Bilateral freezing

It happened a while back, but it's one of those interesting details that make what's happening in the Arctic so unique. In August we saw something that hasn't been recorded before (not on that scale, that is): the detachment of large parts of the ice pack. In October the opposite happened.

In the East Siberian and Laptev Seas freezing started on the coasts before the ice pack had expanded far enough to reach them first. For a couple of days, the freezing on both sides caused a large pool of open water surrounded by ice on all sides:

In a way the detachment had something to do with this event, as a lot of ice floes that had been part of the little ice pack were still bobbing around (just not enough to show up on the Uni Bremen sea ice concentration maps), making freezing easier near the coast. But of course, the main reason for this to happen is the fact that the edge of the ice pack was so far North. I think bilateral freezing on this scale is a first.

So the East Siberian and Laptev Seas froze over relatively quickly, the Beaufort Sea was behind on schedule (causing a nice new record anomaly dip), but is now filling up nicely, Baffin Bay and the Kara/Barentsz Sea are late, late, late (as can be seen on the regional graphs page of the ASI Graphs website). I will soon start to compare those regions, but in the meantime I have extended the concentration maps on the aforementioned ASI Graphs website with November, so you can compare to previous years yourself.

Comments

[Werther]

Hi Neven,
As I'm on line right now, I'd like to give a first reaction.You're right IMO that remaining shrapnel did play a role in the quick refreeze over there. I think excess fresh water through exceptional melt was the other reason.
I'm very interested in coming comparisons on the Atlantic side. It seems unbelievable that November ice is to be found a staggering 300 km North of Frantsa Yosefa!

[Werther]

And for a good nights sleep thought...
Where's the cold???
Check it out on Daily Mean Composites 9 November? Mongolia? Alberta? Oh yes, Mid-Greenland (that's comforting...).

[Apocalypse4Real]

I think there is another factor in the quick refreeze of the Beaufort that is unfolding - that is the wind pushing the sea ice towards Alaska and Canada. See the HYCOM runs for the next five days - that I think the results will cover the remaining open area:

http://www7320.nrlssc.navy.mil/hycomARC/navo/arc_list_arcticicespddrf.html

On the other hand - I suspect the quick refreeze is locking in warmer saltier water below the surface, that will have a reverse impact during 2013's melt.

[Kris]

Apocalyps wrote:

On the other hand - I suspect the quick refreeze is locking in warmer saltier water below the surface, that will have a reverse impact during 2013's melt.

.

Quite right.

And it probably already happened in October-November 2011 too.

[Kevin McKinney]

Interesting. I note that those waters are among the shallowest in the Arctic Basin. How much bearing does that have on this refreeze pattern?

http://geology.com/articles/arctic-ocean-features/arctic-ocean-seafloor-map.jpg

[D]

I'm wondering if anyone is up to speculating on what will be the maximum sea ice area as measured on CT"s Interactive Graph during the 2012/13 season. Because of the events last summer, anything over 12.5 million sq. km. will be surprising. A peak that didn't get to 12 million sq. km. would be another very big red flag for humanity and one over 13 million would have this observer breathing a sigh of relief.

Your thoughts are welcome.

[iceman]

hmmm, it seems that the surrounding ice would reduce wave action in the enclosed open water. This would cause
1) faster re-freeze in that locale;
2) smoother surface, probably even more than is normally characteristic of first-year ice.
If the smooth area survives ice pack dynamics during the winter, it would have different ponding behavior the following spring. See the Perovich & Poshenski study on how FYI differs markedly from multi-year ice in its seasonal albedo and melt pond formation. SekeRob commented on the latter; "...it's nice and smooth and promotes much larger and interconnecting ponding than ... MYI." In a similar vein, Flocco et.al. in JGR found that ice beneath ponds melts up to three times faster than bare ice. Higher melt rate, earlier start. You can see this as a steep downslope in the anomaly graphs.

There's a lesser but still reinforcing factor that Lodger pointed out in the _PIOMAS September 2012 (minimum)_ thread: Earth' solar aphelion comes after the northern summer solstice, so ToA solar irradiance is slightly higher before the solstice than after. This consequence of the Earth's elliptical orbit multiplies the effect of lower albedo early in the melt season.

So there are two trends that appear to go in in opposite directions, on just slightly different timescales. The first is widening seasonal percentage swings in sea ice area, as illustrated by Jim Pettit's graph. crandles noted one possible cause in an earlier post: "One way you can get more ice at maximum in the seasonally ice free version is that it misses out on autumn/fall snow cover which is excellent insulation compared to just ice."

On a y/y time scale, Arctic ice extent (though not volume) has had a one-year negative autocorrelation. For some discussion sans consensus, see the thread _Arctic sea ice loss and the role of AGW_. I think Chris R also addressed the topic in his Dosbat blog.
This rebound behavior - especially following record ice loss years - is one reason Notz & Martotzke ruled out self-acceleration as a dynamic in sea-ice cover. (see their Figures 1(b) and 1(d), paper available here.) The correlation seems to have weakened in recent years. For September extent, if not for March, it may have even flipped to positive. A big question for 2013 is whether the recent pattern continues. The combination of trapped heat (noted above by Apocalypse4Real) and lower albedo from earlier and more widespread ponding seems to favor another extreme low in sea-ice cover. That could be a harbinger of self-acceleration taking hold.

"Accentuate the positive," as the happiness researchers like to say - except here we are dealing with climate feedbacks.

[Peter Ellis]

Oh come on people, scale! Saying the ice will damp down waves in the surrounded area is like saying there are no waves in the Mediterranean because of the surrounding land. The size scale for wind/wave dynamics is smaller than a single pixel on satellite maps.

[iceman]

Peter, your counterexample is flawed on two counts. First, the mean significant wave height averaged over the Mediterranean Sea is less than 1.5 meters, even in winter. Second, if you look at the animation above, it's clear that the ratio of ice edge to water area in the freezing region is far higher than it would be under normal freezing outward from the pole.

Your views seem to be at odds with oceanographer Luc Rainville, who was quoted in Discovery News as saying that sea ice muffles all waves "like a big damper."

[Chris Reynolds]

Iceman, Peter Ellis,

The waves in the Mediterranean are typically smaller than one expects in the Atlantic or Pacific because of less 'fetch' over which wind action can form and reinforce waves. Sea Ice does damp waves. But the early season ice is thin and fragile, so its damping will be less.

D, Iceman,

The problem I have with less snow cover affecting ice thickening is that it seems to me that this factor is most at play in the Arctic Ocean basin itself, whereas the maximum is set outside the basin - feel free to correct me if I'm wrong. I can't hazard a guess as to what the maximum will bring. That's mainly because I never did figure out why after 2007 there was a rebound in winter (CT Area) that lead to three years with roughly the same area, despite different weather patterns.
http://arctic.atmos.uiuc.edu/cryosphere/IMAGES/seaice.area.arctic.png

I'm still pretty sure that next year's minimum won't see a rebound as 2007 caused - I still think the ice is critically thin and we can expect at least a match of 2012, on balance I expect a further drop, possibly even a crash below 2012 of similar percentage magnitude (CT Area).

[Enno]

why would a record low melt (record low surviving ice) have to be followed by less winter ice?

take the thought example of a just full meltout of ice in summer. Then all ice in fall and winter has to grow from scratch. (In reality more complex as it will depend on state of the water/mixing etc but bear with me). In that argument, the amount of ice formed depends on the strength and weather etc. of the winter - and not on the weather of the preceding summer.

There should be no reason to expect that a specially low ice winter has to follow a low ice summer. Unless it can be shown that low-ice summer conditions by themselves prime the atmosphere for less ice growth in winter. As far as I am aware (big caveat), that is not clear.

In opposition, a low ice winter DOES prime a subsequent summer for low melt, if there is simply e. g. less ice thickness.

So I would suggest that the natural expectation should be that summer ice extent has a memory of the preceding winter, but winter ice extent not a memory of the preceding summer. Is anything wrong in this argument?

case in point - last winter´s end-winter growth spurt, happening (as Chris points out) in completely seasonal seas. How could such a thing possibly be foretold from the end-of.summer ice state? Ergo, how could one even expect that winter low should follow summer low a priori?

I do suspect that a summer low can create special conditions in either sea or atmosphere to hamper winter´s ice growth abilities; but then people might better try and figure out what exactly these changeways are in order to learn to predict whether winter will be high or low ice. (case in point - Beaufort upwelling retarding ice growth for several weeks circa; big aha event for me).

[Enno]

... "I think bilateral freezing on this scale is a first. " ...

Is it? Looking at CT ice plots: (Laptev/ESAS)

terms:
seaside: ice cover grows from the basin until it hits land
landside: ice growths all from land until it hits ... ?
bilateral: as in the animation: two growth fronts meet surrounding a middle opening
zipper: shelf sea never fully open in summer, closes like a zipper along coastline

2000 Oct 10 largely seaside
2001 Oct 1 largely seaside never fully open
2002 Oct 15 largely seaside (zipper freeze)
2003 Oct 17 bilateral
2004 uh no images ?
2005 Oct 21 half zipper, half bilateral
2006 Oct 15 bilateral
2007 Oct 20 half zipper, half bilateral
2008 Oct 16 mostly bilateral
2009 Oct 11 mostly seaside (leans bilateral)
2010 Oct 19 bilateral
2011 Oct 18 ... a bit bilateral then zipper
2012 Oct 22 strongly bilateral

to me this looks like a trend:
* progressively later freeze (later date of similar state) (not entirely unexpected this trend, ahem)
* progressive change of way of freeze from seaside/zipper to bilateral.

it also makes sense: if the northern sea route didnt open up in the first place then one cant really have "bilateral" freezing ...

This year does stand out in lateness and clarity of the "new" pattern; even though this is the marginal sea that froze first.

if one looks closely at the animation, then one can see that the ice that grew "out from land" doesnt just do that, it seems to grow near cost parallel towards the east. Best spot to watch: Lyakhovsky island (of the New Siberian Islds), before 20th Oct. Why this directionality? is this some consequence of sea currents?

self answer using NCEP : prevalent west winds there at the time. Still: ice growth in wind shadow areas? Does that make sense?

[crandles]

A while after the summer storm, I did ask if it was crazy to think the freeze might occur from edges inwards rather than from remaining ice outwards. Further discussion about this did suggest that it was most likely for shallow water to freeze from edge inwards.

This revised view of shallow water freezing from edge inwards does seem to vaguely resemble what has happened. But how good an explanation does that make it?

(The idea was that water would be mixed to greater depth than usual so there would be a taller column with more total heat to be lost to atmosphere before freezing could begin. Thus slowing the freeze from remaining ice outwards towards edges as well as effect of minimum ice edge being a long way North.)

If it is a reasonable explanation, presumable the later and stronger the summer storms we get, the more pronounced the freeze from edges inwards effect should become. How strong is that effect likely to be compared to smaller minimum ice areas delaying the outwards expansion of the ice?

Both these effect suggest we will get more of this in future. But maybe there are many other effects that have stronger influences?
(I would be very surprised if this was an adequate explanation without other strong influencing factors.)

[Apocalypse4Real]

The updated sea ice thickness and concentration are updated through Nov 10:

https://sites.google.com/site/apocalypse4realseaice2012/home/sea-ice-concentration-and-thickness-comparison

The methane release data has also been updated through Nov 10:

https://sites.google.com/site/apocalypse4realmethane2012/home/2012-vs-2011-airs-ch4-359-hpa

[Fairfax Climate Watch]

Apocalypse4Real: beautiful data display! I think the zoom on the recent ones is very effective.

And how clear is it that the ice is super thin! 2 to 3 meters thick for a good solid portion in the center, but the rest is hardly there! 2013 could be the lucky year for those betting on sea ice demise.

[Chris Reynolds]

M Owens,

First year ice is typically very thin at this early stage in the winter. PIOMAS averages for November show large areas under 50cm in recent years. It's later in the winter that the ice thickens, though by how much remains to be seen this year.

Anyone know why there's a delay in the PIOMAS figures?

[bluesky]

Maybe it has been already noticed, apologies as I do not follow the blogg regularly.
Anybody would know what is happening North of Franz Joseph Islands very close to the North pole? Arctic sea ice concentration appears in yellowish even green (70-75%):
http://www.iup.uni-bremen.de:8084/ssmis/arctic_SSMIS_nic.png
At the same time the temperature anomaly is at +10 C
http://meteomodel.pl/klimat/gfsanom_np.png
around -14 to -12 instead of -24?
http://meteomodel.pl/klimat/gfssr_np.png

[Wipneus]

PIOMAS update:
Last value: 2012-11-5 6.648

I have updated my graphics at ArctischePinguin for the latest data:

Monthly data
Daily data
Daily data with a "prediction" based on exponential trend
Daily Anomalies

[Neven]

Thanks, wipneus. I'll have a post up as well.

[Aaron Lewis]

Enno,
Recursion to the mean. We all learned that if this summer is hotter than usual, next summer will be cooler. We always think, "Soon the weather will go back to normal."

Many people have not internalized what it means to live in a time of global warming. Now the rule is, " No matter how hot this weather is, soon the weather will be warmer."

[Neven]

PIOMAS post is up.

[Enno]

Aaron, thanks. I agree completely, but my keyboard ate a more explicit reply (sorry). (In short .. I woke up to a sense of unreality about the weather in the same year 2006-7 that led to the first great ice collapse: as we had a 3/4 yr of semi incessant SW winds, from the atlantic, that annoyed me on my daily bike commute ... I realized only later that those winds were part of the same abnormal weather that transferred so much heat into the arctic that year; which led me to turn into an arctic watcher ... as things go.)

Same sense of unreality I had when I looked at the (excellent) thickness maps A4R put up above. There´s something seriously weird there "above" Franz Josef Land, as werther et al also said. it looks like there is a strong west drift pushing the ice nearly reverse of the usual transpolar drift direction, see Neven´s graphs page. But what is cause and effect there? That ice edge seems to behave like a half-polynya? Has that ever happened before? Where is this west-to-east stream (along the north edge of the siberian high) coming from?

Neven: I write actually to point out a development on a related blog,
http://fromtheblueside.blogspot.de/2012/11/new-projects.html

if that blog follows through that could turn into a helpful site, like "ask-an-oceanographer". That could be very useful, and people here could find it interesting to get answers there.

[Werther]

Hi Enno,
I just made an entry on the 'weirdness'-thread that relate to your questions. Another part of it being posted on '2-million mark', it looks that I'm pretty spread-out.
It hurts the consistency of my wordsalad, I'm afraid.

But I find your entry to justify this part of my Daily Composites journal:

'One of the features contributing to an accumulation of air to form the Siberian anomaly is a transport corridor on 500 Mb from the Atlantic near the Canary Islands, through Eastern Europe to the Kara-/Barentsz region. From there, it heads right into the Central Arctic. The corridor is not consistent, but shows a regular anomaly. A similar feature remarkably fits the ‘Sandy’-track and leads into Nunavut.

[opensheart]

What if the BiLateral refreeze that occurred in the East Siberian and Laptev Seas in Oct 2012, and discussed in Naven’s November 2012 blog contributed to/is related to the massive shattering of the Arctic Sea Ice pack in the winter, Feb/March of 2013?

What if the line of Thick multi-year ice that normally piles up along the Greenland/Canadian Archipelago almost to Alaska forms one side of a triangle. The opposite point being the East Siberian/Laptev Seas. A thick region of ice would stretch across the Arctic from Greenland/Canadian Archipelago to the shores of the East Siberian/Laptev Seas . This would form a triangle of sorts that would stabilize the Arctic Sea Ice . Keeping it from being blown/pulled/shifted all over the Arctic Basin. Allowing the ice to grow strong and thick.

But what if this BiLateral refreeze, where the ice grew from the edges, leaving a hole/weak spot in the middle, fundamentally changed the structure/strength of the ice in this area. So that this 3rd point of the stability triangle was weakened to the point where the whole ice pack was no longer stabilized. The ice pack was free to be pushed, pulled, twisted, moved around whichever way the winds/conditions warranted. Thus the ice shattered.

Like packing a pane of glass in a shipping container. Packed correctly, the packing material absorbs the blows of being transported and the glass is likely to remain intact. But loose packed, where the glass can bounce around inside the shipping container, and it will surely arrive at its destination in a thousand pieces.

[Neven]

That's an interesting thought, opensheart (welcome BTW). I think this could be a factor, but the main reason for the cracking event to happen, is that the ice pack is thin on all sides.