PIOMAS July 2013

Another month has passed and so here is the updated Arctic sea ice volume graph as calculated by the Pan-Arctic Ice Ocean Modeling and Assimilation System (PIOMAS) at the Polar Science Center:

The slow start of this year's melting season has had a dramatic effect on sea ice volume. After going steadily upwards from a very low volume at the start of the year, catching up with 2011 and 2012, this year is now well above those two lowest volume years (by 1729 and 1720 km3 respectively) and even slightly above 2010 (126 km3).

Although I knew that the bad weather for ice decrease would keep 2013 above 2011 and 2012, I didn't quite expect such a big gap. It never hurts to emphasize that PIOMAS is a model, and even though it's constrained by observations, the uniqueness of this season so far might cause some parameters to get sidetracked.

Some might say I emphasize this now because as an alarmist I'm not pleased with the results, but I don't believe this is the case (the results are bad anyhow, even if better than previous years). I emphasize that PIOMAS is a model on a regular basis, especially when PIOMAS tends to go low, because many people assume that the PIOMAS graph is showing observed data. It took me almost three years to finally be able to say last year: yes, the ice is as thin as PIOMAS says it is, because area/extent keep decreasing steadily, whereas weather patterns imply they should stall.

I always feared that PIOMAS was overestimating volume loss. For the first time I fear the opposite. That sounds like a conclusion, but it isn't, as I have more graphs to show to complete the PIOMAS update for this month.

Here is Wipneus' version with the calculated "expected" 2013 values (dotted lines), based on the same date values of 1979-2011 and an exponential trend.
A caveat from Wipneus: "Note that the statistical error bars are quite large."

The trend line is well above the dotted expectation, and if my eyes don't deceive me, above the confidence range as well.

Still, the anomaly from the linear trend is about to leave 2 standard deviation territory, continuing the pattern first seen in 2010:

Average thickness is still tracking as low as last year, and that's because CT sea ice area numbers are also remarkably high compared to last year. I crudely calculate the ice pack's average thickness by dividing PIOMAS (PI) volume numbers with Cryosphere Today (CT) sea ice area numbers. If they're both much higher than last year, average thickness remains just as low, of course.

Here's average thickness for June 30th in metres, with change from last month between brackets:

• 2005: 2.47 (+0.13)
• 2006: 2.44 (+0.13)
• 2007: 2.19 (+0.03)
• 2008: 2.36 (+0.07)
• 2009: 2.11 (-0.03)
• 2010: 1.91 (-0.04)
• 2011: 1.75 (-0.11)
• 2012: 1.71 (-0.12)
• 2013: 1.71 (-0.08)

If you want to have a look at the data yourself, you can download the spreadsheet I use and update from GoogleDrive.

Average thickness is on a par with last year. Just like all years since 2010 average thickness has now started to drop, contrary to the years prior to 2010.

The thickness graph from the Polar Science Center is again showing 2013 slightly above 2012 (just like last month) [edit: that's because the graph hasn't been updated! Will replace the graph once it's updated; N.] [edit2: graph has been updated, thickness is a bit higher than last month compared to 2012; N.]:

Right now I'm not sure what to make of that big gap between 2013 and previous years. If it's real, this could mean that 1) the start of the melting season is of paramount importance, and 2) 2013 is a recovery year, the first one in 5 years, volume-wise. More or less the same conclusion I drew a couple of days ago, after reviewing maps showing atmospheric circulation patterns in May and June.

But I think I'm seeing a huge melting potential out there, despite those slow start weather conditions, with holes in the interior of the ice pack and lots of discoloured, thin ice on the fringes. We mustn't forget that this melting season started with the most first-year ice on record. A bad start can postpone much of that ice to melt out, but all bad starts come to an end.

I believe there will be a rapid area/extent decrease in the coming weeks, and it remains to be seen how PIOMAS is going to interpret that. If, like I said, some of those parameters are sidetracked due to the unique pattern of this melting season so far, they just might get back on track once the melting season throws off its mask.

The coming weeks could be the most interesting we're going to see this melting season.

Comments

[Chris Alemany]

It is only going to get more and more interesting as things progress. Thank you for another great update.

[dominik lenné]

Somehow I think since long, that the average thickness is not such a good indicator of whats happening, though very easy to calculate. It should be possible to find a better one.
E.g. divide the area into zones and give graphs for the mean thickness within them. One ends up with a couple of lines, some of them zero for a part of the year, some not, giving an impression of the time development of the thickness field. Or a thickness over cumulated area graph for each month.
This is of course much more complicated to achieve, and I don't know wether i have the skill and the will power to do it...

[Rlkittiwake]

Right now on CT we're where we were in 2008, which went from being "a recovery" to almost tying 2007.

In the last 6 days we've lost a million square kilometers.

If the next 6 days continues the trend, we'll be 200K above 2012, and if the rally lasts for 12 days, we'll be 100K above 2012.

[Tor Bejnar]

2013 is following the "new" PICT path that 2010 started, with calculated (on numbers half modeled and half measured) thickness maxing in June - so at least in this one metric, 2013 is "just like" 2010, 2011 and 2012.

It is so nice to be able to depend on something being stable!

[Wipneus]

Am I the only one that noticed the thickness graph did not update (ends in May) at all?

[Neven]

No, someone else noticed as well and notified me. I edited the text and will update the graph as soon as there's a new one.

[Glenn Tamblyn]

Does PIOMAS produce basin level results? What would be interesting is seeing which basins have done what. Also, just eyeballing the main graph, it looks like most of the difference occurred in May. June has paralleled to 2012 line fairly well.

[jbyrd]

In agreement with the more pessimistic (optimistic?) opinions shown here, I predict the 2013 extent value will cross the 2012 line between the 15th and the 21st, but the curve will level off shortly thereafter and come in just below 2012.

Two other observations/questions though. The interface between cold surface air and relatively warm sea water is the starting point to analyze the situation. Taking into account that by far most of the increased global temperature is being stored in the ocean, and given the immense heat sink capacity of the ocean, while SST's are relevant as a benchmark, the important thing is the actual increase in stored heat over the historical norm. After all, the Canadian arctic continues to have somewhat below normal temps, but for those of us who were watching, Great Bear Lake just went from completely frozen over to essentially ice free in a week. Ice looks about the same at -5 as is does at 30 below, but is has a lot of warming to do to become water from the lower value. In other words it appears that most of the melting is coming from underneath, due to increased heat content of the sea.
The other thing might be a little more controversial. While a longer melting or freezing season can be measured and analyzed
statistically, the beginning and ending are natural occurrences. The year as we know it is an human construction, set up for our convenience. If the melting is significantly from beneath then certainly currents are involved, if currents are involved then the lunar cycle is involved.
If the lunar cycle is involved, there is an 11day difference in the length of the lunar year and the solar year (we'll leave out those variations), so there is at least that much variability in the system.

It certainly is interesting however it works though.

[jbyrd]

Oops, the previous post was supposed to be under "So how slow was the start". Sorry about that,

[wanderer]

Andy Lee, something for you:
http://io9.com/130-years-of-global-temperature-data-converted-to-musi-634152387

Climate artsy ;-)

[Jai Mitchell]

What is interesting to me is that the SIA is paralleling 2012 when the average temperature data shows a much cooler melt year: http://ocean.dmi.dk/arctic/meant80n.uk.php

I think this comparison of previous years is helpful. https://sites.google.com/site/arcticseaicegraphs/concentration-maps/sic0707

If temperatures do stay above the average for any period of time the volumes will crash.

[Mickael Delatre]

About ice thickness measurements, I use this occasion to point you to some recent works from ... seismologists.
Yup, seismologists.

The idea is the following : seismometers are able to measure sea ice inertial oscillations. These oscillations are linked to thickness and extent. The advantage is that it is not a local measurement but an average over a wide area - the default being that this is not a local measurement, and thus you loose resolution

Here is a link to the abstract : http://hal.archives-ouvertes.fr/hal-00752413/

My apologies if you already cover that. But it could be a way to improve PIOMAS efficiency.
Another way with seismometers to obtain measurements on 1 km² area could be to measure the flexural waves travelling through sea ice, these waves are directly linked to the ice thickness and quality. But a seismometer costs some money, and the measure (although not exactly local) does not cover much ...

[Neven]

Hi, Mickael. I never covered this, but remember reading about how measuring the effect of waves on seismic activity could say something about the overall thickness of the ice.

[michael sweet]

According to the most recent NSIDC report, the thickness of the sea ice measured by CRYOSAT was 8% less in March 2013 than 2012. AIR, last year CRYOSAT and PIOMAS were in closer agreement. Will the greater sea ice area this year mean less energy is absorbed by the ocean for melting later in the season?

[TenneyNaumer]

I am just going to go out on a limb here and state that PIOMAS cannot handle rotten first-year ice very well.

If this year had more first-year ice than any previous year, is it likely that there is more volume, especially when we see holes spread out all over the Arctic?

[NeilT]

When I went to visit the Discovery museum in Dundee, I listened to the AV rendition of how Discovery was recovered from the ice locked situation in Antarctica.

They sent a whaler with a captain who had a lot of experience from recovering boats from locked ice.

What they did was to drill holes in the ice. Drop barrels of gunpowder down the holes, blast the ice open and let the seawater do it's work. Yes there were sawing parties and ships butting the ice, but the last two miles remained solid.

Six weeks after blasing the ice holes, the last of the ice broke up and floated away, allowing Discovery to come home.

It was the action of the sea on the weakened ice which broke it up and destroyed it.

I'm assuming that some 6 - 8 weeks after the opening of so many holes in the CAB, there will be consequences which defy normal observed patterns. Including melt speeds and melt locations for time of year.

[Pete Williamson]

I don't know if this is old news because there are no publication dates on this PDF but it seems like an important comparison of Cryosat2 and PIOMAS.

http://www.personal.soton.ac.uk/pgc1g08/grl50193.pdf

What I'm taking from this is there is a large loss of volume in recent years. Also that the PIOMAS model underestimates both the rate of volume loss in the melt season as well as the rate of volume gain in the winter.

(The absolute volumes for PIOMAS in Fig3 of this paper don't seem to match with what Neven has here.)

Anybody know why yhe summer data for Cryosat2 is not included? Does the satellite not collect data then?

[Pete Williamson]

Another curious point from that paper is in regard to the freeze season. If you look at the end of the freeze season in Figure 3 (that's the right hand side b) d) f) h)) then the loss of MYI is obvious. But actually most of the the extent of the arctic basin is covered by FYI, that's true in the period before the great clear out of MYI in 2007 as well. And what's noticeable about that area is the ice is in basically the same condition at the end of the freeze season for the whole period. The point I'm getting at is the for all the talk of rotten ice and so on at least at the beginning of the melt season the ice is at basically the same thickness that it's been at for a decade for much of the basin (i.e. those areas were FYI has been found).

I think to summarize what this seems to be saying is that for all the warming that has occurred in the NH and the many processes that amplify that warming in the arctic and cause the ice to melt, it has yet to have a large impact on the ability to make ice in the central arctic in winter.

As a counterpoint obviously the melt season has been much more efficient in melting ice and clearing out MYI in recent years. And at the fringes of the arctic then there is probably more of an impact in the refreeze season as well.

The difference in how the two seasons are impacted is fascinating.

[Clare]

Further to the drilling of holes by the Discovery crew mentioned by NeilT:
I don't remember where I read this (some sea ice textbook I think) & I made notes about it because I am interested in 'holes' of all sorts as patterns for my still-in-the-designing-stage arctic inspired artworks -
that the vertical brine channels in 1st year ice acted as weak points in the ice, so a row of them could act as a 'zip' to 'unzip' the ice.
But I'm no expert & maybe this is not of sufficient scale to be significant here?
Back to my corner to lurk!
Clare

[Rlkittiwake]

Pete, there may not be an enormous change in area or extent at the height of freeze, but there has been some change and the trend is downward. The volume trend, however, predicts an ice-free arctic sometime within the next 30 years.

[Rlkittiwake]

*an ice-free arctic IN APRIL sometime within the next 30 years.

[Rob Dekker]

Hi Pete, good observation !
Couple of notes :
1) When referring to panes b) d) f) and h) you probably meant Figure 1 from the paper, not Figure 3.
2) PIOMAS records a very significant drop in MYI between the average of 2003-2008 and 2010 specifically, which make up pane a) and b).
http://neven1.typepad.com/.a/6a0133f03a1e37970b0192abe2b76c970d-pi
So there does not seem to be any apparent contradiction between PIOMAS and Cryosat 2 on Figure 1.
3) Indeed most of the Arctic Basin on the Siberian side is FYI, or second year ice, probably through much of the past decade.
With 2013 showing very significant 'holes' in that FYI, the question for the Arctic Basin is if these holes will widen enough to cause an ice-free NP later this summer, or if they will close and create a 2009-like situation.
4) Regarding the Central Basin, and the confirmed loss of MYI in that area and "the difference in how the two seasons are impacted is fascinating." I completely agree.
I also think that the moment the difference will show up unambiguously, that it will be too late.

After all, the Central Basin is the last resort to prevent an ice free Arctic summer.

[Andy Lee Robinson]

Thanks Wanderer!

I had this idea for a long time now, and decided to have a go using PIOMAS volume data after discovering a midifile class for perl. A day later, I uploaded the results, using 500km³ per semitone (what a unit!).

Piano version:
http://youtu.be/bL7D5eRlDxk

Worrying a bit that the results were a little acoustically unpalatable, I made a spectral version modifying comb filters:
http://youtu.be/KfJHmyKYpeE

I didn't advertise them much, but I'm surprised that none of the big sites picked them up like the cello version.

[Clivepmitchell]

I find ice extent confusing. If I fill my bath with near-freezing water and arrange ice cubes to evenly cover 16% of the surface, NSIDC would say it's 100% cover. If I put them in one slab, the coverage is 16%. If I cover 2% and evenly spread out the rest I have 2% cover. Same ice volume, same thickness, same area exposed to the sun. 2%-100% extent.

I now copy CT images into a folder to make a manual "video" with picture viewer to give me a better idea of what's going on.

Or not.

[LRC]

Extent is the area that you can find ice in with 15% being the minimum amount needed to be counted. Note: Some formulas use 30% as a minimum.
Area is the area that you get if you put all the ice in one spot.
Volume is total amount of ice.
The science is not 100% accurate and that is why different groups come up with different numbers.

[Wipneus]

The thickness graph is updated now.

[Neven]

Thanks, Wipneus. Post has been updated.

[Gerg]

Thanks; here's yet another (very slightly different) view: http://gergs.net/2013/07/northern-sea-ice-update/

[Superman]

I understand the interest in tracking the ice melt, but I'm wondering if we're operating on the right time scale. The net energy trapped by the greenhouse gases is allocated among at least four sinks: atmosphere, ocean, land, endothermic processes (like melting ice). There is no rule I see that requires the allocation to be the same for each sink every year. We have seen that the atmospheric temperature has been in somewhat of a plateau for more than the past decade while more energy was siphoned off into the (deep) ocean. If this blog had been devoted to following the atmospheric temperature at short time scales, there would have been much disappointment, and the blog would have provided the fodder for the deniers. Is it really that much different for the ice; isn't the long-term trend the real problem, and don't we have a pretty good idea of where that is going?

[Clivepmitchell]

Thanks, LRC. My confusion is the reason for measuring extent with a fixed cut-off at all as it appears to be wildly independent from both ice volume and exposed sea area and thus has little bearing on thermodynamic processes but more to do with localised weather.

[Jeremy Poynton]

"Some might say I emphasize this now because as an alarmist I'm not pleased with the results"

Give yourself away there don't you? Would it not be far better for the planet and its inhabitants if what we experience climatically were all down to natural variability, with maybe a little bump from co2 (the current **hypothesis**) than that we all fry? Even if it means the premature death of your cult's beliefs?

You need to take a long hard look at yourself, author.

[Fufufunknknk]

Wow, Jeremy, not to respond to trolls on a fairly troll free forum (which must frustrate you) but you really need a reading comprehension course.

[Neven]

Give yourself away there don't you?

Yes, I did! Stupid me. :-D

Would it not be far better for the planet and its inhabitants if what we experience climatically were all down to natural variability, with maybe a little bump from co2 (the current **hypothesis**) than that we all fry? Even if it means the premature death of your cult's beliefs?

Of course, it would be. I personally would cry for joy and book a couple of plane tickets to see the world.

But what if what we experience climatically isn't all down to natural variability? What would be best then? Wouldn't it be best if a critical mass of people wake up to that danger and act on it? Even if it meant the premature death of your cult's beliefs?

[Mikeymannblaster]

Wonder what the Vikings thought as they rowed their boats around a mostly ice free Arctic Archipelago hundreds of years ago?

Do you think they worried about the Carbon emissions from their herds of dairy cattle on their farms on the west coast of Greenland?

[Thank you for your brilliant research and keen sense of humour. No more trolling, please, thanks; N.]