« Yet another broken record | Main | December 2011 Open Thread »


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


Hi Neven,

Find it interesting that that the storms of 1974 and 2011 are at almost exactly the same calendrical date...

Chris Reynolds


I think I'm correct in saying that the storms this time of year are ones that come in from outside the Arctic. I've not followed this one too closely (merely read Neven's posts - thanks), but this seems to have come in from the North Pacific.

There's probably a Pacific mechanism similar to that in the Atlantic where typicaly autumn low pressures are deeper and give stronger winds. I think this is due to the mid-lattitudes still being warm as the Arctic rapidly cools - hence a strong temperature gradient.

Artful Dodger

Neven, I disagree that this storm probably caused sea ice ridging and rafting.

Ridging requires significant mechanical strength from the ice. In this region, new sea ice is less than 40 cm thick and highly saline. Not gonna happen.

Rafting is a compaction process. If this was the case, we would see extent go down and area remain steady. MASIE and CT show this is not what occurred.

So what did happen? Most likely the new salty sea ice melted due to mechanical mixing of the surface layer. There is still plenty of heat in the top 50 meters of the Chukchi sea, and wave heights of 10 meters will easily mix the 0.4 m of new ice with the warm water below.

Look at the SST maps for the Chukchi over this period. You'll see that water temps above freezing have moved North throughout the storm. I suggest that this ice melted.

I Ballantinegray1

I'd tend to agree with A.D. here? Another point being the loss of the old Halocline( meaning a 'new' shallow layer forms each year leaving warmer waters at depth). 'Rafting merely depresses the ice back into the melt zone and we end up with the standard 2-3m depth ( as we see with Antarctic sea ice?).

Chris Reynolds

I agree that there will have been substantial vertical mixing allowing warmer deeper water to mix to the surface. However it'll rapidly lose heat to the atmosphere. I'm not sure how much ridging will occur.

NCEP/NCAR daily composites are available for a few days back, currently they're up to 13 Nov. However I don't know if they include ocean mixing as part of their physics - I suspect this is a detail that wouldn't be modelled.

In short I doubt we'll ever know.

However if vertical mixing has occured that means more heat loss to the atmosphere and less heat to affect the winter freeze/next year's thaw.

Artful Dodger

Hi Chris. That's true in a closed system, but the Chukchi sea is open to the Pacific.

Chris Reynolds

Hi Artful,

It's true in the Chucki Sea too (if you're talking about heat loss) - because the amount of heat in Chucki doesn't affect the amount of heat driven in through the Bering Strait - not directly anyway.


Come on, guys. Look at the animation. There is no way that all of that ice melted due to the warmth from below. Some of it maybe, but I think that most of it has been compacted.

Sure, the ice is thinner, otherwise the storm wouldn't have made such a big dent, but that doesn't mean it can't be pushed together or piled up. By the time the water is mixed, all the ice has been pushed away by those 50-70 mph winds.

That storm has now been replaced by a very big high, and we see a little hot spot where the ice has been cleared by the storm:

It could be that those warmer waters from below are now releasing their heat to the atmosphere, which explains why things haven't frozen over yet.

Anyway, I wouldn't be surprised if a relatively large area of the ice in the Chukchi/East Siberian Region turns out half a meter thicker than it would have been otherwise.

But that's just what I think. It's my first big autumn storm in the Arctic. :-)

Kevin O'Neill

Daily Cryosat data would clear this up ....

Looking back at the last decade of Cryosphere Today images for mid-November, there doesn't seem to be any correlation between ice area in the Chukchi and the following year's extent. Baffin Bay, on the other hand, might be a halfway decent predictor. I haven't tried to run any numbers, just eyeballing the comparison maps.

Kevin McKinney

Very interesting--thanks to all.

"There is no way that all of that ice melted due to the warmth from below."

I wouldn't be *too* sure, Neven--I don't know anything specific about this question, and don't have a strong opinion either way, but do recall that turbulent mixing has a very strong effect on heat transport in water--including the melting of ice. Which suggests to me that general "norms" as to how much ice could melt (and how fast) could be misleading to the intuition on this specific (and 'outlying') event.


OK so assuming mixing has taken place but not enough to melt it all then you have stacked thin ice from compaction in warm water. what is to stop it all spreading out again before the surface water gets cold enough to freeze the mix together? at 40cm thickness presumably it has broken into fairly small sheets has it not?


"what is to stop it all spreading out again"

If it is small pieces on small pieces then it may spread out again if that happens before the pieces freeze onto each other. However, if the small pieces have been thrown onto larger pieces of ice that haven't broken up then how would they spread out?

Bob Wallace

Is anyone thinking that this storm event could have piled up ice enough to be noticeable during next summers melt?

When I look at the extent and area graphs there was an unusually abrupt change in freeze behavior but things are now back on track.

That suggests that there might be some extra thickness in some places, a little extra volume by the end of the winter. Enough to flatten out the downward spiral a bit? Or are we talking small average thickness gain?


A large number of thin sheets would have to stack upon each other before their combined height would have a noticeable effect.

I'm more inclined to think that we might find a small band of stacked ice at the fringes, but that the ice in the center portion of the now open area has succumbed to turbulent mixing / melting, and that the effect will be negligible come summer.


with violent mixing the ice would have become small pieces, thin ice would not survive in large sheets would it? so as Twemoran as suggested a narrow ridge of stacked ice on the edge of the smashed area makes sense but not large areas. what mechanism would create them? either the ice is being moved violently and gets smashed and is no longer stackable or the ice damps the wave action and no stacking would then occur would it?

Rob Dekker

Neven wrote

"There is no way that all of that ice melted due to the warmth from below."

I tend to agree with you, Neven. The waters are too cold for sudden melt, and besides, there is a much easier explanation for the reduction in area and extent : a storm flushes water over ice very easily, which 'fools' the satelite instruments in interpreting (thin) ice as water as long as the waves dominate the area.

Similar to what we saw happen during the "Flash Melt" late summer, (I don't have the link to your post on that ready), I think in a few days, when the water starts to calm down, and precipitation freezes over the ice, it will indeed be as if 'nothing happened'.

As for ridging and rafting, I tend to agree with Lodger : the ice in the margin is too thin now for serious ridging, and even if ridging occured, thermal insulation of ice should eliminate any differences with now open water in a month or two.

Finally, a storm typically stirs up the waters in the upper 10 meters or so, but that effect is quickly eliminated once you get under ice. For example, ITP41 (tethered profiler) is located only a few 100 km from the turbulent waters in the margin, and it shows no noticable effect on deep water heat profile around the time of the storm :


It's clear that there is heat trapped under the Arctic sea ice, but it's apparent that this storm did not disturb the 'stratification' much, at least not under ice...

Here are more profilers :

Chris Reynolds

Rob & Neven,

I don't think the storm will have mixed near to ITP41 as this is well under the ice pack in northern Beaufort.

ITP55 is a better candidate, and this has a salinity anomaly (25m deep) after the storm - could this be the signature of mixed water drifting into the area from the open seas under the storm? This suggests mixing but not deep mixing, and not enough to affect the temperature stratification.

More interesting to me is the blip anomaly on temperature and salinity around day 312, that's 9/11/11 (9 Nov) if they're counting from 1/1/11. It seems to be around the time the ice in Chucki dropped back (from Neven's animated gif). Jeff Masters blog post is dated 10/11/11 and refers to the storm hitting 21:00 the previous night, i.e. 9/11/11.

So it seems to me that at least one profiler shows an episodic deep mixing under the ice, however in the area of that profiler the event did not cause warmer water to come near the surface. That said by definition the profiler is ice-mounted so records what's going on under the sea-ice. I'd suggest that this profiler implies that the area of exposed ocean could have experienced deep-mixing. However the warm patch Neven shows on the NCEP data may merely be due to open water, not due to mixing, which may not be modelled in NCEP.

PS Last night on Attenborough's Frozen Planet pancake ice was shown ridging in the Antarctic - thin ice can ridge.

Artful Dodger

Hi Chris,

Unless the sea ice rafting video you refer to was taken in 50 knot winds with 30 foot seas, it does not apply to the events in the Chukchi sea on 11/11/11. New sea ice simply does not have the mechanical strength to withstand that kind of pounding. When mixed with that volume of water below, that new ice melted.

In fact, this storm ate new sea ice all the way back to the edge of the multiyear ice, North of Wrangel Island at 76N, 180E as shown on Uni-Bremen's SMIS sea ice maps:

Sep 29, 2011
Nov 12, 2011

Rob Dekker

Chris, thanks for digging up ITP55. Indeed it's a lot closer to the margin than ITP41. Still, I see no noticable temp anomalies in the upper ocean profile since the time it was open ocean. The salinity anomaly you note after day 316 (Nov 11) actually seems more a resumption of measurement (black suggests that no measurements were taken, or data was not reliable) rather than a real 'anomaly'.

There are some upwellings of heat to the surface, most notably around day 305 (Oct 31) and some stirring at 75-100 meter deep around day 311 (Nov 6) but very little activity in the profile around the time of the storm (day 316, Nov 11).

The temp profile around the same time actually suggests that the upper ocean water is a fraction of a degree colder after the storm than before, if anything. Either way, no suggestion that the storm caused any heat from below to be stirred up to the surface at all, so there is little reason to suggest any "melt" to have occurred.

I'm a bit surprised myself by the relative indifference of the sub-surface heat to this storm, at least if we look at the profile data. But then again, even ITP55 was under ice all through this storm and we have no real measurements of profile data in the flushed-out area of the Chucki around Nov 11.

Measurements in the Arctic always seem to be few and far in between...

Regardless, I would NOT want to be a polar bear (cub) swimming back to shore during this lethal storm...

Kevin McKinney
The temp profile around the same time actually suggests that the upper ocean water is a fraction of a degree colder after the storm than before, if anything. Either way, no suggestion that the storm caused any heat from below to be stirred up to the surface at all, so there is little reason to suggest any "melt" to have occurred.

What about the opposite effect? In other words, the mixing of water and ice in the turbulent surface conditions of the storm facilitate melting even with the comparatively small temperature differential; the melt lowers the temperature of the surface seawater. If the math works, then the 'possibly colder' water is an indication that perhaps some melting took place?


Hi everyone,

I've really enjoyed following the blog since the summer but haven't had anything to contribute so far.

Here's one observation for others to critique, but looking at the ITP55 salinity data for November 11th, I interpret the black colors before then to mean the salinity was below the scale, i.e. below 25. After the storm passed the salinity was raised to 25. The temperature for that day in that region seems to be about -1.2 - -1.3.

Google leads us to a water freezing temperature by salinity calculator: http://www.csgnetwork.com/h2ofreezecalc.html

At 23 PSU salinity the freezing temperature is -1.255 deg C; at 25 PSU salinity the freezing temperature is -1.366 deg C, so even if the temperature remained constant the changed salinity due to storm mixing might have melted the ice, as it seems to be right on the phase transition.

The next question this raises is energy (enthalpy/entropy?) changes in a unit of water due to dissolved salt versus energy changes due to raised temperature, but I'm afraid my thermodynamics skills were never adequate to answer that question and certainly aren't now.

Anyways, just thought I'd contribute to the conversation. Thanks again to everyone for their erudition, civility and hard work.


Good observation D. Definitely possible.

Rob Dekker

D, your suggestion sounds very plausible !

In fact, I owe ChrisR an apology, since indeed to looks like color black suggests that salinity is less than some threshold (25 PSU?), and thus there really IS a salinity anomaly around the time of the storm.

As for the thermodynamics of salinity change, here is a first approximation to quantify the effect :

The ~0.1 C change in melting point you propose (due to salinity changes) allows water directly under ice to cool down 0.1 C. ITP55 suggests that water cooled down to some 25 meters, which represents about 3 cm of ice melt (delta T * depth * heat-capacity / heat-of-fusion).

So, the salinity change, caused by the storm stirring up water down to 25 meter or so, represents a potential for some 3 cm bottom melt. That's fairly small, but combined with top melt caused by water flushing over the ice and snow cover) it seems plausible that at least some melt should have occurred directly as a result of this storm.

Moreover, the change in salinity, and the cooling of the upper layer, seems to have set in motion some convective forces :ITP55 shows now (more than a week after the storm) that deeper (warmer) water is convecting upward. If (and how much of) that heat reaches ice, the storm may cause more bottom melt to occur, even long after the winds settled down...

Interesting stuff, guys !

Chris Reynolds

Sorry, I posted got busy with other things and clean forgot, remembered on the bus home tonight.

I must confess to not having been following this storm, so may have dates wrong. But as I said above Jeff Masters blog and the animated gif in the main post both suggest there was a substantial impact on the ice from 9/11/11 to 10/11/11. It's around that time that ITP-55 shows a disturbance of the temperature profile, I suspect that the later salinity anomaly may be mixed water from the area of open water pushing under the ice into the area of ITP-55.

If I'm correct this suggests signifcant mixing in the area of open water caused by the storm.

Having had time to reflect on this matter, and despite Ds useful observation, I now agree with Neven that the loss of ice is mainly mechanical in nature - compression and mixing of thin ice, and thin ice being turned into a 'slush' in the water which wouldn't show up on satellite as ice.

Artful Dodger

Hi Rob,

Good first approximation. Your calculation of 3 cm of bottom melt near the Ice tethered buoy, while interesting, is somewhat removed from the edge of the sea ice.

Recall, this instrument was placed in 3+ meter thick multi-year sea ice during the Summer of 2011. This will significantly displaced horizontally from the sea ice edge in the Chukchi on Nov 10, after months of sea ice growth.

I suggest the next step in your successive approximation is to estimate the distance from the buoy to the ice edge at the beginning of the storm-induced wave action.

You may find that the Ice edge moved NW some 200 km during the duration of the storm. Also, buoy tracks should provide the relative location of the temp/salinity sensor during that time.

I don't think anyone can suggest that the mixing observed beneath the buoy is anywhere close to the mixing at the ice edge during the storm, even though it may hint at the magnitude experience there.



Hello gentlemen,

These are very good points and suggestions, and since I'm on the trail I think I'll keep digging a little deeper, so to speak.

Here's an interesting paper that seems relevant to the circumstances: "Storm-driven mixing and potential impact on the Arctic Ocean" http://www.whoi.edu/fileserver.do?id=48251&pt=10&p=33752

I just started looking at some of the other buoy data, and check out ITP54: http://www.whoi.edu/page.do?pid=75638

Not sure what's happening there but something seems to be.

I've read only the first two pages of the paper, but hit some interesting sentences already, for example "In this study we will
show that intense storms could actually force mixing through the Arctic halocline to the thermocline." (p. 2). Also "...a storm in October 1988 intensified vertical mixing, enhanced the entrainment of warm and salty Atlantic water into the mixed layer, and resulted in considerable
melting of sea ice [Steele and Morison, 1993]." (p. 2-3)

As I said, I'll keep digging but thought I'd share what I've found so far in case others are interested in joining in.

All the best,

Rich and Mike Island

How often are strong deep storms in the Arctic ocean in winter? I would think that the strange configuration of a slightly milder Arctic Ocean surrounded by very cold land masses would keep temperature contrasts down. So how deep do storms in the deep arctic get in winter?

Rob Dekker

Lodger I don't think anyone can suggest that the mixing observed beneath the buoy is anywhere close to the mixing at the ice edge during the storm, even though it may hint at the magnitude experience there.

That is very true. It seems quite obvious that in wide open ocean, this storm stirred up water from much deeper than what ITP55 experienced. However, the question was about 'melt', and for 'melt' to occur, you need heat or a change in melting temp (due to changed salinity).

With atmospheric temps well below freezing (-20 C at the time of the storm), and surface water temp close to sea-ice freezing temp, the change in salinity that D proposed (and shown by ITP55) is the only 'immediate' source of melt. That is a temp difference of some 0.1 C, so even if this storm stirred up water down to some 75 meter, there is no more than 9 cm potential melt obtained from salinity change. No matter how far from the ice edge you are.

Now, ITP55 actually shows a delayed response that is very interesting :

Remember that the (peak of the) storm occurred around day 315 (Nov 11). This profile shows that a lot of salty and warm water mixed in the upper 50 meters since then, and the temperature profile shows that a 10 meter blob of -1.0 C surfaced and is replaced by -1.6 C water. That's approximately 5 cm of bottom 'melt' potential ! Obviously, because of the more thorough mixing in the upper 50 meters, stratification was disturbed, and heat was released from the deep. And there may be more to come. Since salinity gradient is now very even over the upper 75 meter, there should be some more convective forces at play that may release another 'blob' of heat from below.

Keep your eyes on that URL in the next few days...

Either way, my point is that open ocean does normally not have much of a stratification layer, thus there is little potential for heat from the deep, or even salinity change to occur. But we will never know the exact data, since we don't have ITP data from the open ocean Chucki sea area at the time of the storm.

Rob Dekker

D check out ITP54: http://www.whoi.edu/page.do?pid=75638

Not sure what's happening there but something seems to be.

Yeah. Not sure what's going on there. But whatever it is, it has been going on since day 296 (Oct 23), well before this storm hit.

Kevin McKinney

"How often are strong deep storms in winter?"

I heard a report on NPR when this storm hit in which an Alaskan meteorologist said the low pressure observed was not unprecedented, and that the storm season there was (IIRC) October through (I think) March. To him, the unusual thing about this storm was not its strength, but its "meridional track"--that is, its south-to-north trajectory.

I tried a quick Google search to see what I could see about the AO 'storm season,' and didn't really pull up much. But there was this UK item on "Polar Lows" which gives a little bit of useful background:


Rich and Mike Island

Here is a really cool video of a 'brinicle' forming, bringing extreme cold from above to the sea floor.


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