« ASI 2013 update 1: a slow start | Main | New map on the block »


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


According to today's Navy Hycom prediction, the Big Thin continues through 02 June 13 (which has observational support) and then out to 09 Jun 13.

The first animation shows the polar cap region within 80ºN; most of the thinning action has affected 2m thick ice in the upper left quadrant. The thinnest ice there predicted for 09 June (magenta) is 0.8m per color scale picking, which is not open water. Modis today under the eye of the cyclone is (self-evidently) cloudy.

The second animation shows the same 14 days for Ascat, enhanced for the melt zone and multi-year ice.

The final figure compares the last four years for today's date; the multi-year ice is configured most similarly to 2010, though microwave loses its inter-year comparability early in melt season.

 photo navy8odegCircB2_zps89ddca8b.gif

 photo navyAscatB2_zps5b046f49.gif

 photo Navy4year2_zpsf75b9566.jpg



I got your point - of course divergence can not be seen in AMSR2. My point was - there is nearly no SIA lost in the wide area arround the storm (only 40-50k since 12 days) - so there is no source for transport of ice from the central CAB. Only arround Svalbard, but that region is not inside the 2.76 Mio km2 circle and I would not consider that as driven by the storm directly. It is more driven by A-Teams gear-wheel.
If you compare 20th May and 1. June
AMSR2 20. May: https://sites.google.com/site/apamsr2/home/pngcby32/Arc_20130520_res3.125_pyres.nc.png?attredirects=0
AMSR2 1. June: https://sites.google.com/site/apamsr2/home/pngcby32/Arc_20130601_res3.125_pyres.nc.png?attredirects=0
-> in the concentration map it looks much darker in the CAB after the storm, but it is only 40-50k less sea ice area. The only ice-movement increasing SIE is again around Svalbard into Barents.
So - SIA-wise not much of an effect yet by the storm. But according to hycom concentration forecast something is still to come in the next days. I will try to get that - just to check the prediction.

Chris Reynolds


I see the Svalbard movement (out into the Atlantic) as part of the general movement and associated divergence. And I suspect that a large factor in the modest losses of area is the divergence. There's a lot of opening of water between newly broken floes visible in MODIS but not in AMSRE, therefore not counted as area.

Whereas in February we saw the entire pack move in a clockwise manner for a long time. Now we have had an anticlockwise shift of a similar nature but shorter duration.

I have also observed the continuing reduction of concentration/thickness in HYCOM. I agree that will be mirrored in the real world. But I think it will be hard to spot because, unlike in HYCOM, in reality much of the opening isn't in one area but distributed between the broken remnants of the floes.

I think the effect will be long term, and will probably result in low concentration throughout the affected area by later in July.


Thanks to Wipneus's work, the last six weeks of high resolution sea ice concentration form ICDC Uni Hamburg can be animated.

Below, I did this for the last 13 days, retaining the original resolution, which meant that only key areas could be shown sticking within our 415 pixel width for Typepad. A contrast-enhanced version is also shown. File sizes can be cut to a third by using grayscale(vs RGB).

The final animation shows the whole Arctic Basin, which cut the width from 1216 pixels as cropped to 415, so losing about 3x of the resolution. These animations are a megabyte and change, so it is barely feasible to serve the full basin at full resolution (9.9 MB).

These are just thrown out there as prototypes to see what people want as product.

 photo wipCrop1B_zps4d8b3e54.gif
 photo wipCrop1enhancedB_zps87c0dd77.gif
 photo wipneusWhole_zpseda04a5e.gif


Very nice, A-Team. I want to have one more of those new Uni Hamburg/Wipneus SIC maps, and will then probably write about it tomorrow. Also as an update to this blog post. It's fascinating stuff.

BTW, on that comparison of 2010, 2011, 2012 and 2013, the latter looks much thicker/white than 2012. If this represents a real difference, I think that this year will have a very hard time breaking last year's record(s).


Neven writes '..Hamburg/Wipneus SIC maps, and will then probably write about it tomorrow'

You might want to write the UHH scientist whose name is in the lower left corner -- I imagine he is quite proud of this new resolution and ASIv6 and could give us the inside story.

 photo beitsch_zps9ec86095.jpg


Hi A-Team , do you have the specs for the bottom microwave pics, perhaps aperture radar? Is there a chance that a resolution setting may affect the outlook of the image? Or were all images calibrated equally?

I have some doubts about the mass buoy data, I believe there is some great action at the interface between ice bottom and top of sea water. I also think that the data does not recognize density differences. Or needs finer settings to describe what is going on through the ice column. looking at 2012G


contrasts sharply with A-Team microwave displays. Although at about the same site, the microwave may be more tuned to a certain ice density? Anyways surface temperatures warmed a lot, and the response by mass buoys thickness does not show any significant event. Looks like the sonar was calibrated to thickness at installation, perhaps that explains some disparities.

Interesting stuff.


In the tropics, it is my understanding that intense lows are fueled by evaporation of the rain, which then rises - fueling the low pressure system.

What is fueling the low here? Newly opened water? Melting?

Perhaps the arctic cyclones are growing to be a feature of the new normal?


Werther noted major Siberian rivers starting to flow: "MODIS r05c05 shows the Lena river flooding south of it's still iced lower bed."

Well, the Mackenzie ended its slumber today too, with heavy melt sediment flooding the delta and working its way under the landfast ice to open water, where a sediment plume is clearly visible. This is mud, not a plankton bloom.

This date, 02 Jun 13, is early but not notably early. It will have quite a pronounced local effect, near-shore and beyond, on ice and salinity.

Width 1173 pixels; 415 showing:
 photo macSediment_zps553d17dd.png


Good effort there, Satire. Try using the magic color pickers more to select areas, especially everything in the image that goes with a particular palette color block.

Gimp has two: the first for contiguous color blocks, the second for every matching pixel. By loosening the color cube radius up from zero, these become more tolerant of departures from the reference clicked pixel.

You might try your hand at volume loss (taking Navy Hycom at face value). It looks more or less confined to the left half of the hemispheres below. So erase the left halves and set the tolerance on the non-contiguous color picker up from zero but not so high it recognizes more than one block in the palette.

The sums of pixels * their volume gives total volume loss for that half hemisphere as a function of time. According to Navy Hycom.

Width 657; 415 showing.
 photo NavyVolumeLoss_zps2f59ab4e.png


Quick question on the direction of the cracks. Looking at the animations above, it appears the winds are counterclockwise. So why do the cracks seem to be as if the outer edges were driven that much faster? Is this thing being driven by wind velocity or is there a component that is more esponsive to fetch? Or some other explanation?

John Christensen

Question on Arctic 80N cold:

I tried asking this question on a different blog entry, but will try again here:
The Arctic 80N temperature on DMI has been low for the past month (probably lowest since 2004), but since the Arctic air temperature relates to heat exchange with the water (at least when the water is not completely covered by ice), I am wondering what this means.
High air temperature in the fall (when above freezing) should be the actual air temperature. When it falls below freezing, it should still stay higher due to the heat release from the water freezing and turning into ice.
For spring, what kind of cold does the melting ice release to the air? Does it keep the air at freezing point, once it melts, or can melting cold ice (e.g. due to bottom melting) release enough cold to keep the air temperature a few degrees below freezing?
2004 and 2007 had similar periods of relatively cold air temperatures above 80N for late spring (http://ocean.dmi.dk/arctic/meant80n.uk.php), but is this 'real' cold or energy release from melting ice?


No further SIA change in the storm affected area again yesterday.

A-Team - I guess you are also refering to your excercise in the other thread with the grey palette for numerical colours. Sorry I am such a noob with gimp - selection and histogramming works fine, but I could not figure out how to do most simplest things like "search and replace" of pixel colour values (instead of using that stupid bucket at every pixel) or putting a pixel value as function of pixel coordinates or some other function - let alone things like your "guillotine" or what.

I will look in the manual today evening again - but right now I have the impression, that programming or lovely MatLab could do the job much easier for me. Simply because I have no artists background ;-) But I will not give up, because gimp could be a possibility also for non-Nerds.


Satire, sounds like you have not set up the toolbox properly. Instead of the manual, google menu commands -- all sorts of explanations are out there, including straight math. No question, a learning curve before you are absentmindly posting intended outcomes.

Science is best done using the method of descent. That is the real reason for using Gimp -- very rapid exploration of ideas. It doesn't pay to dig into details unless you know in advance that it has to work out.

Very little of our imagery or product maps come with a known projection, properly marked north pole, length or area scale (welcome to climate science). There are three options -- spherical earth, oblate ellipsoid of revolution (the radii differ by 21km), or actual earth geoid (equipotential surface). Over water, the ellipsoid is satisfactory; for Greenland mass balance, the actual geoid. Longitude lines are still geodesics on the ellipsoid.

For those maps -- like the DMI mean polar temperature -- which use lat x long cells regardless of their shrinking size toward the pole, the figure below provides the dimensions and areas of those cells from the north pole out to 65º according to the NGIA geoid.

As a practical matter, it's worth noting that the 70º line of latitude just grazes the Siberian shore south of Wrangell whatever the projection. The number of pixels and angle to north pole can be measured with great accuracy in Gimp, yielding scale conversion factors for image superpositioning from all satellite imagery (some provided below). To convert measured pixels on a scene to kilometers, some known geoid distances from the Wolfram widget can help with that estimate:

837.5 km from NP to Alert, CA
2087.3 km from NP to Barrow, AK
2887.2 km from NP to Reykjavik, Iceland
1316.7 km from NP to to Longyearbyen, Svalbard
2051.9 km from NP to Tiksi, Siberia

Using these and stably identifiable reference features on the ice, the fastest moving features above the CAA ice line have moved 127 km east in the last 14 days; for the triangular feature this is 15% of the point of no return (Fram exit at Nord). The region along the CAA itself is not moving coherently.

It would also be feasible to measure the rotational speed of the cyclone (from Modis clouds or meteorology graphics) and that of the ice (from cloud-penetrating microwave). The ice would lag on spin-up, never really reach the angular velocity of the cyclone, and lag the cyclone on spin-down. Once that mass of ice is set in motion, it no longer needs the wind (Newton's 1st law) though slows from damping underneath.

It is quite a step up in difficulty to squeeze out that last 3-4% error from area measurements (topic for forum). That can sometimes be sidestepped by considering only relative change in a time series.

 photo latLongCells_zps0cc49a47.png


Fufu asks about radial velocities. We're seeing a very substantial patch of thinning but that is not the same as melting -- this ice is very susceptible to plastic deformation (and beyond that, tearing). Yet I don't see thickening elsewhere to counterbalance the thinning through conservation of mass (or volume). While that could be masked by background seasonal thinning, which is becoming substantial on its own, the 'control' half hemisphere is not showing that. This implies net volume loss attributable -- directly or indirectly -- to the cyclone. And we can easily measure that on Hycom's terms.

Below is the 02 Jun 13 Jaxa image. It is *not* photoshopped but instead has been subjected to a global transformation in color space that brings out colors that are *already there*, just not apparent to our retinal cone receptors. The most striking features are the counterclockwise rotation with widening of the reddish brown pinwheel streaks (thinning to near-melt) converging at the eye of the cyclone (at its most effective point in momentum transfer).

If you are following along in Gimp, try Colors -> Auto -> Equalize followed by Edit -> Fade... This would be extremely difficult to replicate in matrix number view -- as are some routine darkroom manoevers.

The bottom image shows some displacement arrows drawn on recognizable features on the ice. The end points represent thirteen days ending on 01Jun 13. Velocities are uneven across the ice pack; the streamlines fork as some ice circulates back around the pole while the floe next to it exits toward the Fram.

 photo NavyCircFlowStreaks_zpsb0ee1880.jpg

but right now I have the impression, that programming or lovely MatLab could do the job much easier for me. Simply because I have no artists background ;-)

I know the feeling, being productive in Gimp or similar is not for me.

For lucky me there is Inkscape, basically a vector drawing program but quite able with bitmapped objects as well. And it has an XML file format that I can interface with a program as well. And from within R I can throw in some ImageMagick commands if I need some image enhancing.


BTW, the well know regions map from the Cryosphere Today:

This map uses the same mapping as the Uni Hamburg with pixel scaling of 25km (1:8).

So you can do your regional calculations as well.


"before you are absentmindly posting intended outcomes. "
Outch! Please correct me if there is any error in the SIA calculations using Gimp (Short of the <6% error due to grid area).

Hans Gunnstaddar

Here’s a question in the true sense of a question, i.e. curiosity. I know people tend to presume the person asking a question is taking a stance (which I find very strange and I suppose gets into psychology), but this really is simply just ‘a question’. Is it possible the weather will be different this year in the Arctic (with a higher ice extent minimum than last year) due to what may be a recent transition from La Nina to El Nino?

It’s suggested the switch is currently going on due to dying trade winds in Hawaii. Here is a link:
http://news.yahoo.com/trade-winds-drop-hawaii-gets-muggy-080531040.html ‘Trade winds drop, and Hawaii gets muggy’

And here’s an article on why trade winds die out: http://www.earthlyissues.com/ninilanina.htm

El Niño/La Niña is a naturally occurring 2-7 year cycle of the ocean-atmosphere system in the Tropical Pacific having important consequences for weather and climate around the globe. ‘Trade winds die’ and the warm water that was once pushed westward is allowed to shift back towards South America.

Scroll down until you get to the pic with the caption; ‘This image shows the current El Nino's split personality’ (which shows the trade winds in opposition).

I know it’s too early in the melt season to know what will happen, but it’s good to look at all possibilities.


Hans, I've never been able to see a correlation between ENSO and Arctic sea ice minimum. And the only person I've seen proposing a mechanism, was Wayne Davidson with his anvil seeding theory.

R. Gates

Two key concepts here for the rest of June:

Divergence and then the "Cliff".

As pointed out by several denizens here, we are seeing divergence around the edges of the Arctic, that is creating the what I feel is the misperception that the melt season is in slow motion, with obviously the area and extent both being affected by this divergence. Overall, with the exception of the cyclone, we are seeing generally higher pressure over the extreme NH:


And this leads directly to divergence of the ice, but additionally, as we've seen all winter and spring, the ice is thin enough to easily diverge. This has affected the slower extent and area to a large degree.

But now, after the cyclone, and this divergence, comes the mid-June Cliff. ECMWF is showing a warm band forming in the next week to 10 days, right up along the North American Rockies right up and into the Beaufort and across to the Chukchi-- right across some of the thinnest ice. Additionally, as A-Team and others have so excellently pointed out, we are getting some big increases in inflows from the major rivers. This will amplify the melt in those regions of inflow, as it usually does.

So we've had our first cyclone, we had divergence, and now comes...The Cliff, which I think will involve several weeks in mid to late June of jaw dropping, and likely record setting daily declines in extent as we enter the heart of the 2013 melt season.


Mr R. Gates,

I agree your vision, 2013 melting season is starting slow, but this a disguise of sheer misery.


OK, I've posted an update to this post: New map on the block. I'll probably discuss it some more in the next ASI update, due to be published this weekend.

Chris Reynolds

R Gates,

Couldn't agree more, but the June anomaly crash (what you call the cliff) is puzzling me - how will it manifest this year?

New animated gif and comparing recent SSMIS Bremen images for 2012 and 2013.

The slow area/extent decline is looking more and more irrelevant. If we don't see a massive crash in CT Area anomalies in June I really will be puzzled!

R. Gates

Chris said:

"If we don't see a massive crash in CT Area anomalies in June I really will be puzzled!"


I would be puzzled as well, since all the factors are lining up for this occurance, and if it doesn't occur, it would seem some basic laws of physics are being broken! But then again, nothing in the Arctic seems to surprise me anymore.

But you asked about how it will manifest itself? (i.e. the big melt or "The Cliff")that seems to be coming in the mid to late June timeframe.

Multiple factors, many of course related to each other:

1) Long term decline in volume, with incredibly thin ice across areas that will be prone to melt ponding is a major one. But it is not just the thin ice but the density of the ice that should be of concern. We've heard for years about "rotten" ice, and this mixture of MYI and FYI will really show itself again, as it is not just the thin ice that should be considered but the overall density-- with rotten ice, or this "rotton ice" mixture being less dense and more prone therefore to rapid melt as it can not withstand as much energy being absorbed before the phase transition to water.

So thin ice plus rotten ice.

2) We've talked about the inflow of warm water to the Arctic via the oceans, but we've only talked briefly about the flux of energy via the several rivers that feed into the Arctic. These rivers have been rising in temperature (right along with the decline in late spring/early summer snowpack) for several years. The Arctic ocean regions where we see these rivers are now covered with weak FYI. Expect BIG melt from these areas in the coming weeks, as the thin ice will be hit with huge heat flux.

3) High pressure over key areas at the maximum insolation combined with increased melt ponding from the thinner ice. The energy transmitted through the melt ponds at the time of maximum insolation is more crucial than most might realize. See for example:


4) The early divergence of the ice. I would hope that a better metric can be developed for this, but the thin ice/rotton ice plus early divergence is crucial to what I see as the key related factors leading up to the coming Cliff. I think the divergence has been a bit of a "head fake". We usually see this later in the summer when the ice is thin with lots of melt ponds and major divergence takes place and can cause a "whipsaw" behavior such that we can get a flat or even spike up in extent, but this is then followed rapidly by a rapid decline. I think the thin ice is causing divergence to happen earlier this year. Part of the changing dynamics, and in a few years when it is all first year thin ice, I'm sure late spring, early summer divergence will become part of the normal pattern.

For the coming Cliff-- I expect day after day of big extent drops, with several single day records being set.

Natalia Longway

R. Gates - I agree on 1, 2, and 4. Regarding pooling, I can't see how most of the ice could support pooling after all of the churning of the last week. I expect pooling to play a smaller role.
Perhaps a 5th point - Expect the bottom melt season to start early from the cyclone and wave action getting into the thin CAB. I expect that July to be the month of records featuring CAB loses.


Isn't it this way: Under a low pressure the ocean rises a bit and this causes dispersion. The winds would work in the opposite direction, but the draft in ice is larger. This, with the coriolis force results to a sea ice movement at an angle (was it pretty exactly 45 degrees?), away from the low pressure centre. Just to make this thing clear for newcomers.

On melt pools I've no comment, but the gray rotten ice floes can be as small as 100m in diameter (all the snow turned in to slush on top of the proper ice) Once these break up they become apparently whiter, I guess the slush loses water in the breakup. (this about the Baltic ice)

John Christensen

R. Gates and Chris,

Yes, it will be interesting to see how the melting season develops once it gets started for real.
With all the shortcomings of e.g. CT SIA measurements, this is how CT always calculated SIA so should be comparable year-over-year, just with the difference that with thinner ice, cracks, and divergence, the over-estimation in CT SIA numbers should be on the increase.
Still: Less than three weeks from summer solstice, it seems that keeping the Arctic open water area low just one or two additional weeks should be important with regards to the amount of heat the water will absorb from sun radiation during the summer period. As was found in 2007, the heating by sun radiation of top water layers, caused significant and rapid bottom-melt in the Beaufort, so the risk of such an event reoccuring should be somewhat reduced. But will that make up for the deteriorated state of the ice, or otherwise have any measurable impact? Will be interesting.

John Christensen

Article on 2007 Beaufort melting:


Chris Reynolds

R Gates,

Regards your numbered points:

1) I do think it's worth remembering that the FYI covering much of the pack this year is probably of roughly the same thickness as last year, apart from regional differences (Barents/Kara was thinner last year). The difference this year is that there is no MYI intrusion into the Siberian sector. I argue that such an intrusion caused the region of persistent low concentration which delayed extent loss in 2012, probably protected the bulk of the pack adjacent to it, until it was wiped out by the August storm. This can be seen in Bremen plots for July.
Either way, winter thickness over much of the pack puts most of the ice this year in the region of increasing percentage open water.
Although using profiles from 2010, 2011, and 2012 doesn't predict a new volume record this year, I think the profile of melt in those years was affected by conditions in those years, so this comparison isn't ideal and I don't find the result persuasive.

In short I expect a similarly low concentration state to 2012 by the end of July.

2) The flux from rivers is a major player in the breakup of Siberian landfast ice, whose thickness can be surprising (it surprised Neven back in April). If I fail to mention it it's because I've taken it as known. Thanks for correcting my omission.

3) Agreed. I also think that the role of insolation as the driver of the melt season is something many overlook. Volume loss in PIOMAS (and in reality) is high in June through August, peaking in July. See figures here:

4) I think this early divergence and the creation of a lot of open water within the pack is very important. I'm 70% sure we will see this region play a big role this year. I am tending to expect a massive June drop (cliff in the anomalies) this year.

Natalia Longway,

I don't think the ice supporting melt ponds is a problem. Some of the ponds connect to the ocean, so are at sea level anyway, i.e. no extra weight. Even with the ponds who's surface is above sea level, the density of water is fairly close to that of ice so they shouldn't weigh more.

The ice in the region hit by the May storm is very broken, as seen in this animated gif:

But the resultion here is 250m X 250m per pixcel, so any feature visible has to be approaching 250m in size to be visible. This means that there are still a lot of reasonable size floes on which melt ponds will form. Furthermore there's a lot of ice that's not in this bad a state in other areas.

John Christensen,

Where the ice is in low concentration, as opposed to open water, temperatures will be held down as insolation absorbed by the water between floes will go into melting ice laterally (horizontal). However if we see 'rotten ice' over large swaths of the pack then the area of ice edge is increased and lateral melt is. In 2007 the melt in Beaufort needed to be from the base of the ice because the regions affected still had a lot of MYI.

And so I click, POST, and consign my post to the all devouring maw of Neven's paranoid spam filter. ;)

R. Gates

For the coming Arctic sea ice "Cliff", I would suggest that we really drill down an take a look at some of the particular areas that will be contributing to the "Cliff" in the beginning stages. We can ignore areas like the Hudson, as we know that will melt out anyway, and focus more on areas that provide more of the dynamics discussed earlier. I think focusing right now on the Kara Sea will provide a look at the dynamics. The Kara "Cliff" is poised and ready for a big-time drop. It has held up well here in the late spring melt, with some of the clockwise rotation of the whole ice pack back in March (when we got the cracking near the CAA)really helping to keep ice piling up in the Kara.

Now, that ice has "unpiled" i.e. diverged, keeping extent and area numbers from falling rapidly, but setting the stage for the Kara "Cliff". The Ob and Yenisey rivers are starting to move, and will assit greatly with the warming of the Kara.

If you look at the chart of the Kara:


We can see classic signs of divergence in mid to late May, with sea ice area just flattening out and even bumping up a bit here and there.

The Kara is the perfect test for "The Cliff", as it has all the ingredients discussed, and with not one, but two mighty rivers feeding warming waters. Hold on to your seat belts...

R. Gates

Just to follow-up on the river discharge into the Kara, here for example is information for the Yenisey:


A few years old, but shows the long-term averages of how strong the June influx isof the Yenisey into the Kara sea, and one more reason why the Kara sea ice has been living on borrowed time before taking the June plunge over "The Cliff".


And things are turning blue since yesterday in Beaufort, Chukchi and ESS.

Chris Reynolds

R Gates,

Yes, I'll be watching Kara.


Thanks I hadn't noticed that.



shows ice conditions till June 6!

I must be missing something?


Wayne, that was the forecast made on May 29th. The ACNFS model puts out different forecasts (concentration, thickness, drift) every day.


Thanks Neven, the model needs some work!


Agreed, but I think the consensus is that it is still useful. :-)

Al Rodger

So is it just through stormier weather that AGW is increasingly fracturing the Arctic's Sea Ice.
This article (hat tip David B. Benson comment @ RealClimate) suggests to me there is another effect at work. It appears higher levels of atmospheric CO2 increase the brittleness of ice. The article talks of glaciers & ice caps being affected to some yet-to-be-determined significance. But Sea Ice too is surely up for it, perhaps with brittleness having more effect out on the oceans.

Jai Mitchell

Since Jennifer Francis was mentioned earlier in this blog, I wanted to share that she will be doing a group presentation with a U.S. Meteorologist, Stu Ostero to share their views on atmospheric blocking in an arctic amplified environment. Stu apparently has developed a separate view of the process that is complimentary to Dr. Francis' work. It will be located via livestream at the following link:


The presentation starts at 4:30 PM east coast U.S. (New York) time today (6/6). (GMT-0:400)


Cool! That will be 22:30 my time, just what I need after a day of hard work. Thanks a lot, Jai!

George Phillies

I undoubtedly missed the earlier link to this.
The June 5 Bremen map


seems to have some noteworthy features near the pole. A respectably large area supposedly of limited ice coverage.


HYCOM is predicting another hole being smashed in the CAB this time right in the MYI



Better resolution



Best resolution (but ignore that "smear" due to water vapour): https://0c35ba35-a-62cb3a1a-s-sites.googlegroups.com/site/apamsr2/home/pngcby32/Arc_20130606_res3.125_pyres.nc.png
and suitable for your own SIA measurement e.g. to check that CICE prediction (will that 2nd hole show up?): http://www7320.nrlssc.navy.mil/hycomARC/navo/arcticicen_nowcast_anim30d.gif

John Christensen

The early comments on this thread are very interesting given how PAC-13 has played out this far..
We live and we learn, and will continue learning in the coming weeks.

The comments to this entry are closed.