NSIDC Arctic sea ice news September 2012

A couple of days ago the NSIDC released its latest analysis for the month of September.

For the exact numbers with regards to the shattered record I refer you to the article itself. This graph says it all:

The analysis then moves on to some interesting comparisons with 2007. First off, the differences in atmospheric patterns, something I wrote about at the end of July (Arctic atmosphere June-July 2012):

Weather conditions prevailing over the summer of 2012 were quite different from those in 2007. The summer of 2007 featured unusually high sea level pressure centered north of the Beaufort Sea and Greenland, and unusually low pressure along northern Eurasia, bringing in warm southerly winds along the shores of the East Siberian and Chukchi seas (3 to 5 degrees Celsius, or 5 to 9 degrees Fahrenheit above normal), favoring strong ice melt in these sectors and pushing the ice away from the coast, leaving open water. The pressure pattern also favored the transport of ice out of the Arctic Ocean and into the North Atlantic through Fram Strait.

These images compare sea level pressure and temperature anomalies (at the 925 hPa level) during summer 2007, the previous record low extent year, and summer 2012. Anomalies were less pronounced in 2012 than in 2007 (as shown in reds and oranges). While weather was a factor in the 2007 record low extent, the 2012 record extent occurred during near average weather conditions. Credit: NSIDC courtesy NOAA/ESRL PSD

In contrast, the summer of 2012 saw unusually low pressure along the Eurasian coastal seas and extending eastward into the Beaufort sea, most prominently over the East Siberian Sea, with unusually high pressure centered over Greenland and the northern North Atlantic. Air temperatures for summer 2012 were above average over most of the Arctic Ocean (1 to 3 degrees Celsius, or 2 to 5 degrees Fahrenheit), most prominently over the Beaufort Sea, where, because of the pressure pattern, winds were anomalously from the south. Melt began two to three weeks earlier than average in the Barents and Kara seas, leading to earlier retreat of sea ice in the region; however, air temperatures remained below average during summer in this region. This points to the impact the continued loss of old, thick ice is having on the ability of the sea ice cover to survive summer melt. Other than the August storm, the pressure pattern in 2012 does not appear to have been as favorable in promoting ice loss as was the case in 2007, and yet a new record low occurred.

And then it's time for the yearly Arctic sea ice age map and graph:

The end-of-summer sea ice cover was not only the least extensive in the satellite record, but also very likely the lowest volume, based on combined model-observation estimates from the University of Washington, and inferred from ice age. The extent of ice of nearly all age categories declined from last year and remained at record low levels. The only category that increased was 4-year-old ice. This is ice that has aged since the previous record low minimum extent in 2007, when substantial amounts of first-year ice were lost. This 4-year-old ice will now age into the 5+ year category as the ice-growth season begins. However, even with this replenishment this winter will see only approximately 20% of the old (5+ year) ice compared to the 1980s. Because of the record summer ice loss, this winter will see the Arctic Ocean region even more dominated by the thinner first-year ice. As shown in Figure 5, the amount of ice in nearly all age categories has decreased since 2007, particularly the oldest ice.

The analysis ends with some background info on the Antarctic sea ice maximum. It's very nice of the NSIDC to provide the background information lacking in the yearly fake skeptic diversionary action, aimed at downplaying what's going on up North (anyone still falling for that?). This piece on the New York Times blog by Justin Gillis also puts things nicely into perspective: Running the numbers on Antarctic sea ice

Comments

[Andrew Xnn]

Useful to remember that multi - year ice is still a very significant portion of the total during the September minimum.

Wondering if multi-year ice will disappear all at once, or if there will be a gradual reduction in the oldest ice first.

That is could we expect to have a time in the near future where only First, 2nd, 3rd and 4th year ice is present or will it all be gone in 1 summer?

[Nightvid Cole]

Andrew, I think there is still some very old ice (like 10 yrs +) up there, but it's broken up and probably covers less than 50,000 km^2 . Extrapolating, I don't think the ice will vanish completely in an "oldest first fashion" exactly, but of course I'm just guessing...

[Kris]

Andrew Xnn wrote:

multi-year ice is still a very significant portion of the total during the September minimum

.
On the contrary my dear fellow, on the contrary. End of August was left less than 13 % multi-year ice as shown in this NDIDC map.

Even stronger, this winter there won't only a sort of recover to 20 %.

Suggest you update yourself by reading the NSIDC 5th Oktober "Old, thick ice dwindles; young, thin ice prevails" news flash

[idunno]

FWIW, IMHO 2007 is/was a great distraction from the long term picture.

I think that the Arctic is melting mainly due to an increase in the amount of warmer water flowing into the Arctic from the Atlantic side. I think this has been steadily increasing, year on year, due to Anthropogenic Global Warming (AGW) and the peaking of the Atlantic MultiDecadal Oscillation (AMO). A process that is steady, inexorable and predictable.

I think that the collapse of 2007 had little to do with this. I think it was caused, along with the local Arctic weather, by an unpredictable surge of heat entering the Arctic from the Pacific side, due to the El Nino.

I think that the amount of heat heading up to the Arctic on the Pacific side is much smaller. But I think that Pacific heat input can fluctuate wildly from year to year. And in 2007, it was unusually high.

This might be better expressed if I make up some fictional numbers:

Solar heat: 5,000 units per annum, increasing by 5 per annum (0.1%) due to AGW (but +/- 50 (1%) due to weather)

Atlantic heat: 5,000 units per annum, increasing by 5 p.a. due to AGW, and a further 5 due to a peaking AMO (0.2%), (but +/- 50 (1%) due to natural variability).

Pacific heat : 2,000 units p.a., increasing by 2 p.a. due to AGW (0.1%) (but +/- 200 (10%) due to natural variability, associated with the El Nino cycle).

Baseline Increase Variation

Solar 5,000 +5 p.a. +/-50

Atlantic 5,000 +10 p.a. +/-50

Pacific 2,000 +2 p.a. +/-200

And I think that 2007 was a +9.98% year on the Pacific side.

(I have thought all this for some time, but have not posted on it before, aware that I was going to have difficulty making myself clear. So it proves.)

Comments much appreciated...

[dabize]

http://www.americanwx.com/bb/index.php/topic/37053-climate-sensitivity-and-timing/

When it comes to tuning out the significance of the Arctic changes, nobody beats some of the characters here.

I feel like I've been wrestling with a bunch of unusually stubborn alligators.............

[Twemoran]

dabize

Amphibians not alligators :-)

Terry

[dabize]

You mean ebullient-thermophile ranids?

[Neven]

There was definitely a lot more heat flux through Bering Strait in 2007 than any other year since measurements began. I've asked around a couple of times to find out how 2012 stacks up, but no answers. I'm going to try again.

[LRC]

The bigger story in the Antarctic is explained very well in this article IMHO http://www.wunderground.com/climate/Antarctica.asp?MR=1. Do not know when this was posted but suspect sometime in '11. Should not think that '12 numbers would have changed things much.
As for that MYI I really think we are taking about a vastly different MYI then 10 yrs ago. I suspect that a very large % of that MYI is rotten and only held together because it is protected from the edges of the pact. once exposed, that 'MYI' will behave no different then FYI ice.

[Chris Reynolds]

Idunno,

I disagree: The evidence of a role for Atlantic Water (AW) in ice loss is not strong. I'm not saying it plays no role, but I am far from convinced about a pre-eminent role.

Polyakov finds a role [1]. 28 to 35cm after 50 years of applied heat flux is found in a model they state that this could account for the observed thinning of landfast ice off Siberia. However their figure 4 shows the broader situation with regards AW, as it is more saline than Arctic water as it reaches the edge of the abyssal deep poleward of Svalbard/Franz Josef/Severnya Zemlya (map) it drops to form a deep layer. Indeed this is the reason why whereas the ice edge has receded massively in the Pacific/Siberian sectors, such recession is not as great in the Atlantic - the ice edge remains pinned to the bathymetry. The AW layer shows warming but is seen to be some 100 metres below the surface [1].

This is confirmed by Korhonen et al [2] who state in their abstract:

Particularly the heat content carried by the Fram Strait branch is nearly conserved as the warm and saline Atlantic Water becomes decoupled from the atmosphere north of Svalbard, where a relatively fresh surface layer is formed by mixing with ice melt. The passage between the Arctic and Pacific oceans is more restricted and the volume transport through the Bering Strait is only one tenth of that from the Atlantic. Nevertheless, the Pacific waters are, at least during summer, relatively warm and fresh (Stigebrandt, 1984) and, due to their low density, import heat closer to the ocean surface and sea ice compared to the Atlantic Water.

This outlines the significance of Pacific Water, which while lesser in volume is not as saline as AW so is more able to actively engage with sea ice at the surface, except in certain regions like the Eurasian coast.

Compare this with 2007 and the atmosphere.

Since 2007 there has been a persistent Arctic Dipole pattern during the months of maximum melt, June, July and to a lesser degree, August, blog. And Arctic Dipole patterns are shown to be a driver of record low sea ice extent, blog. A negative mode AD (using Overland's definition) draws Pacific Water in through Bering, so it is reasonable to assume that there has been an increase in this since 2007.

After 2007 there has been a change in the seasonal cycle of area, ref, after 2010 there has been a change in the seasonal cycle of the PIOMAS volume, ref. That both changes in the seaonal cycle follow large step losses in volume suggests a role for volume reduction in these processes; both 2007 and 2020 were large volume losses, from which there has been no recovery ref.

It's been claimed over here by others that only the AW influx can explain the preferential thinning of thick ice, this is important as almost all the volume loss until recently has come from loss of thick ice (PIOMAS and Kwok). However this neglects the key point that thick ice has a slower response time than thin, so thin ice can rebound each winter, leaving thick old ice to carry the signal of reduced volume forward.

With regards your comparison of insolation and AW heat, Francis and Hunter [3] find that the ice edge is negatively correlated with insolation (up to 2006) (i.e. insolation increase implies less melt), supporting their finding that downwelling infra-red contributes a significant amount to the recession of the ice edge. Downwelling IR is positively correlated (if IR increases then the ice edge recedes)

Links marked blog are link to my blog, at the bottom of each page there's the references.

[1] Polyakov et al 2010, Arctic Ocean Warming Contributes to Reduced Polar Ice Cap. Abstract - sorry can't find paywall free copy.
[2] Korhonen et al 2012, Time and space variability of freshwater content, heat content and seasonal ice melt in the Arctic Ocean from 1991 to 2011. Abstract & PDF
[3] Francis & Hunter 2007, "New Insight Into the Disappearing Arctic Sea Ice." PDF

[idunno]

Hi Chris,

Thanks for your reply.

I take your point about bathymetry; though I think that warmer AW (Atlantic Water) that is hundreds of metres down will, over decades, come to affect the ice at surface.

Wikipedia's "Fram Strait" entry has "In the past century, the sea surface temperature at Fram Strait has on average warmed roughly 1.9°C (3.5°F),..." IIRC, there are estimated to be 6 Sverdrups of AW flowing Northwards through the Fram Strait. This is quite enough extra heat to make my head hurt if I tried to do any maths to it. And still my favoured explanation for this...

http://psc.apl.washington.edu/wordpress/wp-content/uploads/schweiger/ice_volume/BPIOMASIceVolumeAnomalyCurrentV2.png?%3C?php%20echo%20time%28%29%20?

Where I know that you have been doing some very interesting stuff on recent seasonal changes in the rate of decline, (which I greatly commend) I still think the most important point is the bleeding obvious: a 30+ year inexorable decline in all seasons.

I still think that 6 Sv of water at 2C above normal is a primary driver of this long term change.

[Chris Reynolds]

Hi Idunno,

In my most recent blog post I discussed research that, amongst other things, finds that the spread of infra-red forcing in models translates into a 1 to 10m spread in equilibrium ice thickness. That's for a 40W/m^2 spread.

Be careful with the decline in all seasons as the maximum is set outside the Arctic basin, where AW is not an issue. The issue is complex, research shows different processes are at work in different regions at different seasons. However AGW drives ocean and atmosphere warming outside the Arctic affecting influxes of air and ocean. Also increased CO2 drives increases in IR, so that will be a factor, which is the reason for the first paragraph of this post.

Within the Arctic itself the summer losses are now being amplified by feedbacks; AD, ice-albedo, IR from increased water vapour due to open ocean.

I'm not saying the AW warming hasn't a role. But I don't think it's anywhere near the main driver.

[Jim Williams]

Idunno: "I think that the Arctic is melting mainly due to an increase in the amount of warmer water flowing into the Arctic from the Atlantic side."

I think you are right here, but probably in a much more subtle way than would be easily recognized. It's the increased Atlantic Heat which underpins all other changes. Without it I think the signal would be largely hidden in the noise. The Air certainly brings higher Summer temperatures, but also less heat and Winter cold. The insolation is also seasonal, and actually a bit below climatology. The Atlantic is bringing a fairly steady stream of heat into the Arctic -- both directly and as weather.

I also think that even now there is more mixing and heat transfer than is obvious, and that with more open water this will increase greatly. Much of the fresh water in the CAB is due to Summer thaw, and with reduced ice available the halocline will be closer to the surface and more subject to wind induced destratification.

[wayne]

THe NSIDC sea level pressure anomaly has been imprinted on the Arctic Ocean sea ice over time making it spring and summer minima look. Cyclonic heat incursions in total darkness of the long night have always been temporary but had largely the same paths. In the case of spring and summer 2012 it was more like a constant stream. I deal with this top of my blog. http://eh2r.blogspot.ca/ . The North Atlantic and North Pacific being warmer was a pathway and a cyclo-genesis enhancing area like the Northern inter-tropical convergence zone for hurricanes and typhoons.

[johnm33]

I'm trying to write something further about my take on kinetic energy and what it contributes to the melt, without accounting for any losses my calcs [very unreliable] show that any volume of water going from 60N [southern tip of greenland] to 80N [SZ] can melt 7% of its own volume of ice by kinetic energy alone.

[Chris Biscan]

I am with chris Reynolds on this.

In my daily obs of 1-3 hours per day or more the last 18 months, solar radiation is the main killer by far.

Another major factor has been heat transport from the Rockies and SW United States downsloping winds heading N or NNNE into the Yukon and those regions that bring tons of heat.

In 2012 for instance we saw HP set up form the Beaufort warm pool, then July 27th or so to August 4th warm Southerly flow slammed into the MYI.

However the Beaufort was so warm there was no buffer 20-25C air temps plowed into the CA and Canadian Basin.

I'd put Heat flux from the Atlantic at 10-15% over the years with GHG forced feedbacks typically solar insolation as the main cause. lowered albedo and such

[idunno]

(Actually, the point of my first post was that I think there is much more interannual variation in the Pacific heat input, but this is still interesting).

I think the Atlantic is much more important. Grabbing the nearest map that shows various lines...

http://www.athropolis.com/map2.htm

...by February/March, I expect that the ice edge on the Pacific side will have reached down to somewhere around 60N. On the Atlantic side, around Svarlsbard and Franz Josef Land, it is often closer to 80N.

This after months of darkness, well within the Arctic Circle, so insolation cannot be a factor. It is clearly caused by Atlantic heat. Which frequently causes a retreat of the ice in this area in midwinter to latitudes which would be close to record territory elsewhere in midsummer.

It is clearly a huge amount of thermal energy; and if SSTs have increased in the Fram Strait by the quoted 1.9C, it is an increasingly huge amount of thermal energy. 6 Sverdrups equates to six times all of the fresh water currently entering the seas from all the world's rivers combined. That's climate change. Winds off the Rockies is weather.

A couple of specific points: Chris R "Be careful with the decline in all seasons as the maximum is set outside the Arctic basin, where AW is not an issue." Nah, max area and extent are set outside the AB, but volume is Pan-Arctic. And it is volume, as per PIOMAS, that is steadily declining, over a thirty year span, in all seasons. Area and extent show a much more pronounced decline at their minima.

And Chris B "solar radiation is the main killer by far" contradicts Chris R, I think.

[Kris]

Idunno stated:

And Chris B "solar radiation is the main killer by far"

Do allow me to have you all remembered the absurd high 7 °C temperature at Svalbard early February 2012. And the parallel melting of the ice in the same period in the Kara Sea between Nova Zembla and the continent. With as a result ultra thin ice at the end of April, leading to the actual situation of even now not being an ice field around the islands of Severnaya Zemlya.

We are sure, absolute sure that had nothing to do with solar radiation.

"Solar radiation" is only one instrument of the real and only killer: that is the climate change.

[Kevin O'Neill]

idunno & Chris (R & B)

I think there's a difference in your reference points. If we take the arctic at it's defined boundaries, then Atlantic Ocean warming is one of (if not) the largest reasons we have an arctic nearly free of ice in September.

On the other hand, if we're considering recent reductions in arctic sea ice, then the significant changes have occurred in the Pacific sector and Atlantic waters have had minimal effect.

The Atlantic does not have seasonal sea ice and we're seeing it stretch those salty tentacles deeper into the Barents and Kara Seas each year. And the area that melts out closest to the pole is usually on the Atlantic side. But if we view a map of each year's minima it is the losses in the Pacific sector that really stand out over the past decade.

[LRC]

@ Kevin: Could not part of the 'melt ' on the Pacific side be the effect of the gyre. In past years the gyre would hit the MYI on the Atlantic end and get pushed back toward the Pacific. When the Atlantic MYI got pummeled and started to get flushed out the Fram, the gyre sending the Pacific MYI onto its Atlantic spin got spun out because the Atlantic MYI no longer was there. IT replaced the Atlantic side MYI and the Pacific side lost its return feed of MYI. That then setup the possibility of even more damage being done on the Pacific side. So this all means following this argument (that could have more holes in it the a spaghetti strainer) That although it appears that the Pacific side has had more heat damage then the Atlantic side. The truth is actually the opposite. The Thicker ice on the Atlantic side keeps on being knocked down enough by heat so that the gyre instead of hitting a wall of ice hits something like loose rubble on the side of a road and pushes everything into the Fram.
If you watch the movement of ice over time, in the past few years you are seeing less and less ice coming back around the CAA, a lot seems to being sling shot out of the Arctic.

[Yuha]

Chris R,

The Francis & Hunter paper you cite finds a correlation between atmospheric humidity, downwelling infra-red and ice loss, but they seem to completely ignore latent heat flux between atmosphere and ice/ocean. Latent heat flux is certainly strongly correlated with humidity; it could even change direction. Do you think latent heat flux could have a significant role here?

[Lewis Cleverdon]

Johnm33 - I'd like to help but I've been stumped by a parallel problem on the GIS, which may be far simpler in having fewer components but is beyond me, and that I'm hoping that you or anyone else might perhaps find intriguing.

The issue is to identify the volume of ice melted within the ice cap by the kinetic energy delivered by a supraglacial lake suddenly emptying when its moulin unblocks. Teams from both Woods Hole and Aberystwyth Uni have witnessed such events, with the former reporting a "2kms diameter lake". . . "holding over 11 billion (US) gallons" . . . "emptying in two hours".

My maths reliably gets me as far as an average of 6,000 tonnes/sec off 11Bn US Gls in two hours, but then as far as I can see from their report the lake was at about 700ms over bedrock that was about 200ms above sea-level, on the interior slope of Greenland's 'encircling ridge' [ER]. (The bedrock altitude numbers are as hazy as the poor best topographical map I can find). The height above sea level [asl] of the summer internal water table is unknown, as it the sum volume of myriad air-filled closed-top moulins above that water table, as is the sum area of open-topped moulins that had already drained their lakes.

To disappear like that the water has to have gone into an extant void (equal to a 350m cube) or to have pushed other water out of the way, with 6,000Ts/sec flow being in part the measure of the resistance. The nearest low point in the ridge (100ms asl ?) was about 180kms north at Jacobshavn Isbrae, where some fraction of the island's summer water table finds its escape by pushing the glacier to go faster.

Glaciologists refer to a 'conduit system' for these transport routes, but given the myriad melt-lakes emptying within the encircling ridge [ER], and the far greater distances that a good fraction are from any of the four low points in the ER, and that the lowest of these points (at Petermann Gl) appears to be ~50ms asl, perhaps 'hydro-vascular system' might be a more descriptive title.

Recent research showed a mean diameter of melt lakes at over 1.0km, and a mean depth of 4 to 5ms, and that almost the full potential albedo loss effect occurs in water over 1.5ms deep, so there is substantial solar heat intake, but most will presumably go directly into melting ice on the lake bed, rather than down the moulin.

So the larger issue here may be just what melting the kinetic energy of melt lakes' drainage would achieve, using the example of the Woods Hole 43.2MTs of water under the above conditions. Given that there are tens if not hundreds of thousands of melt lakes within the ER on the S,W,N & NE coasts,
- and that as annual warming raises the altitude at which they form and as the slope they form on declines with height, this is generating both more numerous and larger additional lakes,
- and that a large fraction of them and their drainage canals/gorges are being re-used and thus enlarged each year,
- and that those collection systems will channel rainfall and its surface meltwater down the moulins as long as they remain open,
- and that soon after the start of each melt season the standing volume of water in the 'arteries' below 50ms asl will have been flushed with a fresh intake to transfer its heat to the base of the ice cap,
it appears that just how much melting is achieved by the kinetic energy of the melt lakes' drainage may matter rather a lot.

It seems to me that we're looking not at 2.9Mkm^3 of ice just waiting to be exiled, but at 2.9M gigatonnes of ice at a mean height of ~2,000ms asl, with all of the potential energy that entails. From this perspective global warming and its feedbacks and multipliers are not so much melting the ice-sheet's surface as unlocking its inherently destructive potential energy -

My maths won't get anywhere near a useful answer on this issue - such as they are my skills are in other fields - so I'd really appreciate any help available as it has me rather troubled.

Regards,

Lewis

[Espen]

Healy is on her way back to the Arctic Sea??
http://www.sailwx.info/shiptrack/shipposition.phtml?call=NEPP

[johnm33]

Lewis , if you have a better map of greenland i'd like a look. http://en.wikipedia.org/wiki/File:Topographic_map_of_Greenland_bedrock.jpg
Just from the topography Peterman/ Humbolt look the likeliest breakout points but if the geology to the south, inland of Ilulissat, is limestone or some other porous rock which is at least hinted at by the number of 'lakes' then we might see a catastrophic collapse there at some juncture, similar to the one that seperated Britain from mainland europe. Not knowing the geology of either area I, for now, assume smaller periodic floods, creating sub glacial channels, but share your concerns about the possibilities in the near future. Though i think that the kinetic energy in this case would be expressed in the erosion of a new drainage channel, rather than heat.
Since i don't seem to be able to write anything sufficiently coherent about [my view of] kinetic energy and its role overall i'm going to do it bit by bit.
First the effect of turning right of sea currents in the northern hemisphere is easily explained by their relative rotational motion. A deep cold current emerges from the depths where it was travelling marginally slower than the surface water and heads begins to drift west meets resistance slows and heads north/right. If a current flows north it eventually finds itself with more eastern inertia than the local rotational speed, so starts to head east/turns right,[think north atlantic drift], to remain at equilibrium. Once it's heading east it is clearly moving faster than the local rotational speed so 'wants' to turn south/right. When it heads south it meets resistance from the water already there, heats up/slows down, heads west/right. Heading west it again meets resistance, heats/slows and heads north/right All of these effects are due to the distance from the axis of rotation of the various lattitudes and consequent surface speed. Nothing to do with coriolis so far.
http://www.thevlecks.net/rmj/earth.html

[Steve C]

Lewis,

You raise some fascinating questions about melt lakes and what happens to the water and the warmth once a moulin forms and the lake drains (at mind-boggling speed).

One bit information may inform our thinking about what happens. The mass of ice below the surface is bitterly cold, in the rough ballpark of -20 C. Warmer near the top, and warmer at bedrock level, but mostly very, very cold.

Once a given liter of water disappears down a moulin, it either keeps moving fast, or it goes to a stagnant place (cracks, crevices, crevasses with dead-ends) and freezes.

If that liter of water freezes before reaching the sea, the heat of fusion transferred to depths of the ice sheet dwarfs the kinetic energy it may have (at least, I'm pretty sure on this).

I don't know how much has been discovered about whatever network of persistent drainage channels may persist year after year, but I'd wager that they're essentially all sloping continuously towards the sea. Any dips that could form would form pools after the lake finishes draining that would then freeze solid.

None of the above begins to answer the crucial question of what percent of melt lake water finds its way to the sea, and what percent of water fills cracks, crevices, and transfers all its heat to a warming of the ice sheet itself.

Looking forward to future decades, it's a crucial question. If the depths of the ice sheet warm quickly from -20 to near 0 C, the whole ice sheet becomes far more plastic/fluid, and sea level rise will accelerate massively.

What we need is some research into what actually happens to the water in those melt lakes that form every summer. I can imagine golfball-sized devices put into melt lakes that are designed to have a density of fresh water, that can record when the lake drains and when the device hits sea water. Maybe when hitting sea water it can wake up and transmit the information by radio. A hundred each in a sampling of melt lakes could tell us a lot.

Best wishes,

Steve C.

[Seke Rob]

Once again reminds me of the 10-20 yellow duckies that were dropped on top of GIS in hopes that 1 or more would show up at sea somewhere. Never heard if any did float up.

http://www.outpostmagazine.com/2010/05/21/rubber-duckies-dropped-into-glaciers/

[P-maker]

Steve:

Maybe a technology similar to the one described here may do the trick:

http://www.webstech.dk/

Monitoring climate change will be King both in the grain stacks and in the ice-sheet.

[Espen]

P-maker,

Looks like a solution, although modified, but that should be simple, placing them will be much more expensive though!

[Chris Reynolds]

Idunno,

Yes the Atlantic water (AW) heat flux is crucial to maintaining the ice edge so close to its summer boundary in winter. But that still doesn't answer the question of how that AW then actively interplays with the ice. Which is a crucial question given that research shows it's a deep layer with a gap of around 100m between it and the surface.

But the decline in volume has been from the thickest ice categories. The evidence showing interplay between AW and ice shows it's occurring off the Eurasian coast, not in the central Arctic basin.

Chris Biscan may implicitly disagree with me. If so I disagree with Chris Biscan. I've not read any study that shows insolation is going up, indeed the studies I've read find that the Arctic is getting more cloudy, not less, and that the IR from this has a role in ice loss.

Yuha,

Latent heat will be a factor, maintaining deeper more convective liquid water clouds, which back-radiate more IR than thinner clouds with a high ice content. But the back radiation is what enables interaction between layers above the surface layer of the atmosphere and the surface of the ice.

Kevin,

I think the disagreement is simply about the role of AW in the ice loss, for me the fact that the ice edge hasn't receded so far in the Atlantic sector shows the significance of bathymetry in removing AW from direct interection with the ice.

[Jim Williams]

Chris, if you are saying that AW isn't the dominate factor in any particular year then I think I could agree with that. It looks to me that the changes in the Atlantic flow northwards is more of an underlying urge than a high frequency pulse. The effect of AW is more notable in the ice missing because the wind blew it into the Barentz than in the water's incursions north of the islands.

I'd still call the AW the dominate factor, but not necessarily the proximate cause. When some other force temporarily reduces the ice coverage it is the Atlantic heat which prevents that ice from recovering.

(I also think that AW will become mixed water in the high Arctic shortly after the end of the Summer Ice Cap, and that a year-round ice free Arctic will follow shortly after a seasonally ice free Arctic.)

[Lecollie]

The potential energy of the ice looks like a fairly small effect. A kg of ice dropped 2000 m gains (mgh) = 1 kg X 9.81 X 2000 = 19.6 kJ. The latent heat of ice is 334 kJ/kg , so the fall melts 19.6 / 334 = about 6% of the ice, if it was at 0C to start with and if all the energy goes into melting.

Kinetic energy of the moving seawater is also not that large. At 2000 km from the pole it is rotating with the earth at 145 m/s, so moving to the pole and stopping provides 1/2 * 145*145 = 10.5 kJ/kg, eqivalent to warming the water by 2.5C. But most water doesn't go to the pole, it goes round in a circle and back out at depth.

Hoping I haven't misunderstood the conversation here...

[Chris Reynolds]

Jim,

The problem I still have with AW playing a large role is the physical seperation of around 100m between it and the surface where the ice is. When I say 100m I mean that's the depth of the top of the AW layer. Polyakov gives reason to think there's a role off the Eurasian coast (where the ESS clathrates are BTW), but I've simply not found convincing argument of a large role elsewhere.

In contrast atmospheric processes are demonstrated to have played a substantial role, from Fram Strait export to the AD.

As for what happens after the transition: I'm not sure. Will there be a warm cap of bouyant water from surface warming, or will the deep AW mix upwards?

We have a mechanism for a warm cap - basic bouyancy, the same thing that allows tropical oceans to be warm despite frigid cold in the deeps below. But what about the mixing? Surely a warm cap will act to further stratify the ocean keeping the deep AW down there. But come the autumn there will be massive production of freezing saline water as hitherto unseen amonts of new ice grow, this frigid saline water could be dense enough to drop through and displace/mix the AW upwards.

How will newly open ocean and possible changes to winds interact with the bathymetry?

I suspect we may see a rapid transition to a year round ice free arctic; rapid being after a couple of decades of regional warming. But that's mainly due to the work of Abbot & Tzipperman on cloud radiative feedback.

Which kind of dumps me back in the view of the atmosphere as the primary driver. ;)

[johnm33]

lecollie i'd appreciate it if you could take the math one stage further and calculate how much ice a specific volume of water raised by 2.5deg could melt %wise, my effort shows about 7%, but that was from 60degN to 80degN

[Peter Ellis]

Yes, but that only applies if the water moves northwards and then stops dead. This is not the case, as can be readily observed from the fact that we don't have a tower of water thousands of miles high stretching out into space from the North Pole.

The amount of water moving northwards is exactly balanced by the amount moving southwards, it has to be. There is thus no net energy change.

Honestly, the amount of confusion required to postulate Coriolis effects as a heat source beggars belief. It's not even wrong. Heck, use a rotating coordinate frame and the apparent force just disappears. Where's the heat source then?

Coriolis effects change the apparent direction water currents (in reality they keep going while the Earth moves under them). They don't change temperature.

[Artful Dodger]

Peter,

It is not confusion. It is obfuscation.

Please DNFTT.

Cheers,
Lodger

[Lecollie]

OK, to complete the sums 1 kg water at 1C melts 1/80th kg ice, so 2.5C melts 2.5/80 = 1/32 kg. But as Peter Ellis wrote you don't get net energy transfer this way. The Earth's energy of rotation cannnot be tapped by just moving things around on its surface

[Donald]

The Earth's energy of rotation cannot be tapped by just moving things around on its surface

Another way of looking at this is that if one could get energy this way, where would it come from? Would there be a slowing of the rotation of the Earth?

My two cents on Exponential v Gompertz: ice extent and even volume are not the primary measures; they are incidental to the warming process. What we should be looking at is some measure of the aggregate energy in the oceans and atmosphere. In that realm, there likely is some type of sigmoid function that describes warming over time due to GHG forcing. If for some reason we need to focus on Arctic ice, perhaps because of the feedback effects, then an exponential function seems to be appropriate.

[Kevin McKinney]

Confusion... obfuscation... either way, it's so daft it's funny.

I mean--where would the Coriolis Force increase come from? Talk about 'hide the incline!'

[Lewis Cleverdon]

My thanks for the responses concerning the kinetic energy injected to the GIS by melt lakes' drainage.

Johnm33 - if the bedrock were soft enough for flowing water to erode, surely we'd have seen the results being flushed out at Jakobshavn, Petermann, etc ? From this perspective making arteries wider in ice on bedrock would seem a more likely event, allowing larger surface areas for heat transfer while the water backs up waiting to reach an exit.

Steve C - I'd agree that ice near the bedrock likely isn't very far below freezing given the constant upflow of heat energy from below, but we'd differ over the speed water needs to move at to avoid freezing - we've had -20C in the last few winters here and as long as the last tap is allowed to trickle, the whole system stays liquid.

If we could rig it I think I'd bet a good bottle of wine against your confidence that the moulins would feed channels essentially sloping continuously to the sea. I'd expect them to have made routes down to the lowest point, and then to fill all voids until the level of an exit was reached, after which repeated annual flows would expand the routes leading to it. Given the south to north annual development of melt lakes, and the fact that the lowest point of the depression is far to the north with numerous feeds along the way, arteries running full length if the ER seem predictable, but to what extent they keep going or stop and freeze up each winter is debatable. If the latter, how much heat energy is transferred to the surrounding ice?
And what happens to incoming melt water next season ? Perhaps fresh channels are cut beside and over the old ones each year ?

Given that a mean of 2,000ms of ice exerts a pressure of around 2,800psi, I'd suspect that major change will be occurring around arteries in ice that is nearing melting point and has lost significant structural integrity. The reduction to slush, being crushed ice plus melt water, would of course take far less heat energy than getting through the phase change.

With this concern in view, I'm still looking for the numbers to describe the kinetic energy's input, on top of melt-lakes limited heat energy, to achieve that degree of warming.

Lecollie - I'd well agree with the small actual melt ratio you enumerate, but the issue is a bit finer - it's about what fraction of that energy gets expressed as heat - by compressing air, friction, slamming into dead ends, pressure waves, etc. I'm sorry I should have made clear that getting through the phase change of unfreezing is not the primary issue.

With thanks to all, any further thoughts would be much appreciated,

regards,

Lewis

[wayne]

Thanks in part to Neven displays of SMOS and NSIDC reports, I have possibly found a reason why sea ice models have failed. In order to prove this, one must answer a simple question, what is the surface air temperature of models when sea ice freezes with sst's at -1.8 C?. More on this on my blog.. http://eh2r.blogspot.ca/

[johnm33]

Lewis i was thinking more of cave systems, through hills, being eroded by melt water until they collapse.
Peter i talk about kinetic energy you talk about coriolis, i didn't know the first thing about coriolis so since my last effort i've looked into it, the only significant influence i can see, and i'm not sure it isn't angular momentum, is in driving the denser water out through the bottom of the fram strait.[and thus being indirectly responsible for the sucking in of atlantic water] I've tried to find some work on the subject, such as the difference in density directly below the axis and the outflow water but can't find anything about that outflow at all this side of a paywall.
To circle the pole the atlantic water,which starts south of 60N has to go to 80N,at every point where it is driven north it wants to turn east and south, in all those places it does, there exist temperature anomolies, why kinetic energy, why now? i've no idea.
Donald exactly http://www.swa.ca/Publications/Documents/3%20Whole%20Basin.pdf
http://www.unistuttgart.de/gi/research/schriftenreihe/quo_vadis/pdf/schmitzhuebsch.pdf
http://www-das.uwyo.edu/~geerts/cwx/notes/chap11/day_length.html

[johnm33]

Anyone whos interested in this kinetic energy thing for mackenzie bay should take a look at this, http://web.mit.edu/raffaele/www/Publications_files/annurev.fluid.40.111406.102139.pdf
go to section 5, look for EKE. eddy kinetic energy, it fits the bill quite well and goes a long way to explaining why the temps graded higher towards the shore.

[Steve C]

John33

Warming of the Mackenzie river outlet area waters is simply NOT from kinetic energy of waters moving via Coriolis forces. AS was already pointed out, you cannot tap the rotational energy of the earth by moving around mass on its surface.

Angular momentum of water that moves towards the pole is simply transferred to the angular momentum of the water moving away from the pole. Those movements are ultimately driven by solar heating and wind, not the rotational energy of the earth.

In principle, those waters could be warm because of warmer pacific waters moving there -- but sea surface temp data don't support that interpretation well, from what I can glean.

[johnm33]

SteveC the whole thrust of my argument was that it was kinetic energy, nothing to do with coriolis,which is, so far as i can tell a narrow view which only exists if you are locked into a mindset that assumes the water percieves the apparent motion in the same way we can, what i clearly state is that the water always moves to a point of eqilibrium. That said the water is induced in to flowing into the arctic, way beyond its comfort zone, by other forces, and retains an excess of kinetic energy which is expressed by it trying to head south, as well as east. Thus you get hot-spots wherever that happens, also because the water rapidly approaches the axis of rotation[ie at 90deg to what it was at the equator] another type of kinetic energy also comes into play, which is angular momentum, i dont believe this 'belongs' to any particular body of water, [just as electricity doesn't belong to any length of wire] but that it gets expressed as turbulence [+heat] whether the current moves north or south. If it heads into open water it dissipates and is lost among the background, however if it heads to shore it becomes more 'frantic' as it approaches the shallows. If you want more scientific terms read the paper.
I did originally think it was the kinetic energy of pacific waters that drove this, but have now formed the view that atlantic waters reached around to the beaufort early in the season, lifting the pacific waters and amplifying the effect.
Coriolis may be a real effect at a quantum level, and no doubt we can see it reflected in the 'real' world, but what i'm talking about is kinetic energy,inertia, momentum and angular momentum.
Heres an interesting example of what the rotational energy of the earth does http://www.esri.com/news/arcuser/0610/nospin.html
Please, read the mit paper then argue with them.

[johnm33]

OK I've wasted an evening looking for any inference on my part that the energy expressed around mackenzie bay had anything to do with the coriolis force/effect, found nothing, couldn't find anything on " Angular momentum of water that moves towards the pole is simply transferred to the angular momentum of the water moving away from the pole." either, do you know of a paper?
This, http://web.mit.edu/raffaele/www/Publications.html
is a list of publications, the most relevent of which i referred to above, on the subject of kinetic energy, by http://web.mit.edu/raffaele/www/Home.html of http://mit.edu/

[Mike Constable]

johnm33
Having sort-of ploughed through http://web.mit.edu/raffaele/www/Publications_files/annurev.fluid.40.111406.102139.pdf from above
(way too technical for me). I noticed that kinetic energy of eddies is about 10^-3 of the potential energy (do they mean KE?) of strong ocean currents! (Bottom of p.16)
Eddies are more likely to bring deeper, warmer water to the surface than provide much heat from mixing (I do not know how much heat could be liberated by fresher water mixing with salt water)?
My feeling is that in the arctic imports of warm air and water, together with energy from the sun will dwarf any KE inputs coming from the water.
One other mixing factor which I have heard is significant in oceans is the movement of the deep scattering layer (plankton/fish/etc.)(ref. unknown!) on its diurnal cycle, but this is not likely to become important in the arctic because of the ice and continuous daylight in summer!

[johnm33]

MikeC I was only looking for an explanation for the anomoly around mackenzie bay before the solstice, i was previously totally unaware of the potential of kinetic energy, and hadn't heard of eddy kinetic energy, but once you look for EKE with google there's loads of papers some much more relevent, i didn't save them, thinking i'd laboured the point here long enough.