« Similar melts from 1938-43? | Main | More details on PIOMAS volume loss »


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


"They have their models which have not performed well..."

Does anyone know if there is any serious consideration that what the shorter-term ASI models are attempting to achieve is even possible?

This is an honest question. (And this question could belong in this PIOMAS thread as well as anywhere else, I believe, but correct me if I am wrong).

For example, if you asked a hurricane researcher to predict, in any way at all, whether a particular depression just a few days out into the Atlantic from Africa would be a land-falling hurricane anywhere in the USA, he would likely say, "nope, strictly impossible to make such a prediction in any way that has real value."

He simply knows, as well as this area has been studied, there are far too many variables to even start to pin down such a prediction.

Honestly, is there any solid basis to think we have a grasp of the necessary variables to give meaningful predictions about the Arctic at this time? Say anything shorter than 50 years?

If we are so wrong even over shorter periods of 10 years, has anyone dared propose that this system is beyond our reach? It is no big deal when seen from the POV of the hurricane researcher, and just assumed that we may need hundreds of years more work and computer power to ever make the longer range forecast such as the example above, and even then it may still be impossible.

We want to know, badly, what we can about where things will be, and soon.

Is this desire simply blinding us to the fact that what we are trying to achieve is totally impossible in any way that has real value?

If we stepped back, took a deep think, and added up all the factors that are going into this particular model, on the particular geographic and time scale we want it to work on, it could be very obvious we just can't get there from here. We, perhaps, just don't want to admit that yet.

Lewis Cleverdon

Given the host of iterative interactions that ice loss is driving and is driven by, my request earlier for more posts on the consequences of ice loss - for the northward migration of rainfall, the Jet Stream, the mega-feedbacks etc, might have been misunderstood as looking simply to raise the sites' educative capacity on those consequence for society at large.

In reality those dynamics are also of course significant contributors to ice loss/prevention - in the rising volume and temperature of continental rivers' runoff, in a more convoluted Jet Stream bringing intense heat waves to the far north, in the local late summer CO2 & CH4 concentrations, in ESAS methane plumes obstructing ice-crystal formation, and in the ocean currents' transport of increased heat into the arctic basin.

One aspect I should really like to see discussed further is the point at which sufficient sea temperature and/or loss of salinity due to Greenland meltwater prevents freezing and so robs the Gulf Stream of its downwellings - and just what is the outcome for ice-cover of a cut in the latter's heat transport to the basin ?

Providing better information on the critical significance of the loss of arctic ice cover is certainly crucial for arousing the public and politicians to demand an end to the reckless bipartisan US policy of a 'Brinkmanship of Inaction' with China, but I wouldn't want to see the site's unique value as a public forum for scientific discussion of arctic ice cover diminished in any way.

Hoping this adds a bit more clarity than complexity to my earlier request,


Jim Williams

BlackDragon, I don't think we have any choice but to keep our eyes on the ice. The world is going to do the computation for us well before we can.



Yes, agreed, we should watch it intensely with every tool at our disposal - and more as fast as we can deploy them. This gives us heads up into any fast-breaking developments that are seriously important.

It is also very valuable to try and model things as best we can, but only as long as we are being honest that we have enough understanding to create something worthwhile, or are close to achieving this. Resources of all kinds are going into such endeavors - are they well applied?

It is one thing to say "oh, we missed this sudden change, but probably because we left out factor Y, and didn't make factor X large enough."

It is quite something else to be deluded into thinking there is only X and Y and maybe also Z to get right, when in fact there are many hundreds of things that we cannot grasp in any meaningful way.

The IPCC 2007 included an ensemble of many models to make a long-term prediction about where ASI was going, with what scientists thought were good error bars. This ensemble, all our best efforts combined, was grossly wrong! Not even close.

That is lots of time, money and effort. It could realistically tell us to not even try again.

That doesn't have any impact on the value of keeping an eye on things - all the great data this site presents so well.

Al Rodger

Predicting the future progress of a hurricane is impossible because the process is chaotic.
Predicting the future progress of Arctic Sea Ice levels in a warming climate is very simple. With enough warming, the Arctic will become first ice-free in summer and then, with enough further warming, ice-free all year. It is long established that AGW has the ability to achieve both. What is not yet known is how quickly. What is rather bizarre is that ice-free summers will probably arrive before the models are built that would have accurately predicted it.

The usual gauge of our limited knowledge in this matter is the timing of an ice-free summer which now looks only a very few years off when only a decade ago the best guess was still post-2050.
I am presently awaiting the snow cover data for August to see if an end-of-century 'worst prediction' will be 'worsted' this year!! It seem the Snowman is lighter on his feet than the Iceman.

PS I consider expanding the scope of this website would be a recipe for its destruction. Its USP is being polar. Keep it that way.


I'm wondering how the work of Dr David Barber ties into the huge drops in the PIOMASS readings.


If, in fact, as he believes, the older rotten ice is masquerading as multi year ice, then it is very likely that we actually have much less volume than the instruments would have us believe.

In this case when the thin ice melts out what we wind up with is lots of slush with increasingly separated floes. Something which allows much more radiation to be absorbed and allows the sea to heat more than expected.

I have observed a lot of this on the Arctic mosaic overheads.

Personally I don't think it is as simple as viewing the numbers and making a prediction. Arctic ice melt is subject to far too many factors to be a simple calculation and we are still discovering side effects (algae on the underside of ice).

Given solar, weather and methane impacts, plus unknowns with the thermal balance in winter/summer, late onset of seasonal ice, opening ice in winter giving back heat and moisture to the atmosphere, warmer winters (20C or more for protracted periods), I don't believe that any really solid, year to year, prediction can be made under these circumstances.

Science is about observation, analysis and repetition along with all the calculations, maths and physics. Simply put, we've never observed this before and in a system with so many parameters, it's highly likely that we will see new and unknown impacts and variations which impact the predictions.

The only really reliable prediction is that we are going to lose all the ice in summer.



Probably this has been posted before. Here is a mar '11 interview that actually brings up and answers many of the questions/discussions that have come up recently. http://www.sciencepoles.org/articles/article_detail/david_barber_arctic_sea_ice_in_a_changing_climate/

Seke Rob
In this case when the thin ice melts out what we wind up with is lots of slush with increasingly separated floes.
Well, Dr. Walt Meier already said "Parts of the Arctic have become like a giant Slushee this time of year. ... That's a goliath year — the greatest melt since satellite recording began in 1979". 5 days ago in the NYT.
Artful Dodger

Indeed, LRC. This does sound eerily prescient (even from 21 Mar 2011)...

"There have been predictions that there will be more open areas of ocean in the Arctic and that this will lead to more low pressure systems and storms in the Arctic. Could you explain why?"

Well worth the read. Recommended!


Chris Reynolds

Black Dragon, Jim Williams,

[Bangs head against desk and mumbles 'why oh why oh why...']

Why not try reading some of the science?

Just a suggestion. Thought that being informed might help you to figure through the issues.

Perhaps you might find a stack of science into models deficiences and why they have them. Perhaps you might see that the assimilating models (PIOMAS/NPS) are giving much better agreement with observations. You might then wonder if there is something in the atmospheric situation that's adding to ice loss, an element the models are missing.

You could have read into a bit about chaos and prediction of physical systems and the importance of initial conditions.

Having now read a fair amount you might be able to see why comparing tropical storm behaviour to sea ice is fatuous - hint; sea ice is an integrator of impacts - a memory. Storms don't have such a memory.

But it's a lot easier to wave your hands about and (I've read your comments Jim) arbitrarily dismiss the modellers. Those modellers are people who, intellectually speaking, could wipe the floor with you and I.

I know how intelligent these scientists are and how stupid it is to dismiss them. That's because I read the science as a hobby.

Did someone mention a bet?

You claim that the Arctic will be sea ice free in winter after one more crash?

Think about it. I'm prepared to offer you out on a short term bet, like Crandles. I've not got much cash at present, so no sums held in escrot I'm afraid. But I'll bet you tens of thousands of GBP we won't see a sea ice free winter this decade. Hell, why not millions - win the lottery recently have you?

I know I won't lose, so not having the cash is no problem.


Well, looking at the last page of comments, this thread is beginning to look more and more like one you might find at HuffPo. I am beginning to feel as if I'm wasting my time. I did not feel this way several months ago as I tried to absorb, as much as possible, the detailed analysis I found.

Signing off for now. Will check in and lurk later.

Chris Reynolds


"I'm wondering how the work of Dr David Barber ties into the huge drops in the PIOMASS readings."

It's intrinsic.

Wind the clock back to the 1980s and earlier and the ice pack was a vast mass of thick MY ice. With the transpolar drift and Beaufort Gyre acting as a flywheel system, ageing the ice and moving it throughout the pack, not just a remnant off the Canadian Arctic Archipelago (CAA).

Now the ice pack is a totally different system. The remnants of thick old ice are mainly off the CAA but aren't a cohesive mass any more, after 2010 there is very lttle of it.

Chris Reynolds



L. Hamilton

With so many threads it's hard to know where to post this, but in response to several comments above and elsewhere I added a note to my "Naive predictions" thread, comparing recent 1-day values with the Gompertz predictions and confidence intervals.


L. Hamilton

"The photos are breathtaking - any idea which glaciers?"

Terry, I did not have a map so I couldn't say which glaciers those are. The photos are taken looking N along an E-W stretch of coastline in SE Greenland, more or less on direct line from Reykjavik to Nuuk.

I've sometimes looked down on this area from 37,000 feet, but last week's trip was on a turboprop much lower, affording a closer view.

Lake on the ice sheet:

Lonely iceberg:

Big picture:


Older matter, but still very relevant. 2 Interviews with Dr. Barber from my fav science program. Especially the 1st note the date!
See Pt.:

See Pt.: Frost Flowers Pump Pollution

Seke Rob

Frost flowers... good example what those beautiful high smoke stags achieve...Iniut disposing of dead polar bears as toxic waste. Would it rain out too when an east coast coal plant billows it high and it then coming down over the west coast, or runs into the Rocky Mountains, then precipitating out somewhere on that home vegetable garden? Just the thought.


In therms of the snow flowers I was also thinking of the added insulation facture, not only the air pockets they would create adding insulation, but also the fact that they take out the slat from the ice under them therefore making it denser and not allowing the cold to penetrate as deeply.

Jim Williams

Chris Reynolds: "Did someone mention a bet?

"You claim that the Arctic will be sea ice free in winter after one more crash?"

Science is modelling hundreds of hurricanes and adjusting your models until you have good predictive ability; which they are just beginning to achieve. After that you might develop a theory based upon your models, but that isn't Science. For science you then have to once again observe hundreds of hurricanes and demonstrate that the theory predicts the data.

The modeling of Climate change isn't science. :P

What I see in the Arctic is a chaotic natural system in the early phases of reorganization, and with no indication of homing on a new strange attractor. Unless there is some combination of forces which will support a new meta-stable state (a strange attractor) after depletion of sea ice the system will continue to rapidly move towards warmer conditions with the removal of the damper of persistent ice. Nowhere have I seen a good argument that there will be a winter loss of heat significant enough to compare to that involved in converting ice to water. Show me that replacement damper and I'll change my mind, but right now I see a system on the cusp of catastrophic change as it attempts to find a new meta-stable state.

If the Arctic is essentially ice free by the beginning of August the currents and insolation ought to be sufficient to prevent complete freeze-over during the following Winter -- and almost no freeze-over the Winter after that. That is about how fast a natural system of this size in failure mode would be expected to change. (The "less than a decade" which evidence seems to indicate for past such events.)



You mentioned that Cryosat-2 data is in agreement with PIOMAS. Do you have any links to new Cryosat data?


Chris Reynolds:

"Why not try reading some of the science?"

I have been reading the science, published research papers, since just a few years after I was old enough to read.

Being born into a family of hard-core scientists makes that rather easy. I helped my parents feed stacks of punch cards into now-ancient university computer systems in order to complete statistical runs on their PhD work when I was very young.

I was also programming on machines that were just a bit less ancient than that.

Having extended convos with my dad on many of the ASI issues, and pretty much all of the extended AGW issues, being discussed here, when I was seven, also gave me a bit of a heads up.

I was also aware of what Hansen was researching a while before his first presentation to congress.

Model deficiencies, assimilating models, initial conditions, chaos, etc, are all very old hat for me.

"Having now read a fair amount you might be able to see why comparing tropical storm behaviour to sea ice is fatuous"

Hint: Why do you think I said the hurricane model prediction I was talking about is "strictly impossible" ?

(Answer: Chaos and initial conditions.)

"You might then wonder if there is something in the atmospheric situation that's adding to ice loss, an element the models are missing."

This is exactly the point I was making. Why do you think it could be just "an element" ? There are currently dozens of elements feeding into this, many of which we have missed entirely until very recently, many of which we are probably still missing, and yes, probably many of which are interacting chaotically.

Memory in physical systems appears in many ways. Storms leave various traces behind that serve as memory that affects storms after them, as one example. Still the overall situation is chaotic and unpredictable past a certain horizon.

Yes, it will all melt via AGW. That is certain. If we know that already, what, then, are we trying to achieve with the Arctic models, and is it realistically approachable?

(I completely respect the people and their efforts going into this. That's beside the point.)

When and how it comes across to the larger public that Arctic changes are drastically impacting us, they will look at the work the modellers have done, and say, what the hell?

If you think a few city and state governments and their citizens can get worked up about hurricane forecasts and their fallibility, you ain't seen nothing yet.

It would be wise to get a handle exactly on what we are trying to achieve, what we are capable of achieving, and get this clear to the public in advance.

Bob Wallace

Jim, please take your tedious arguments to an appropriate thread.

Currently "Why Arctic ice..." is serving as the general discussion thread.

This is a thread about Arctic sea ice volume. And I really want to learn as much as possible about Arctic sea ice volume. Off-topic comments, please, elsewhere.

Bob Wallace

Someone with GIMP/Photoshop skills.

Would it be relatively easy to do a count of pixels by color from the ice thickness images that A4R is posting? Then use these counts to graph the area of ice for different thicknesses.

I think it has to do with "writing a script", but I've never gotten that deep into graphic stuff.


If people interested in SIV (volume) haven't been checking these images, I'd recommend taking a look. They give one a three dimensional view of the ice over time.

Espen Olsen

Anything is easy in Photoshop as long as you know how to do it, and know where the f...... button is!

Chris Reynolds


Sorry, I read pseudo-scienntific handwaving. Yes we're in the aftermath of at least one bifurcation. I think 2010 was another. I also think we're on the edge of a big one. But that doesn't mean a winter situation with metres of winter ice growth now can plausibly turn into blogger-all in a few years.

And yes I have read Dorian Abbott and Eli Tzipperman's work!


I said 'an element' imlying one of a set.

Storms don't have memory like sea ice! Just look at the steps in the volume trend.

What are we trying to achieve with models? Understanding. That's the primary reason for using models in science - all models are wrong, but some are useful.

You can reply how you want, I'll leave it at that. But as several people have already asked, could you take it to another thread so those of us who want to discuss the current PIOMAS volume record can do so with a low signal to noise ratio!

Bob has suggested an appropriate thread.

Jim Williams

I'm sorry Chris Reynolds, but Bob Wallace will bitch and moan if I respond.

Chris Reynolds

Hi Bob, what's the purpose? i.e. what are you thinking?

I've access to PIOMAS data from 1938 to 2011, and will shortly be giving out a link to graphics and some useful spreadsheets.

I'm going to see what I can do in ImageJ.

Chris Reynolds

Forgot to add - a major problem is that for a flat image centred o the pole the area of pixcels moving away from the pole reduces. If you don't account for that you'll have increasing error the larger the ice gets.

A rough and ready solution would be to save the images as bitmaps and directly read from a program - if you can work out the pole pixcel you can then work out the scaling for pixcel size.

Bob Wallace

Thanks Chris. What I've seen so far is graphs of EOY ice age, which I assume is actually ice thickness. Thickness interpreted as age.

It just seems to me that it would be interesting to watch the thickness shift over the season and season to season in a more detailed version than the PICT graph.

I'm afraid that I don't have "deep thoughts" to add, I'm operating at the incoming freshman level. I guess I'm reacting to the claims previously made by some about how the very thick ice against the CA would keep the meltout from happening and my eyes tell me that all that very thick stuff is quickly disappearing and graphing would make it clearer.


Bob Wallace

"If people interested in SIV (volume) haven't been checking these images, I'd recommend taking a look. They give one a three dimensional view of the ice over time."

Do they have these images for previous years? These only tell me that ice melts in the summer. I know there is more to the story.



Source for prior years, January 1 2011 to date:


Go to "Global - Global Ocean"
Go to "Sea Ice Thickness"

I reset the range to 6 meters and 0 meters.

For May 1, 2010 to Sept, 2012,

Go to "Global - Arctic Ocean
Go to "Sea Ice Thickness"

Reset the range/scale to 6 meters and 0 meters.

Bob Wallace

Dj - I don't think so. Perhaps A4R can chime in here.

I find them fascinating. The ice isn't just melting in from the edges to the last bastions along the top edge of Ellesmere and the CA. It's disappearing in place in great amounts. Melting in place or sliding out the Fram/into the CA channels.

We should be able to do two-year comparisons next year.

This goes back to my (unanswered) question about whether there is any data that gives us the relative top/bottom/edge melting amounts.

I think it's also going to be interesting to see how the ice thickens this year. How high will it pile up along the coasts? Being overall thinner will it be easier to move and, thus, pile higher than the previous year?

I think it might tell us something about the shape of the melt curve during the last one or two years - continued rapid melt or a leveling out....


Chris, others:

As I said: "this question could belong in this PIOMAS thread as well as anywhere else, I believe, but correct me if I am wrong"

Of course I meant that. I respect the unspoken rules here, and apologize for being OT in larger and smaller ways as I have been. "If I am wrong," fine! - there isn't any problem dropping it or taking this elsewhere.

PIOMAS is an important model, and one I have been following a long time. It looked like PIOMAS was doing well. Now it seems it may be off a bit. Possibly a big bit. I am personally disappointed about that.

So thinking about that more, I started thinking even more about the value of what even the best Arctic models are predicting.

Chris: "all models are wrong, but some are useful." - completely agree. The distinction has to be made where models fall on the

(not even wrong)<------------>(very very useful)


Volume is key to larger models, and my thinking about PIOMAS still possibly getting it way wrong leads me down this train of thought.

"Storms don't have memory like sea ice!" Thanks for pointing that out. Both memory and chaos work on a variety of scales, in time and space - that was my point. How does the larger-scale memory of sea ice work with the larger-scale chaotic factors?

Any solid basis for thinking we have a handle on that? Even close to having a handle on that?

Personally, I have come to the conclusion that we are very possibly just dorking around in the land of folly with Arctic models that are trying anything better than "hey, the ice will melt!" - on the 10-100 year scale. And that is with all due respect to my intellectual superiors - and also does not include my feeling about global AGW models.

There is a reason Pauli came up with that brilliant turn of words. Please read:


Especially the part: "cannot be used to make predictions about the natural world."

For my last word on this, I will leave it with God.

Or at least Feynman. Close enough for me.

"I can live with doubt and uncertainty. I think it is much more interesting to live not knowing than to have answers which might be wrong."

Al Rodger

A first stab at analysing the difference between the PIOMAS thickness graph & thickness via PIOMAS/CT.Area.
I had a quick look at the last 6 years & there is no trend to be seen. The only rogue year is 2008. So ignoring 2008, I used the mean of these years to convert PIOMAS/CT into a possible PIOMAS thickness. It appears on first sight that many pre-2007 years may resemble 2008 rather than the 6 most recent (non-2008) years. Thus a trend may still be hiding in there.
I anticipate a second stab in coming days.

Chris Reynolds

Hi Bob,

Try here:
under the thickness folder. The chap running the server has just given me the OK. It's not yet ready for public release, I've got to do readmes and tidy up some macro spreadsheets.

If you have Excel you'll find a spreadsheet I used as the spine of my most recent two blog posts under 'Spreadsheets'. It'll be ThicknessCats.xls - give me 45 minutes.

The thickness folder on that server is a series of maps from 1978 to 2011 for all months showing sea ice thickness. It's been calculated using visual basic macros in excel from gridded thickness binary files available here:
I'll explain the rest in a blog post later this week.

Actually - you might find the anomaly plots useful - they're difference from each grid point's long term average (1980 to 1999).

Bob Wallace

Thanks Chris. Looks like you've made my evening (busy).

Chris Reynolds

The Excel spreadsheet of volumes of different thicknesses of ice is available here:

It runs from 1978 to 2011, the source data only goes up to 2011, and is monthly.

I've stripped out my notes and left it free for people to ammend as they see fit.



Thanks for sharing the coming thickness maps and excel data - what a gold mine.



I put A4r on a layer in CAD, scaled it to the adjusted known surface area of all arctic seas (a crude but effective projection when account is made for bias up to 5% regionally), and mapped the areas between the lines of even thickness (…isowhats??...).
What I come up with is a reasonable match to PIOMAS when account is taken within the different areas for the concentration.


Sorry A4r, you didn't fit the digitizer...
Of course I should have written A4r's linked thickness graphs.


FWIW, I just can't figure why there's a difference between the structural info visible on MODIS and the position of these lobes on the graphs. This troubles me in the same way as recent PIOMAS values do.
I'd like to comment on this later.
BTW I do respect what the experts produce...

Peter Ellis

Al Rodger: That looks very much like the inverse of the CAPIE graph.

That makes complete sense given that (if I understand the PIOMAS graph correctly) it's treating each grid cell as either ice-covered or non-ice-covered, and thus their denominator will have a measurement that behaves more like extent than area.


The point I was trying to make earlier is clearly stated by Dr Barber himself here

A very key point from his talk is that the satellites were incorrectly classifying MY ice. Something he reiterated in March 2011 and I have not heard any more about the training they are trying to do.

So it would not be surprising if PIOMAS was actually highly optimistic and it would be no suprise if the melt being seen in MY ice areas is actually the annual ice melting out leaving the rotten ice behind.

When trying to calculate the volume of ice remaining, this information is critical.

Sadly his lecture in Norway "On Thin Ice: The Arctic and Climate Change" can't be accessed now.

Artful Dodger

Webcast - Oslo Science Conference - Dr. David Barber
"On Thin Ice: The Arctic and Climate Change"


I'm late to the party, but anyone who wants to talk politics, the gaia principle, and the incipient collapse of society and who isn't a right-winger or afraid of a little healthy debate is welcome to come to the Environment/Energy group on Democratic Underground.


(You don't have to be an American to join; we have people from all over.)


"Webcast - Oslo Science Conference - Dr. David Barber
"On Thin Ice: The Arctic and Climate Change""

Been there done that. Get a server not found stream error.

Artful Dodger

Hi, NeilT. I'm watching it now...

Hopefully others the link will work for others as well.





"Hopefully others the link will work for others as well."

It worked for me.


Thanks Lodger Getting clear feed. He does talk about what my feeling is that the quantity is becoming a major issue. Incredible talk.


Good from here - will watch later.


Just a point. Unfortunately most of the funding he talked about has been almost totally axed in the last couple of years. IMHO a very big mistake we will pay heavily for.



"Sorry A4r, you didn't fit the digitizer...
Of course I should have written A4r's linked thickness graphs."

No offense on the "missing link"...smirk!

I think it neat that we all bring so many resources to a complex problem, and each makes a contribution.


Al Rodger

A second stab at the question why PIOMAS thickness is so different from thickness calculated using PIOMAS volume/CT area (PICT), but I haven't even wounded it.

The whole set of graphed PIOMAS thickness years graphed here as PIO-THICK/PICT ratio shows 2008 is a bit of a rogue. Also the pre-2000 years are little different to recent years, unlike 2003-06.
As for 2008 - it saw PIOMAS volume up 20% on 2007 but (unlike 2009) 2008 CT Area was not greatly different from 2007. That explains why the 2008 PICT thickness numbers are higher than other recent years but not why 2008 PIO-THICK are so similar to recent years.

And it doesn't explain generally the difference between PIO-THICK & PICT, graphed here. I was expecting the falling ave thickness would give some clue to it. Or perhaps it would show individual transitional years. But nothing has appeared. It does all seem too large just to be caused by a different calculation of area.
So more work required. I shall sharpen the knife!

Peter Ellis

Al Rodger:


The volume cancels out, leaving you with PIO-AREA / CT-AREA. The difference mathematically has to be a difference in how they're calculating the area. Volume is the same in both calculations (since it's taken from PIOMAS) and we've divided it out! Volume is provably irrelevant to that graph. I can't say this any more clearly.

Given that, then when looking at the graph, PIO-AREA / CT-AREA maxes out at ~1.4-1.5 for most years. Turn that upside down, you'll see that CT-AREA is ~65-70% of PIO-AREA. That is exactly the same range as CAPIE in recent years, meaning that the discrepancy vanishes entirely if you assume PIOMAS are defining thickness as volume / extent rather than as volume / area.

Given that their stated methodology flat out states that that's what they're doing, I really don't see the benefit of further numerology.

Peter Ellis

Typo - wrote the calculation the wrong way up.



It seems that also IMS has made a new record.


The numeric values are not updated yet but the graph is.

Bob Wallace

"It seems that also IMS has made a new record."

Oh, somewhere in this favored land the sun is shining bright;
The band is playing somewhere, and somewhere hearts are light,
And somewhere men are laughing, and somewhere children shout;
But there is no joy in Mudville — mighty Tony has struck out.

Artful Dodger

Hi Al, hi Peter

One more factor to consider between CT and PIOMAS is the domain for sea ice. CT defines it very broadly, including places like the Gulf of St. Lawrence and Sea of Okhotsk. PIOMAS is pretty much just the CAB and contiguous peripheral seas.

So CT Area only approximates PIOMAS area once the sideshows are over, and the curtain comes up for the main effent. ;^)

Maybe July to Oct? Would also vary by year. Should be useful at min SIE though.


Peter Ellis

Lodger: Sure - but we can get a guesstimate for the maximum difference in domain size by looking in midwinter. The graph bottoms out at ~0.9, suggesting that the CT domain is (at maximum) 10-11% larger than the PIOMAS domain. In comparison to the summer difference that's pretty small beans.

Al Rodger

Hi Peter Willis,
Sorry to have missed your initial warning over my numerological excersion. It always did have a certain feel of being in the wrong direction.

It was talk of assimilation ice consentrations than most convinced me that PIOMAS would be using Area not Extent. But a quick look at (PIOMAS.VOL)/(NSIDC.EXTENT) shows substantially the same thickness curve as (PIOMAS.THICKNESS).

I note you hardened your view on PIOMAS's use of Extent between your two comments. Do you have a particular reference in mind that may have strengthened your view? So far I find nothing newer than the ubiquitous Zhan & Rothrock 2003. It talks of "extent" but perhaps more in the general sense of the ice edge. It also appears to say it defines the ice edge as greater than 10% concentration. Figure 9, where Extent (ie >15% concentration) is explicitly mentioned, shows PIOMAS simluated extent substantially higher than NSIDC extent. And calculating an area/extent value from the excellent PIOMAS thickness table of Chris Reynolds (linked to up thread) does appear to give a value higher than NSIDC extent. (Then the average bucket thickness I use may be needing tweeking here.)
So any pointers would be useful.

Peter Ellis

I may have been overly emphatic there. I'm basing it on the statement "The average thickness is calculated for the PIOMAS domain by only including locations where ice is thicker than .15 m"

That reads to me as simply "Select every grid cell with thickness >0.15, average the thicknesses for all selected grid cells"

That isn't exactly extent or area, since it's based on thickness rather than concentration. However, since it's based on a binary present/absent call, it will inherently be more similar to extent than area. To get a "true" thickness, i.e. volume / area, they'd additionally have to weight each individual grid cell by the concentration for that cell. If they did that, I believe they'd have mentioned it.

OK, it's not fully conclusive, but given the very nice alignment of the graph with CAPIE, I'm pretty sure that's right.

Chris Reynolds


I think you're right here. This is why some of the features most extremely seen in Extent calculated thickness but not in Area calculated thickness, are seen in the recent thickness release from PIOMAS. When I compare the PIOMAS thickness graph to both extent and area calculated thickness graphs - it looks more like extent calculated thickness.

Peter Ellis

Of course, given that Chris has the technical mojo and access to look at their actual outputs, he can tell us whether we're on the right track :-)

Peter Ellis

It won't be exactly the same as extent though - if I'm right, then the PIOMAS calculation not only uses an extent-like denominator, but also selectively overweights thinner ice during the summer melt (because thin ice is likely to be low concentration, so a 15% cell of 15cm ice will artificially inflate to a 100% cell of 15cm ice).

Al Rodger

Using Chris Reynolds spreadsheet figures linked up-thread I produced a bar chart of the extent of the September ice thicknesses & thought to put on NSIDC Extent values as a comparison.
NOTE that I use the mid-value of the thickness buckets in Chris Reynold's spread sheet. Thus 0.0m becomes 0.125m in my calculation of extent, etc. (I think I used 4.0m for the final bucket.) That the most recent years don't exhibit an over-estimation of extent (in fact show the reverse) sort of vindicated this decision.
Graph here (usually two clicks to 'download yur attachment')


New PIOMAS update: I have updated my graphics at ArctischePinguin for the August data:

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


Thanks for your excellent graphs, Wipneus!

So PSC came out with their August update today. It gave 3400 km3 for 2 september. Wasn't it 3590 on 25 august? So 200 km3 down in about a week.
It is also mentioned that the 600 km3 difference with 10 september 2011 is within the models estimated uncertainty....

Kevin McKinney

Wipneus, that "daily anomaly" graph is interesting. Such a different shape to the curve since 2010--a sharp downward 'elbow' that one doesn't see so much in earlier years--and not at all, really, in the early years.

Regime change writ plain, perhaps--but what is the physical interpretation? Just that thinning has gotten to the point where late season melt does the 'it just suddenly melts away' thing?

Artful Dodger

Hi Kevin,

The full non-linear model proposed in Eisenman & Wettlaufer (2009) suggests during the initial stage of warming, Arctic sea-ice thickness varies seasonally between 0.9 and 2.2 m: (click the image to go to a larger version)

In the three points below, "ΔF" should be read as "change in forcings":

  • Under a moderate warming (ΔF0 = 15 Wm−2), modeled sea-ice thickness varies seasonally between 0.9 and 2.2 m.
  • Further warming (ΔF0 = 20 Wm−2) causes the September ice cover to disappear, and the system undergoes a smooth transition to seasonally ice-free conditions.
  • When the model is further warmed (ΔF0 = 23 Wm−2), a saddle-node bifurcation occurs, and the wintertime sea ice cover abruptly disappears in an irreversible process.

PIOMAS analysis suggests we were already down to 1.1 m average thickness on Aug 1, 2012. It would be interesting to see a PIOMAS thickness trend plot to speculate further.

We may shortly reach that critical 0.9 m threshold when 'it just suddenly melts away'. Not 2012 of course, but the thickness trend should be the clue we're looking for...


P.S. Bonus quote from Eisenman & Wettlaufer (2009):

"Because there is significant incident solar radiation during both the maximum and minimum periods of the seasonal cycle of E (Fig. 1), the ice–albedo feedback ensures that all seasonally ice-free solutions will be unstable (Fig. 2)."

P.P.S. "This ain't your father's Arctic"



The anomaly graphic displays the same data as the "official" PIOMAS anomaly graph.

The three spikes that have appeared since 2010 are now shown in greater detail, which was my purpose to create the graph in the first place.

What is interesting that not only the melt in May and June is greatly increased, but most of it is compensated by a slower melt in July-November.

I hold reckoning that this may be a quirk of the PIOMAS model. PIOMAS may be generally confirmed by other data, but not on this detailed level.

The comments to this entry are closed.