 1.24 million km2, and the 1979-2012 average has been 1.26 million km2.)

--CT sea ice area decreased by 971k km2 from 1 May through 26 May. That's the smallest extent drop for that period in the past 16 years (since 1996), and it is, in fact, smaller than the average 5/1-5/26 drop for the decades of the 1980s, 1990s, and 2000s. (The average over the past ten years--2003 to 2102--has been 1.3 million km2, and the 1979-2012 average has been 1.27 million km2.) Only three years in the record have seen smaller drops for that period: 1989, 1990, and 1996.

I'm not implying anything; I'm just bringing up some info I found interesting...


As GreenO notes it is "just another day in watching the Arctic circle the drain."

Below I've updated the slow motion breakup of the Nares-like arc bridging Cape Bathurst and Banks Island. It began about a week earlier than 2012 but has not been as dramatic.

I've included a conventional map for place names and local km scale. The biggest ice island calved off is 77 km in length and 1081 sq km in area or 18 times the size of Manhatten Island (58.8 sq km).

The very top of the animation shows the Mackenzie River and its delta. There is some breakup of landfast ice, inland snow melt and suggestion of an early freshet under both Mackenzie and Kugmallit bays. The Mackenzie will put some 330 cubic km of fresh water into the Beaufort in coming months from its 1.8 million sq km collection basin.

Full-on sediment blow-outs will show up in the next week or two, going by past years. As shown on the AMAP chart, this is a very specialized area with bottomfast ice, piles of ice grounded on shoals (stamukhi), salinity variances and very shallow continental shelf, making it an unsuitable gauge of overall melt season.

 photo amundsen2B2_zpsd5817b4c.gif

 photo stamukhi_zps9d4bc705.jpg

 photo canMap2_zpse60dc104.jpg

Hans Gunnstaddar

Jp, I agree it is looking like a rebound year. Not that the trend is down, just that this season may have more ice extent than 2012.

Although it is unintentional on the part of nature, the net result of a trend-line that undulates is it's as if nature is sneaking up on us like a good Indian. Every time a new record occurs climate change is given the mantle, and every time a rebound occurs it provides ammo for the recovery crowd. All the while we edge closer to a major tipping point.


Islandr, the best resolution I could get for public domain Narcissus -- 844 x 1033 -- is here. It would have to be resampled up to 300 dpi to print adequately.

It is about my favorite too -- and would make a good cover photo for the next IPCC report.

The problem is, Narcissus already thinks this whole blog is about him. Each mention reinforces that, like feeding the dog at the table. What's next, Neven posting his favorite song ("You're So Vain" by Carly Simon) as background?



The Ice and Snow data page from the KlimaCampus of the University of Hamburg marks as "New New New..." some AMSR2 ice concentration images and data. Sample image:


(AMSR2 Arctic sea ice concentration on a 3.125 km grid, highest resolution anywhere AFAIK)


Don't throw yourself under the bus based on this, but here are the best matches to the today's breakup date for the Mackenzie Bay ice shelf (defined, somewhat arbitrarily, as first big loss past Kay point). With the exception of 2010, the dates are ten days or so later (15, 10, 6 resp.).

I am finding 2010 to be quite odd across the Canadian Archipelago; it seems the land interior was systemically warmer early on. Chris has noted anomalies in Piomas that spring as well.

And note the mud flows from the river on the ice in 2011 (and seeminly under it in 2009). The mud can be seen in upper Mackenzie distributaries in the others, including 2013 (arrow). The Mackenzie is the only North American counterpart to the great rivers of Siberia.

While their cumulative impact is quite important to Arctic Ocean salinity etc, what happens with the Mackenzie in early spring will have a limited local impact on the Beaufort Sea.

 photo mckCompYears_zpscaf906f6.jpg

Hans Verbeek

Like your paraphrase of Joe Bastardi's "Enjoy the weather".
Slow start of the melting season is something to be cheerful about, I guess.
The longer than usual winter in many parts of the Northern Hemisphere may be a factor in the slow start.


Like your paraphrase of Joe Bastardi's "Enjoy the weather".

I'm glad someone noticed. If I make more vids, I think I'll use it more often. :-B

Slow start of the melting season is something to be cheerful about, I guess.

Definitely, but I'd be more cheerful if it stays like this all melting season, and then a couple of years in a row.

The longer than usual winter in many parts of the Northern Hemisphere may be a factor in the slow start.

Could be. It's crazy, BTW, how cold it has been here in this part of Austria in the past couple of days. But in great parts of the rest of Europe as well, or so I've heard.

(AMSR2 Arctic sea ice concentration on a 3.125 km grid, highest resolution anywhere AFAIK)

Wow, that's amazing! Thanks, wipneus! Pretty big file (20 MB)...

What's next, Neven posting his favorite song ("You're So Vain" by Carly Simon) as background?

I will if there's a new record! :-P

Hans Gunnstaddar

"Could be. It's crazy, BTW, how cold it has been here in this part of Austria in the past couple of days. But in great parts of the rest of Europe as well, or so I've heard."

Sprinkling rain & 61.3F at noon here in No. California today a few days short of June! Usually low 90's where we are this time of year.


@HG - I think you may be overestimating typical late May temps in N CA, which has fairly frequent shifts back and forth this time of year.

That said, the heat is still there. The question is, where is it going?

Kevin McKinney

Slow start to melt presaging a 'rebound?'

Maybe, but I'm not altering my prediction for September. If I'm going to be wrong--and odds are I will--at least I'll be steadfastly wrong. Oh, wait, that didn't sound right...

Anyway, it's still too early to tell what the melt season will bring by July, let alone September.

Hans Verbeek

It will be a miserable summer in the Netherlands, Neven.
North Sea SST is still more than 1°C below the 30-year average for the last week of May. It will take a lot of sunlight to bring the temperature back to normal.


Remember when everyone was excited about the cracks and predicted severe melting?
Now the start is a bit slow, so what?
It can change within days.
Temperatures are finally back to positive anomalies, let's see how this will change the current situation.
And... of course, PIOMAS will hopefully show us soon, what was really going on these last weeks.


Continuing laments on cold anomalies… I experienced ground frost around my house in the French Dordogne region last Sunday. Which is not impossible at the end of a clear night on 430 m ASL. But I have no recollection of that over the last eight years.
I can’t stave an opinion, as I’ve not been following NH weather as close as I used to. But I’m inclined to think it’s all about changing patterns, weird atmospheric wave behaviour.
As I suggested before, AGW could well be manifest in other niches than as pronounced in the Arctic as through the last few years. I noticed the Antarctic to be often anomalously warm these weeks.

Still, MODIS shows this loose mesh structure all around the CAB. Don’t be fooled by extent in the periphery. Four weeks of unusual weather could wreck the pack.


I usually look at the barrow webcam regularly in May as it is a indicator of how things are going.

I noticed, earlier in the month, that it was down to 11F. Which is pretty cold compared to the last 5 years or so, so no surprise that the ice wasn't changing.

Then it jumped to over 32F for a few days and melt pools began to form.

I've noticed that the AMRS2 concentration maps from Uni Bremen are now beginning to show quite comprehensive change and melt beginning over large areas of the pack.

I'm guessing that a late start with such thin and fractured ice is not going to be a herald of a re-growth. In fact it could even be that all the extra moisture, from that anomalously warm sea and all those open leads, is keeping the sun out at the moment. That won't last though...

Time will tell. I expect it will be a record, or close to it, ice loss for June unless the weather intervenes in a big way.


It'd be great though if it would last another 10 days or so, and Goddard writes something silly about it, that then gets picked up by Watts. But I guess they wouldn't be that stupid after what happened in 2010...

Jai Mitchell

I have been watching the progression from the Naval Research Laboratory HYCOM/CICE images and it appears that there is significant mid-floe fracturing and thinning which has been progressing very differently from last year. While sea ice extent appears to be about the same, the rate of thinning indicates that we will begin to see a rapid acceleration of the loss of <1 year ice in the next week or so.


If this does occur it will be about 35 days ahead of last year's melt.


the slow melt so far is no surprise given the DMI 80N temp chart - this May has been quite a bit colder than other recent years (http://ocean.dmi.dk/arctic/meant80n.uk.php). weather is weather. well, and the less said about earlier predictions of a complete melt-out this summer the better - you know who you are


Add to my previous posted comment:

Greater 'in situ' melting, due to the simple fact of exponentially more ice surface area being exposed, because of the more fractured nature of the ice this year, as opposed to previous melt seasons. But, this will have a localized negative-feedback effect, keeping some arctic waters temporarily cold.


Making a sharp contrast with a month ago backwards in time , we have elegant Cyclones dominating most of the Arctic Ocean, I expected this, because there was unbelievable continuous observations of adiabatic processes at the surface to air interface, for most of the winter. The reason for the anticyclonic earlier on dominance, creating more ice accretion, was likely from the remaining pack ice melt in september 2012 prompting a natural base for anticyclonic activity standing over ice creating colder air heading towards warmer air from the wider open Arctic ocean surrounding the surviving pack. One must not forget the greatest melt in 2012 left an "adiabatic legacy" , which is lesser sea ice volume, making the sea ice ice overall warmer thermally. This has created the the apparent cooling at the center of the pack. But an adiabatic process causes more clouds, Low clouds which are essential for a great thaw. The root causes for a great melt need be more understood, while extent is a good indicator of the pending future to come, it can fool just as simply as an adiabatic process lifting warmer air to rise.

Recently I observed a great turn around, from cold to warm , like someone pulled the freezer plug. I don't quite understand this switch, it was not a warming gradual process. But it is all of the sudden very very warm for most of the CAA (3 d wise not only the surface).


Thank you, Neven for this thorough overview summary.
I knew these must take you a lot of time to write in 'pre-digesting' all the info for us, am not surprised it has been 3-4 hours a time. I'm just hoping now there are sufficient funds trickling in through the donations tip jar to assuage Mrs Neven, even a little bit, over all the time you are spending on us here!

Also belated congratulations on the now +500 blog posts, and its good to see other posters contributing this blog, as well as all the quality comments. Thank you to all.
This all makes for fascinating reading for me, much way over my head, but I am learning A LOT lurking here.
And I have just 'bitten the bullet' & started the Coursea Climate Literacy course, may no longer just glaze over at mention of units of W/m2 etc......and its only week 2!

& Thank you A-team for this beautiful image which I've saved as inspiration for my arctic inspired artwork.

Rob Dekker

After Larry Hamilton's "predict the Sept ice extent this year", I've been doing a bit of analysis on the relation between snow cover in May/June (by Rutgers snow lab data), and ice extent in September (NSIDC).

Of course, since early summer low snow extent absorbs a great amount of heat during this time when the sun does not set over the Arctic, so one would expect a positive correlation. But how much ?

Now, it turns out that the linear regression suggests that 1 million sq.km. snow cover reduction at this time of year will cause abouot 0.5 million sq.km. of ice extent reduction in September.

The correlation between these data sets is pretty good (R=0.8 for May snow extent, and higher for June) for prediction but by itself still results in a standard deviation of about 400,000 km^2 in Sept ice extent.

Just wanted to report that, as I find it interesting that a physical effect (reduced albedo in May/June due to snow) indeed appears to show up in the September ice extent record.


Rob Dekker: that's interesting as there could be a physical mechanism too. Early snow melt may lead to warmer river waters, as the permafrost regions are more likely to be on upland, so very small changes on tributary river areas could converge at the outlet.

Remko Kampen

"It will be a miserable summer in the Netherlands, Neven.
North Sea SST is still more than 1°C below the 30-year average for the last week of May.
Posted by: Hans Verbeek | May 27, 2013 at 22:07"

Invalid conclusion. There is no connection. No North Sea 'exists' during easterly or southerly winds. Also there is no relation between spring and summer weather. So the cold spring of 1962 was followed by a cool summer, the cold spring of 1983 was followed by a very hot summer.

Btw, the two fair days we are experiencing now are from air that had travelled over the entire North Sea from Norway. Originally this air was from the first Moscow heat this May.


Remko, I had read somewhere on a Dutch news site that the cold was caused by an Arctic outbreak. I looked a bit at the weather maps (Europe not too far from the Arctic), and though I saw something that could be causing it. Was all of the cold attributable to cold Arctic or North Atlantic air? Did it also cause the cold as far as Austria, or further, as someone who visited Dubrovnik (south Adriatic) last week, said the weather was absolutely horrible.

Rob Dekker, that sounds very interesting. If you have any extra info for me, I'd be all ears. I'm keeping more of an eye on NH snow cover this year, having added one graph to the ASIG as well (at the bottom of the page). Unfortunately Rutgers update their maps and graphs with a separate file name each day, instead of a generic file name, otherwise I'd have included them too.

R. Gates

Wayne said:

"Recently I observed a great turn around, from cold to warm , like someone pulled the freezer plug. I don't quite understand this switch, it was not a warming gradual process. But it is all of the sudden very very warm for most of the CAA (3 d wise not only the surface)."


I really hate to keep harping on this, but as mentioned a few weeks back, we had a late season "soft" SSW event take place, that brought down high pressure and some higher temperatures in the stratosphere that descended very slowly to the troposphere. This was the result of a weak planetary wave. The AO did go mildly negative and bounce around the negative/positive line. The effects on the surface have been muted because this SSW was so late in the season.

Here's the animation at 10 hPa that shows this unusual late season SSW:


And you can watch the temperature anomaly quickly fade away over the Arctic.

And in this chart you can see the AO index go slightly negative for a few days as this mild SSW hits in mid-May.


Again, this was a very unusual late season SSW, and now the index has gone positive again. What all this means for the melt season is hard to gauge. Perhaps nothing, though the negative AO index for those few days and slightly higher pressure did serve to shunt a bit more of the colder air out of the Arctic, allowing the DMI Arctic temps to rebound.


I looked into whether it has been noticeably warmer or colder this year relative to 2012 using the DMI 'mean' temperature within the polar cap (80ºN to pole) for 2m height. As the figure shows, this cap is rather off-center with respect to the Arctic Basin (covering about half of it) and the Cold Pole.

This data has various issues, which DMI explains quite well, notably the 0.5 degree lat long grid used favors extreme northern grid cells -- so they say "do NOT use this measure as an actual physical mean temperature of the arctic" though inter-year relative comparisons are justifiable.

Actually it makes more sense to compare sea ice years starting in mid-September rather than calendar years, so I butted together DMI graphs for 2011-2013 and excised 2012 and 2013 ice years. I then numerically integrated (counted pixels) areas above and below 1958 - 2002 climactic mean (green line below) to obtain excess and deficient 'degree-days' (orange and blue fill resp.).

Both years are well above the 44-year mean. While within 10% of each other, net temperature deviations are not necessary comparable in their effects on the ice because they occurred at different times during the season, for example fall 2012 (belonging to 2013 ice season) is noticeably warmer than fall 2011 (belonging to 2012 ice season) and vice versa for mid-winter.

Since these aren't actual physical mean temperatures, the thermodynamic impact on the ice is a bit of an unknown.

Full width 971 pixels, only 415 pixels show:
 photo TemperatureComp_zps0650017a.png


Just to follow up on Jai M's observation of coming accelerated melt (according to Navy ice thickness prediction), I pulled out the relevant frames and compared them with a bizarre pre-melt streaking I'd been watching on Jaxa 36HV 18V, which is also quite conspicuous on the Jaxa 89 Ghz polarization ratio (not included below). Multi-day, multi-channel support tends to suppress interpretive artifacts.

This is mostly happening in the polar quadrant diagonally across from Greenland's but also includes the pre-Fram region. The bottom row of the'film strips' below enhance the gold streaking feature.

The width of these features is very small but nonetheless picked up exclusively starting from known open water color using a color cube radius of 93, suggesting melt confirmation had the resolution had been better.

The 28 May 13 image is just being swathed in now so is not included; it is continuing the trend but so far without the full drama of the Hycom projection.

A remarkable event shaping up or just an algorithm glitch? I'm in wait-and-see mode on this.

Pixel width 1511; 415 pixels show:
 photo navyEarlyMelt_zps36092291.jpg


Very interesting R.Gates. But 10 mb is hardly as influential as 100 times the pressure on the surface. Theory prescribes a colder stratosphere with a warmer surface. I believe this SSW is more like the response from the switchover.

Another big event to watch is the displacement of the Cold Temperature North Pole from Northern Ellesmere to North East Greenland giving the expected results:


But there is no doubt that every level of the atmosphere plays a significant role in what causes your weather. http://eh2r.blogspot.ca/2013/04/cold-temperature-north-pole-ctnp-coming.html


Nice spotting there by wipneus of the new "AMSR2 Arctic sea ice concentration on a 3.125 km grid, highest resolution anywhere".

As Neven notes, its 20 MB file size precludes routine use -- the image is too big for computer screens at 3977 x 6134 pixels (an iMac G5 21" monitor is 1920 x 1080). Animations are a top priority but serving them is out of the question at this file size.

Two-thirds of the file size results from an unfortunate choice of base map -- the Arctic Basin only occupies a third of the image. Where is the urgency in monitoring ice in southern Virginia, the threat from icebergs in the Straits of Gibraltar, or possible re-glaciation of Puget Sound?

The background colored-relief land map is a huge distraction, requiring a separate palette for elevation (not provided) and vastly increasing file size. People who want this can simply paste it in, presumably with that day's snow cover rather than the static mid-summer look.

Land masks are normally used (or at least provided in a metafile, missing here) because they greatly reduce file size, the reason being compression algorithms really like big blocks of pixels exactly the same color (normally pure black is reserved, making it unique).

It is contradictory to provide high resolution data, followed by ruination from lossy compression -- jpg should never have been used here as file format. It dithered the image badly and now the colors on the map no longer correspond to anything in the ice concentration palette.

This palette has many oddities, including white between two pinks at the high concentration end. The palette blocks are uneven in their concentration extent and most are hardly used anywhere. The melt interest really resides in the upper end and that is where visual resolution should be maximized.

All this purported resolution and implied accuracy -- yet here is Cape Cod and even Manhattan shown surrounded by ice at the end of May. Puget Sound is iced up too, while the Juan de Fuca Strait looks impassable to shipping. None of the above. I'm dubious too about what it is showing in the Lincoln Sea and upper Nares. Some of these artifacts might drop out if they had resampled up, say to 6.25 km. That would cut the file size by a further factor of four.

Lat long grids should never be laid over data from a billion dollar satellite, instead held out as an optional alpha transparency channel. Photoshop is now in its 25th year.

There being no meta-files or explanation, I was left wondering what satellite, what instrument channel, what was collected in initial data acquisition, what post-processing occurred, what projection was used, what scale. Above all, where is the url to the 3.125 km microwave data that they used?

I suspect the whole thing is based solely on the Jaxa 89 Ghz channel -- Jaxa just announced calibration on 17 May 13 as blogged earlier, followed shortly by this product. This satellite observes at 6 wavelengths at 2 polarizations -- the idea that one channel alone is optimal for sea ice concentration strikes me as the height of naivety.

The algorithm used is provided in the lower left: ASIv6. This stands for the ARTIST (Arctic Radiation and Turbulence Interaction STudy) Sea Ice (ASI) algorithm, described most recently in free full textsKaleschke 2001 and Spreen 2008.

I don't know what all has changed in the sixth version v6, nor whether the code is offered online. (This would required for publication in an American scientific journal -- how else could it be replicated?) If so, it might be feasible to pull out explanatory lines -- code cannot be maintained for this long without extensive commenting.

I can applaud early release of their new product. It has a few minor issues but we can hope these will be quickly fixed via constructive collaboration, presumably with the person whose email is embedded in the image corner.

The ultimate issue though is whether the higher resolution has really improved our understanding of the ice physics going on on the ground. Beyond what you can see, without the algorithm, simply looking at the PR89 Ghz; the comparison at blog resolution is below.

 photo 89GhzComp_zps5727ebb9.jpg



The data is also provided as netcdf files. A landmask is implicitly provided by marking landpixels with special values, like 115%

There is some explanation (eg the grid) in the files as well, i am using the tool ncdump on linux for that. I think they are using nsidc grids, stereographic projection on a plane intersecting the globe on a latitude circle near 90 degree.

Unfortunately there is a mistake in the files ( specifying the scale as a string value). I did spend my freetime today for finding out the problem and a workaround to read in the files in R.

I am on the road tomorrow, can get you a png with landmask the day after.

Chris Reynolds

May 1 to May 26;

So 2013 is a bit colder than 2012 for May average so far. However the Atlantic sector is notably warmer than 2012.


Here's what a recovery looks like.

This is PIOMAS volume broken down into thicknesses contributing to volume (March).
In 2006 there was a winter thinning over the pack, leading to ice of 3m and above increasing the volume of ice between 2 and 3m thick.

Then in 2007 there was a massive ice loss. You can see that in the 2.5m and thicker ice categories.

In 2008 the thicker ice bounced back, this was because the 2007 event took the ice below it's pseudo-equilibrium state.

Then in 2010 there was a massive loss of volume caused by a MYI export into the Siberian seas. After that there was NO RECOVERY. Unlike 2007/8.

Now in 2013 due to the large amount of FYI there was an abnormal amount of 1.75m thick ice volume. But after the record of 2012, once again there has been NO RECOVERY over winter. Just compare it with 2007/2008's March state if you think I'm talking c**p.

In 2007, 2011 and 2012 there was a massive (apparent) loss of area as seen in CT Area anomalies over June. Graph of CT Area anomalies.
This may be due to anomalously large appearance of melt ponding at that stage. Whatever the cause, when the June anomaly crashes we have a new area record, based on three successive events.

However this is now two weeks away, and the melt in the Arctic Ocean hasn't really got underway because it never does in May. It is after the first week of June that things will start to happen.

After detrending by interannual differences, the correlation between area at minimum and area on May 20 is, 0.2022. That's not just a low correlation, it's not even statistically significant.

The ice will be the judge of the matter. But I'm seeing rather a lot of evidence-lite talk about the significance of this 'late start' for the melt season to come.

There is no significance.

It means nothing.

Chris Reynolds


Thanks for the link to the high resolution AMSR-2, that will be very useful.

Aaron Lewis

Note the low pressure in the center of the Arctic Basin, the cyclonic air flow around it, and the drift / speed of the sea ice under it.

I declare the 2013 Arctic Sea Ice Melt Season -- open for business.

While it took some extra time getting organized, I expect that the 2013 melt season will now operate with breath taking efficiency. Cyclones are heat pumps, and this is a big one.

I assert that it is driven by water vapor from the North Atlantic Drift being condensed on sea ice in the Arctic Basin. Most of the heat will go to melting sea ice, with a small amount going to drive the wind field. The film of water on the surface of the ice will also lower its albedo.

I also draw your attention to the differences between the modeled wind field from the weather models and the wind field observed by MODIS and AVHRR.

R. Gates


I agree with you about "all levels" affecting the weather. The "weak" SSW event that worked its way down from the upper stratosphere in early to mid May certainly did not have the kinds of effects that these events have in the winter, as there is now quite a lot more energy in the lower atmosphere for this kind of event to contend with. But there is no doubt in my mind that it did affect the AO index in mid May, but again, very weakly. As it was, the event seems to be more interesting for the late nature of the timing, rather than the overall strong effect. You can see the descent of this pressure quite readily in this chart:




Thank you for the Narcissus image. I will let you know how this turns out! (but it may take some time)


That was good, Chris, I agree with what you are saying.

I cannot agree with PIOMAS volume figures since I saw no evidence of recovery during the winter -- to the contrary.

For the past day or two there was a nice sized cyclone over exactly that weak spot, around 85N 120E. Will be interesting to see how its effects show up in the various graphs.


ICDC's landmask for the 3.125 km grid AMSR2 maps, extracted from the NetCDF data files here:


John Christensen

Chris and Aaron,

I completely agree that 'recovery' would be a stretch in any sense of that word.

It seems to me the cold snap we saw mid-late February (linked to the SSW event?) with the extensive cracking at the same time did lead to a extensive degree of volume gain, and caused areas like Beaufort, Chukchi, ES, Laptev, Kara, and Barents to be at a comparable or improved state compared to 2012. In winter/early spring of 2012, the SIA was primarily held up by positive anomalies in Bering and Baffin, which quickly melted away during late spring weeks.
Mean temperatures above 80N are now at -3 - -4C, but still 1C below normal so 'so far, so good' is probably the best label for the situation. This winter a weather event seemed to have helped us, not climate returning to some 'normal state'.
Now, the current low pressure will either bring too much heat to the ice, or the clouds will protect the ice from the sun now that we are nearing summer solstice.
Again this is weather, but for June I would prefer the low pressure and clouds and then for August I would prefer high pressure and quiet weather.
Let's see.

Espen Olsen

Repost from the Forum:

11,797,031 km2 (May 28, 2013). And the highest since 2002.
+11,875 km2 more than May 27.
287,234 km2 more than 2000s average.
and 135,344 km2 less than 1990s average.

Remko Kampen

"Was all of the cold attributable to cold Arctic or North Atlantic air? Did it also cause the cold as far as Austria, or further, as someone who visited Dubrovnik (south Adriatic) last week, said the weather was absolutely horrible."
Posted by: Neven | May 28, 2013 at 14:11

Short answer: yes.
Elaboration. If you want a cold spring in Europe, you want flows from the continental northeast until beginning or half April when the land warms up; at this time you want the flow to back to the north so as to utilize the maximally cooled sea and ocean waters.
This is basically what actually happened. The uniquely long run of easterlies by what I called 'The Thing' - the hyper Arctic high in March. Thereafter we've seen cold air pools drop in from the north steered by a northwards/northwestward displaced 'Azores' high.

These Arctic operations have moved south deep into North-Africa. Meantime, though, north of Scandinavia is into a prolonged spell of extremely warm weather, rendering places at 70° NL warmer than the Canarian Islands :)

On the SkS article by John Mason on the jet stream there is a great animation of what happened this year. Bring yourself in a trance looking at it :)
(unfortunately I can't link to the image now as SkS seems off the air).

John Christensen

And great comment from A-team above on the validity of DMI above 80N mean temperature:

It seems the calculated mean is kept down by the cold at the central Pole area due to the low pressure in place (so the heat has not hit the surface yet, or is negated by cloud effect?). The actual 80N mean temperature should be higher and at or slightly above normal.


11,797,031 km2 (May 28, 2013). And the highest since 2002

and CT SIA : 10.94 million on Day 4000.
We have to go back to 2001 to find higher SI area on that date.

Guys, whatever is happening up there, we need to revise our minimum estimates upwards. I revise mine to 5.5 million + ( extent) and 3.9 million + (area)


Was this the cyclone? I guess so...

See the "blue cracks" in the central arctic

You can also see them here:

The outlook is even worse:

Perhaps A-team could make one of his great animations of this new cracking event!

John Christensen

On the cold European spring:

Yes, it has been cold, as also the Giro d'Italia riders can testify with cancelled mountain ascents, shortened stages, etc. (I am riding La Marmotte this summer and fear the top of Col de Galibier at 2645m, which is still closed by snow).

However, with Arctic SIA at decent levels since the beginning of February (ROOS has Arctic SIA within 1STD of '79-'06 monthly average since mid-Feb), it does not seem reasonable to blaim current western European temperatures on disturbances caused by lack of Arctic sea ice. Therefore, regional temperature anomalies seem to be resulting from local weather events.
Or would people disagree on that?


Wipneus, thanks so much for extracting that land mask. Below I compare it to the one provided by JAXA for their 89 Ghz. The 3.125 km mask is evidently not derived from it given the peculiar rotation and rescale needed.

Do you know the url for the satellite image itself that they are using? Before they apply the sea ice concentration algorithm. I'm not seeing anything at this resolution at Jaxa AMSR.

Or does the image not exist except implicitly as a netCDF text and numeric meta file?

Are you able to make a landmasked (grayscale?) image of the Arctic Basin from the netCDF in NSIDC or other common imagery projection? Greenwich 0º lat (or +90º multiple) down. Thanks!!!

 photo landMasks_zps9d0ac93a.png

Kevin McKinney

"Guys, whatever is happening up there, we need to revise our minimum estimates upwards. I revise mine to 5.5 million + ( extent) and 3.9 million + (area)"

"'We, Kimo Sabe?'"

I think Chris Reynolds has the right of it, above:

"After detrending by interannual differences, the correlation between area at minimum and area on May 20 is, 0.2022. That's not just a low correlation, it's not even statistically significant."

Accordingly, I ain't revisin' nothin'.


This new 3.125 km resolution would be a big deal for microwave monitoring of the melt season, never mind their particular processing of it.

Meanwhile we have even sharper images from Modis, but I am not seeing the ice features from the 3.125 km on the 1 km resolution visible or near-IR images of the same day.

So are they looking deeper (certainly possible) or just generating artefacts? For example, the goat's head cannot be seen in Modis or AVHRR or 89 Ghz but only in Ascat and 36HV 18V.

 photo Banks3125_zps6148210d.png


"I revise mine to 5.5 million + ( extent) and 3.9 million + (area)"

... ... ...

an unusually large amount of the ice area is thin FYI. it will melt.


I'd wait with the revising as well. Like I said:

I feel the Arctic sea ice pack could soon go POP under the right conditions.

Yes. Respectfully, those who want to "revise" can speak for themselves. Fundamentally, the situation in the Arctic should give no strong reasons for any substantial turnaround, if any. As far as I'm concerned, it's like the "weather is your mood, climate is your personality" analogy. The week to week weather is going to cause the ice to ebb and flow, and then ebb again. But the Arctic basin from Beaufort, to Chukchi, and crossing over to Laptev is a mess of splintered ice awaiting the brutal June solstice. MODIS shows a lattice of splinters and ice floes on May 28th, suggesting a ghoulish amount of what very much so appears to be FYI getting ready to "go POP." As far back as 2009, there doesn't seem to be an overall comparison to what's happening this year. I don't think even 2012 for the same date can compare with the amount of splintering happening on the Pacific side. This isn't going to be counted in the area or extent numbers, but there is a very significant weakening taking place that I strongly suspect will soon appear in the area numbers. Just my two cents.


Minor melt mysteries: I've been puzzled for some time by the sharp triangle on the Ascat and Jaxa color microwaves. At first I dismissed it as satellite sensor glitch or swath overlap artefact (there's no shortage of these) or perhaps an ice fracture that pivoted around a resistant pivot point in thick ice.

However the 70-day animation below establishes that it arose as a region of first year ice near the neck of the goat's head got pushed by pack motion up against the core CAA multi-year ice over a ten period beginning on day 86 (27 Mar 13). It is still with us today, somewhat displaced by pack motion towards the Fram but still holding its emissivity shape in both polarizations, 36 Ghz vertical being more pronounced. That rules out sensor malfunction and swath compositing.

However it is hard to distinguish ice wedging itself into the older ice (pushing it apart or deforming it) from ice riding over the older ice. I favor the former because the original region separated first into three regions and the other two are rounded, lacking the sharp triangular edge of the middle piece, which is not conducive to penetrating or overriding. Even the oldest ice is quite susceptible to plastic deformation.

However the stability over two months is surprising given subsequent viscous deformations of the overall ice pack.

 photo mysteryTriOverview2_zps8465f8de.png

 photo mysteryTriangleB2_zpsb1814459.gif


Wipneus, the jpg dithering problem on the new 3.125 km resolution sea ice concentration is illustrated below. I picked a small square at random (Laptev Sea) and checked to see if it the individual pixels used the colors of the palette. They didn't.

This makes it all but impossible to determine the number of pixels (ie areas) in each of the palette classes. Presumably somewhere in the netCDF file, before the data got corrupted in making the jpg, they provide summary statistics for each bin of ice concentration.

While netCDF is a 'standard' file format (one of dozens for geospatial data), it is totally opaque to most potential end users. I noticed that they have no google analytics going to monitor outside interest. It would be better to get the product out in a universally accessible format (lossless, minimal MB).
 photo colorCube2_zps79cf516d.png


Only half heartedly, here is an odd ball suggestion:

Search the ice flows for a picture of Jesus.

If you find a resemblance, that motivates a lot of publicity in an area that probably wouldn't otherwise pay attention.

Bill Fothergill

Revision Time???

It was around 5th/6th June last year that all the measurement systems (NSIDC, CT, Jaxa, Univ Bremen, Arctic ROOS) went into freefall: unless 2013 goes way past this date without the bottom falling out the top of the world, we're in for more of the same. (Or worse)

As several others have already commented, given the fractured state of what is overwhelmingly first year ice, the June losses are likely to be frightening.

No ice in my whisky please.


Chris Reynolds

John Christensen,

It seems to me the cold snap we saw mid-late February (linked to the SSW event?) with the extensive cracking at the same time did lead to a extensive degree of volume gain...

Check out this graphic, it's PIOMAS volume broken down into contributions from different thicknesses for March.

Along the bottom are thickness bands, go along to 2m, you'll find that March 2m thick ice has the lowest volume in that series and the lowest volume since the PIOMAS data starts (1978).

Above 2m thick 2011 onwards are pretty similar, with less thick ice volume than 2010 and earlier.

Below 2m thick 2013 shows more ice volume.

Now look up to 2006, you'll see that for some reason volume of the thickest ice was low, while thicknesses between 2.5 and 3m were higher volume than adjacent years. This is what happens when the bulk profile of the ice gets thinner. There's a deficit of thick ice and because the ice is thinner volumes of ice in thinner ice increase.

Now look at the shift of the bars (blue bars in each cell).

The thickest ice declines from 2000 onwards until in 2010 there is very little left.

But the thinner ice, in the range 1.5 to 2.5 shows a general pattern: Volumes decline into lower thickness bands. The increase from 1.25 to 1.75 in 2013 shows that 2013 in March there was the thinnest March profile in the PIOMAS data.

The differences are small, and PIOMAS uncertainty may be argued to overwhelm the differences. But the general distribution of MYI and FYI is radically different this year than in previous years.

But having gone over this in detail I hope you now understand that the weather in February and March did not thicken the pack more in 2013 than in 2012.

Chris Reynolds

Bill Fothergill,

I've recently blogged on the June crash.
If there's a time to consider revising upwards it will be the end of June at the earliest.

Espen Olsen

It is too early for Bounty mood yet!


The key to these discussions about our mutual prognostications is not to be able to "Brag" about why we were right come September. It as about being able to "Learn" why and where we were "Wrong" come September.


Thank you, OldLeatherneck for re-injecting some sensibility. Meanwhile...

Is that open water I see about 200M or so out, from North Pole webcam 1?!?!

Certainly is sunny.

Chris Reynolds

Old Leatherneck,

This year is primarily exciting for me because I see sea ice thickness as the key driver of the last ten years of sea ice decline.

If I'm correct that 2012 would have been much worse but for the wing of MYI that caused the Chukchi/ESS low concentration region. The region that was hit by the August storm. Then this year, with unimpeded FYI should produce a much greater melt out.

If I'm right and we see a massive new record this year then, as a question will have been answered, I'll move on to more interesting questions.

If I'm wrong and this year doesn't bring a new record then one conclusion I'll be drawing will be that we'll need to see much more winter thinning before a drop to under 1M km^2 CT Area (daily) minimum will be possible. i.e. the pack could last well into next decade.

What matters is mechanism. Being right for the wrong reasons is being wrong.

Bragging is for children and idiots.


Re: minor melt mysteries.

OK, I am not saying this is what it is, but in early 2012 (before the sensor quit on the Envisat), you could see where methane burps left a memory in the ice reflectivity. Generally, you could see where the methane began to accumulate under the ice (it was February), then when the pressure was strong enough, it broke through. These events were rounded.

Other events were more like seeps, and yet others were more like sudden jets. Jet events might explain the triangular shapes.

Again, I have no proof. But the most persistent memory in the ice was left around the Franz Josef islands, after an event on the subsea shelf that they are on. I have these images saved on my hd if anyone wants to get in touch with me. Also pdf files of false-color images showing what are likely methane burps. Can't prove it, though.


Cyclonic storm still having a big party up there at 150E, and seems to be growing:



Watching MODIS r04c04 (CAB Laptev Side) and the adjacent tile r05c04 (north of Chukchi Sea), it’s clear that the grinding Low is laying out the bad quality of the pack. Not just rubble-filled, broad and traversing leads. A web of open ones too. There’s no ice formation anymore. The black open waters, sometimes 5 km wide, seem to indicate an upper ocean layer hostile to volume.
This has to show in SIA soon. Maybe the large extent in the periphery hid the trend during the last few weeks.
I agree with Chris Reynolds. The behaviour this spring is insignificant for the outcome. These variations fool us through the volatility of a system out of balance.

I see others are noticing this, too... Green Octopus, Wanderer. Let's watch this close, it will soon unfold (I love some bombast).

Kevin McKinney

Chris R., what are the units used in your interesting PIOMAS graph? (Not meters, of course, just the y-axis.)

I think perhaps you explained that and I missed it--maybe in an earlier link to it? If so, sorry--but I'm curious.

Fairfax Climate Watch

I've been following the cyclone - and posting updated photos of it here: http://www.fairfaxclimatewatch.com/blog/

From the Navy CICE model of thickness and concentration, it looks like there could be a large amount of open water soon on the Eurasian side - pressure maps indicate the cyclone will last till Monday: http://www.esrl.noaa.gov/psd/map/images/ens/z500anom_nhsm_animation.html

Susan Anderson

One feels one must buckle one's seatbelt. Thanks Tenney and M. Owens, every picture tells a story, don't it? Why do I want a front seat at this, pray tell?

(seriously, great stuff)


Neven: thanks for linking my first Arctic article of the season. You say: "I feel the Arctic sea ice pack could soon go POP under the right conditions." I agree.

Hans Gunnstaddar says:
"Every time a new record occurs climate change is given the mantle, and every time a rebound occurs it provides ammo for the recovery crowd. All the while we edge closer to a major tipping point."

Just so. I have just finished a new article covering this very topic. Citing two excellent in-depth resources I show that the current warming is not 'just natural variation' as some would have it. There is a clear and continuous warming trend from the late 1800s through to today, with superimposed natural and mainly regional cycles.

The Arctic's Warming Islands

Check out this image from the mosaic:

Left of center near the bottom you will see a sharp boundary between thicker and thinner ice running from Banks island to the mainland. There is a just-perceptible crack running almost parallel to the boundary. Watch for a very rapid break-up there. Will the NWP be open to navigation early ? Time will tell, but I am confident that it will be open again this year.

Stephen Pekar

Hi there,

This is my first post here. I do want to thank you Neven for all of your excellent posts.

I just saw this and would like to hear your and others thoughts (see link below). It looks like the ice is thinning quite rapidly.


Kevin McKinney

Nice article, 'logicman!'

Yet another own goal to Mr. Watts' 'credit.'

Chris Reynolds


The units are 1000's of km^3. If you were to tally up each figure you'd find they add up almost exactly to the average PIOMAS volume for each month. The slight difference is because the grid areas (calculated by Wipneus) differ slightly from the grid areas used by the PIOMAS team (which aren't publicly available).

Chris Reynolds

Tenney, M.Owens,

The impact of the storm in HYCOM is striking.

This is being discussed over at the Forum.

Rob Dekker

About spring snow cover and fall ice extent, Neven said :

I'm keeping more of an eye on NH snow cover this year, having added one graph to the ASIG as well (at the bottom of the page). Unfortunately Rutgers update their maps and graphs with a separate file name each day, instead of a generic file name, otherwise I'd have included them too.

I noticed that, and thank you !

Snow cover in spring and high summer has a tremendous impact on the amount of heat absorbed in the the Northern Hemisphere (and the sub-Arctic specifically), as was also noticed by Tamino last year :

It seems that snow cover melts out much faster over the last decade, most likely due to overall warming of the planet, leading to vastly reduce snow cover in summer snow cover over the past decade, and a record -6 million km^2 in June 2012, which was also promptly followed by the lowest Sept ice extent in recorded history.

The amount of heat that the Arctic absorbs during May/June due to reduced snow cover cannot possibly be underestimated, and neither can it's effect on sea ice melt. Last year, I presented a quick back-of-the-envelope calculation that suggested that more than 1000 TW is absorbed during the June 2012 6 million km^2 record snow cover "anomaly" in 2012.

May 2013 seems to run at a similar large snow cover anomaly, so we can expect the same tremendous amount of energy being absorbed during this May/June as last year.

On snow cover data, Rutgers seems to be the best reference, but it is indeed disappointing that they do not present their data in such a way that we can monitor it on a day-to-day basis.

Regarding the remarkable correlation between May/June snow extent and Sept ice extent, I would love to write a guest post on that at some point. If I could only find the time to do so....

John Christensen

Chris Reynolds,

I did not claim the weather in February/March thickened the pack more in 2013 than in 2012, only that the weather did lead to an extensive degree of volume gain, as Neven also concluded in 'PIOMAS March 2013'.

If you exclude Bering, Okhotsk, Hudson, and Baffin, central Arctic SIA was actually extremely low during the winter of 2012 until some very late freeze improved the situation in March and April, so no wonder that it would melt out very rapidly as well. Beaufort was already melting at a higher rate 12 months ago than it does right now - as well as Barents, Kara, Laptev, and ESS.

As long as the water is ice-covered, you will not see rapid SST increases, which is the reason for my 'so far so good' comment. I am not saying the ice will hold long, and I am not trying to give any estimate of where we will end this summer compared to last year.



To make sure you have the same information as I do, this is the header of the NetCDF file (the metadata part):

netcdf Arc_20130201_res3.125_pyres {
        x = 2432 ;
        y = 3584 ;
        time = UNLIMITED ; // (1 currently)
        double x(x) ;
                x:standard_name = "grid_longitude" ;
                x:long_name = "x coordinate of projection (eastings)" ;
                x:units = "none (index variable)" ;
                x:axis = "X" ;
        double y(y) ;
                y:standard_name = "grid_latitude" ;
                y:long_name = "y coordinate of projection (northings)" ;
                y:units = "none (index variable)" ;
                y:axis = "Y" ;
        double time(time) ;
                time:standard_name = "time" ;
                time:units = "days since 1-01-01 00:00:00" ;
                time:calendar = "proleptic_gregorian" ;
        short sea_ice_concentration(time, y, x) ;
                sea_ice_concentration:standard_name = "sea_ice_area_fraction" ;
                sea_ice_concentration:long_name = "daily averaged total ice concentration" ;
                sea_ice_concentration:units = "%" ;
                sea_ice_concentration:scale_factor = 100. ;
                sea_ice_concentration:description = "Sea ice area fraction is area of the sea surface occupied by sea ice." ;
                sea_ice_concentration:source = "AMSR2 Sea Ice Concentration, Institute for Oceanography, University of Hamburg" ;
                sea_ice_concentration:range = "0-10000" ;
                sea_ice_concentration:scale_factor_comment = "divide by 100 to recieve 0-100%" ;
                sea_ice_concentration:date = "Daily averaged sea ice concentration for 2013-02-01" ;

// global attributes:
:CDI = "Climate Data Interface version 1.6.0 (http://code.zmaw.de/projects/cdi)" ;
:Conventions = "CF-1.4" ;
:history = "Wed Apr 24 20:12:09 2013: cdo setdate,2013-02-01 -settime,12:00:00 /scratch/local1/u241125/SEA_ICE_CONCENTRATION/AMSR2/3.125km/Arc_20130201_res3.125_pyres_temp.nc /scratch/local1/u241125/SEA_ICE_CONCENTRATION/AMSR2/3.125km/Arc_20130201_res3.125_pyres.nc\n",
"Created Wed Apr 24 20:12:09 2013" ;
:grid_resolution = "3.125 km" ;
:description = "gridded ASI AMSR2 sea ice concentration" ;
:algorithm = "ASI v5" ;
:title = "Daily averaged Arctic sea ice concentration derived from AMSR2 L1R brightness temperature measurements" ;
:landmask_value = "12500" ;
:tiepoints = "P0=47 K, P1=11.7 K" ;
:gridding_method = "Nearest Neighbor, with Python package pyresample" ;
:Comment1 = "Scaled land mask value is 12500, NaN values are masked 11500" ;
:Comment2 = "After downscaling (divide by 100): land mask value is 125, NaN values are masked 115" ;
:hemisphere = "North" ;
:contact = "xxxxxxxxxxxx" ;
:netCDF_created_by = "Alexander Beitsch, xxxxxxxxxxx" ;
:datasource = "JAXA" ;
:offset = "0" ;
:grid = "NSIDC polar stereographic with tangential plane at 70degN , see http://nsidc.org/data/polar_stereo/ps_grids.html" ;
:missing_value = "11500" ;
:sensor = "AMSR2" ;
:cite = "Spreen, G., L. Kaleschke, G. Heygster, Sea Ice Remote Sensing Using AMSR-E 89 GHz Channels, J. Geophys. Res., 113, C02S03, doi:10.1029/2005JC003384, 2008." ;
:geocorrection = "none" ;
:CDO = "Climate Data Operators version 1.6.0 (http://code.zmaw.de/projects/cdo)" ;



I am happy with the netCDF files, just because I happen to be familiar with them.

The landmask that I posted yesterday was the result of a handful of R commands in the terminal:

3:landmask[landmask != 1250000]<-0
4:landmask[landmask == 1250000]<-1

Line 2 creates a 2432 x 3584 array with the concentration data, which is discarded and only the entries marked as land are kept.

Rotation over 45 degrees is not trivial, and will not be lossless even if I can figure out how to do it.

First I will work on a non rotated color png, I share your dislike of jpg's.

John Christensen


Thank you for the table though - I did not get to this back in March.
And 2006 is very interesting as you noted regarding the thickest ice. CT is showing a similar picture, indicating that the 2007 meltout really started in 2006. Where the 2006 events thoroughly analyzed anywhere?



Do you know the url for the satellite image itself that they are using? Before they apply the sea ice concentration algorithm. I'm not seeing anything at this resolution at Jaxa AMSR.

I don't know any other data. But I have just read the introduction of Spreen et al 2008, where they are calculating ice concentration data from swath data and thereafter interpolated into the desired geographical grid. Those results from swath data are compared with calculations on Level 3 gridded brighness temperatures from NSIDC.

I short: what you are looking for may not exist.

Espen Olsen

Repost from Forum:


11,801,094 km2 (May 29, 2013)
+ 4,063 km2 May 28,
+ 329,766 km2 more than 2000s average,
and 101,031 km2 less than 1990s average.


I forgot to mention:

The ASMR2 maps from May 11-14 are both missing from JAXA/IJIS and the Bremen websites. However these (from Hamburg) are present, and from the file times where updated at the times. This is consistent with something happening with the processing at JAXA while low level data still being made available.

This in contrast with speculations that the spacecraft was in some safe mode for incoming solar flares.


A few extra commands:

# rotate so Greenland is south
# create an array with three planes
# for R,G and B 0<=x<=1
# map 0-100 to 0-1 for grey's
# mark land as brownish
for(i in 1:3)
#write the result



File size 1.7 MB, ten times smaller than the jpg.

Is that getting useful?

Artful Dodger

A-Team wrote | May 29, 2013 at 16:35

"the neck of the goat's head got pushed by pack motion up against the core CAA multi-year ice over a ten period beginning on day 86 (27 Mar 13)."

Hi A-Team,

This is likely the very beginning of breakup. The date is near sunrise for this latitude. Have you compared the ASCAT images to MODIS?

If you want to delve further, perhaps contact the folks at NASA Icebridge. They flew four separate low level sorties over the Lincoln sea beginning on day 80. Flight tracks here:


Their sea ice thickness and concentration data is already(!) available for download online. They also had some human observers on-board their Lockheed P3C Orion research aircraft, so you may get some anecdotal evidence too.

BTW, there's an entire thread on this topic over at the ASI Forum (the very first thread opened ;^)

Forum user 'Hunter' is a member of the NASA Icebridge team and contributed many visually stunning images from the flights in near real time. This image was taken over the Lincoln sea during the North Pole transect mission on March 21 (this is the joint calibration mission with Cryosat2 overhead simultaneously). Hunter posted the image on the Forum on March 22, 2013 (login required to see forum images).


Already, Arcticio has downloaded and processed the March 2013 Icebridge data. His thickness plot is also published on the forum here Again, you'll need to login to see Forum images.


Verify your Comment

Previewing your Comment

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

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

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

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

Having trouble reading this image? View an alternate.


Post a comment