« The last ice expedition | Main | PIOMAS December 2012 »


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


Everything I read here seems to get more and more troubling. I don't see any indications that this trend won't continue...albeit with possible stops and starts.
I do have a few questions which I have seen alluded to in this blog previously: How much of the water from the melted ice is still trapped under the GIS? Will this water eventually be released in large avalanches and floods or will it gradually escape as the ice melts. Due to the history of the Missoula Floods I lean towards large avalanches and floods. What is the thinking of the ice experts. If you could direct me to information about this topic I would much appreciate it. Thanks.

Artful Dodger

Hi VaughnA,

The GRACE satellite measurements for Greenland involve MASS loss, not melt volume. While there may indeed still be melt water working its way down slope, the 574 Gt net value reported for 2012 has, like Elvis, left the building.

Follow the PDF link provided by Neven above for more details.

Bouke Van der Spoel

The caption for the first graph doesn't correspond to the graph: 2012 is red according to the graph and dark blue according to the caption.


Hi Neven & Lodger

Thank you for bringing this up. However, your version of Fig. 8 from the Tedesco et al. paper does not address one of the key aspects of the problem:

The Accelerating Arctic Ice Loss!

Both GRACE data and PIOMAS data display accelerating ice volume loss this century. If you could add these two (and even at a later stage add the accelerating snow loss, and possibly also the NH permafrost and local glacier volume loss), then such a diagram would help raise awareness about the challenges we are currently facing in the Arctic.


Neven, Lodger eo, good morning.

I was a little surprised by the initial GRACE data adding up to app. 574 Gt mass loss 2012.
Above, Lodger suggested the difference between mass and volume loss, mentioning the possibility that an amount of melt water hasn’t yet left the GISS.
I hinted at the same in the ‘Unprecedented…’ thread.

While I haven’t yet finished a CAD survey of differences in the icesheet limit between ’09 and ’12, I had a look for some other indications.

Here’s what I found for the ‘equilibrium line altitude’ at Swiss Camp. You may remember the little graph by Konrad Steffen et al showing 4.85 m lowering up to 2011.

CIRES, ‘From ice to water – watching Greenland melt’ by Jane Palmer
(pub. app. April 2012?)


“Over the years we have lost first 30 centimeters of ice, then 50 centimeters of ice, then a meter,” Steffen said. “Overall we have lost 6 meters—so the losses are very visual.”
Last sentence:
In spring 2012, when Steffen steps out of his helicopter and looks at the camp, the evidence will be staring him in the face.
End of quote.

Steffen hints at 1.15 m lowering during 2011? So what could that have boded for 2012? Not much good, I guess.

A lot of meltwater originating on the ablation zone must have left the icesheet (corresponding with the GRACE results). Remember the ‘pulse’ on Watson River.
That zone was much larger during 2012, affecting ice layers that had never experienced melt before.
I’m not referring to the rare two/three days of 100% melt even up to Summit, because that release would have trickled down only into deeper firn layers.

I have made a raw assumption on CAD, showing very large area over 1100 m high, experiencing sustained melt for weeks.
My guess is that that release is captured in crevasses, voids and partly still on it’s way to the sheet base. It is hard to prove, but I think that could more than double the number given by GRACE.

What does it do? It weakens the ice, raises its temp and prepares accumulating melt through the coming seasons.
BTW, I found this link to a CIRES journal, very interesting:




in support of your line of evidence, I will once more refer you to this link:


These daily videos have for the past couple of months shown a "river of icy water" exiting the Ice Fjord at Ilulissat. Thus, it is also my impression, that the runoff from the ice sheet has continued long after the melting stopped at the surface. Apparently the GRACE value for September this year does not tell the full story of this year's volume loss.

Al Rodger

Surely the graph above does exactly what you ask - presenting the "key" acceleration of ice loss, at least for Greenland.
The data of the Tedesco et al Figure 8 (also presented in fig 5.19 of the NOAA Arctic Report Card and linked here for easier access - http://www.arctic.noaa.gov/reportcard/images-terrcryo/g-fig5.19.jpg ) when converted from Mass Anomaly to Rate of Change in Mass Anomaly (as the graph in the post above does) provides a measure of the Acceleration of the Mass Anomaly through its slope. Thus the regression line labelled 30 Gt.yr^-2 is a measure of the acceleration in Greenlan'ds ice loss over the decade 2003-2012.

There is still a question of whether this acceleration is constant or not. A regression on the data 2003-2010 yields an acceleration of 18.5 Gt.Yr^-2. This isn't a conclusive measure of an increasing acceleration. But I'd suggest it does indicate that an increase in acceleration of some value is occurring. So hold on to your hats.


Can someone less lazy than I do the math? If the accelerated mass loss is 30Gt/yr^2; when has half the Greenland Ice "left the building" (i.e. raised the sea level)?




thanks, but I was not asking for a linear trend line based on relative changes. I was actually asking for real, comparable volume numbers (cubic km's) stacked on top of each other, year by year since the turn of the century. If you need a conversion factor for snow, you may try with an initial average snow depth of 1 m and a density of 0.3, just to get started. Considering the permafrost and local glacier ice loss, I was thinking about a similar rough conversion of the annual runoff from major rivers to the Arctic Ocean.


This is not a final paper, just a review draft submitted. I have reviewed it and those comments are published and I did recommend publication. You are correct Werther that some of the meltwater is still retained in the ice sheet and that the numbers will get worse. However, this meltwater does not weaken the ice, this is a part of the normal system in Greenland, just more of it. This meltwater can warm the ice locally which makes it flow a bit easier, though not significantly. In southern Greenland this summer Qaleraliq Glacier retreat indicates that even with glaciers that are not calving much, melting is taking its toll.


Or someone can check my math:

(wiki says ice sheet has "2,850,000 cubic kilometres" of ice)
1 m^3 = approx 1 tonnne
1 km^3 = 1 approx 1 Gt (ignore ice less dense than water)

So the whole thing is 2,850,000 Gt? I.e. a cubic kilometer is roughly a Gt of ice (take 10%?)
How long to lose half?

Integrate -570-30y = -1425000 and solve for y

570y+ 15y^2= -1425000
= (-570 +/- sqrt(570^2+4*15*1425000))/(2*15)= approx 400 years.

My expectation is that the acceleration is accelerating as suggested by Al. If an exponential fit is accurate. This is much more alarming. For instance:

1. if the rate of ice loss is doubling every 5 years, using the numbers above, half the ice sheet will be gone in 56 years and - of course -
all of it is gone 5 years later; or

2. if the rate of ice loss is doubling every 3 years, all of it is gone by 2050.

That would really suck. Ice levels would rise by 3.6 meters by 2068 (2045) and then by another 3.6 meters between 2068 (2046) and 2073 (2049).

The horrors of exponential growth. Let's hope there is another feedback loop yet to kick in and save us.


Speaking of crowdsourcing: Project Darksnow is seeking to raise $150,000 online to determine whether Arctic fires played a role in the record Greenland melt this summer.

Needless to say, the frequency of these fires is not unrelated to the loss of ice sheet, loss of June land snow cover, and overall Arctic temperature amplification.

For this, Jason Box enlisted the help of Thomas Painter, a JPL expert on the effect of soot and dust on premature snow melt. Box has mounted some 23 previous Greenland expeditions; Painter has published high profile papers on the effect of dust storms on the Colorado snowpack (aka water for Phoenix, Tucson, Los Angeles and the principle US winter vegetable crop in Yuma and Imperial Valley).

(These dust storms are really something to see -- I got caught up in one east of Flagstaff which closed the freeway for many hours. Not to mention the Phoenix haboob which is getting to be an annual event. These are due to over-grazing and abandoned ag fields, as are our spring dust storms from Mongolia.)

Also coming along are Bill McKibben and Peter Sinclair of DeSmogBlog and various Ohio people capable of running a public awareness campaign.

Two-thirds of the money goes to the Dash-6 Twin Otter ski plane rental which will make 3 landings between Iqualuit and Kangerlussuaq to dig snow pits in one very arduous day.

Pretty good cause I would say.


Dust and Soot


Hansen suggested doubling every 10 years this century was possible and that was viewed as potentially alarmist. So doubling every 5 or 3 years possibly sounds a little too extreme. Nevertheless, it looks like it has doubled a couple of times in last 9 years (100Gt/y to 400Gt/y) so perhaps 5 years looks supported by the data.

Before pushing such a calculation we need to check it sounds sensible. Where is all the energy to melt half of it in 5 years going to come from? There is only so much solar radiation and you have to lose quite a bit of that back out to space. I think you would have to be cooling the oceans large distances away from Greenland to provide that sort of energy. That requires icebergs to drift long distances but once the glaciers have retreated well inland most icebergs are going to stay within Greenland basin. Is the volume of warm water that would have to flow in via glacier outlets into Greenland basin plausible?

Therefore, I think the growth could well reach some limit to growth before half the ice can disappear in 5 years. I suggest we need to calculate a reasonable maximum rate else we end up doing silly calculations. Any suggestions for likely minimum solar radiation that goes back out to space and other necessary numbers?


Peter Sinclair noted above is not with Desmogblog, he has his own ClimateCrocks.com blog and makes videos for the Yale Climate Forum.


Although half the greenlandic ice disappering in 5 years sounds like way too much, one shall, as NLPatents points out, never underestimate the power of exponential growth. Not only will the reflectivity feedback that will create extensive melting, but also remember that for each hundred meters you go down temperatures usually increases something like 0,8 degrees kelvin. Assuming that this works the same way on Greenland, you would increase temperatures with about 5-10 degrees kelvin "only" by melting away half a mile of ice. And, as you probably already are aware of, if you combine this effect with albedo increase and the temperature increase created by global warming itself you get an inferno of melting. To melt away the entire greenlandic icecap within a century doesn't sound to unfamiliar to me, although I rather think it is more realistic to melt away about half of it in a century. I haven't done the math for either of theese scenarios, but think they are far from impossible.

Ron Mignery

PIOMAS data for the last 10 years show a decline in the summer minimum of 7500 km3 or 750 km3 per year. http://img.photobucket.com/albums/v224/Chiloe/12_Climate/sea_ice_PIOMAS_min.png
At that rate, zero will be reached by 2018. Extrapolating the annual curve http://psc.apl.washington.edu/wordpress/wp-content/uploads/schweiger/ice_volume/BPIOMASIceVolumeAnomalyCurrentV2_CY.png down by 750 km3 per year sees zero ice arriving at the end of July around 2021 and at the summer solstice around 2030. From the same graph, the heat of August appears to melt about one km3 of sea ice and the heat of a full summer about five km3. When the sea ice is gone, that heat must then find something else to warm. Would not a large fraction then likely end up melting Greenland increasing the melt flow manifold?

Ron Mignery

Oops, I meant 100 km3 and 5000 km3.

Ron Mignery

Oops again, 1000 and 5000.

Al Rodger

Me and my abacus make it roughly 289 years, 9 months & 21 days. In other words, not many centuries.
Whether or not the increased melt is something like a smooth 30Gt per year or not, half of Greenland is a contribution of some 3.5m to SLR. This is far from outlandish.

Al Rodger


A couple of points.

But before that, I am not Lodger. (Perhaps in China or Japan it may sound like I am!) On this site especially, Lodger is someone else entirely.

Firstly, is the volume/weight convertion that much of an issue? Average Sea Ice density is something like 0.91 sg. I would have thought Greenland snow quickly tends to ice with an sg of >0.917 as it cools below freezing (0.94 at -200°C so not by much). But then given the accuracy of the data to hand, an sg of 1.0 is probably accurate enough (or 1Gt=1cukm.)

Secondly, I am still not entirely sure of your requirements. The one value that perhaps could be your aim may be something along the lines of the Arctic's heat content or annual rise thereof.
The talk of ice that has melted but re-frozen on its way "out of the building" so not appearing on the GRACE data could be put in context by considering heat energy. The energy to melt or re-freeze ice is 334 kJ/kg. The GIS gets down to temperatures of no more than -35°C. Even warming of this coldest ice all the way up to freezing/thawing temperature requires less than 73 kJ/kg. So it would take a lot of even the coldest ice to re-freeze significant volumes of melt water.
Thus, assuming there are no vast under-ice lakes of melted water accumulating, I would assume that re-frozen melt water probably isn't a major factor in any energy balance calculations.

Beyond Sea Ice & Greenland, while there is a steep decline in NH snow cover in early summer, the winter snow cover has not declined so (assuming average depth is unchanged) a year-on-year increase in NH energy through snow melt doesn't mirror the year-on-year loss of Arctic Sea Ice volume which is down throughout the year.
Melting permafrost: well that is another matter.

Aaron Lewis

Last summer, as late as mid-November, the SW coast of Greenland was subject to above freezing East Winds. The fact that they continued so late in the fall proves that they were not driven by local solar heating. Greenland is using energy from the south to make its own weather!

There is still unusually warm water in the shallows of some of the fjords. At this point, it is not likely solar heating, but the result of the conversion of potential energy to thermal energy as water drains out of the GIS.

With a substantial loss of Arctic Sea Ice, we can expect the more and more of the latent heat in the Northern Hemisphere to condense out on the GIS.

The old rules of thumb on the transport of water vapor in the atmosphere were based on different atmospheric circulation patterns driven by different temperature gradients. Now, we are more likely to see atmospheric rivers of moisture running from the Caribbean to Greenland. Or, atmospheric rivers flowing from the tropical Pacific to the Bearing Sea.

Latent heat imported from the south is likely to be the dominate factor in the melt of the GIS. Even continuous sunlight with zero albedo cannot deliver as much heat as a warm, moist wind. And, the wind can blow after the sun goes down. The sea ice was important because it condensed latent energy that would otherwise go to melting the GIS, and then provided a large surface that radiated heat off through a dry atmosphere. And sea ice was important because it reflected energy that will otherwise go to melting the GIS.

The sea ice was important because it maintained a dry window that facilitated radiation of surface heat, which facilitated the sea ice. Now that window is covered with water vapor, CO2, and 1,900 ppb of CH4, and the Arctic atmosphere holds more heat near the surface. The other side of this is that more latent heat is imported from the south.

Thus, any geo-engineering to restore Arctic sea ice would have to overwhelm changes in Arctic albedo, changes in the Arctic atmosphere that trap heat, and the import of heat into the Arctic from the south.

Our climate models are not adequate to provide a basis of design for such engineering. We also have to consider side effects such as effects on ecosystems and crop yields. At this time, we do not have the tools to support these decisions. To say that it can or cannot be done, shows a lack of thought.

We do not need to worry about half or a quarter of the GIS melting. Now, we need to worry about that first foot of rapid sea level rise, and the storm tides it will produce. Then, after lunch we can worry about the next 2 meters of sea level rise (half from GIS and half from WAIS?)

Chris Reynolds

Neven, Thanks for the post. And thanks to Lanevn.

Geopotential height (GPH) is the height at which you find a certain pressure level. So for 500mb pressure the typical height over Greenland was about 5000m or 5km.

NCEP/NCAR shows that by this year that had risen to nearly 5580m, an increase of about 10% - that's huge! PLOT OF GPH

This cannot continue indefinitely.

I think what we're seeing is a gear change, a step increase to a higher rate of loss. The coincidence between the June/July/August new circulation pattern and the same period of acceleration of SMB loss in the first graphic of Neven's post is too much to bear as a mere coincidence with something like smoke being a driver of the process (sorry A-Team). Although smoke becoming entrained in the blocking ridge may be an additional factor, I think that the driver is increased insolation due to open skies under the high pressure due to the GPH ridge.

As for whether this continues. The plot of GPH above suggests that underlying the 2007 jump, and associated Arctic Dipole and bullseye of Greenland surrounded by low pressure regions, is a general trend towards higher GPH, this is in part due to general high latitude increase in GPH with warming. However adjusting using both Arctic GPH N of 70degN, and GPH for same latitude band as Greenland, shows that the post 2007 increase is not an artefact of wider warming, as does the emergence of the new circulation pattern since 2007.

If the new circulation pattern continues then increased Greenland mass loss will continue. Overland et al think rapid retreat of snowline is the probable cause of this pattern. I have sympathy with this but cannot shake the suspicion that sea ice retreat is having a supportive feedback role on this pattern. I expect the pattern to hold in future years, but I can't argue that forcefully.

You're right to carry the doubling argument to its ridiculous conclusion, losing half in 5 years is inherently implausible. All I think we can say is that we have an increase, to bound the maximum increase I think we have to fall back on papers like Tad Pfeffer's.


The earth's geothermal gradient under central Greenland is quite a complex topic (eg tectonic rifting on both sides) -- and has not been experimentally measured in bedrock at a sufficient number of sufficiently deep boreholes as far as I can tell.

I did locate some actual ice borehole temperature profiles from an older article. Note the ice is not all that cold to begin with.

These temperatures are not to be confused with the more commonly discussed paleo surface depositional temperatures, inferred from the ∂O18 proxy (which makes assumptions, some only recently corrected, on the source of the snow's water vapor).

Here is what life is like at the new Eemian drill camp at 77°N (where they brought in a rock drill bit for the first time):



Greenland glacial history, borehole constraints, and Eemian extent
J Geophys Res 108 2143 doi:10.1029/2001JB001731, 2003
L. Tarasov and W. Richard Peltier

free full text: http://www.agu.org/pubs/sample_articles/cr/2001jb001731/2001jb001731.pdf

Chris Reynolds

A Team,

Thanks for that. It gives a qualitative sense of the sort of warming needed to melt all that ice. What's going on now is the pre-warming of those depths of ice that will count towards loss in centuries to come. Every Moulin is a heat source going down through the sheet.

Aaron Lewis

It is Dec. 10, 2012, and today it was 47F with NE winds of 30 mph in Julianhaab, GL.

So, it is December, and there is still a wicked lot of heat blowing around Greenland.

Data from 2003 does not tell us much about what is going on there now. A pool of super-glacial melt water can punch a moulin and change the temperature profile of the ice, from top to bottom in a matter of hours.

Chris Reynolds

Aaron Lewis,

Agreed a Moulin will warm the ice, but with the bulk being that cold the Moulin will only warm around itself. The winters are still cold, you've got adiabatic cooling with height, so I doubt if the plot snipped by A Team will be massively different todays typical profile for the bulk of the sheet.

I'm pondering whether it's possible to quantify inward atmospheric heat flux into the Greenland area using NCEP/NCAR...



I like your optimism. My little calculation does sound absurd.

And worse, as soon as you assign some non-zero probability (even one chance in a 10,000) that I am correct, you get the potential for insurable claims on the order of net present value of 100 Trillion in the next 50 to 70 years - almost all of the happening in a very tight timeframe.

It would become impossible to insure any building that did not take 7 metre sea level rise into account.

Will bet me 10,000 to 1 that I am wrong. I'll need to bet $100, because my house is within 10 metres of sea level, but all access is below.




sorry about the name confusion. I think we agree that the conversion from volume to mass is not a big issue considering the other uncertainties.

The reason I was asking for positive volume estimates of lost ice, was exactly to get every measure on the same footing - not to translate into energy.

The idea of stacking the numbers (from glaciers, sea ice, snow cover and permafrost) was exactly to get a handle on the actual acceleration of total ice loss in the Arctic.

Once we have a robust acceleration estimate, we can start making projections, such as:

1) September snow cover could be lost in a year or two

2) Multi-year sea ice could be lost within a handful of years or so

3) Tidewater glaciers could be lost within a decade or so

4) Mountain glaciers could be lost within the next couple of decades or so

and so forth...

The thinking behind this is, that excess energy in the system will be used to melt some other type of ice, once the "low-hanging fruits" are gone.


A hopefully informative map on the second July 2012 situation:


Anyone interested can get a larger copy on my photobucket dump.

The icesheet limit is rather detailed, to be compared to earlier and following years.


On Qaleraliq Glacier, South Greenland...
In one of these details I checked my ice/snow boundary to Mauri Pelto's latest post.
Mauri, I got the Nunatakker you describe as 'D' in a narrow, 10 km deep incursion. No snow/ice. Maybe it is a deep, moraine filled trench and could still count as 'glacier'. But my impression is that the retreat there is even more pronounced than your post suggests.
Thanks for your work, it is invaluable and helps me adjust what I'm doing.


It is Dec. 10, 2012, and today it was 47F with NE winds of 30 mph in Julianhaab, GL.

I had to look it up -- http://www.weather.com/weather/today/GLXX0004

Will be more or less the same all week.


Thanks. Artful Dodger and others. I an thinking it only takes a small amount of water to float a large amount of ice especially in the right terrain. I have done the experiment of freezing a tapered jar mostly full of water then melting just enough to float the ice. It takes less than a tablespoon of water to float a pint of ice...not saying that a valley full of ice could be floated quite that efficiently however.

James Lovejoy

For maximum GIS loss, I did a back of envelope calculation that suggested that 1 watt/m2 over the entire earth,used solely to melt land ice would increase sea level by about 1 meter in 8 years.

So if Greenland is about 7.2 meters, at 1 watt/m2 forcing, the maximum melt would be at 2% a year, and that's if the entire watt/m2 were all concentrated on Greenland.

So there are limits, just not very comforting ones.


Werther (or anyone really),

Do you have a good reference, almanac, or graphic with all the key names of glaciers, fjords, inlets and the like for Greenland? Ellesmere?



Hi Sam,

I'm too greedy to get topographical maps from geus.dk, which presumably are the best.
What I might do is fill in on my above map any relevant locale and thus expand what is rapidly becoming my memory training game.



you could start in the NE corner using this new feature:


Cheers P

Artful Dodger

VaughnA, you're welcome. GRACE data just shows how much mass has been lost from Greenland. That mass loss is the sum of melt/runoff and glaciers calving ice directly into the ocean.

So whether from melt or calving, Greenland is shedding it's icy coat. And the effect is still accelerating.



Sam, here's a good map: Greenland Ice Velocity Map.

Werther, your CAD image is awesome. Could you send me a big resolution version? I'm sure I can fit it into a blog post, perhaps this one.


Hi Neven,

I'll send it toninght, after checking the names I've given with your map. I mix up FI 79 and Zachariae.

Climate Changes

" Now, we are more likely to see atmospheric rivers of moisture running from the Caribbean to Greenland. Or, atmospheric rivers flowing from the tropical Pacific to the Bearing Sea."

I have noticed this already happening. I check regularly the global flow and for the last 4-5 winters water vapour from the Equator has shifted from a fairly regular North East direction to truer North sending wave after wave of warm tropical moisture.
I believe this is down to the Jet Stream getting weaker and it is more obvious over the Autumn/Winter months (till re-freeze is achieved I suppose). A daily check here:
can help follow the flow and spot those pesky JT loops that are becomming more common bringing Cold snaps, rain, etc.

Just Testing

Since I started following this blog i have been feeling an urge to stock up on rice and pasta. This latest thread makes me wonder how I'm going to move that stuff to a new location without letting it get wet.


Chris, a couple of clarifications.

First, I should have made clear that Project DarkSnow appears primarily targeted at explaining the essentially unprecedented all-Greenland melting event this summer, not at an overall revised climatology. Greenland has had lots of sunny days over the observational-record years yet this hasn't happened before.

From the timing and tracking, they suspect a particular fire in nearby Labrador. It is easy to get runaway albedo tipping effects with snow -- wind blunting crystals, soot, salt, dust, warm weather from the south, and now rain (see those camp logs). Rain-on-snow events can cause spectacular flooding, depending on the topography. Lots of experience with that along the North American cordillera.

Second, since Greenland ice core melt particulates over the last 125,000 years have been studied since Day One, why make another trip? The answer may be that the conventional ice core record doesn't really get its act together for the first 80 years. Thus for the NEEM core, they started at the bottom of a ten meter trench. DarkSnow has not disclosed wheres and whys for the three sampling sites (perhaps they are somewhat opportunistic) nor how any soot layer in recent snow will be dated and attributed to source.

From my perspective, actual data on soot is infinitely preferable to further years of climate modelers blowing smoke, ignoring positive feedbacks solely because they're inconvenient. What we've gotten from that is models that grieviously undershoot the time scale of climate change. And so delay (or eliminate) a timely policy response.

Third, the DarkSnow scientists already have government grants (NASA, NSF, CWC) so why do they need to pass the hat? A new grant involves an extended time frame; the public engagement components are important but not fundable. It's fair to say a lot of prior climate reseach has just been the proverbial tree falling unheard in the forest.

And I think they have a strong policy case here: while wildland fires are unstoppable and getting worse, cooking stove soot, slash fires, arid land agricultural dust etc have rapid and affordable remedies with outsized effects -- soot, anthropogenic methane, urban ozone, HFCs account for ~40 percent of current warming according to the 6 Dec 12 NY Times piece, "Going beyond Carbon Dioxide".

Finally -- without getting into the Amazon going up in smoke -- boreal and peatland fires go way beyond their tipping effects on Arctic sea ice and Greenland ice sheet: permafrost albedo collapse, direct, messy and rapid atmospheric injection of a vast carbon store and so forth. With all these mutually reinforcing feedbacks going on, I'm not convinced that any of them is that usefully singled out as driver.


Al (and others)

I did some number crunching on the NH snow cover extent. Please check, correct and use as appropriate.

The numbers are here:

Year, mio km2,km3 (x/1000*0.3), Loss (km3)

2000, 0.68, 204, 096
2001, 0.70, 210, 090
2002, 0.99, 297, 003
2003, 0.71, 213, 087
2004, 0.55, 165, 135
2005, 0.65, 195, 105
2006, 0.56, 168, 132
2007, 0.42, 126, 174
2008, 0.45, 135, 165
2009, 0.32, 096, 204
2010, 0.36, 108, 192
2011, 0.24, 072, 238
2012, 0.34, 102, 198

Source: http://climate.rutgers.edu/snowcover/table_area.php?ui_set=1&ui_sort=0

I selected columns Eurasia + N. America (-Greenland) for August (minimum) extent each year since 2000.

I did a simple rescaling (divided by 1000 and multiplied by 3) to get remaining ice volume in cubic km. I then subtracted this number from 300 to get a time series of positive ice loss values. Essentially, when this number reaches 300, all remaining snow fields in the NH have “left the building”.



Just discovered a mistake. 2011 should be 228 km3 instead of 238.



For those interested in aerosol, black carbon, soot or albedo impacts on Greenland and elsewhere in the Arctic, Giovanni has just released an interactive report view using Google Earth.



Thanks everyone. Those are some awesome links. Wow.


Does anyone know where the borders for Cryosphere today and MASIE regions are defined? Someone from google wants to know for some analysis of PIOMAS data.


I have found Masie boundaries, just CT I am looking for.


Ha Crandles,
Here's a chance to get some use out of my work, made this two years ago;


It is CT scaled to known sea areas/boundaries.


Thanks Werther.

You have digitised that? Does that allow you to find the lat and long of each corner? Or did you find the lat and long of the corners specified somewhere? Would an email to William Chapman at CT be a better route? Sounds like you have checked the areas so you are pretty confident you have the borders in the right places?


Crandles, those are a lot of questions...

It depends on what one wants to do (quality wise).
I digitised it using the CT map, G-Earth and MODIS in CAD. It is not 'high cartography'.
No, I didn't fit lat/long on the map. I have the PC and 80dN on one of the layers, but the projection fans out further to the south, distorting FI Hudson Bay because I held on to the known area.

I am very interested how Mr. Chapman fitted the map on CT. So mailing him would be a good idea.

It makes me wonder... is PIOMAS info also available in a graphic format? I'd be very interested to get that in CAD too.

dominik lenné

IMHO the fear of exponential growth is exaggerated, because although there are positive feedbacks, none of them will grow infinitely. E.G. albedo will never decrease below a certain value, and of course solar irradiation won't. Atmospheric heat transfer is limited by both circulation and heat transfer coefficients at the surface, and so on.

As the mechanisms are complex and contradictive no simple function is distinct in its predictive capability and earns more trust than a linear or at most a quadratic. Only functions based on a - however simple - physical modelling are of any value.

That said, I put my penny in with the value of sea level rise above now in the year 2050, supposed the quadratic holds for that time (which is of course questionable):

delta_t = year - 2012
melt rate i Gt/y ~ 450 delta_t + 30 delta_t²,
integrated over 40 years = Delta_V
= 450*40+30*40²=66000 Gt = 66 e3 km³,
earths water area = A = 361 e6 km²,
sea level rise = Delta_V/A
= 66/361 m = 0,18 m.

Espen Olsen


December 10 2012 at record low for the date at: 10.935.313 km2


Ron M,

I don't know how you do your numbers, but here is a better estimate based on the link you provided. Maybe someone could verify it and use it.

Year, Daily Min. (1000 km3), Loss (km3)

2000, 11.0, 1200
2001, 12.2, 0000
2002, 10.8, 1400
2003, 10.2, 2000
2004, 09.9, 2300
2005, 09.2, 3000
2006, 09.0, 3200
2007, 06.5, 5700
2008, 07.1, 5100
2009, 06.9, 5300
2010, 04.4, 7800
2011, 04.0, 8200
2012, 03.3, 8900

Source: http://img.photobucket.com/albums/v224/Chiloe/12_Climate/sea_ice_PIOMAS_min.png

Again, once this number hits 12200 cubic kilometers, the Arctic Sea Ice "has left the Building".

Chris Reynolds


Seems we've been contacted by the same person via Neven. I have some masks directly derived from CT's regions image - copied CT region image saved as bitmap and coded to convert regions into region numbers stored as a bitmap - ie. comverted to a greyscale bitmap with new greyscale colours as bytes - ocean, land and regions have different numbers. Compared to the resolution of PIOMAS they region edges will be blocky, but it could be upscaled and smoothed. Bill Chapman may be their best bet for a more accurate region mask.

NSIDC Masks are here.
But they're of much lower resolution than the mask I have, and each are much lower than PIOMAS gridded data. Also CT regions are not the same as NSIDC.

Chris Reynolds

A Team,

Thanks for the clarification.

Over at Tamino's I posted a link about pollution smoke in British Columbia from Russian wildfires.

I've just recalled a paper from INDOEX about mid tropospheric warming from black carbon. I think it could be this, if not there's stacks of similar pages in Google Scholar.
That paper describes warming in the Himalayas due to BC, the same thing may be happening over Greenland. But the paper discussed in Neven's lead post of this thread uses a model (MAR) to calculate cumulative SMB losses - whether that model is capturing BC warming is doubtful from my reading of the paper. See section 2.3 of the paper.


Thanks Chris,

I have replied suggesting contact with Bill Chapman.

CT different to NSIDC. Yes, I got the impression CT grid was 80 by 58 and provided link to
though I don't see a region mask being available there.

448 by 304 would be a better starting point than 80 by 58 resolution. So I will send that link, unless you have already done so.

Chris Reynolds


Send it anyway, I was kind of butting in on your conversation (Simon?). I've sent him a copy of the mask I intend to use, when I kick myself up the backside and get busy again. However Google will want to get it right. Had a chat about how to get PIOMAS gridded data, but I think he'll have got in touch directly with Dr Zhang.

P Maker,

Do you have Excel? - I've just done a spreadsheet on Rutger's snow data, your linking to it reminded me to get round to it. Very interesting. My email chris886222 at btinternet.com for anyone with Excel who wants a copy.



I do have Excel - and thank you for your kind offer - but I prefer to walk this "Trajectory of the Commons" together with you in the open.

Cheers P

Chris Reynolds

P Maker,

I wasn't suggesting a private discussion, try tabulating by month/year, and calculating monthly anomalies. There are some intriguing drops in 2007 and 2010. I'm still wrestling with the role of the new circulation pattern in sea ice loss since 2007, and vice versa. Overland et al suspect that snowline retreat is driving the pattern, the pattern is then driving sea ice loss. However I suspect that sea ice has a role in the new circulation, perhaps via impact on snow (Lawrence et al and the warming in models during RILEs). In that sense it is intriguing that PIOMAS shows massive loss of volume early in 2010, Shakhova says that in 2010 they had problems setting up their summer Siberian sea camp due to the extreme thinness of the ice - and Rutgers snow lab shows a drop in anomalies in May/June from 2010 onwards, with 2012 - having a very strong example of the new circulation - having extreme low anomalies of snow cover over Eurasia.

The more I look at the data the more I see tantalising suggestions of feedbacks upon feedbacks. I suspect things are substantially accelerating since the volume loss of 2010.

Aaron Lewis

The GIS is a structure, with ice at lower levels supporting the ice at higher elevations. The ice at the perimeter serves as a dam or buttress to support the thicker ice in the center of the structure. When the low altitude ice dams/ buttress fail, the high altitude ice falls to a lower altitude and the argument of adiabatic cooling with height is void. Ice sheets undergo structural failure from the bottom up. And, the bottom of the GIS is 1.6 km below sea level. Ice at the bottom of the structure is under a lot of pressure and has a significantly higher melting point than ice at the top of the sheet.

If there is a weak spot in the dam or buttress, ice behind that weak spot will force through the weak spot - rather like water leaking through a dam or dike. There was good reason for the Dutch Boy to stick his finger in the dike to stop the leak. Leaks always get bigger - fast.

We also know that in the bottom of the llulissat Fjord is old, warm ice that flows more easily than the newer, colder ice above it. The llulissat Fjord area is one of the places where moulins form. It is a place where moulins are warming the newer, stronger, colder ice that in the recent past prevented the flow of the old weak ice. That old ice is under pressure, and ultimately moulins will perforate its cap, and that old weak ice will squirt out.

Look back at the images of the GIS in 2003, how many moulins do you see? Not many! Now, they are common at altitudes of over 1,500 m and 100 km inland. That dramatically changes the volume of affected ice. After water stops falling down a moulin, the ice flows to fill the void. As ice flows, internal work is done on the ice, and energy/heat from that work remains in the ice until it can be conducted out, and thermal conduction in ice is very slow. The result is a volume of ice from the top to the bottom of the ice sheet that is warmer and weaker. The more moulins form, the more of this weak ice we have. And, it is cumulative.

The GIS is an old cathedral of ice. When a buttresses on an old cathedral fails, the cathedral falls down. When the buttress on the GIS fails, the GIS will fall down into the sea as water/ice slurry with large horizontal velocities driven by the large potential energy of the ice. We know from the debris fields in Disko Bugt that there have been large (sudden) flows out of the llulissat watershed in the past.

The tail of the last flow left a dike of debris in the fjord. Considering the potential energy of the extant ice above it, the dike of debris in the llulissat fjord is ephemeral, and can easily be transported by water/ice slurry into Disko Bugt. (I know, you want to assert that the debris in the fjord is moraine material marking the maximum advance of the glacier. However, that does not explain the deposits in Disko Bugt.)

GIS ice does not have to melt before it can raise sea level, it only has to break up into pieces that can be carried by a high speed flow of water - a flow of water that is fast enough to rapidly erode solid granite. You may not believe this can happen because you have never seen it, but I trust physics. Physics has never lied to me.

I am not trying to be an alarmist. I am trying to state the applied physics of ice. I am tired of having the physics of ice ignored. Geologists treat ice like a glass, and the atmospheric guys treat it like a reflective surface. It is both, and it is neither. Ice is ice.

The physics of ice tell us that it is worth a huge effort and great treasure to stop putting CO2 into the atmosphere - NOW! The IPCC has understated the GIS issues by 2 or 3 orders of magnitude. Thus, all of our economic analysis of the issue is wrong.



"a flow of water that is fast enough to rapidly erode solid granite"

I feel that you are right, but do you have any modern examples.

The landscapes of Canada & Greenland were carved by ice and certainly the badlands features in the US west were formed by catastrophic failures of ice dams. I'm sure any features not part of the granite bedrock will be flushed away & that grounding lines could be altered that way.

To lose large amounts of ice I think fjords need to be expanded & while deepening seems possible depending on the bottom material, the widening and straitening of the granite walls might take a much longer time.

Are there any modern examples that can be pointed to where solid granite structures have been altered by sudden outflows?



Do concrete bridges washed away count? I guess that is more a matter of undermining foundations.


Crandles wrote:

that is more a matter of undermining foundations

Do allow me to correct a bit.

The concrete foundations of bridges are usually if not always pillars at/to which the rest of the bridge has been soldered, bounded, jointed, bended ecc... ecc... .

Meaning "the rest" is the real weak point, especially the road, which can't be attached to thight for the good reason there must be room for the temperature bounded expansion and shrinking of the materials.

Bottom line, when there are huge floodings, the concrete is te last to go and the road is the very first to go.
Well, we have seen this phenomena this Summer in Greenland, haven't we?


Aaron stop with the ice sheet is swiss cheese that will fall apart. Having worked on the Jakobshavn Glacier where there were plenty of moulins even in the 1980's and on other glaciers where there are streams and moulins every few meters, the physics of ice just does not support your hypothesis. We have observations and models of current and former glaciers that do have to reasonably approximate glacier behavior and they do, and there just is not a swiss cheese example.

Aaron Lewis


The best modern example is the Lake Missoula floods that formed the Potholes area of Washington State and deposited material where the Hanford Repository now resides (US DOE-RL)

Conventional wisdom (http://www.glaciallakemissoula.org/ ) says it was a canyon of water with an ice dam. However, an ice dam cannot dam a head of more than ~6 meters of liquid water. (It may take a while to come to equilibrium, so the head may be larger for a very brief period.) It is the same physics as moulins. Thus, the Lake Missoula floods were caused by the progressive structural collapse of a canyon of ice, with shallow super glacial ponds causing the wave benches that are so visible in the canyon.

If ice dams do not hold water, then the lake Agassiz models must also be reviewed.

One group of folks who do know about ice dams are the USACE because they maintain rivers. Look at the data. River ice dams rupture at ~40 feet. Look at modern glacier ice dams. They rupture at a head of less than 30 meters. This is a long, long way from the large heads that geolgists assign to Lakes Missoula and Agissiz.

Rapidly moving water tends to swirl in vortexes that produce pressures low enough to form water-vapor filled bubbles in the water. When those bubbles strike a surface, they collapse, and the back wall of the bubble hits the surface at hyper-sonic speed. Such a vortex can eat through granite like a hot knife goes through warm butter.

Donald Michael Graham

Greenland warming of 1920–1930 and 1995–2005

Petr Chylek

Los Alamos National Laboratory, Space and Remote Sensing Sciences, Los Alamos, New Mexico, USA

M. K. Dubey

Los Alamos National Laboratory, Earth and Environmental Sciences, Los Alamos, New Mexico, USA

G. Lesins

Department of Physics and Atmospheric Science, Dalhousie University, Halifax, Nova Scotia, Canada

We provide an analysis of Greenland temperature records to compare the current (1995–2005) warming period with the previous (1920–1930) Greenland warming. We find that the current Greenland warming is not unprecedented in recent Greenland history. Temperature increases in the two warming periods are of a similar magnitude, however, the rate of warming in 1920–1930 was about 50% higher than that in 1995–2005


The scale of the Bretz/ Missoula floods is stunning. These were episodic cataclysmic floods. They varied in size with depth at Missoula Montana ranging from several hundred feet to two thousand feet deep at rupture. The lake contained ~ 250-1,000 cubic miles of water and drained in something like two days.

The flow down the Columbia River gorge scoured the canyon basalt flowing 500-1,000 feet deep a mile across at 70 miles per hour. Four massive ephemeral lakes formed down the length of the flow lasting for a few days, not least of which is today the 75 mile long 50 mile wide Willamette valley, home to most Oregonians..

And this repeated every 55-60 years for 1,500 years. The tribes remember the floods well and tell a lot of stories from those times nearly 10,000 years ago. In one, they tell of the floods coming in the fall after the fall salmon runs when the rains began.

It is impressive to drive down the very scenic gorge realizing that the canyon overflowed in most parts and ran half full in others. That is even more impressive in the canyon looking up at the rims 1,000-2,500 feet above.

Yes water carves rock.


PIOMAS update:
Latest value: 2012-12-1 9.76

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

Monthly data
Daily Anomalies
Daily data
Daily data with a "prediction" based on exponential trend
I have added the "predictions" for 2013 this time. As for other years, the 2013 values are computed from
an 1-year extrapolation of the values of previous years on that same date, fitted to an exponential declining time-line.
Note that the uncertainty of the predictions made this way is quite large: at minimum the volume
could be anything from zero to 4,000 km³ even assuming the exponential decline continues.


Aaron, Sam, I think you are a little optimistic about water wearing down granite and basalt quickly. Beacon Rock bore the brunt of the deluges in the gorge which washed away the pumice and tephra yet had minimal affect on the basalt core which stands about 600' high along the Columbia River near Bonneville Dam today. Even though a fair percentage of the water was rocks mostly 1' in diameter and smaller the basalt was barely affected. I live on the outflow deposits of these floods and rocks 1' and smaller form at least 100 layers a few inches to 10' thick in gravel bars up to a depth of 200'on top of an underlying clay layer and bedrock.
I agree that if large floods from ice dam ruptures happened here over 100 times, then it will likely happen numerous times in Greenland as well. Soft material and loose gravel will be scoured out but I do not see the granite being affected much in time periods of a few days. I expect to see the flood like happened in Greenland last year repeated in numerous places in the future with the larger floods being several magnitudes larger as the breadth of the melting increases.


On Greenland swiss cheese, melt physics and SLR…

I did my first simple calcs somewhere in 2006. Imagining the warming trend strong enough to lead to a general retreat of the ice sheet margin up to 4 km within a stretch of 10 years, a SLR of 15 cm could result.
When would that occur? As soon as the yearly mass loss would reach 4000 Gt.

I read above a calc by our friend dlen, mentioning 66000 Gt over 40 years. The number is not that far off my rough cumulative mass loss for 10 years.

I mean, now that I digitalised the sheet boundary on app. 57000 km1, a general retreat of 1.3 km1 isn’t alarmist to suppose, is it? It is already noticeable on pixel-distance when the MODIS images are well studied. At least, on the SW side of the sheet.
The Steffen Group confirms this through decreasing height, Mauri Pelto does recently on Qaleraliq Glacier.

The 570 Gt mass loss reported through GRACE is still a lot separated from the alarming 4000 Gt mark. Again, is it?

I supposed the amount of melt water to be double as large, most of it still underways or collected in firn/voids. Mauri Pelto reacted that to be a ‘normal’ process. He’s right, but I would point at the much larger ablation zone now compared to earlier years.


Most of that melt water from this ‘new’, 1100-1300m high ablation zone terrain shown in the pic above is underways. Creating new ‘plumbing’.
I see what Aaron so vividly describes. But I don’t imagine that leading to ‘sudden catastrophical collapse’. I see it capable of getting to the 4000 Gt benchmark soon, which is in itself worrying enough. And physically feasible.

So what can be the SLR pattern? Chris Reynolds describes the coupled ocean/atmospheric changes, which are profound and rapid. They are fully capable to impact the southern part of Greenland, which is already hovering in a marginal glacial niche. A substantial part will be lost by 2050 (hey…there’s the 40 year stretch by dlen!).

Does it go in an exponential fashion? I think not, being subject to geomorphologic bounds.
I think the 4000 Gt benchmark can be reached soon, the main provider being the SW side (Nuuk-Upernavik). Say two, three years. Then, the 10 year period 15 cm pulse follows. Wearing down 1.3 km mean over the whole 57000 km1 long ice limit. Most of it, again, on the SW side, the easy part (there’s enough warming ‘in the pipeline’).

Then, the process continues, taking longer in the high southern and eastern mountain parts. By 2050, most of the central/northern sheet will still be there (1,1 million km2). The south dome too, confined to its high mountain ridge (400.000 km2).
SLR may reach app. 50 cm then, hopefully stabilising a bit while on its way like FI Church and Rahmstorf have predicted.


I feel urged to express what I feel about this all.

I think it sucks…

Never having seen the majestic beauty of Greenland, I am (was) happy to live in a world where that beauty is present. I know its there, even through the stories of the natives that come through and by the great works of art of those artists having had the privilige to work there.

I pledge the scientists to be artictic too. Please, get your stories out, not just raw data.

One of the occasions I had to see the beauty of the World of snow and ice was in 1985, near Grossvenediger in Austria. I never revisited it and I won’t. Googling the place a few days ago, I saw that FI Mullnitz Kees, one of the glacier tongues in a beautiful setting that I have clear memory of, is almost gone, retreated 1.25 km in these years.

It breaks my heart to imagine The Alps ruined without their icy crowns and rubbled by weathering.


Wipneus, thank you for the 2013 estimate. A few more years like this and we are done with the MY sea ice.

Moving back to the GRACE data, I took the liberty of reading September (minimum values) off the graph provided by Arctic Report Card:

Year, Loss (Gt), Change (Gt), Loss (km3)

2000, ?, ?, ?
2001, ?, ?, ?
2002, -0130, 130, 143
2003, -0250, 120, 132
2004, -0410, 160, 176
2005, -0700, 290, 319
2006, -0820, 120, 132
2007, -1130, 310, 341
2008, -1350, 220, 242
2009, -1560, 210, 231
2010, -1980, 420, 462
2011, -2400, 420, 462
2012, -2870, 470, 517

Source: http://www.arctic.noaa.gov/reportcard/images-terrcryo/g-fig5.19.jpg

If someone could help getting reasonably robust numbers for 2000 and 2001, as well as check and correct gross errors, it would be highly appreciated.

Looking ahead, I was pondering how to get my foot on some NH runoff/discharge data from the many great rivers running into the Arctic Ocean (Ob, Yenisej, Lena, MacKenzie, Yukon etc.). I was considering using the observed August and September base flow values for each year. My assumption is that any runoff before 1 Aug will be dominated by snow melt, whereas runoff after 1 Oct will be dominated by lakes/permafrost melt, since surface melting on mountain glaciers in these basins must have come to an end at this time of the year.

It would also be nice to have a rough estimate of the remaining glacier ice volume in these large drainage basins, so we can eventually estimate, when the remaining glaciers will have “left the building”.

Steve Bloom

Sorry, Donald, that paper has been superseded by more recent work finding that the current rate of ice loss is indeed higher than the '30s. The details are readily available on the web. Better luck next time!


30 Nov
2011 day 335 10.859
2012 day 335 9.760

down 1.1 km^3

31 Oct
2011 day 305 6.999
2012 day 305 6.188

down 0.81 km^3

So despite less sea ice, the gap to last year has grown 0.3 Km^3. That isn't good.

Maybe that 0.3 Km^3 is just weather variability and there is still quite some time to maximum during which gap of 1.1km^3 might be reduced.

But if maximum keeps dropping at anything like 1km^3 a year then it is hard to see the sea ice surviving summer for long even if other things like ice over deep ocean taking longer to melt due to less heat from below try to delay the ice free state.


Indeed, another way to to look at it is the 'official' PIOMAS average thickness plot:


Now that extent and area are not so much below average anymore, the pack is the thinnest for the date. A bit over 1 meter now, 4 month to grow the needed additional 70 cm or so.


Aaron, Sam

All rocks are not equal. I've seen what flash flooding can do to sandstone & now live quite close to Niagara Falls where limestone has been eaten away by fast water and ice. I also live in a river valley where extreme ice dam flooding, that until recently was a yearly phenomena, hasn't changed the limestone valley contours over at least the last 5 millennia.

I suspect that much of the alteration to granite is caused during freeze/thaw cycles when water entrapped in cracks and fissures expands during freezing and flakes off a section of rock.

If collapsing bubbles were enough to tear through granite, wouldn't these have had a much stronger effect on Niagara's soft limestone causing a much faster upstream collapse than what has been recorded?

I'd expect sills to be eroded allowing warm water to enter fjords and eat away at the exposed bottom of marine terminating glaciers & I think it's possible that gravel deposits presently underlying portions of the ice sheet will have channels cut through.

Greenland had less ice than present during the HTM when the Flade Isblink melted away. 79 was far back from it's present position and carbon dated drift wood and clay deposits at Blase indicate the last opening to the ocean to have been ~ 4k BP.

This year Flade Isblink seemed close to bifurcating, but there's still a lot of ice there to melt. I don't believe that there is any evidence of large changes in the topography of Greenland occurring during the HTM so I doubt that we'll see any looking forward, certainly not until we've melted down to those levels.

Greenland is melting rapidly & accelerating. I think most of the present predictions are far too conservative, but I think that at some point the underlying topography will slow the damage.

As an aside, one or two meters of additional tidal surge could cause so much damage to infrastructure that world wide communication might be permanently disrupted. We may never be aware of when the last ice has floated free in Greenland.




If a picture is worth a thousand words, your graphs are worth a million.

November's freeze hasn't done much for volume & I can't imagine anything but a rapid melt this spring. Your spreadsheet shows last month as having the greatest drop from 2011 for the year. Not good looking forward.


Al Rodger


Re snow cover thickness.

I had a memory of a map of snow thicknesses for a couple of months in a paper of some age. I don't remember seeing actual observation data of snow thickness elsewhere so it has stuck in my mind. And hey presto, it is figure 2 in The effect of Eurasian snow cover on regional and global variations. Barnett et al 1989.

This may be useful to you in scoping snow cover water content (but I'm yet still unclear as to your objectives).

Stu Ostro

Thanks for the post and the link to the paper.

One question ... In the first graph within the blog and Figure 5(b) in the paper, the caption says "2012 (dark blue), 2011 (light red)," whereas on the graph itself the legend has 2012 in red and 2011 in blue. Is it a correct assumption that that's correct, and the caption is erroneously reversed?


It breaks my heart to imagine The Alps ruined without their icy crowns and rubbled by weathering.

It's not just rubbled by weathering. In the past two decades there has been a dramatic increase in massive landslides in the Alps, changing the entire faces of mountains in an instant as permafrost melts.


Hi Al

Thanks for the reference. I also remember reading about the snow extent - Indian monsoon link in those days. It could have been the Hahn & Shukla (1976) paper. However, you refer me to a NH spring (fresh snow) map, whereas I was thinking about lingering snow fields in August.

Concerning my objective, it is still very clear: Produce one single diagram together with you (and with the help of others), which tells the whole story about "Accelerating Arctic Ice Loss".

Chris Reynolds

Stu Ostro,

I've taken it to be as the graph is labelled, in view of the record melt in 2012, implying 2012 is the record low for cumulative SMB.

On a different matter - I was wondering - have you read Overland 2012, "The Recent Shift in Early Summer Arctic Atmospheric Circulation." You'll find the new pattern as seen in NCEP/NCAR SLP here, 2012 was a textbook example:
Have you written about this work anywhere? In view of your work on anomalous trough/ridge patterns and changes in the Arctic I'd be interested in your take on this issue.

Chris Reynolds

P Maker,

"Produce one single diagram together with you (and with the help of others), which tells the whole story about "Accelerating Arctic Ice Loss"."

I've pondered the same thing, but the complexity keeps leading me to adding a time dimension - i.e. a video.

Aaron Lewis


Does your physics predict events like: (http://www.guardian.co.uk/environment/video/2012/dec/12/chasing-ice-iceberg-greenland-video?intcmp=122) This is exactly the process I describe, only in miniature. It is gravity driven, progressive structural collapse.

I am simply saying that as the GIS warms, over the next 50 years, such events will move inland, and get larger, and larger.

I would say that we are 5 years into this process, and we can expect it to go faster and faster as more heat is transferred to smaller areas of NH ice.

Aaron Lewis

How water wears rock depends on how fast the water is moving. I have waded the chest deep Columbia at Richland, WA many times, but been knocked off my feet by swift flowing, ankle deep water in the Clark.

We used faster moving water to cut the tops off of oil wells after we blew the oil fires out. Those well heads are the best tool steel that money can buy, but a high speed water/ grit slurry cut right through them.

Much of the energy of water coming over Niagara Falls is dissipated by the pool of water at the bottom of the falls. Thus, there is less energy to erode the soft rocks of the region.

The object lesson is where a stream seems stable, and then a flood event comes along and reshapes the stream. That is the result of deeper water moving more swiftly. The bad lands of Washington state are what you get with multiple, very large, flood pulses.

Now, this will upset some people, but I assert that the very deep fjords of Greenland were the result of erosion by high speed hydraulic flows, rather than being gouged by solid ice. The snowfields at the top of the GIS were not channeled into narrow glaciers as occurs in mountain glaciers. Therefore a glacier on the GIS that was big enough to dig a trench 1.3 km below sea level, would have been wider than the width of the fjords. A glacier digging a trench 1.3 km below sea level is not going to push rock up hill. (Moraines are formed by moving rock downhill.) On the other hand, a high speed hydraulic flow will pick up material and throw it sideways as the Potholes in Washington State demonstrate.

No, the GIS filled trenches that had been excavated by earlier hydraulic flows. The source of the hydraulic flows was progressive structural collapse of previous accumulations of ice.


Chris Reynolds

You'll find the new pattern as seen in NCEP/NCAR SLP here, 2012 was a textbook example:

I'm speaking from ignorance but it seems that ice covered Greenland is supplanting the pole with regards to arctic high pressure.


The polynia near Cape Morris Jesup has opened again closer to Flade Isblink - but still to the north of it. Temperatures at Nord have risen to -9.6C as a result.

Aaron - I'm not ignoring you & will respond when I get back or tomorrow.



Twemoran wrote,

Temperatures at Nord have risen to -9.6C as a result.

Wouldn't it rather be the other way around ???

[Temperatures have risen to -9,6 °C and as a result the polynia has opened again]


Since there have been many large ice dam collapses(a hundred meters in height or more), it seems prudent to consider the possibility that ice dam structural collapses like these will happen in Greenland even if we don't think so or quite understand how they happen. One of the things I have learned is that if it has happened numerous times in the past it will happen under similar circumstances in the future whether or not anyone understands the process. It certainly looks to me like it won't be too long until parts of the GIS becomes similar enough to past conditions to allow for large ice dams to form.
Also, I have read about ice structure in deep ice on this blog in the past or links from this blog. I searched for this/these reference(s) but was unable to find anything. If I remember correctly, ice under more than about 800 meters of top ice acts quite like a pliable solid and any plumbing that opens up by water pressure quickly closes as soon as the water drains. Dams would logically form at a critical interface from shallow ice where the water continually drains away to where the water floats the deeper ice because of the dam. I know this does not explain the upper portion of the deep ice where the water could conceivably drain away nor exactly how the process works. Nevertheless, large ice dams have formed and collapsed hundreds of times.

james cobban

Yesterday I went to see Chasing Ice, the film chronicling James Balog's Extreme Ice Survey project (trailer at:


), and since then I've been trying to come up with some way of visualizing the yearly addition of GIS, WAIS and glacial meltwater to the world's oceans. What would it look like if all that meltwater was averaged out to a steady flow from a single source? Would it look like a river? How big a river? How many fire hoses would it take? How many garden hoses? What I've come up with so far seems unbelievable to me, so I would appreciate it if someone could check my math and point out where I may have gone wrong, because if my crude calculations are correct then it would take 8.6 Niagara Falls flowing for an entire year to equal 500 cubic kilometers. Can that much ice really be melting now?

My first assumption is that about 500 cubic kilometers of water is being added to the world's oceans every year. I know that this figure is higher than the average since 2000, but it seems reasonable to me given this year's 574gT loss from Greenland alone (I'm assuming a gigaton of ice is roughly equal to one cubic km.of water). Am I right to think that the real figure might be much higher, say at least 600 km^3?

So, 500 cubic km = 500,000,000,000 cubic meters, divided by 365 days = 1,369,863,013
divided by 24 hours = 57,077,625
divided by 60 minutes = 951,293
divided by 60 seconds = 15,854 cubic meters of water per second.

Wikipedia says the average flow rate of Niagara Falls is 1,834 cubic meters per second, so
15,854/1,834 = 8.65 Niagara Falls! Can that be right?

If this is correct, then 15,854 m^3/s could also be visualized as:

water spewing out of a one-meter-square culvert flowing at a rate of 15.854 km/s or 57,074 km/h (!)(the jet of water would exceed escape velocity and continue into space), or:

1,014,331 2-inch fire hoses pumping out 938 liters/sec. or:

14,783,387 common garden hoses pumping out 64.34 liters/sec. or:

water flowing out of one of the 7.6 meter diameter channel tunnels at a rate of 349 m/s or 1,258 km/h (nearly ten times faster than the trains!), but:

only 0.075 Amazon rivers, at 209,000 m^3/s

Or have I made some obvious mistake?

Jim Hunt


As a somewhat ageing surfer I have an unfortunate habit of always seeking out offshore winds.

When it comes to attributing cause and effect, where do you suppose 50 mph southerlies fit into the scheme of things?

To my way of thinking wind => open water => temperature rise, but I'm open to suggestions. As Chris put it recently, I've never been closer to the Arctic than my laptop!



I'm envious of your dexterity, every time I tried to stand up the board hit me in the chin.

Jim & Kris

The strong Southerlies must be a factor, and my understanding is that as the open water in a polynia interacts with the air it transfers a large amount of oceanic heat to the atmosphere. The more normal Northerly winds abet the formation of the NEW polynia south of Flade Isblink & I'm assuming this is almost a mirror image of that, on the other side of the obstruction.

I don't think warm air temps are needed for a polynia to form, but they often result from the exposed roiling waters.


Otto Lehikoinen

Yeah, sounds about right james cobban. It's about an additional Mississippi flowing out of the glaciers, like I saw stated somewhere. Those atmospheric rates of moisture transport in tropics to Amazon basin are quite astonishing. According to http://en.wikipedia.org/wiki/List_of_rivers_by_discharge the glacier outflow would stand in the top ten currently, I guess.


Come on guys!

Has the "Elf-ice" already "left the building"?

If you are waisting your time on the headlines in the right-hand column of this blog, you might as well consider early retirement...

Chris Reynolds


I remain to be persuaded on that view - it's essentially what Wayne Davidson and others seem to think.

I suspect causal roles for Eurasian snow and Arctic sea ice. Two points against what you say - when you look at absolute (not difference from mean) geopotential height the centre of action is still clearly the Arctic ocean, i.e. it's not shifted. And There wasn't a similar pattern centred on the Arctic that's then shifted to Greenland after 2007.

John Hursan

Hello.This article was really fascinating, especially because I was browsing for thoughts on this issue last week..... i really enjoy your blog


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