« TOPAZ data by Cryosphere Today regions | Main | There's an Ice Floe in my Garden »


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


No tipping point?

The deniersphere has made a great deal of noise over a recent paper which suggests that the Arctic will not exhibit a tipping point and will recover in 2 years after an ice-free summer.
"We examine the recovery of Arctic sea ice from prescribed ice-free summer conditions in simulations of 21st century climate in an atmosphere–ocean general circulation model. We find that ice extent recovers typically within two years."

The main source is subscription only. Fortunately, a free pdf is available - thanks, Harvard!

Although the deniers-in-chief purportedly do not trust models, they are seizing this particular model as if its output is robustly proven and widely accepted. It isn't and it isn't.

It appears to me that the model neglects multiple feedbacks which could accelerate warming and heat-trapping effects and makes unwarranted assumptions about future greenhouse gas levels.

I may write an article on this, so I would welcome everyone's take on this topic.

My best wishes to all here,
Patrick Lockerby.


Hi, Patrick. It reminds me a bit of an article on the New York Times website we discussed in one of the Open Threads: No tipping point for sea ice in polar bear's future. I was a bit too harsh at first (also due to some ignorance on my part, wrt the negative feedback of thin ice as an insulator), but still...

There's a short quote in there by Dirk Notz, one of Tietsche's co-authors, and a link to another paper of his on tipping points in the cryosphere. If I remember corectly Notz also wrote guest blogs for RealClimate.

Of course it's too bad that pseudo-skeptics jump on this, but it's good for science that different groups go for different explanations. We'll know soon enough about tipping points etc.


BTW, IJIS hasn't been reporting the last few days, but all other graphs (area and extent) are shooting up. It looks like we can expect a higher, later maximum.

Gas Glo

>IJIS hasn't been reporting the last few days

"A slow-moving coronal mass ejection (CME) launched from the sun on March 3rd should reach Earth later today, possibly sparking polar geomagnetic storms. High-latitude sky watchers should be alert for auroras."

Obviously IJIS satelite has been put into a protective mode. Are other satelites not as sensitive or do they do 5 day averages so there appears to be data as long as the outage doesn't last more than 10 days?

Hopefully they can start reporting data as usual from Monday or Tuesday though I haven't watched to see what the delay typically is from CME arrival to reporting data.

Peter Ellis

Hi Patrick: We discussed that paper in the first page of comments on Open Thread 6.

My take is that it's a good paper that doesn't address what the denialsphere wants it to address. Specifically, it's looking at the short-term response to stochastic variation in conditions - i.e. a single year where the melt is much higher than anticipated, such as 2007.

What they show is that after such an unusual melt (including one sufficiently severe to remove all ice), there is a rebound back to the levels dictated by the "general climatology". This is exactly what we saw after 2009: ice extent rebounded upwards to the general trend line. However, that trend itself is accelerating downwards due to changes in the "general climatology"

It shows that there is no "latching effect" whereby a weather-driven perturbation can lead to permanent loss of the ice pack. It does not address general mechanisms of melt, and does not address how the ice responds to changes in overall climatology. But that's OK, because it never claims to, so criticising it on those grounds is missing the point.


This may be elementary for many readers of this blog but let’s talk about the basic physics involved with sea ice. If you take enough sea ice to yield a spoonful of water and put just enough heat into the ice to melt it and then once it is water put an equal amount of heat into the water you will have raised the temperature of the water 80 degrees C. When the authors removed all the sea ice they removed an incredible heat sink from the arctic. More than 14,000 km^3 of ice melts every summer. So if we take the area of ice extent to be 14 million km^2 we can figure out the heat flux into the arctic. If you do the math it turns out that the temperature of all 14 million km^2 of sea surface to a depth of 10 meters is raised about 8 degrees C. This does not include any additional heat from the albedo change. Now it turns out that sea water has a heat capacity by volume of over 3300 times the heat capacity of the atmosphere. This heat can raise the temperature of the atmosphere 8 degrees C to a height of 1 km over the entire 14 million km^2 area of the arctic 8 degrees every day for 33 days. It is hard to see how this heat can all go away and the sea ice fully recover in 2 years. We can already see that the ice is thinning every year just look at the graphs produced by William. So if the loss of ice volume is occurring right now with only a minor decrease in albedo I can’t imagine what hoops the authors of the “recovery in 2 years” must have gone through to obtain their results. Whenever somebody talks about greater heat loss due to thinner ice remember the 3300 times more heat capacity in sea water by volume than sea air. Maybe a reader of this blog can help by showing how much additional heat is produced by the albedo flip of the entire Arctic Ocean.

Andrew Xnn

The discussion of tipping points is contentious and somewhat academic.
A true tipping point would have to be a change for which there is no recovery.
However, as most people recognize, 2008 was a recovery year compared to 2007 in terms of both area and volume.
So, have still not reached a tipping point with Arctic Sea Ice.
However, with respect to volume recovery, 2008 is within the range of measurement uncertainty and 2009 and 2010 have shown a resumption of the underlying trend.

When the arctic reaches it's first ice free summer period, I suspect most of us would also not be too surprised if the following year didn't have an ice free summer period. In other words, it's entirely possible that there could be a recovery. However, there is also a good chance that the following years may all have ice free summer periods as well.

Either way, within a few years after the first ice free summer period, unless there is a tremendous volcanic eruption in the tropics, all the summer will have ice free periods of progressively greater duration.

It's only the most staunch bigots that are going to deny the existence of tipping points and only be being overly anal or ignorant with the definition. IMO, these type of people are best ignored. However, it also a good idea to careful with predictions and not go overboard.

Steve Bloom

Andrew, I think there's a lot of confusion in your comment and others about the "tipping point" concept. A change from one climate state to another (as in ice to no ice) involves crossing a tipping point only if the new state has persistence. Persistence is in turn defined as a change that will tend to remain in place even if the forcing that resulted in it flips back to its prior state. In the case of the Arctic sea ice, that would mean summer ice-free conditions not only occurring, but then continuing to occur even if it gets colder again. There doesn't seem to be any basis for thinking such a thing is likely or even possible with the Arctic sea ice, meaning that it's a change of climate state but not a tipping point.


The Tietsche paper is interesting, but I side a bit closer to Artful Dodger than Peter Ellis, above.

The Max Planck Institute for Meteorology GCM is described in links from here:
I'm not terribly impressed with the results of their reference run 21st century projection, using a 50–200 km horizontal resolution of their ocean model in the Arctic, with 40 vertical levels, that leads to "the Arctic Ocean is typically ice free by the end of summer from 2070 onward".

It's an interesting thought experiment though- take away all the Arctic sea ice, see what happens in the model. This is standard procedure for understanding a complicated model - push the boundaries of parameter space. As Artful Dodger says, and I had the same thought, you could also magically stick in thick sea ice on July 1 and see what happens in the next few years.

I liked their definition of "Arctic energy unit (AEU)" which is the energy accumulated when a heat flux of 1 Wm^2 acts over the area of the Arctic Ocean domain (8.4 · 10^12 m^2) for one average month (30.5 days).

Note that they start their discussion of the Arctic energy budget by mentioning oceanic heat transport (combination of advection of warm water from Atlantic and Pacific and export of ice bergs out of the Arctic) is about 4 AEU's in March, 7 AEU's in August. Although their thought experiment doesn't change this oceanic heat transport, this is the fundamental factor that will cause all the Arctic sea ice to disappear in the summers. This is where the global warming rubber meets the Arctic road. The preconditioned thinning ice, the ice albedo feedback, these will allow the summer melt to be fast enough to remove all the ice - but the reason the ice will no longer be perennial is the warming Arctic ocean, even though the halocline layer delays the effects.

I liked that their model captured the mixing of the warmed surface waters (newly exposed to sunlight after some passing alien mothership beamed away all the sea ice on July 1) down to 50 meters depth - that shows sophistication with the expected winds and waves caused by removing the sea ice.

Like Peter Ellis, I think this paper had some good points - but I don't think GCM's developed before 2005 handle regional predictions too well, and this is the reason all GCM's are progressing towards higher resolution, faster supercomputer platforms, and are working on better software/physics models of key processes.


If we have a new open thread, it must be time for a new dumb question from me. Following on from Runincircles heat budget comment...

Is the following stream of consciousness correct?
First year sea ice is saltier than older ice which has undergone a process of brine rejection. Effective, old ice melts into fresh water, while young ice melts into salt water (perhaps a bit less saline than sea water though). Freshwater freezes at 0 Celsius, while salt water can remain liquid at a lower temperature (as low as -2.something C?). Therefore when comparatively saline first year ice melts, it does so at a lower temperature than comparatively fresh multi year ice.

Given that we have now lost virtually all multi year ice and what we currently have is about 30% second year ice and 60% first year ice (with a bit of older ice), is it true to say that a given volume of ice today (on average) takes less energy to melt than an equivalent volume of ice, say, 10 years ago (on average)? That current ice is not only lower in volume, but less resistant to melting?

Was I asleep when we covered that?


@ RunInCircles, regarding the heat flux into the Arctic, and the role of albedo effects therein, I made a few posts at William Connelley's blog at stoat :
after Gareth posted an analysis of ice volume developments over the past years :
By the way, this is similar to what Prof. Maslowski posted a while ago :

I share Gareth's (and Maslowski's) concern with the rapid reducing ice volume numbers, but when I adjusted heat flux by calculating albedo changes due to open ocean replacing ice cover, I was shocked. My first assessment showed that if this year's ice extent remains about 400k km^2 below last year's extent, that a collapse of Arctic sea ice by the end of the melt season may be imminent.

The reason is simple : when ice melts, it changes the albedo of the surface from some 0.66 reflection to virtual black-body absorption. In the Arctic summer, where the sun is out 24/7, and brutal on a clear summer day, this seasonal change in albedo causes a tremendous amount of extra heat being absorbed by the Arctic ocean. With no major currents going through the Arctic, this extra heat will build up and mostly go to melt of additional sea ice, which will then decrease albedo which will allow more solar heat to be absorbed. That is a very strong positive feedback, which will be most prominent when ice is thin (single year ice), which happens to be the major constituent of current Arctic sea ice. I was surprised that neither Maslowski nor Gareth account for this albedo effect. Maybe they consider their predictions to be too agressive already, but I find no reason to believe that physics will stop working just because we believe that our assessment is already more than we can handle emotionally.

My initial albedo change assessment is that if 2011 ice extent remains 400k km^2 below last year's extent, that this will cause an additional 3,000 km^3 ice to melt before Sept minimum, even considering an 'average' Arctic spring/summer. Since last year there was little more than 4,000 km^3 left over, so a 3,000 km^3 reduction would result in essentially total elimination of 2011 summer sea ice.

Now please note that my initial numbers are pretty 'rough'. I am very new here, and I would hope that you guys can help figure out if this concern can be validated.

In short : I request your help in finding accurate data.

For one, I am not sure if I used the correct Arctic insolation numbers. I used average insolation numbers for the Arctic obtained via a wiki site (about 100 W/m^2 yearly average insolation) :
If anyone has better Arctic insolation (irradiance on the ground) data, then please post it.

Second, I am not sure where to obtain accurate ice volume numbers. I've been eye-balling PIOMAS graphs, but can't find monthly data in a table. Besides, PIOMAS seems to be down for a month already. I found William's TOPAZ analysis, and will surely download and analyse that data, but at first sight TOPAZ seems inconsistent with PIOMAS. Especially the winter volumes from TOPAZ don't seem to match PIOMAS winter numbers.

Also, for ice extent, I'm using JAXA, because they publish their data in a table. However, I'd prefer to use sea ice "area" data, which I find on Arctic Roos, but there I can't find the data in a table, only in a graph. Besides, Arctic Roos' 'extent' numbers don't seem to match JAXA's, so now I'm getting very confused.

Where do you guys get the best ice volume, extent, area and other data ?

I would really like to solidify my calculations, preferably build a 2D model for albedo effect in the Arctic, but I'm at this point really lacking solid data input.

Either way, it looks like this year's melt season is going to be very, very interesting. Especially if extent remains below the 2010 line until the real brutal Arctic summer heat starts kicking in. Ice is so thin that albedo changes due to reduced extent, and a couple of weeks of nice weather in summer could set the stage for a spectacular collapse in late summer.....


I'm sorry guys. Seems my login shows a weird link. But my name is Rob.


Hi Rob, welcome to a place where it's less quiet. ;-)

Here's what I know (which isn't much): First, you can go to the Arctic sea ice graph page and on top you'll find some of the daily and monthly extent and area data out there. You'll probably want to use Cryosphere Today Area numbers and for that you'll need the third column (from left to right, first column is date, second is anomaly from the fourth column, which is 1979-2008 average).

With regards to albedo and insolation: don't forget about clouds. Last year's melting season was heading for ice carnage when all of a sudden the weather flipped and things were very cloudy and cold during a period of 6 weeks when insolation was highest.

My initial albedo change assessment is that if 2011 ice extent remains 400k km^2 below last year's extent, that this will cause an additional 3,000 km^3 ice to melt before Sept minimum, even considering an 'average' Arctic spring/summer. Since last year there was little more than 4,000 km^3 left over, so a 3,000 km^3 reduction would result in essentially total elimination of 2011 summer sea ice.

That's based on an average thickness of about 0.8 m, yes? Maybe the ice that remains is thicker than that, so volume numbers will not necessarily crash so hard?



For someone who has argued the same lines qualitatively here and elsewhere, even I find your albedo change results scary. Firmly buckled in for a wild ride....

One thing to watch out for with extent numbers is that different sources use different threshold concentrations which can lead to some pretty big differences. 15% and 30% are the most common thresholds - which is used where is usually flagged prominently.

For PIOMAS numbers, Wipneus and I generated monthly data using slightly different methods - not perfect, but it might meet your needs. Take a look at Open Thread 3 here for the details of how the data was derived, and links to the monthly datasets.

Bill Fraser has kindly posted some excellent TOPAZ data for extent area and volume - how consistent they are with other datasets, I have not yet determined, but they should be internally consistent at least (although I have found some spurious datapoints that
need correcting). The description and links can be found on the TOPAZ thread.


Gas Glo

Probably a dumb question, but people are saying the extra heat trapped by water instead of ice is scary. That heat presumably goes into higher temperatures before it can melt ice. So what proportion of that heat energy is radiated away to space due to the higher temperature before it gets to melt ice? It obviously at least partly depends on how far away the ice is, so maybe the question is like asking how long is a piece of string. But it still seems worth asking whether I should be thinking of 10% or 90% radiated away to space before it melts ice?



If you want to do the albedo calculation properly you need to account for the angle of incidence since the sun will not be directly overhead. I've attached a site which shows how it depends on angle for a smooth surface, let me know if you need any help.





To get the actual equations you'd need to calculate the dependence of albedo on angle see here: http://hyperphysics.phy-astr.gsu.edu/hbase/phyopt/freseq.html


Greg Wellman

There's a whole article (discussed on another open thread) about the reflectivity of seawater as a function of incidence angle *and* surface texture (i.e. how wavy it is). Waviness is very important.


IJIS is back up. Currently extent is at 13.733 K, just 50K below the 2006 record low maximum.

Pete Dunkelberg

Neven, that's another great picture. Where is it from?
Patrick, there is also Zhang et al 2010: Arctic sea ice response to atmospheric forcings with varying levels of anthropogenic warming and climate variability.
RunInCircles: It is hard to see how this heat can all go away and the sea ice fully recover in 2 years.
Evidently it is easy for the models to 'see' this. There are ocean currents, and heat in the air can just rise and blow away.

All, I have one simple interest for today. Note the new topic at RC. I'd like to stimulate/provoke the experts to comment on Why ignore Maslowski? How can this be justified? (seeking a better answer than "Use Global models. It's a rule.")
I think many here would be interested in the answer. You are welcome to help pose the question ....


Pete, I found at some blog by an actor who went to the Arctic to shoot something (a movie, not polar bears). I know I should source those images, but I couldn't be bothered. Only do that with graphs etc. :-)

Good idea about the question on RC. I'll see if I can manage to get a question in.


Patrick, if you're still reading (I forgot to put this link up): there was a good discussion of tipping points and hysteresis a year or two ago on Michael Tobis's blog.


Tipping points? How about a picture Wayne Davidson put on his EH2R blog two days ago. NOAA VHRRS picture north of Greenland 0303.
You'll also find it on the DMI site. Wayne never saw fracturing reaching that close to Nares Strait and Ellesmere Island.
We used to think that these regions would be a sort of 'last refuge', may the summer ice free period arrive.
But what you see on the pics is very, very vulnerable sea ice. It makes you guess wether the 'late winter freeze uptick' is just about spreading and advection. Not about volume recharging
BTW I see a big difference between the VHRRS and the ASAR pics. Explanations?


Yes, I saw that too, Werther. It won't be long before that white ball in the middle of the MODIS Arctic Mosaic will retreat enough for us to have a look and start comparing with last year. If I've understood correctly an ice bridge has formed in Nares this year. I wonder how long that will hold.


Well the weather is so ugly up there that Michele Pontrandolfo and other arctic explorers teams were forced to cancel their expeditions. http://pontrandolfo.wordpress.com/2011/03/05/resolute-bay-05-03-2011-finito/

This year, the last date to get picked at the Pole is April 26. They were supposed to start on Feb 26, having just 60 days to reach the pole... Catlin expedition should begin soon (but they won't go to the Pole this year).


Greetings, everyone! I've been a lurker since last year, but this is my first comment! I just wanted to alert everyone that PIOMAS updated today to Feb 28th. It appears that the negative anomaly has grown to 8.5.


Oops, forgot the link:


Impeccable timing, maltose.

So this means total volume is currently approximately 20.000 cubic km? Maybe 19.000-20.000 km3 is more like it.

To remind everyone, here are the unofficial PIOMAS data as compiled by Wipneus.

And FrankD also produced some similar numbers.

Greg Wellman

Yes, I get about 19,000 km3. In an average year would have another 1000 growth left before volume starts to decay into summer.

No statement at the PIOMAS website about validation with CryoSat.


No statement at the PIOMAS website about validation with CryoSat.

Greg, if I've understood correctly this will take at least another year.

Greg Wellman

We're also almost 1000 km3 less than this time last year (as one can see from simple inspection of the PIOMAS graph). The question is whether we will see that May-June "fall off a cliff" effect in the anomaly. If we do, then I boldly :-) predict that we will continue to see such a seasonal cycle (i.e that there's a new average seasonal cycle, so using the old average results in a seasonal residual).

Daniel Bailey

@ Patrick

IIRC, Gavin's take over at RC on the Tietsche paper is that their presumption was that forcings were held magically at zero. Since this is an impossiblity (AGW+Arctic Amplification) he didn't bother to go into any more details on his thoughts on it. Which is telling.

@ Neven

Farscape dude? Some actor? Dude, that's John Crichton! (/Fanboy) :)


IJIS just updated to the current date as well.


Hehe, I don't have a clue, Yooper! Chucked the TV out almost 3 years ago. Never was much of a SF fan anyway.

Daniel Bailey

TV's over-rated...The meltdown of the Arctic Polar Cap is a SciFi story in its own right...No network exec would ever approve this storyline.

The Yooper

Pete Dunkelberg

Greetings maltose, somehow you have the same red symbol as me, and you're putting it to good use.

Artful Dodger

Here is an audio file and transcript of Adm. David Titley (U.S. Navy Task Force Climate Change) on March 4 with Ira Flatow on NPR Science Friday: Navigating In A Changing Climate


@Neven, thank you for the pointer to the Arctic Sea Ice Graphs. Very helpful, and the first thing I looked at was the 'sea ice area' from Cryostat. I was shocked to find that today we are 669k km^2 below last year same day. That's way below my critical line, and thus very, very concerning....

Regarding clouds and the weather, I agree with you that 2010 was probably saved by very cold weather in summer, and the cold spell that started in March (now), which increased ice extent to the point where WUWT was reporting victory 'recovery to mean' postings. No such 'recovery' in sight this year. Just average weather..

Regarding the weather, my first data variable is solar insolation during Arctic summer. I assumed 3.1 GJ/m^2 downwelling shortwave radiation at surface level over the year, based on a graph posted on Wikipedia, by none other than William Connelley himself. Based in HadCM3 model data. Today, I checked the satellite data from ISCCP-D and collected a first rough overview of actual solar irradiation "on the ground" (ok, on the ice/sea surface). Here is a plot for the "shortwave downwelling flux", over the Arctic basin, plotted over the year :


Also, here is an excerpt from Serreze's book, which shows 2D plots of the Arctic with irradiance numbers (1985-1993) as observed by ISCCP-D for each month March-September :


Assuming that this (1985-1993) data still constitutes "average" weather over the Arctic, I compiled the following monthly insolation numbers, which can be used as a first order approximation for solar irradiance over the melting season :

March : 80 W/m^2
April : 150 W/m^2
May : 225 W/m^2
June : 250 W/m^2
July : 250 W/m^2
Aug : 150 W/m^2
Sep : 50 W/m^2

Integrated over the entire 7 month 'illumination' period, the Arctic receives about 3.0 GJ/m^2 of energy. And this is pretty consistent over the entire Arctic basin area.

For Phil(scubageek), I know what you are saying, but that angle was already accounted for in the wiki top-of-atmosphere insolation graph. Now, this ISCCP-D shows us how much of that radiation actually gets to sea level, considering even the (75%) average cloud cover and other weather conditions. You see confirmed in the 'June' graph in Serreze's book that indeed the North Pole area still receives the highest insolation anywhere in the Arctic, as the wiki top-of-atmosphere insolation graph already suggested.

Back to the monthly insolation numbers.
First let's look at that 3 GJ/m^2 that is accumulated over the melt season. If that hits ice, then 66% of it will be reflected. If it hits sea water, it will pretty much all be absorbed. That means that there is a difference of 2 GJ/m^2 accumulated for every square meter of ocean that was frozen last year.

So if the difference between this and last year's ice extent is 400k km^2, then the Arctic will receive 2 GJ/m^2 * 400k km^2 = 8 * 10^20 Joule of EXTRA heat over last year. That extra heat is enough energy to melt some 2400 Gigaton (km^3) more sea ice by the end of the melt season. By itself, that would cut last years remaining ice volume already more than in half, and probably destroy the majority of sea ice left.

But it gets worse :
Here is how you can see how the albedo effect actually amplifies melt as the summer proceeds :

Imagine one m^2 of open ocean on the west side of the Arctic Ocean (e.g. the sea of Beaufort or Chukcki sea) that was frozen last year. Heat accumulated there has no where to go but to the ice at it's bounday. With an insolation of 250 W/m^2 in late May/early summer, 1 m^2 of open water will accumulate enough heat to melt a nearby 1 m^2 of first-year (80cm) ice in approximately 12 days. Albedo of 0.66 for ice extends that 'extra' melt-to-double size to 18 days. So in 18 days, the difference between last year and this year will double. 18 days after that, it doubles again. And again, until the insolation starts to fade out at the end of July.

In my simple model, I assessed other factors that prevent all that heat getting to the remaining ice (like heating of ice free ocean, and reduced heat transfer from larger distances) and conservatively assumed that the 'multiplication' effect above only occurs for heat accumulated in the West Arctic Ocean and only operate on first year ice and nowhere else. I ended up with some 3000 km^3 more ice (than in 2010) that should be melted at the end of September if the 400k km^2 ice area difference exists by the end of spring.

That would almost certainly knock 75% of ice volume that was there last year, and most likely more than 75% of ice area. Only some multi-year ice on the north coast of Canada and Greenland would remain. Which means a virtually ice free Arctic.
The way it looks now, we are even ahead of that assessment...

Next post, I'll do a regional albedo assessment. Heads-up: keep your eyes on the Chukchy Sea and the Beaufort Sea. If open ocean starts to fom there earlier and larger than last year, and ice flow is faster, then it's gotta be a hell of a ride this summer..

By the way, I may go off-line for a couple of weeks. My wife is about to give birth to our baby girl. That overrules everything else, even the prospect of an ice free North Pole for the first time since the beginning of the Holocene...


Gas Glo

IJIS extent is now above the record low maximum. Still a good chance of record low area maximum on CT. IJIS extent seems likely (7 of 8 years movements) to stay below 14m km^2.


By the way, I may go off-line for a couple of weeks. My wife is about to give birth to our baby girl.

Good luck, Rob!


It's your call, Glas Glo... but as every measurement has its' limits, that call is within a margin of uncertainty. I guess you interpreted the IJIS number comparing to 9 march 2006: 4600 km2 less! But Uni Bremen also counts on AMSR-E and still leaves 74K to go!
What's interesting is that the extent is 1.75 MK under the average 1078-2008, and 500K under last winter (Uni Bremen).

Daniel Bailey

Family's the one ball we jugglers cannot afford to drop! Good luck with the addition Rob & best wishes!

What the melt season will look like upon your return...stay tuned.

The Yooper


And FrankD also produced some similar numbers.
Of course, I would say that Wipneus produced some similar numbers to me ;-)

Just for the record - I posted my graph of my data, then Wipneus posted his dataset, then I posted my dataset - and now I'm thinking of Tom Lehrer's Lobachevsky...

Anyhoo, regardless of primacy, I've updated the PIOMAS interpolated numbers. I note that they have rebaselined their graph (the big dip last year ran down to -11.2 and is now only 10.9), which is a bit of a bother. The graph scales have also changed slightly, changing my process for interpolation.

Rather than a complete do-over, I've reevaluated all of 2010 (the changes were fairly slight) and added in the first two months of 2011. Since the purpose is to produce absolute volume data, the results should be pretty consistent, as long as one is consistent with which anomaly from which baseline is used: http://snipt.org/wnoP

The changes for 2011 compared to 2010 are -880 for January and -824 for February. Both are below the quadratic fit linked at the beginning of open thread three (ie, the intercept for the new curves for January and February are ~6 months earlier when 2011 data is added.

So far, so bad.


Blast, I always forget something - on the new dataset, the top row after the months is the 1979 - 2009 average for the mid-point of each month, that is, the baseline against which the anomaly is read.

As I say, PIOMAS have rebaselined their graph. I have not rebaselined my data, but I have adjusted the 2010 rework and the new 2011 data to conform to the old baseline.

The graph shows a current anomaly of around -8300. My anomaly for February is around -8600, because the old baseline I am using is about 300 higher than the one used on the new graph.
(I hope that makes sense)


Yes, Frank, that makes sense. Good catch realizing that they re-baselined their data. I certainly didn't notice that. (Though now that you mention it, I recall someone on here complaining about it last time they did that.)

So far, so bad, indeed.

Gas Glo

>"I guess you interpreted the IJIS number comparing to 9 march 2006: 4600 km2 less!"

Don't follow where the 4600 comes from nor the 9 March 06 date. I have record low maximum as being 11 March 2006 13,782,344. When I posted 6 March 2011 was 13,811,719 and preliminary for 7th March was 13,817k so we appeared at least 29,375 km^2 more not 4600.

Yes, you are right, that may well still be within measurement uncertainty. We are now another 36,000 km^2 further above record low so are now much closer to being able to rule out a tie within measurement error. Could part of the 74k difference to Uni Bremen figures be because of a longer delay or longer averaging period? (The graph line looks barely past 1 March, not that the IJIS line looks much further on.)


For Phil(scubageek), I know what you are saying, but that angle was already accounted for in the wiki top-of-atmosphere insolation graph. Now, this ISCCP-D shows us how much of that radiation actually gets to sea level, considering even the (75%) average cloud cover and other weather conditions. You see confirmed in the 'June' graph in Serreze's book that indeed the North Pole area still receives the highest insolation anywhere in the Arctic, as the wiki top-of-atmosphere insolation graph already suggested.

Rob, you misunderstand me, what I'm referring to is the reflectivity of the water which is a function of the angle of incidence, for example if you look towards the setting sun over water you'll see increased reflection.

First let's look at that 3 GJ/m^2 that is accumulated over the melt season. If that hits ice, then 66% of it will be reflected. If it hits sea water, it will pretty much all be absorbed.

That's the key, at lower angles more reflection, at ~20º above the horizon ~15% will be reflected.

All the best with the baby.


Lord Soth

Althoug sea ice maximun extent is almost entirely based on weather, its fun to make predictions.

Based on the eight years of observations that we have, which is a small sample size, here is what I have come up with.

The sea ice maximun appears to either occur with a single peak or a double peak

2003 and 2010 were double peak years, while the rest was single peak years.

In both cases the difference between both peaks was 40K or less.

Discounting the double peak and taking the earlier date; the average date of the minimun is March 7th, and counting the double peak makes the date of the minimun as March 13th.

So on average we have reached the average date for maximun ice growth; however I do not expect more than 50K of further growth, if 2011 becomes a double peak year.

Based on this sparse data set, I can gladly predict that 2011 will be the second lowest maximun sea ice extent; with a total lack of confidence :)


re albedo and predictions for a rapid collapse, if you run the prediction model for last year's numbers, what happens? Does it come close to matching what we saw?


@Artful Dodger - thanks for the link to the Adm. Titley interview - I used to listen to Science Friday quite often.

I'll give you a couple of - just one example of what we are doing. We are working with the National Oceanic and Atmospheric Administration and other federal agencies to fundamentally retool our ocean and atmosphere and ice prediction systems to make them, frankly, more relevant to sort of the human scale, not just global scale, but the human scale and human times - so seasonal, decadal times of times, the kinds of times that families, businesses, industry, government, military actually make budgetary decisions on. And we need to do that to, frankly, have credibility.

As I'm sure your audience knows, while the large-scale effects of climate change are quite easily understood, when you come down to regional and specific effects that you're actually going to budget for and mitigate against, the science is much less certain.

Sounds like the oceanographer of the Navy, and director of the Navy's Task Force Climate Change, is saying that the the current ocean/atmosphere/ice-prediction systems are just the beginning, and that making climate predictions about specific regional effects is much harder than broad stroke global predictions.

I imagine that when the Arctic summer sea ice disappears in the next five years, this view will be trumpeted among cautious groups that predicted perennial Arctic sea ice until 2070 or so...

Rob Dekker

No baby yet, so still time for some notes, and a couple of thank you's :

@Phil, : I'm sorry I misinterpreted your post. You make a valid point. The angle of incidence makes a difference for reflection, no doubt. In defense, a turbulent sea surface will probably reduce the 'mirror' effect that you refer to, and on top of that, for albedo effect, the real interesting number is the 'difference' between reflection from ice/snow versus (turbulent) ocean water. I don't have any numbers on that, but since every scientist mentions the albedo effect of the Arctic, there must be some significant difference. Do you know of a study that compares reflection from ice/snow versus turbulent water for low-angle incidence light ?

@FrankD, Wipneus, Bill Fraser : Thank you so much for the work done to get the Topaz numbers out. I used Bill's table to calculate a couple of interesting points:

First of all, the TOPAZ numbers suggest that (both inside and outside the basin) that ice area is fairly constant (see Open Thread 3), but that ice volume is going down the tube. Ice is getting thinner, and thinner fast, both in and outside the basin.

For example, ice in the basin on 1/31/08 (after a disastrous 2007) was on average 1.47 m thick, while last year it was 1.66 m, while 1/31/08 it was 1.88 m. At the volume minimum, 10/31/2007 showed 1.53 m ice in the basin, while 10/31/09 showed 1.23 m and 10/31/10 had only 0.96 m on average. The steady decline is obvious and concerning.

So I wondered how low this volume can go without affecting the ice area.
For that, I subtracted the basin numbers from the total, to get the "not in the Arctic basin" ice numbers. Outside the basin, we have thinner (and younger) ice at the minimum, so maybe a better indication of what the basin will look like in a couple of years. Is ice area collapsing there when the ice gets too thin ? Let's check out the ice area minimum outside the basin (based on the TOPAZ numbers posted by Bill):

10/31/2007: outside basin area : 3007077 volume : 4778431 thick : 1.589062
10/31/2008: outside basin area : 3723601 volume : 2283939 thick : 0.613368
10/31/2009: outside basin area : 3158821 volume : 2140378 thick : 0.677588 10/31/2010: outside basin area : 3153626 volume : 1379818 thick : 0.437534

So here we see that the area of the ice does not show a significant trend, while still the ice volume goes does down by a factor of 3. Even down to 43 cm of ice last year.

If that is the pattern for the basin also, then the ice will get thinner and thinner, but it will not reduce in area. Either somehow Nature prefers to thin out ice when it melts, but refuses to break it up, or there is something wrong with the TOPAZ numbers...

PIOMAS shows the same difference in trends (volume down while area stays the same) so now I'm starting to be very skeptical. Are we maybe getting all concerned/excited just because of a model that may contain flaws ?

Where will we be able to process Cryosat data and get actual observed "by-the-day/week/month" ice volume data, rather than rely on a model that makes guesses in the dark synchronized only on multi-year intervals ?

(Neven, I finally registered ! I'm very happy to be here. I'm looking forward to many good discussions with the great people that post here. Your blog is the best undiscovered secret in the 'climate' blogosphere. I hope it stays like this.)

Rob Dekker

One correction (regarding trend in Jan 31 ice thickness), sentence should read like this :
"For example, ice in the basin on 1/31/11 was on average 1.47 m thick, while last year it was 1.66 m, while 1/31/08 it was 1.88 m (after a disastrous 2007) ."

Kevin McKinney

"If that is the pattern for the basin also, then the ice will get thinner and thinner, but it will not reduce in area. Either somehow Nature prefers to thin out ice when it melts, but refuses to break it up, or there is something wrong with the TOPAZ numbers..."

Or the extent 'crash' is highly non-linear and we haven't hit the knee of the curve yet.


Based on physical observations, I would be very surprised if area and thickness were a linear relationship. On the Great Lakes, it's not unusual to see a large area of ice disappear in the matter of a few days.

It's often a quick ice out, Lake Erie can essentially lose the ice cover in a week. From observing ice out on smaller lakes, it more often is such that the ice continues to thin over a whole lake until winds break it up in a day. Although the seas are not the Great Lakes, it would seem it would be similar, with the ice thinning over large areas until such point it was thin enough for a weather occurrence to break up a relatively large area in a very short period of time.


Just saw your note about having a baby girl - same thing happened to me about 3 years ago. The first 3 years are pretty busy, daughters need a lot of attention, but after that, they're pretty independent and you get your life back.

At least, that's what I'm hoping.


Christoffer Ladstein

So GREAT with new potential Arctic Ice "freaks"!!

Anu: Though I understand what you're hinting, I have to admit that all my 3 kids are quite independent, but I find the preteen years just as labourous as the early years. This is mainly due to a continuous internal fighting, just like high and low pressures are trying to gain land and territory. Also, like us in this forum, we need to be motivated, at least to a certain degree, sometimes hit hard by anti-scheptism from those still denying to see the changes taking place ( http://www.calgaryherald.com/technology/Receding+tundra+will+change+North/4391350/story.html), and for a child this can be totally chrushing, so they need being pushed over a certain treshold, then they probably achieve an inner glow, and most likely girls are better or faster at getting there?! But most of all I feel like a taxidriver, these modern kids are attending all to many activities, not at all just playing around in the local forest climbing trees....

The PIPS for tomorrow will cause a massive flow of ice out in the Fram-strait, most likely this will lead to this years maximum, and also pushing higher than 2007, which had a maximum ( 13945625 km2) 10.March....
This mean we have a double peak this season?


Rob Dekker:

1 - a hearty welcome to the 2nd best 'Arctic Watchers R Us' blog.

2 - congratulations!

3 - here's a link to a range of albedo numbers:

And the best site is ???
Modesty forbids. LOL


Almost-but-not-quite completely off-topic as regards Arctic ice:

For those who missed it before, I have re-published Neven's excellent article on infinite growth -

It is currently in the featured articles box on the home page at -

Please do read it - and pass on the links to your friends.

Rob Dekker

Neven, Daniel, Phil, Anu, logicman : Thanks guys for the best wishes and the congrats. It's a very exciting time. I'm an engineer but I'm completely blown away by the magic of life and the deep rooted feelings of responsibility and caring that come with that. This is our second child, but the feelings do not change. It's almost like life shows the purpose of itself again.

It's so clear now, and so easy : Live to create a future. When it's your time, make sure that you feel that you contributed something that makes the future a better place for those you leave behind. That's all that counts.


My thanks go to Patrick for re-publishing my infinite growth thingy. He also included the Churchill quote ('the era of procastrination...') that somebody suggested here at the time!


IJIS made a bit of a nosedive. Let's see what remains of it after the revision. Was this the maximum, I ask tentatively?


"IJIS made a bit of a nosedive." Neven | March 10, 2011 at 11:16

What are the areas IJIS includes in Arctic Ice
- versus -
the areas NSIDC use for Northern Hemisphere Ice Cover, as at MASIE

I like what Bfraser has provided in the "TOPAZ data by Cryosphere Today regions"

I'm concentrating on regions approximately North of the "Arctic Circle" but fudging to a Latitude of 65-N instead of 66.x-N because 'lines on maps' provided makes it easier to mask and then do a so called "pixel count" for ice above 65-N.
Work in Progress. Also, excluding Hudson Bay Sea Ice because of the narrow width of water in the Fury and Hecla Strait between Baffin Island and the Melville Peninsula of Nunavut connecting the Gulf of Boothia and Foxe Basin.

The Maximum Extent, IMHO, the higher it goes the better. I think we will see that a considerable portion of the Maximum Extent will "dive off a cliff" within a couple of months after the "March Equinox." The more 'peripheral ice' the more dramatic the rate of early season melt. Should be a reasonable attention getter in "other media."


Couple of questions.
I have been thinking. First up will some of the water that melts on the surface refreeze deeper in the glaciers again, effectively causing cracks in the glaciers and breaks in their strength. Secondly we know much of the water flows underneath the glacier to eventualy get to the sea, but if it melts where the underlying geology is not able to easily move it to the sea and given that much of Greenland is below sea level, will there not be large sub glacial lakes and even eventually a small sea under the ice sheet?


Hi all,

"Night, and as the rosy fingers of the Dawn

reach out to caress the face of the wine-dark sea..."

Beautiful isn't it? The maps on Cryosphere Today, I mean, not my loose paraphrasing of Homer. Sobering too to think that the old blind poet stood there reciting rather less than 3,200 years before present...

That's about enough time for the great ocean currents to send their vast bodies of water round the globe twice. 1,600 years per circuit. Two circuits ago predates any Western written records. Its very little data to work with...

Back to Homer. Also within the Odyssey, he tells the tale of the twin ocean perils of Scylla, a group of surface rocks, and Charybdis, a whirlpool. To steer a boat too close to either spells disaster...

I am beginning to wonder if we are not all here at risk of paying too much attention to the events on the surface of the Arctic ocean, (Scylla) at the expense of failing to address the danger posed by a possible melt in the depths (Charybdis).

We have now examined a lot of aspects of the surface melt, and shared a lot of information about the ways in which this is a serious concern. I agree with all or most of this. But a litre of frozen sea water, at surface, produces after it melts a litre of liquid sea water.

If we examine the water that is frozen at depth, the seabed permafrost, the chemistry is quite different. The water here is frozen at high pressure, and in the presence of abundant dissolved methane. It does not therefore form sea-ice, but a chemical known as "methane clathrate" - in which a lattice of ice molecules trap methane molecules within a chemical cage.

When a litre of methane clathrate melts, it produces just 0.8 litres of liquid seawater... and 168 litres of methane. I suggest that this is quite an ugly looking "Charybdis".

I don't really claim to know too much about no science book, and my methodology of lying in the bath, getting ratted and thinking about things that I don't understand seems to have passed largely out of the scientific method round about a revolution and a half of the oceanic currents ago, with Archimedes, but...

Given that this methodology, I admit, may have led me to assign undue importance to the danger and drama of subaqueous methane release, I wondered if there's anybody else out there who's beginning to worry about what is, for me, fast becoming the (potentially flatulent) elephant in the room/ocean?


Idunno, I mentally blocked the one about methane clathrates when I first read about it two years ago. Have been in denial about it ever since.


Hi Neven,

Its kept me completely shtum for the last ten days or so.

@Patrick Lockerby, if you really want to blow Tische et al out of the water, try this...

After the alien mother ship removes the sea ice on July 1, this prevents any furtther cooling of the sea water by sea ice at surface during the rest of the melt season.

Under normal circumstances, a further 6 million km cubed of surface ice is melted by the heat arriving in the Arctic between July 1 and October 1.

Rather than assume that this extra heat hangs around with its hands in its pockets doing nothing, lets assume that it is all available to attack the submerged methane clathrate stability zone.

So instead of 6 million cubed km of seawater, you get (0.8 x 6 =) 4.8 million km cubed of melted water and (168 x 6 =) 2004 million cubic kilometres of methane.

Of course, this is very crude and unrealistic. Its probably a lot worse, as the methane clathrates are acting as a cap on the further deposits of methane which exist in gaseous form in the sediment below.

Neven, I understand your mental block entirely, and probably this belongs in another thread with a skull and crossbones over the door and a warning to those of a nervous disposition, etc.

At least we're not quite as denialist as Dubya, who appears, on hearing of this, to have appointed one of his PR goons to prevent anybody at NASA mentioning it in public again.

Jon Torrance

Scylla and Charybdis were on opposite sides of a narrow strait, with the challenge being to thread the extremely narrow safe path between them. Disappearing arctic sea ice and the potentially meltable methane clathrates at depth are, so to speak, on the same side of the strait, so the metaphor rather breaks down there since avoiding their dangers isn't a question of staying at the sweet spot between them but of giving them both as wide a berth as we can manage.

In any case, the methane clathrates may not have made it to the big time in terms of wide public awareness of the danger of reaching a tipping point where they melt and keep greenhouse gas concentrations rising despite anything we do but there are plenty of scientists sounding the alarm. It received a fair number of pages in Hansen's '"Storms of my Grandchildren", for instance.

Gas Glo

>"I am beginning to wonder if we are not all here at risk of paying too much attention to the events on the surface of the Arctic ocean, (Scylla) at the expense of failing to address the danger posed by a possible melt in the depths (Charybdis)."

Regardless of whether both are on the same side, before we should worry about risk of one while failing to address other, you have to show we are addressing one in the first place rather than just doing nothing other than staring at the oncoming headlights/rocks of the one we are watching.

What should we do about a seaonally ice free arctic? It is too late to curb CO2 emissions to prevent it happening. We need to get modelling groups to address the implications of an ice free Arctic for extreme weather. With information on effects we can then think about adaption. It seems to me that CPDN is the ideal candidate to lobby as a result of its potential to run large numbers of models needed to explore extreme weather. Its validation and attribution experiment seems ideal to me:


Hi all,

in my last post, the paragraph beginning "So instead of 6 million..." I have used the wrong figures. Delete "million"; insert "thousand".

I might go back soon to my preferred designation of "shedloads".

While I've resurfaced again there are a whole host of thanks and acknowledgements;

@Bfraser! You little beauty! Fantastic work.

@Anu, sorry, I dropped our correspondance because I was so strongly in agreement with all you posted. No chance then of you sharpening your wits in argument with me;)

Several great links from lots of sources; too many to list.

Also, unmentioned here and of interest are two recent papers, one showing that Arctic microorganisms are blooming over 50 days longer per year, and another showing that some of the little critters produce a kind of natural antifreeze.

@Jon Torrance. Point 1 taken. "one thing that literature would be very much better for/ is a much more restricted employment of simile and metaphor/..." Ogden Nash

As regards Hansen, I'm just left hoping that he's quite wrong, and Dubya's a genius. Or Nash again -

"A Rule of Thumb.
A rule of thumb
Too clever is dumb."

I wouldn't bet the farm on it though, if I had any choice in the matter.

@GasGlo - Very possibly so; but it seems that the powers that be and the denialosphere are entirely happy with a seasonally ice-free Arctic.

They are currently working out how many mega-tonne modern cargo vessels they can send across the East Siberian Shelf, at a time when it is already showing signs of massive clathrate melt and possible imminent shelf collapse.

So, I would suggest that they can send as many as they want, as long as they don't expect to get clearance for any vessel with:

a. an anchor (It really won't work as soon as the permafrost turns to sludge, and there are very possibly very large very shallow gas fields down there).

b. any form of propulsion which is dependent on combustion,

c. any form of crew which is dependent on respiration.

Methane, is after all, as well as being a strong greenhouse gas, both potentially poisonous and potentially explosive

Patrice Monroe Pustavrh

Well, I do not think that Tietsche's paper is really that bad. However, it should be understood correctly. For me, it is an excellent additional proof that reduction in Arctic sea ice cover is indeed triggered by global warming and it is not event of some unknown variability. Or, in other words - if world wouldn't be warming, than we cannot excpect to observe such Arctic shrinkage as we do now - ice would return to previous levels and this is not what we are seeing right now and we shall not see that. At least, Goddard cannot claim that observed ice loss is just due to extreme weather/wind patterns.

Christoffer Ladstein

Idunno: Suppose this is the article your referring to about the microorganism...?!

In my opinion the debate about the risk of methane release is by "all means" a thread to keep alive here, after all this is very much linked to the lesser ice covered Arctic....
Anybody able to explain just how severe a greenhouse gas methane actually is? I know we're talking serious matters here....


Christoffer Ladstein | "Anybody able to explain just how severe a greenhouse gas methane actually is? I know we're talking serious matters here...." Methane has an official global warming potential of 20 times CO2 over 100 years, though because of its short half life in the atmosphere of around 8 years, most of that 100 years is spent as CO2, so over 1 year it is about 120 times as potent as CO2 for the same mass.

Daniel Bailey

@ dorlomin

It is certainly quite possible for sub-glacial lakes to form (water at high pressure can still in liquid state despite being significantly below its freezing/melting point) beneath the ice sheets; witness those in Antarctica & the Great Lakes of North America (formed underneath the Laurentide Ice Sheet). I would not expect refreezing water inside moulins or underneath the sheet to seriously impact the structural integrity of the sheet much. Ice is essentially a viscous fluid & can deform very easily over time due to the varies stresses upon it (such as gravity).

Where accumulations of liquid water lie below sea level they will stay there as long as they are discontinuous from the sea. Once a linkage from the sea forms, the overburden of ice masses atop the water will hydraulically force the water out, even if uphill, while the ice-mass settles downward, filling the void space once occupied by the water. The subglacial exit channels will also either fill with ice or with re-frozen water.

The net result will be slumping of the ice in those areas & dynamical thinning, such as GRACE is currently measuring going on in both Greenland and Antarctica.

Hope that helps,

The Yooper

Patrice Monroe Pustavrh

Regarding Tietsche's paper, I have one objection though. It is a pitty that they haven't run simulation with double ice cover and with same parameters. I think that would reveal some more things (and it would be verifiable as we do experience warming - actually if we have more ice than current heat input situation, their model should show us results similar to situation we are observing now) and it could clarify a little bit more short term dynamics in warming environment.

Christoffer Ladstein

Thanks Dorlomin! Means more action in a shorter period of time, good for us shortlived humans!!
I know my next link is 10 days "old" news, but the information about more severe snowstorms due to a warmer climate, lead to the question: More snow in the arctic lead to more reflection (albedo) BUT also make the growth rate for the ice to drop, so what is the end result here, plus or minus for the icecover? If the snow come heavily from the start of the freezing season that obviously have to lead to thinner ice, more fragile and most likely to be crunched during a winterstorm, which then will salinate the snow and again reducing the reflection.... I know this is not the easiest equation to solve!

Neven: Just as curious as you, have be crossed the "peak"??? 1 day earlier than 2007 that would be.....

Greg Wellman

The effect of increased snowfall question is very complicated. More snowfall on open ocean decreases salinity and helps sea ice formation. More snowfall on sea ice earlier in the winter insulates it from the (cold) air and slows the rate at which it thickens. More snowfall on ice late in the winter increases albedo, protecting the ice from surface melt in the summer. More snowfall on land ice is always "good for the ice".

The relative strengths of these processes are close to impossible to predict from first principles (and in the case of Arctic sea ice, appear weaker than the effect of warm Atlantic water). Observationally we know the answer so far, and so far it seems to be sea ice retreat faster than expected, land ice retreating roughly as expected (with caveats about dynamic processes below sea level).

Of course that article is specifically about mid-latitude winter snowfall. I believe high arctic snowfall should also increase, but I don't know if by as much.

Gas Glo

Could we have the first century break 13,860,156 to 13,749,688? It may well be revised upwards to miss a century break.

7 of 8 years increase less than 111k from this date so 8th March 13,887,188 is now looking more like the maximum.

Patrice Monroe Pustavrh

Don't know for extent, but march 11th is the day of maximum average sea ice area according to CT.


Gas Glo, it's morning here and it looks like IJIS has already revised to 13,779,844. I don't know if there will be another revision later today. CT area went up about 100K from the preliminary maximum as well yesterday.


The CT growth over the last couple of days has been mostly due to a recovery in the Barents and Kara Seas to their levels of a couple of weeks ago. The process, for the Kara at least, has been interesting to watch.

On Open thread 6 we noted a large polynya opening in the lee of Novaya Zemlya (ie to the east of it). That drove down extent totals in the Kara Sea. Over the ~10 days, grease ice covered over the open water, thickened up (or got snowed on) and brought the extent total back up.

I note on todays MODIS mosaic that the polynya in the lee of NZ has opened up again (there was some warm water moving NE through the Barents in the last week, and I guess it arrived and did its job). There are also strong westerlies in the region which will help move the ice further away. Watch for extent in the Kara to drop.

Since the Barents isn't bounded, I find it a bit harder to say what will happen there, but I think extent will continue to rise for the next few days as ice gets blown to the SW (open ocean). The increase should be greater than the Kara decrease, so totals in that zone should rise. Pacific side, Baffin and Greenland Seas are all dropping a bit.

Gas Glo

Nasty Earthquake, best wishes to those affected.

Seems amost inconsequential to wonder if it is a bad time of year for a tsunami to open up more area in the Bering sea to sunlight and churn up saltier water. I doubt much of a tsumami could get through Bering Straits. I can't see it doing anything other than help ice retreat so best we can hope seems to be that it is inconsequential. That does at least seem plausible with least wave movement in North/South directions.

Daniel Bailey

Off topic, but fascinating:

Tsunami video here

Burning buildings & ships being carried over fields & roads; cars driving frantically away...mesmerizing, horrifying stuff.

The Yooper


Neven: your essay on economics is getting 5 times the hits of my latest Arctic update.

Can readers please correct this horrendous imbalance? ;-)

Infinite Growth And The Crisis Cocktail - A Guest Article
Arctic Ice March 2011 - Update #1

Rob Dekker

I share your opinion. It's a pitty that Tietsche did not simulate the opposite scenario, of starting with a very thick ice pack. In general, it is disappointing that they do not show how much ice 'volume' is created in the simulation, and how that volume compares to simulations with PIOMAS, TOPAZ, PIPS.

The trend in TOPAZ volume numbers reported by Bfraser (in open thread 6) are very concerning, and it would have been good to get some sort of assurance from the general circulation models that Tietsche used that this volume trend is going to slow down very soon, before they collapse into an ice free Arctic ocean.

Rob Dekker

Back to basics (and speculations about an ice free Arctic being near) :

As we have seen from posts by Prof.Malkovski, and by Gareth and others, the steep reduction in sea ice volume hints at an ice free Arctic (in summer) around 2016 (+/- 3 years).
One spectacular image of that decline in ice volume over the past couple of years is shown in the TOPAZ data, as posted by Bfraser in the "TOPAZ cryosphere" thread here at Neven's.

However, all these volume numbers are based on models (like PIOMAS, TOPAZ and PIPS). These models do get calibrated with actual satellite altimetry data but not very often. As far as I can see, most of these models have not been calibrated in years, so they could be way off for trends and certainly for absolute numbers of ice volume.

Now, Cryosat2 is up and running, and data is freely available.
I've been looking into the data processing tools, and found that with a Level 2, 3 or 4 tool from ESA, we should in principle be able to calculate ice volume on a monthly basis with a very high degree of accuracy (Cryosat is absurtly accurate in vertical resolution, and is designed to cover the Arctic and Antarctic for sea ice).

Does anyone know if Cryosat data has already been used to calculate sea ice volume numbers ? If not, then I'll look into setting up the tools to get that data out and up and running, preferably in a 'near realtime' application. At least, if I have some time left over after my baby is born (still not there).


Neven: your essay on economics is getting 5 times the hits of my latest Arctic update.

But it stands no chance against articles on penises and earthquakes...

I've clicked on your March update and left a comment. Great stuff to read, Patrick, like a breath of fresh Arctic air. I'll link to it in the first SIE update I'm going to write today.

Does anyone know if Cryosat data has already been used to calculate sea ice volume numbers ?

Rob, I've hard it will take at least a year to properly calibrate those numbers. If you think there's a way to do it quicker, I'll be happy to come and babysit for free. :-)

Artful Dodger

Neven, Rob: The CryoSat2 calibration phase is already complete. While the folks at UWash may take a year or longer to recalibrate their Model, don't misunderstand: CryoSat2 is producing Data, not Models. It is completely independant of PIOMAS calibration.

Rob, I'd go for it. There are some other folks here who can help too, if you ask. I'd just attempt to integrate it to the Topaz data, from which we are already getting per basin data, thanks to Bill. Cheers!

Rob Dekker

Thanks for the offer Neven. But I think I'll wait until that ice in the baby bath in your backyard has melted...and you removed the gravity-defying flower pot on the wood shelf above it ;o)

Regarding Cryosat, ESA released data starting last month. It's still under a disclaimer (could be adjusted later), but the mere fact that they release data to the public means that the data has at least passed a wide spectrum of scrutinisation. For our purposes, it should be good enough. I just wondered if anyone has already attempted to calculate sea ice volume from that data.

Peter Ellis

My understanding is that the various products do not currently include the "freeboard" field (I.e. the amount of ice above the water, and this will not be given for about a year. That time is needed to build up the model of the underlying water levels, which is subtracted from the raw height measurement to derive the freeboard - itself around 1/10 of the actual ice thickness.

Rob Dekker

Thanks for the confirmation that Cryosat is up and running, Artful !
There are several ways to get into the Cryosat data, and depending on the level you choose, it can be difficult or easy. My best bet right now is to go in with the BRAT interface, but we may have to just write a level 2 parser. Neither one of that should be difficult, but I just need some time to figure out the best strategy and build a area grid and volume estimator. Once I know what the best method is, I won't hasitate to ask for help.
But if any of you find a strategy before I do, please let me know.
Either way, I offer my company's linux servers to run the application and post it 'realtime' online.
This is gonna be fun !

Artful Dodger

ESA's ice mission goes live -- 22 November 2010

With the commissioning of ESA's CryoSat now complete, the mission has been officially transferred to the operations team. This milestone marks the beginning of the satellite’s operational life delivering ice-thickness data to understand the impact of climate change on the polar environment.

Artful Dodger

Peter, the freeboard problem was solved during NASA Icesat commissioning. It's simple bathtub physics.

Artful Dodger

Rob, I think NSIDC makes their arctic grid overlay info available on their ftp server. You definitely do not want to recreate the wheel here. Also, look into how NASA Icesat data (2003-2009) was arranged. Further, the U.S. Navy will continue it's Submarine ice thickness program, SCIEX in 2011. There are lots of data sources to validate, and existing frameworks to work with. Cheers!

Artful Dodger

This year scientists from the U.S. Navy’s Arctic Submarine Laboratory will collect critical environmental data while riding submarines on their routine Arctic transits as part of a program known as Science Ice Expeditions (Scicex)
Nuclear Subs Taking Scientists on Secret Arctic Missions

Rob Dekker

Artful, thanks !
Do you have a link to that arctic grid overlay from NSIDC ? We are going to need at least the map of the Actic (shoreline boundaries) and preferably a method to drop observations into the grid cells, plus a way to run algorithms between these cells. Also, ESA hints at all kind of frameworks (CODA and DORIS and such) none of which I have any experience with.
If you would build this system, which framework would you use and why ?

Artful Dodger

Icex Blog.

March 9 -- ICEX blog: This spring, we are conducting another submarine exercise in the Arctic with a large ice camp. A team from Applied Physics Laboratory, University of Washington (APL/UW) is already on the ice building the camp. Right now, the camp is drifting about 160 nautical miles north of Prudhoe Bay, Alaska at about 73N/147W...

Peter Ellis

From the official specs:

Field 31) Freeboard : SAR mode computed freeboard of the Sea Ice.
Initially (e.g. for launch plus 1 year) this will be set to ‘-9999’, when there is greater confidence in the knowledge of the Arctic basin sea surface height then freeboard will be computed in the L2 products. Freeboard will be zero when the SAR Discriminator (in field ) indicates ‘Ocean’ or ‘Lead’, and ‘-9999’ when the SAR Discriminator indicates ‘Unknown’. Note that Freeboard can be a small negative value when there is sufficient snow-loading on thin ice. Set to 0 in SARin and LRM modes.

I may be misunderstanding, but this seems to say the freeboard data will simply not be provided for the first year - i.e. you'd have to compute it yourself from the raw height data (grid the data, look for leads to find the sea level, derive the underlying water surface geoid, parse several months of data to determine appropriate corrections for tides, sea level air pressure etc...)

If I'm wrong, and they've already started including the freeboard data in the products a few months ahead of schedule, then great! Does anybody here have access to the feed to check whether there's actual thickness data coming through or whether it's still a bunch of "-9999" ?

Artful Dodger

Rob, i think you can find the NSIDC metadata from here:

Also bookmark:

If I was doing this, I'd try to integrate my product with NSIDC data since they have the longest continuous dataset, and new per basin area and extent products.

Also, most mainstream press reports quote NSIDC data. Finally, the prestigious SEARCH Sea Ice Outlook is a collection of Community predictions for Sep Avg Sea Ice Extent, based on NSIDC data:


If you had volume data, it'd be fish in a barrel...

Artful Dodger

Hi Peter, I see what you're on about... Certainly some of this work has been done, with the release of the "Ocean dynamic topography and currents" map on Dec 20, 2010:

Cryosat ice mission returns first science

It's been 11 months since Cryosat2 was launched (April 6, 2010) so the time should be neigh for freeboard data...

Thanks for pointing this out, Peter :^)

Rob, can you see the data now to answer this question?


Artful Dodger


I like the changes on the Arctic sea ice graphs page.

Nicely Done!

The comments to this entry are closed.