PIOMAS Version 2

As some of you have noticed the people at the Polar Science Center of the Applied Physics Laboratory of the University of Washington have been very busy upgrading their PIOMAS model. As stated on their website:

This time series of ice volume is generated with an updated version of PIOMAS (June-15,2011). This updated version improves on prior versions by assimilating sea surface temperatures (SST) for ice-free areas and by using a different parameterization for the strength of the ice. Comparisons of PIOMAS estimates with ice thickness observations show reduced errors over the prior version. The long term trend is reduced to about -2.8 10³ km³/decade from -3.6 km³/decade in the last version. Our comparisons with data and alternate model runs indicate that this new trend is a conservative estimate of the actual trend. New with this version we provide uncertainty statistics. More details can be found in Schweiger et al. 2011. Model improvement is an ongoing research activity at PSC and model upgrades may occur at irregular intervals. When model upgrades occur, the entire time series will be reprocessed and posted.

They have also upgraded their graphs, such as the one that features on the Daily Graphs page, which now has error bars:

There has been another big drop since the last update at the beginning of May. How this relates to other years can be seen on the other upgraded graph:

According to the model this year ice volume is much lower than that of 2007. If this is correct and weather conditions ressemble those of 2007, there is a very good chance, a certainty even, that there will be new extent and area records.

To the joy of many the scientists of APL/PSC have made the data used to generate these graphs available. Wipneus was very quick to produce a graph full of trends based on different statistical methods:

Wipneus also made a graph showing the PIOMAS monthly average Arctic ice volume with exponential trend:

This is great stuff. My gratitude goes out to the Polar Science Center for making all of this available and to the many commenters who analyze and theorize.

Again: I recommend reading Schweiger et al. 2011 for anyone interested in ice volume models.

Comments

[Wipneus]

With daily data now available, it is easy to plot ice volume in the same way as the familiar ice extent data. Using the same layout and colors as the IJIS graphics:
http://img714.imageshack.us/img714/4533/piomastrnd4.png

What strikes me that with volumes, it is not 2007 but 2010 that is the exceptional year. It is not clear that 2011 will be able to beat that.

[Gas Glo]

Wipneus, are you sure about 2010 being more exceptional? You need to ignore the 2008 and 2009 lines that came after 2007 to compare 2007 to 2006.

[Lord Soth]

Exceptional ?

Since area is less than 2010, and the ice is already thinner over most the arctic, and unless there is a dramatic change in the weather, 2011 should beat 2010 for area.

[Lord Soth]

"correction"

2011 should beat 2010 for volume.

[Wipneus]

Gas Glo:

I just meant that 2010 is the bottom, where in extent 2007 is still unbeaten.
2010 minimum volume was 68% of the previous record year 2007.
2007 minimum volume was 72% of the previous record year 2006.

[Gas Glo]

Wipneus, fair enough to consider 2010 more exceptional then. I tend to think in km^3 per year rather than %depletion (or simply the lowest) as a constant %depletion never gets to truely ice free.

CT has 2011 area more than 2010 and 2007 and I am not sure I want to trust the figures when there is clearly blue ice in large regions shown at under 70% concentration. Maybe it was like this in previous years?

IJIS seems the most alarmist of the various sources having extent and area at record lows.

Anyway, with volume at a record low and area perhaps not at a record low, we must be at a record low thinness for time of year. 18.647Km^3 /9.957Km^2 = 1.87m. So some melt needed before getting to that 1.1m limit.

[Artful Dodger]

Wipneus:

Thanks for the awesome response with the new graphs. Well done!

I'm sure you're just using a short-hand, but just to be clear PIOMAS v2 is not "daily data":

Firstly the results are monthly means, just like NSIDC SIE it's based upon.

Secondly, it is not data. It is an estimate (much like a hindcast) because it incorporates observations and models.

[Artful Dodger]

Gas Glo: you've used IJIS Extent in your calculation of sea ice thickness. The appropriate value to use is NSIDC monthly sea ice area.

[Wipneus]

Dodger:

The data that is now available is one value per day. The data file will be (if I understand correctly) updated each month.

I don't know whether the calculations were run for each day, or a few times per month and then interpolated. It is more then once per month, that is sure.

For the monthly average graphs I had to calculate this average for all (28,29,30 or 31) days in the months.

I agree on data being an ambiguous word here. It is data from the perspective of my little R scripts.

[Gas Glo]

Lodger, the last piomas volume number available is day 151 18.647. I make that 1 June so I was trying to use CT area for 1 June

[Artful Dodger]

Wipneus and Gas Glo:

Thanks for the clarifications. PIOMAS uses NSIDC Sea Ice Concentration (that's the IC-SST note at the bottom left of their Chart).

Still, CT Area is based on the same AMSR-E data so it should be a valid comparison, especially if we are getting dailies from PIOMAS v2.

Arctic Mosaic - Terra 4km True Color 2011/151 (05/31/11)

So Day 151 is May 31. CT Area was 10.0586348

I still prefer the use of NSIDC monthly SIA data, as daily CT Area is EXTREMELY noisy due to the Cloud effects on the synthetic radar.

Case in Point: Look at the CT SIA for 2011 from May 28 to June 2. Wild swings, and revised twice if I recall... That's why NSIDC prefers a 5-day moving average over CT's 1-day or IJIS's 2-day average for data release.

So, can we determine the Monthly average sea ice thickness for May 2011 with the data we have now?

Cheers, and have fun!
Lodger

[Rob Dekker]

Wipneus,
Thanks a lot for quickly producing the (absolute) volume graphs after the PIOMAS upgrade. I think it is fair to say that your work and FrankD's have been instrumental in visualizing the profound decline in ice Arctic sea ice volume. And for Neven to put up numerous blog posts on volume models. You guys are awesome.

Just curious : Did you do a comparison between the old numbers and the new ones ?
The old PIOMAS numbers are still on Neven's sea ice graphs under this link :
http://snipt.org/xwgn
I think that's your FrankD's table, right ?

Would be great to see where and how much the data changed now that the new PIOMAS numbers are out.

[Gas Glo]

May average is 20.173 k km^3
https://spreadsheets.google.com/spreadsheet/ccc?key=tF-Db0tyg3jtmSfeZ6PZqlA&authkey=CO2ht_8P#gid=0

not really useable like that (but if you want to copy to your own spreadsheet rather than downloading and unzipinng gz file).

NSIDC average area is 10.34 M Km^2

Average thickness in May 2011 1.94m

For 2010 21.203/10.48 = 2.02m

[idunno]

Hi Neven and Wipneus,

One glitch in the original graph, in the main article here;

Unless my eyes deceive me, the horizontal axis has slipped. It seems to give, for example, a reading for ice volume in September 2011. This is clearly not data.

[Anu]

What strikes me that with volumes, it is not 2007 but 2010 that is the exceptional year. It is not clear that 2011 will be able to beat that.
Posted by: Wipneus | June 17, 2011 at 16:06
------------------------------------------------------

Nice graph, thank you.

It seems that for this time of year (the endpoint of the red 2011 line), the order of the yearly lines determines the minimums in September.

That is, the lowest full graph for this time of year is 2010, then 2007, 2009, 2008, 2006, 2005, 2004, 2003, 2002 - and that is the exact order of the September minimums.

This would imply that since 2011 is lower than 2010, it will most likely wind up with a lower minimum in September. For the information in this plot, this has never NOT been the case.

(The September Arctic ice volume has been decreasing in exact order every year from 2002 to 2010, except for 2007 being the only year that "jumped" ahead out of sequence past 2009)

It might help to put some faint gray or dotted vertical lines on your graph to make time matchups across years easier to see.

[Wipneus]

Rob Dekker: Just curious : Did you do a comparison between the old numbers and the new ones ?

Actually I did, graphically:
http://img818.imageshack.us/img818/9798/bpiomasicevolumeanomalyz.png

In this graph (modified from the website) the line colored:
"black" is the new V2 anomaly;
"blueish" is the same but adjusted for insight in the effect of modelling;
"red" is the old (version 1) curve.

As you can see the versions are quite distinct. By coincident (?) the long-term linear trends of the old version matches the adjusted version of version 2.

idunno:

One glitch in the original graph

Sorry I checked and do no see it. The ticks in the graph are at the beginning of the year (1 Jan). The only points plotted in 2011 are for Jan, Feb, March,April and May.

[Bfraser]

@idonno. I think the 9/2010 data is located where september, 2010 would be (i.e. closer to 1/2011 than to 1/2010 -- which are the tick marks).

For comparison, I computed average thickness based on the Topaz data.

Based on Topaz 4:
2011 15.674 / 10.498 = 1.493m

Based on Topaz 3:
2010 17.151 / 10.417 = 1.647m
2009 18.547 / 11.020 = 1.683m
2008 22.729 / 10.853 = 2.094m

[Wipneus]

Anu:
It might help to put some faint gray or dotted vertical lines on your graph

More gray lines on this version:

http://img607.imageshack.us/img607/4533/piomastrnd4.png

[MikeAinOz]

@Anu, I like your logic, I'd been looking for a pattern and you seem to have solved my issue. 2007 was an outlier, I'll look again with enlightened eyes at the trends.

[Anu]

@Wipneus thanks, that's perfect. I don't need to use a sheet of paper on the screen anymore :-)

@MikeAinOz glad to help. I expect 2011 to be the lowest on PIOMAS record - it would be nice if 2011 jumped ahead by a few years too, because of the political ramifications, but we'll see. I also hope to see someone make Cryosat-2 thickness and volume information available as a quick check of the new PIOMAS info by the end of this summer.

http://www.ijis.iarc.uaf.edu/seaice/extent/AMSRE_Sea_Ice_Extent_L.png
http://iup.physik.uni-bremen.de:8084/amsr/ice_ext_n.png
http://ocean.dmi.dk/arctic/icecover.uk.php
The extent is falling below 2010 to record lows - it would be nice if it ends up below 2007 this summer. It will eventually, and the sooner the better, from a "people better wake up soon" perspective.

[Piotr Djaków]

PIOMAS from 1979:
http://gfspl.rootnode.net/klimat/piomas.png

(unfortunetaly it's not updated daily)

[Bfraser]

@Anu, it is really best if 2011 jumps a few years ahead?

There'd be no way to know that it was a few years ahead until 2012, and it would let the deniers point to "recovery" for a few years, just like they did following 2007.

[Anu]

@Bfraser
The summer will unfold the same regardless of what I wish for, but yes, I think politically it would be best if 2011 jumps ahead a few years, like 2007. Yes, we wouldn't know this was what happened for a few more years, but given an unusually large drop in PIOMAS volume for September, I expect CryoSat-2 will confirm this, and Arctic ice extent and area will also be unusually low, and all this will make the news and be in peoples minds as the 2012 US Election kicks into high gear next Spring and Summer. (And yes, next October, if 2012 is behind 2011, there might be some confusion and denier screaming about how everything is back to normal, but there will be only a month before voting, and the long year of alarm will outweigh the one month of confusion among the distracted, casual observer)

Perhaps it is provincial thinking, but I think the US 2012 Presidential Election is fairly important for near-term climate change efforts. If natural variability rings the alarm bell loudly for summer 2011, I think that will make it harder for Republican "Tea Party" candidates to pretend global warming is a "hoax" proven by some leaked emails. I don't want to see the GOP candidate running on a platform of opening ANWR and going to war to control Arctic drilling to secure 'millions of jobs', and promising coal-to-liquids of 3 million barrels of synthetic oil per day to keep the cost of gasoline lower for the large-SUV driving voter... Which is quite possible for this ignorant 'climate change is a scam' crowd.

[Janne Tuukkanen]

I really can't get it, how AGW has became so politicized. There should be no reason. In some alternate universe, where Soviet Union did not collapse, the global warming might be "a capitalist conspiracy to hinder the development of The Workers' Paradise." And American Right would pursue alternate energy sources in the name of freedom and free enterprises.

And those are many, who know the situation quite well, but for short term personal political gain are ready to sing the songs some of their voters want to hear.

[logicman]

"I really can't get it, how AGW has became so politicized."

Janne: ever since the rise of the guilds in medieval times, the 'producers of wealth' - i.e. the monied classes - have sought to protect against all argument their self-declared right to amass money. The argument that enterprise x shouldn't be made illegal has been raised in support of retaining many money-spinners, on the grounds that 'unwarranted government interference' with free trade will damage the economy and cost jobs.

Some examples of a money-making enterprise which - it has been argued - government should not obstruct:

state-sponsored piracy;
the tobacco trade;
slavery;
the handling of mercury by hat trade workers;
the handling of phosphorus by small girls in the match trade;
the use of children in mines and factories;
low cost gin, aka 'mother's ruin;
cellulose nitrate film stock - a cause of many cinema fires;
tetra-ethyl-lead in gasoline;
whaling;
sealing;
deforestation;
and many hundreds more.

Here are some things which were opposed on the grounds of 'unwarranted' cost:
paved roads;
enclosed sewers;
piped drinking water;
basic education for small children;
public baths;
mesh guards over moving machinery;
fences alongside railways;
roofs on railway coaches;
windows in railway coaches;
seats in railway coaches;
brakes on railway wagons and coaches;
loading restrictions on cargo ships;
sufficient lifeboats on passenger ships;
insulation on electric cables in public spaces;
enclosed electrical switchgear;
and many hundreds more.

In all of the cases I have researched, false arguments were presented to counter the scientific evidence of real or projected harm.

The only difference between the old political arguments and the new is in the numbers affected. Way back when hygiene was a novelty, if any single community failed to adopt clean water and proper sewage systems then the resulting epidemics were unlikely to become completely global.

But now, just in case our worst fears are realized: failure to adopt measures to control GHG emissions could lead to a global population crash.

Maybe human megadeath is nature's way of reducing GHGs. Or maybe I am just frightening myself with the products of a fevered brain? I hope I am wrong. Time will tell.

Back on topic:
It looks likely that the NWP and Eastern route will both be navigable earlier this year than 2010. In most of the areas defined by Cryosphere Today, the decline in ice extent is very rapid. The only area which lags well behind 2010 is Nares Strait. That ice is already well on the move.

The normal trend is for ice to move across or around the Arctic basin and to either exit via Fram Strait or be pressed against the coasts along the western lands. Ice flow through Nares and the Canadian Archipelago would relieve some of the pressure which builds ice thickness and conserves multi-year ice.

The new PIOMAS graphs confirm my suspicions that ice volume is much lower than indicated by most estimation methods. Take a look at what I call 'panel ice' - panels of blue separated by white lines in bays and fjords. All over the CA the ice tends to be bluer than at this time last year. Bluer tends to imply thinner. Thinner implies ... Heck, we are talking volume here and I'm writing this for an intelligent audience. Go check last year's MODIS images of the NWP area. Now guestimate this year's volume. Now debate quietly amongst yourselves. :-)

My projection is that by September, by far the greater part of remaining ice will be the first year ice which has failed to melt or to exit the Arctic. As to extent, as the freeze commences ice will remain in Baffin Bay, Nares Strait and the passages and fjords of the CA. That ice may well comprise from 33% to 50% of the remaining extent. Somewhere in the main pack will be the last remnants of the Ellesmere ice shelf - finally detached from the coast where it has been a feature for 3 to 5 thousand years. Our planet's night storage cooler is broken. Times will soon start to get much more interesting.

[FrankD]

Woot!

As I have been hoping for six months, Tamino has finally tackled sea-ice volume and thickness.
http://tamino.wordpress.com/2011/06/23/sea-ice-3-d/

He's used the new data PSC have release so there are differences in the detail, but the methods and results will be amusingly familiar to regular readers here. No h/t though :^(

Still, imitation is the sincerest form of flattery... :^P

[Dominik Lenné]

For comparison of the old, graph extracted and new, upgraded piomas data based plot look here:
http://imageshack.us/photo/my-images/696/piomasvolumeoldnewversi.png/

[Neven]

Thanks a lot, Dominik! I have saved this one.

[Artful Dodger]

PIOMAS chart for June 30, 2011 is out:

http://psc.apl.washington.edu/wordpress/wp-content/uploads/schweiger/ice_volume/BPIOMASIceVolumeAnomalyCurrentV2.png

The v2 anomaly is down to -9.75 million km^2.

[Tor Bejnar]

-9.75E6 km^3 - one more extreme that supports the possibility of a new minimum SIA in September.

[Tor Bejnar]

woops -9,750 km^3 anomaly…

[Wipneus]

Updated absolute ice volume graphics, here is the daily ice volume:

http://img97.imageshack.us/img97/4533/piomastrnd4.png

Monthly mean ice volume with exponential trends:

http://img835.imageshack.us/img835/2294/piomastrnd2.png