By Steve Goddard
In my previous PIOMAS verification piece I noted that PIOMAS volume trends don’t correlate with extent trends over the last three years. PIOMAS has gone down since 2007, while extents have increased.
The diverging trends themselves prove nothing, because it possible (but unlikely) for it to occur. This time I directly compared calculated volume measurements, which should be more definitive.
In their current graph (below) PIOMAS shows a record negative volume anomaly.
This appears to be incorrect, because we can see from the PIPS blink maps (below) that thicknesses were generally lower on this date in 2008. If the visual impression is correct, it would be impossible for the current anomaly to be greater in magnitude than the 2008 anomaly.
Quantifying this further, I numerically integrated May 31 pixel count vs. thickness since 2000.
As you can see, May 31, 2010 volume is currently higher than any year since 2006. It is also higher than 2003. Remember that 2003 had the highest minimum of any year in the JAXA record.
http://www.ijis.iarc.uaf.edu/seaice/extent/AMSRE_Sea_Ice_Extent.png
So why does PIOMAS show a record volume anomaly at present? Something is wrong either with PIPS maps or published PIOMAS volume data. PIOMAS trends are widely quoted and it is important for them to be correct.
Willis has also made some interesting observations.
Discover more from Watts Up With That?
Subscribe to get the latest posts sent to your email.
Bud, stop warping things around. I never denied 3.0 was being developed, I said it wasn’t yet deployed, they were still using 2.0. THE FREAKING US NAVY SAYS 3.0 is in final development.
Old news !
Open the html source code !
The PIPS 3.0 is presently going through its final development… –and will begin its adaptation for operational use in 2003 with a scheduled transition into operational use in early 2004.
Today there is a flood alert for the Colville River Delta thru noon June 2, as reported on the Anchorage news (KTUU ch2). All due to snow melt. High today on the Arctic Slope was around 38F. Highs first consistently went above freezing last week. There is still a lot of snow in Turnagain Pass (elev +/- 2,500′) on the Kenai even though temps have been more like July than May.
Whatever may be happening in the Arctic ice extent or volume up until end May is not related only to temperature if Alaska arctic shoreline observations are any guide.
2008 summer was exceptionally cold in AK, 2009 was warm, and 2010 is starting warm.
from previous post:
“Well, Steve you were wrong about Venus, but do not let that stop you here.”
Steve post was very good argued, different from the book… But what I’m worried about is how is Venus so cold, not so hot. I think must be becouse CO2 dont in fact is a gas, is a supercritical fluid, in so much pressure. http://en.wikipedia.org/wiki/File:Carbon_dioxide_pressure-temperature_phase_diagram.svg
Terry Jackson,
Barrow temperatures are below normal, below freezing, and they are expecting a few inches of snow overnight.
http://www.wunderground.com/history/airport/PABR/2010/6/1/MonthlyHistory.html
http://seaice.alaska.edu/gi/observatories/barrow_webcam
Bill, what mathematical errors do you refer?
Steve,
“The complaints about PIPS 2 accuracy are a smokescreen.”
But your volume measurements are way out of whack with those derived by the Navy team from PIPS 2. Even if the Navy’s model is accurate, your values are not. In fact a comparison of your numbers and the Navy’s for May show a high NEGATIVE correlation, -0.6. On that basis, if you claim the ice volume is increasing, a proper calculation from PIPS 2 may well show the opposite.
jorgekafkazar says:
June 1, 2010 at 6:22 pm
Be reasonable. If someone sets themselves up in a very public way and acts like (or indeed is) a complete idiot, they are fair game. It’s just a bit of fun, ease up!
If development of PIPS 2.0 stopped in the early noughties, I would not bet my shirt on it being the more correct. From the Polar Science Centre
http://psc.apl.washington.edu/ArcticSeaiceVolume/IceVolume.php
“PIOMAS has been extensively validated through comparisons with observations from US-Navy submarines, moorings, and satellites. The example on the left shows a comparison of PIOMAS-derived ice volume anomalies with anomalies measured by the NASA ICEsat Satellite. More details on the model, assimilation procedures and validation results can be found here”
“here” = http://psc.apl.washington.edu/zhang/IDAO/index.html
From: Warm on June 1, 2010 at 11:29 pm
*snicker*
Yeah I see it. That part you emphasized is what is known as a comment. It doesn’t display, therefore it doesn’t count.
First, look at how the comment was done. Looks like an edit, they just chopped off a part and left it there to show what the previous version was.
Second, there is proof that is inaccurate. Back at the other article, R. Gates tried to cite this as proof of 3.0 being operational, specifically pg. 33:
http://www.hycom.org/attachments/101_F.Bub.pdf
It actually is an old powerpoint presentation dating from around 2005, perhaps 2004, showing “works in progress.” The referenced chart shows 2.0 still operational in 2005, with “Development and Transition” of “PIPS 3.0 (G-NCOM)” in 2005, with D&T of “PIPS 3.0 (G-HYCOM)” in 2008.
So you have a comment in some HTML talking about stuff that will be happening in 2003 and 2004, going against a released document saying the G-NCOM part will happen in 2005. Clearly that comment, which doesn’t count as it’s hidden, is older than the non-hidden released document. Thus the comment indicates nothing, save that possibly the earliest scheduling of the PIPS 3.0 deployment was, shall we say, somewhat optimistic. 😉
Heh. Some published pretty pictures from 2002 showing how great the output looks, a web site where the pages have a Last Modified date in 2003, an old powerpoint presentation saying it’ll be finished and installed in 2006… If PIPS 3.0 was meant to be a commercial product, it would have been written off as vaporware years ago.
I find the warmist comments here highly amusing – these are people who: i) will swear blind that tree rings are the answer to determining historic climates, and ii) ignore how GISS and CRU have manipulated historic temperature figures. Yet, these same people question one of the world’s most scientific systems for forecasting ice thicknesses because of minor inaccuracies, which according to the references in the posts here appear to cancel each other out.
My bet is they work as a minor cog for some government or quasi-government organisation.
Steve, why do your ice volume estimates differ from the ones actually published by PIPS2? Are you not actually downloading the thickness data and computing volume as a function of ice area and area per pixel? You have to make sure you are doing the computation correctly if you want to compare with PIOMAS.
Both PIPS and PIOMAS are models that assimilate real time data to produce ice thickness. But they are only model predictions, not real data so it’s not wise to put too confidence in them. More actual thickness data are needed to verify both models.
There is some data from radar altimetry from ENVISAT that can be used to estimate ice thickness up through present (though it doesn’t cover the entire Arctic). Soon with CRYOSAT more validation will be available that can be used to see which model is more realistic.
In the meantime, the ice extent remains the best metric, and given the current rate of decline and the fact that it is now the lowest in the last 31 years, speaks to thin ice that is easily melting out even though Steve continues to point out that SSTs are cold and air temperatures are cold (yet still warmer than normal).
Why why why are Anthony etc giving ammo to the trolls on a daily basis by banging on about models and estimates?
If we do not trust models as scxeptics, why are we suddenly going down the road of comparing them against each other???
What really worried me is that everyone is ignoring the obvious. Look at the AO/NAO data and compare it to the historical ice levels generally. They correlate.
Its the NAO stupid, and its just gone negative.
WOW.
I found something truly impressive.
Searched for PIPS info, found this presentation (note: FTP link, not HTTP) that apparently was used for a conference, from NOAA, NSIDC et al. Note on pages: “Arctic Frontiers – Tromsø, Norway – 24-29 January 2010”. No mention of PIPS 3.0, just 2.0, it’s still interesting to look at. In it was mention of an ice forecast system that provided thickness charts, TOPAZ. I Googled, found it… Wow.
Beware, those with slow connections. Here is the home page. It has an animated GIF, 8.5 Mb. Here is the animation, if nothing else you MUST download and watch it.
From the beginning of 2010 to the end of April, it shows ice thickness. In high resolution. Height represented by a spectrum, not a few fixed colors. You can watch the Arctic sea ice grow, twist and turn, move and flow, like a living creature.
It is beautiful.
Forget PIPS 3.0, whatever it looks like, whenever it’ll show up. It is the past, it is history. I knew that when I saw the specs saying it’ll use SSM/I data when AMSR-E is newer and better.
TOPAZ is the future. And the future looks marvelous.
Now, can someone explain how to get data from it? The “Latest Forecasts” section is not user-friendly for me. The page says it also has plots for temperature, salinity, ice transport… If only I knew how to get to them!
If I remember correctly the so-called ice experts predicted 2008 ice extent would be lower than 2007. They also predicted 2009 would be less than 2008. I suspect the PIOMASS model had something to do with these predictions. It appears they are once again predicting a large melt. Could this be 3 strikes and you’re out?
PIPS 2.0 may not be accurate to the actual ice thickness, but this post isn’t a comparison to actuals, it’s a look at the trend indicated by the model. It’s an apples to apples comparison of the model results. AGW proponents LOVE MODELS, all of the predictions of dire future temperatures are based on inaccurate models. So much so, that a group of verifiably inaccurate models are grouped and averaged to form a trend.
The PIPS 2.0 data indicates a trend toward greater ice thickness between 2008 and 2010. That is a simple fact, demonstrated by the blinking plot. This is to some extent verified by the reference paper that published increased average ice thickness –from 2.4 to 2.8 m.
PIPS 3.0 data can be compared to other PIPS 3.0 data to evaluate a trend. All you PIPS 3.0 fans; please post the link to the PIPS 3.0 model data for the last 5 years, so we can see that model’s trends.
Question for the PIOMASS fans. If the average ice thickness increases by 15% and you have a recorded short term maximum of area and extent, how do you get a decrease in volume for that month? Shouldn’t the monthly volume be greater due to the simultaneous increase in BOTH of the underlying parameters?
Jacob Mack
You make a blanket statement that he is wrong on Venus and as “proof” link to a page in a book? A page which says little or nothing about it and certainly doesn’t prove anything.
A whole raft of criticism from the AGW fanatics above and only one valid question or point to which they offer no real data, ice volume.
Btw, Anu, not to fade your laurels or anything, but it is worth wondering, given the closeness of the results you’re pointing at, what impact if any JAXA’s recent change in the May 20-June 11 results could have had here. They did this recently to get rid of the “blip” (which I totally support –it caused the same tiresome question to be asked over and over again about what was causing it).
Here’s how they describe it:
“Previous version of data processing had made an erroneous blip of sea ice extent on June 1st and October 15th which was seen in the graph of sea ice extent as a small peak on these dates. We improved the processing to make the graph much smoother. The apparent blip had arisen due to a switching of some parameters in the processing on both dates. The parameter switching is needed because the surface of the Arctic sea-ice becomes wet in summer due to the melting of ice which changes satellite-observed signatures of sea-ice drastically. By this improvement most of the sea ice extent values are not affected at all except for the period of May 20-June 11 and October 8-26 of each year.”
It sounds very much like they replaced a “all on one day” adjustment with a smoothed over several weeks adjustment. It’s not real clear to me what impact that would have relatively. Visually, they appear to have done it to the historical data too, so maybe none.
stevengoddard says:
June 1, 2010 at 8:43 pm
Anu,
Regardless of the accuracy of PIPS predictions, their models are continuously updated with available real time data. The complaints about PIPS 2 accuracy are a smokescreen.
I think it’s good that people are finally talking about Arctic sea ice volume. Your approach is fine – ice thickness times ice area is ice volume. Things like the grid resolution, 0.28°, varies from 17–33 km depending on the location of the grid square within the spherical coordinate system (Figure 3), ice concentrations, etc. are second and third order effects:
http://www.tos.org/oceanography/issues/issue_archive/issue_pdfs/15_1/15_1_preller_et_al.pdf
The details are interesting, but not crucial at this point.
Too much time is spent attacking data sources – HadCrut, GISS, NSIDC, etc. If we can all agree that Navy PIPS data is OK, then we can start talking about what the data shows, not the “motivations” of the organizations producing the data.
richcar 1225 says:
June 1, 2010 at 8:57 pm
Anu,
The pip 3 thickness map illustrated in the pdf you referenced appears to show a uniform thickness compared to the pip2 data. This seems suspect considering that the polar ice group emphasizes the use of the age of the ice to make thickness estimates..
Notice that the color scales are different in the PIPS 2.0 and 3.0 images (one steps by 0.25 meters and maxes out at 5.00 meters thick, the other steps at 0.5 meters and maxes out at 8.0 meters thick). I wouldn’t say it is “uniform thickness” in the PIPS 3.o image – the thickness ranges from 0.0-0.5 meters to 3.0-3.5 meters thick. And this was right after the record summer melt of 2007.
And which “polar ice group” are you talking about?
I wouldn’t say the thickness estimates “seem suspect” – this is Navy PIPS data.
As pointed out above PIPS is a model based predictive system. However it can also be used to hindcast the sea ice volume using reanalysis of historic data such as the NAO data. This is evidently how the sea ice volume graph on the retro page of the PSC site was calculated.
http://psc.apl.washington.edu/zhang/IDAO/retro.html#Satellite_ice
If we believe it then the 10 % decline per decade of sea ice volume the last thirty years merely melted the volume gained from 1955 to 1969 when the NAO ran negative. It is interesting that the model data predicts that the ice grew at three times the rate during this period compared to the recent (last thirty years) melting rate. The hindcast begins in 1948 but there seems no reason that it can not be hindcast back to the beginning of the NAO data.
Looking at the historical data I would estimate that the sea ice volume currently predicted is normal and the sea ice volume melted the last thirty years was unusual in the sense that it resulted from the historic low NAO from 1955 to 1969.
http://www.cgd.ucar.edu/cas/jhurrell/nao.stat.winter.html
Anu
Glad that you agree that the PIPS data is a reasonable source for comparison. That makes it possible to have an intelligent discussion.
Tom P
June 2, 2010 at 12:58 am
Odd math that is that Arctic ice has rapidly increased since 2007 but you say volume could have decreased. Odd that. You fell in the rabbit hole?
richcar 1225 says:
June 1, 2010 at 9:09 pm
Anu,
Based on the clear correlation in the past of the NAO and the sea ice volume. Do you think sometime from 1910 to 1930 the sea ice volume was less than today based on correlating the 1955 low with the NAO at the time?
It’s hard enough to model the sea ice volume in 2010.
I think even sea ice extent is unreliable for 1910 to 1930; even 1955.
Amino Acids in Meteorites,
It’s not so odd. Firstly, the ice area is hardly increasing rapidly: http://arctic.atmos.uiuc.edu/cryosphere/IMAGES/current.anom.jpg ). Secondly, the ice thickness has been falling fast: http://www.arctic.noaa.gov/reportcard/figures/seaice2009fig4.jpg.
Hence the total arctic ice volume would indeed be expected to drop.