In today’s report
- Arctic Sea Ice on the rise again, presently in the range of normal levels
- Antarctic Sea Ice is at slightly above normal levels
- Why is early satellite data for Arctic and Antarctic Ice extent referenced in the first IPCC report missing from today’s data?
- Is revisionism going on with the date of the famous USS Skate photo in the Arctic?
- Bonus – it seems NOAA is taking Arctic soot seriously
First the Arctic from NSIDC:
Source: http://nsidc.org/data/seaice_index/images/daily_images/N_stddev_timeseries.png
After being out of the ±2 STD area since before peak melt last year, Arctic extent has spent most of March in near normal territory. After what looked like a maximum earlier this month, it was false peak, and ice is on the rise again.
NORSEX SSM/I shows the current value within ±1 STD
Source: http://arctic-roos.org/observations/satellite-data/sea-ice/observation_images/ssmi1_ice_area.png
A caution, as we saw in 2010, extent hugged the normal line for quite awhile, and that didn’t translate into a reduced or normal summer melt. So, forecasting based on this peak might not yield any skillful ice minimum forecasts.
Antarctic Sea Ice is at slightly above normal levels, as it has been for some time:
Source: http://nsidc.org/data/seaice_index/images/daily_images/S_stddev_timeseries.png
Why is early satellite data for Arctic and Antarctic Ice extent referenced in the first IPCC report missing from today’s data?
In a post last week, Steve Goddard pointed out that in the original IPCC FAR in 1990, there was an interesting graph of satellite derived Arctic sea ice extent:
This is from page 224 of IPCC FAR WG1 which you can download from the IPCC here
And here is figure 7.20 (a) magnified:
The IPCC descriptive text for these figures reads:
Sea-ice conditions are now reported regularly in marine synoptic observations, as well as by special reconnaissance flights, and coastal radar. Especially importantly, satellite observations have been used to map sea-ice extent routinely since the early 1970s. The American Navy Joint Ice Center has produced weekly charts which have been digitised by NOAA. These data are summarized in Figure 7.20 which is based on analyses carried out on a 1° latitude x 2.5° longitude grid. Sea-ice is defined to be present when its concentration exceeds 10% (Ropelewski, 1983). Since about 1976 the areal extent of sea-ice in the Northern Hemisphere has varied about a constant climatological level but in 1972-1975 sea-ice extent was significantly less. In the Southern Hemisphere since about 1981, sea-ice extent has also varied about a constant level. Between 1973 and 1980 there were periods of several years when Southern Hemisphere sea-ice extent was either appreciably more than or less than that typical in the 1980s.
I find it interesting and perhaps somewhat troubling that pre-1979 satellite derived sea ice data was good enough to include in the first IPCC report in 1990, but for some reason not included in the current satellite derived sea ice data which all seems to start in 1979:
Since the extent variation anomalies in 1979 seem to match with both data sets at ~ +1 million sq km, it would seem they are compatible. Since I’m unable to find the data that the IPCC FAR WG1 report references so that I can plot it along with current data, I’ve resorted to a graphical splice to show what the two data sets together might look like.
I’ve cropped and scaled the IPCC FAR WG1 Figure (a) to match the UUIC Cryosphere Today Arctic extent anomaly graph so that the scales match, and extended the base canvas to give the extra room for the extended timeline:
Click image above to enlarge.
Gosh, all of the sudden it looks cyclic rather than linear, doesn’t it?
Of course there will be much wailing and gnashing of teeth over my graphic, and the usual suspects will try to pooh-pooh it, but consider the following
- Per the IPCC reference, it is data from NOAA, gathered by the American Navy Joint Ice Center
- It is satellite derived extent data, like Cryosphere Today’s data
- The splice point at 1979 seems to match well in amplitude between the two data sets
- The data was good enough for the IPCC to publish in 1990 in the FAR WG1, so it really can’t be called into question
- If Mike Mann can get away with splicing two dissimilar data sets in an IPCC report (proxy temperature reconstructions and observations) surely, splicing two similar satellite observation data sets together can’t be viewed as some sort of data sacrilege.
Of course the big inconvenient question is: why has this data been removed from common use today if it was good enough for the IPCC to use in 1990? Is there some revisionism going on here or is there a valid reason that hasn’t been made known/used in current data sets?
If any readers know where to find this data in tabular form, I’ll happily update the plot to be as accurate as possible.
Is revisionism going on with the date of the famous USS Skate photo in the Arctic?
It seems our favorite photo of the USS Skate has had it’s date revised.
Since yesterday was the anniversary of the March 17th surfacing of the USS Skate, WUWT contributor Ric Werme was interested in what the photographic conditions might look like on March 17th 1959 when the sun was just below the horizon, and so found a sub and attempted to recreate the photo conditions himself to see if the photograph was actually possible.
See: http://wattsupwiththat.com/2012/03/17/submarines-in-the-winter-twilight/
Turns out it was, but then he stumbled on something he didn’t expect to find. The date for the surfacing has been changed from March 17th, 1959 to August, 1958 (with no day given) in Wikipedia and in NAVSOURCE. He at first thought I’d made a mistake in citation, but it turns out dates have been changed since I wrote my original article on the USS Skate on April 26th, 2009.
I wrote about how the original date remains on NAVSOURCE in the Wayback machine
Anthony Watts says:
Navsource, in the Wayback machine, had it stated as March 17th 1959, just days before my original article. This is the April 18th 2009 snapshot from Wayback:
http://web.archive.org/web/20090418161606/http://www.navsource.org/archives/08/08578.htm
The caption then reads:
Skate (SSN-578), surfaced at the North Pole, 17 March 1959.
I remember checking NAVSOURCE for accuracy before publishing, my caption then says:
Skate (SSN-578), surfaced at the North Pole, 17 March 1959. Image from NAVSOURCE
History on that photo changed there at NAVSOURCE since then, probably due to alarmist pressure from Wiki etc. and other folks like Neven who went ballistic over the picture when I highlighted it. It is “inconvenient” in March (during peak ice season) but soothing for them in August (during near peak melt season).
The picture may have been taken a couple of days after the funeral photo in March alluded to upthread.
Se EM Smith comment in my original thread. http://wattsupwiththat.com/2009/04/26/ice-at-the-north-pole-in-1958-not-so-thick/#comment-122932
Oddly, NAVSOURCE now shows a caption of:
So what had been certain and unchallenged for years now all of the sudden is uncertain and may be in August 1958. Seems like a case of the tail wagging the dog.
Obviously there is a need to pin this date down, but I’m amused that so much attention has been brought to this photo since I first blogged on it.
BONUS: I’ve always said that the current drop in Arctic Ice Extent might have roots in soot from the industrialization of Asia causing an albedo change which really took off in the 1990’s, would show up in the summer melt season when solar irradiance is at a peak in the Arctic. Now it seems NOAA is taking Arctic soot seriously:
From the video description:
Small, new, remotely-operated, unmanned aircraft are being flown in the Arctic to measure black soot. The soot is produced by burning diesel fuel, agricultural fires, forest fires, and wood-burning stoves. It is transported by winds to the Arctic, where it darkens the surface of snow and ice, enhancing melting and solar warming. See http://saga.pmel.noaa.gov/ and http://www.pmel.noaa.gov/edd/manta.html
As always, check the latest sea ice conditions on the WUWT Sea Ice Reference page.
UPDATE: Robert Grumbine disputes some the the points related to the IPCC1 report and sea ice with EMMR equipped satellites here. – Anthony
blackwhitewash.com says: Skimmed the thread and am surprised nobody has asked:
Where on earth is the ICESAT data?
It was released with great fanfare last year and is allegedly the answer we are all looking for on sea ice questions.
So why is it not answering those questions?
ICESat is no longer in operation (and wasn’t last year either). I think you have it confused with CryoSat. Folks processing that data did release some preliminary thickness estimates last year to the media, but they are still processing and validating the results.
If you are interested in how the ICESat data compared to earlier thickness estimates from submarines you can take a look at the figure here from the Kwok and Rothrock 2009 study: http://nsidc.org/cryosphere/sotc/sea_ice.html
E.M.Smith
March 19, 2012 at 8:11 am
###
As I pointed out earlier, the photo of the Skate, in clear water was taken before the “sail” modifications that were made to allow it to bust through ice as shown in the photos known to have been taken during March.
Oh good. So when I assert it is stupid to merge data from different instruments (thermometers) and splice segments from different eras and locations via ‘homogenizing” I’ll just cite you as the authority… Thanks!
(Or can ‘data from different instruments’ only be merged when it suits the warmers agenda…)
Smokey says:
March 18, 2012 at 7:07 pm
Arctic sea ice
Antarctic sea ice
Any questions?☺
Merge those two charts and what do you get ?
I estimate it’s close to a Level Flat Line.
And in spite of the CAGW climate misinformation following us,
my soul is peaceful and satisfied…
@DesertYote:
I’m not seeing any difference in the sail. There is a reflection of the one “open water” that make it look slightly squarish, but I believe that is an artifact of the light line off the round surface. Compative hights, etc. all look the same to me.
Perhaps you could describe what you are seeing?
To me, the the ‘3 man’ photo and the ‘open water’ photo are strikingly similar with only the boat moving back a bit between them (moving the open water from behind to infront and a minor change of viewer height. Similar ice deposits on deck, similar background water to snow in distance at bow, just way similar. All other shots are ‘way different’.
Add in the nose white that is present in 1958 and gone in 1959 and it’s even stronger.
Figure 3 of the Walsh paper reveals a mean for 5 % sea ice extent from approximately 7 million sq km to 14 million sq km. Today 15 % is the number used therefore Walsh et al would seem to be overestimating the extent. Julienne wrote that she reduced the Walsh minimum from 8 million sq km to six million. It looks to me that it should be reduced to at least six million. Then the 1962 low indicates an anomalous reduction of about 400,000 sq km putting a possible 15 % calculation at 5,600,000 sq km or less. This compares with the 2007 15% sea ice extent low of about 500,000. The question is what is the proper area adjustment for converting from 5% sea ice concentration to 15%.
Why use UAV’s to measure soot? Why not use all the publicity-camping-trips to take ice cores and measure soot content directly?
Richar1225, to be clear, the numbers I quoted are based on the Had1SST data set, but using a 15% thresh hold for the ice extent calculations. Same with the Chapman and Walsh data set. These data sets are provided on a 1 degree lat/lon grid, and they are based on blending earlier satellites such as ESMR (for the microwave), as well as visible and thermal satellite imagery even earlier than 1972 (e.g. TIROS), ship and aircraft observations and any other observations of the ice edge) available. The data record from both Had1SST and Chapman and Walsh starts in 1870, the accuracy of which is questionable prior to about 1953, and afterwards, I have noticed some inconsistency problems that the Had1SST folks are working on correcting.
The Nautilus and Skate are featured in British Pathe News reels of the time.
The Nautilus “first to cross beneath the pole” is the subject of this from 1958: http://www.britishpathe.com/video/the-nautilus-crosses-the-top-of-the-world-aka-naut/ (8.25 minutes).
The review of 1959 mentions the Skate surfacing at the pole. The commentary says it is the Nautilus, but you can see “..ATE” on the sub in one shot.
http://www.britishpathe.com/video/review-of-the-year-1959/ (7:40) The Arctic shots appear from 3:44 to 3:59 and show, among other things, her having broken through ice about 6-8 inches thick, and steaming along the surface through such ice.
Good graph of 1953-present ice from NSIDS: http://nsidc.org/cryosphere/sotc/sea_ice.html which confirms that Anthony’s splice of the mid-1970’s “decline” was a bit too low on the chart.
Rabe says, March 19, 2012 at 12:07 am:
” The photo of Skate sticking through the ice cannot be from 17 Mar 59
because it’s sunny there.”
According to Navsource, the photo of the sub surfacing through ice was
taken “above the Arctic circle” in 1959, presumably during the March cruise.
They say so now, and they said so back when they were saying the
surfacing in essentially clear water was at the North Pole, March 17, 1959.
But not necessarily at the pole
Here is a photo of the funeral at the North Pole, March 17 1959:
http://www.vintagehikingdepot.com/tag/uss-skate/
Also,
http://library.osu.edu/projects/under-the-north-pole/images/wilkins35_5_4.jpg
at http://library.osu.edu/projects/under-the-north-pole/afterwards.html
It looks quite icy to me there at that time.
Here is a photo claimed to be the March 17 1959 polar surfacing:
http://library.osu.edu/projects/under-the-north-pole/images/wilkins35_5_1.jpg
It is not either of the two photos that have been drawing a lot of comments.
It does show plenty of ice.
If the picture of the USS Skate in the picture with open water was during March, wouldn’t the below zero conditions be producing at least some sea smoke across the top of the water?
“NSIDC has a merged ESMR/SMMR/SSM/I data set here: ftp://sidads.colorado.edu/pub/DATASETS/nsidc0192_seaice_trends_climo/esmr-smmr-ssmi-merged/
These are daily and monthly extents that span 1972-2002 (data source is NASA GSFC, using
the NASA Team sea ice concentration algorithm and filling in with ice chart data when there was no satellite overlap). While not current, it would allow anyone to do their own sea ice anomaly plot to see how the earlier years compared to the last couple decades. Comparisons with NSIDC’s sea ice index for the month of September for example show good agreement between the two data sets, so if one wanted to, they could extend it to 2012 (Note I only looked at September, other months may differ).”
In september, there is a mean difference of 130.000 sq. km betweenn merged ESMR/SMMR/SSM/I and Sea Ice Index.
In march, the difference is larger: 1.110.000 sq km.
Adjusting this differences, the result could look like this, eith uncertainties in the merging: http://images.meteociel.fr/im/1717/image001_ntz0.png
What is so complicated about this whole Skate picture?
The Skate is famous for surfacing at the North Pole, so someone inadvertently assumed that picture must have been from that event and said so on the Internets. Then someone who really badly needs to refute the existence of AGW comes by, copies the whole thing and goes: “Hahaa, there you go, warmistas! There was open water at the Pole in 1957! No AGW! Hahaa!”
Of course, this half truth travels around the world a couple of times and gets so much attention that folks find out that this picture wasn’t made at the Pole in 1959 (not that it matters, it was a piss poor argument to begin with), and they change the caption.
As can be expected, this changing now becomes the focus, accompanied by a casual “I’m interested in nailing down the date on the one in question”.
First the Goddard conspiracy theory du jour, and then this, together in one post. I don’t know whether to laugh or cry. As always, top notch science, Anthony.
@ur momisugly Günther Kirschbaum on March 19, 2012 at 12:36 pm:
Thank you Mr. Cherrytree, for the official reply from the drive-by Warmist Wapid Wesponse team. It was indeed very vapid.
===
From E.M.Smith on March 19, 2012 at 9:47 am:
At the Navsource photo gallery, the white bow band is very noticeable in this photo as well, captioned “Skate (SSN-578) moored to the ice at floating ice station Alpha.” By the GlobalSecurity info you linked to (I had seen it before btw), that was in 1958.
So white bow band in 1958, wiped off by 1959 as shown in several photos, it doesn’t reappear on the Skate in photos from later years. Thus the b&w photo in the original post is from 1959.
Gunther’s been pwned.☺
The sketch on the envelope clearly shows a penguin near the North pole, giving a polar bear the flipper. Obviously this poor species has been now exterminated from its northern range. Yet another Species Crime to lay at the feet of dastardly Hoomons, irrefutably documented!
Or not.
😉
Cui bono asks:
“On historical sea ice extents: Is it not possible to take just one year from the first half of the 20th Century (preferably one in which journalists were warning of a Hansenesque submerging of New York) and collate all the shipping observations to try to get a good snapshot of that year? This would surely cost a small fraction of the funds consumed weekly by the Computer Models, and would just take a few slaves – oops, grad students – to research. As you point out, the absence of real data used prior to 1979 is a travesty. Especially when we are told about the ‘unprecedented decline’ of sea ice, and we can’t go back more than 30-odd years.”
This has already been done for every month of the 20th century by Walsh and Chapman 2001:
ftp://128.208.240.87/incoming/PolarFridays/2-walsh_2001.pdf
Updated graph:
http://arctic.atmos.uiuc.edu/cryosphere/IMAGES/seasonal.extent.1900-2010.png
It has also been done back to 1870 by Kinnard et al 2008, as reported in Polyak et al 2010:
http://www.cgd.ucar.edu/staff/mholland/papers/Polyak_2010_historyofseaiceArctic.pdf
Graph:
http://tamino.files.wordpress.com/2010/10/polyakfig2.jpg?w=500&h=340
Indeed Kinnard et al, 2011 has extended the summer sea ice record back 1,450 years by the use of proxies:
http://www.nature.com/nature/journal/v479/n7374/full/nature10581.html
Graph:
http://www.skepticalscience.com/graphics/Kinnard_2011_sea_ice_med.jpg
Now I wonder why Anthony doesn’t use any of those graphs to show us the cyclical nature of reductions in summer sea ice extent in the Arctic?
E.M.Smith
March 19, 2012 at 9:47 am
@DesertYote:
###
I did some hunting around and found a photo of the Skate in Portland England shortly after its shakedown. This was before any arctic operations. The tower looks the same as in the March 17 photo. So what I am probably seeing is the effects of slightly different lighting conditions on a subtlety complex shape. I had an eye exam with pupil dilation so looking at too many photos is not possible, also why it took so long to respond.
Tom Curtis,
Thanks for the additional papers estimating the pre satellite history of arctic sea ice. Although Walsh’s paper was published in 2001 He appeared to be skeptical of the anthropomorphic component of the sea ice decline but rather pointed to the domination of positive NAO. The arrested decline of sea ice the last few years appears to correlate with the resurgence of negative winter NAO. If this trend continues we may expect to see ice volume rebuild. The earlier reconstructed ice history is based on proxy’s that must differentiate between minimum summer anomalies and maximum summer anomalies which have not changed as much. I will read these papers carefully but I remain skeptical of all past reconstructions. It appears to be clear however that the little ice age likely represented a high for sea ice development during the Holocene and from which we are luckily receiving a reprieve.
From Tom Curtis on March 19, 2012 at 7:39 pm:
That graph? The data is available here:
http://arctic.atmos.uiuc.edu/SEAICE/
Concerning the 1901-onward data, read the documentation text file:
http://arctic.atmos.uiuc.edu/SEAICE/arctic.historical.seaice.doc.txt
Thus that is clearly not what Cui bono was asking for. Plus, there are clear warnings about the Walsh and Chapman data, namely that it may not be real measured data but from assorted types of infilling, and be careful with the pre-1953 data.
And what do we see in either Figure 4 in the paper or in the updated graph? At the warned-about place, a step change between 1952 and 1953. While winter shows a small increase, spring, summer, and fall show a strong drop, which shows up in annual. This indicates something may be off, and this is where the warning is, thus it is appropriate to treat this graph and the early “data” with suspicion, not to be used for serious work. (BTW this graph is incompatible with other graphs due to defining the seasons differently, as with defining winter as Jan-Feb-Mar, instead of the common meteorological seasons with winter being Dec-Jan-Feb in the Northern Hemisphere.)
Wow, you’ve done something amazing, you found something that’s even less than being not what Cui bono asked for. This massive compilation piece, bringing together proxies and other things, does yield up as source for “recent” Arctic sea ice extent info, 1870 to 2003, in the Figure 2 caption, as Kinnard et al 2008. Which one can find in the reference, and Google found the paper:
ftp://128.208.240.87/incoming/PolarFridays/9.1-Kinnard-2007GL032507.pdf
Which leads me to conclude you either didn’t check your sources or are engaging in deliberate obfuscation, as that paper clearly says, section 2, Data and Methods:
There it is, back to Walsh and Chapman, with caveats on the data laid out, including:
Reliable ice concentrations are only available from historical sources after 1953, and from satellite imagery since 1972. Prior to 1953, only the ice edge position is reliable.
A paywalled article? If you wanted to bring in proxies you could’ve just stopped at the previous paper link you gave.
Oh well, Google found it:
http://gizmo.geotop.uqam.ca/devernalA/Kinnard_et_al_nature_2011.pdf
And on the second page we find:
And what is reference 3? Walsh & Chapman 2001. What was reference 16?
Polyakov, I. et al. Long-term ice variability in Arctic marginal seas. J. Clim. 16, 2078–2085 (2003).
A. What you presented keeps going back to the same source, Walsh and Chapman, who have caveats on their work, a strong one for pre-1953. On the last one, that questionable source was used for proxy calibration. Calibration to a self-invalidated “standard”? That wouldn’t work from a laboratory to a manufacturing shop floor.
Why would Anthony want to definitively use work derived from such a self-cautioned source, especially the sources you gave which present themselves with an authority the originating source doesn’t even claim?
B. If you actually read Anthony’s post, you might have noticed he wanted pre-1979 satellite data. Got any?
C. To my reading, Anthony noted the cyclic nature in about the same way that Dr. Roy Spencer includes the third-order polynomial with the monthly UAH global temperature anomaly update, “for entertainment purposes”, mentioning how it looks cyclical without outright stating it as cyclical.
But now that you mentioned it, reference 16, the source of the Russian data used for calibration in the last paper you linked to, was found by Google. Here’s the Abstract:
Cyclical patterns, lack of statistically-significant long term trends. Got it?
Well, I see everyone’s glossed over the purloined letter: Sir Hubert Wilkins’ ashes. Hello? Anthropogenic soot? The science is settled. You’re welcome.
Working on some slogans… Cremation’s a crime! Give a hoot, don’t make soot! Don’t be an ash, send me cash!
Please, do your part and report soot deniers — but if you’re wrong, don’t come swimming to me!
What’s that? The ice is on the rise? I thought I heard a noise. Didn’t I tell you about the 17-year data analysis rule that starts right…now?
I’m sure my Nobel Prize is already in the mail.
Thanks for posting on the state of sea ice in the Arctic and the Antarctic! Please keep us updated through the NH melting season, especially in September.
Oh dear! I seem to have upset kadaka by using non-AGW skeptic pre-approved data sources. Lest anyone should look at the data and start to think, he quickly brings out the climate skeptic three step, which goes:
1) Does the data show me want I want to see?
2) If no, is the data perfect in every way?
3) If no, ignore data in its entirety as being irrelevant and flawed.
kadaka gives an example of state of the art three stepping.
For those who believe you should pay attention to the data available, and not just the data you want to be available, I note the following:
1) Walsh and Chapman construct a sea ice concentration record based available data, which prior to 1978 is based on reports from shipping and aircraft. Where such data is not available (and only where that data is not available) they assume that Arctic sea ice in the first half of the twentieth century behaved in the same way as it did in the last third of the twentieth century to extrapolate from known observations.
2) This is what Cui Bono asked for, despite kadaka’s protestations to the contrary. That is, it is a record of sea ice extent from shipping observations where-ever those observations are available. As noted previously, climatology is only used where such observations are not available, so unless Cui Bono wanted the impossible requirement that every grid cell have a ship board observation of ice extent for every month of the year so that no extrapolation is required, Cui Bono has got what he asked for. It should also be noted that it is possible to extract all extrapolated data from the data set quite easily to compare regional ice limits. Because the coverage is not complete, however, doing the precludes plotting a sea ice extent.
3) With regard to the reliability of the sea ice extent, Chapman and Walsh say in supporting documentation:
“It appears that the SMMR and SSM/I data contains significant differences
poleward of the ice edge for most months. Ice concentrations are
generally lower in the central Arctic for the these data than for
other data sources. Ice extents appear to be consistent across datasets,
ice areas derived from pre-1978 data may be significantly higher than
those calculated from the satellite period. The figure contained in
icearea.ps provided with this data illustrates the rather abrupt jump in
total northern hemisphere ice area around October 1978. The figure
contained in icextnt.ps, ice extents calculated assuming 100% coverage
everywhere ice was observed, illustrates that the extent data is more
consistant between data sources.”
Kinnard et al 2008 write:
“Reliable ice concentrations are only available from historical sources
after 1953, and from satellite imagery since 1972. Prior to
1953, only the ice edge position is reliable.”
So contrary to the impression kadaka tries to make, data on the limit of the ice extent is fairly reliable, particularly in the summer months when there was most shipping in the Arctic. There is a distinct exception to this point during the years of WW2, when restricted shipping resulted in very few observations of the Arctic sea ice limit.
4) Walsh and Chapman 2001 do not include data prior to 1901. Therefore the data in Kinnard et al 2008 from 1870 to 1900 is new data not included in Walsh and Chapman. Therefore kadaka’s claim that Kinnard et al 2008 contains no knew data is simply false.
5) kadaka’s objection to Kinnard 2011 appears to be only that it is a proxy study that does not come up with the right (from his point of view) result. It is certainly not that it was calibrated against the “unreliable” Chapman and Walsh data because the Chapman and Walsh data is not universally unreliable. Without determining the data sources for the calibration periods explicitly, which kadaka has not done, he is not able to make that determination. Specifically, he is not entitled to assume the August data in particular contains as much “climatalogical data” as the annual average, and hence that it is as uncertain as the annual average without explicitly looking at the sources of the August data cell by cell.
Tom, thanks for following up with all those links to extended time-series of ice concentrations. The Had1SST data set also relies on the earlier work by Chapman and Walsh. And the Figure that Anthony showed in this posting, was based on these earlier efforts of trying to blend the satellite observations from various sources together with ship, aircraft and other observations of the ice edge that were available prior to the launch of continuous data record from multi-channel passive microwave sensors. With these earlier data records, the ice concentrations would not be as reliable as the total ice extent, which is why most folks only show the extent.
One comment from the modeling results from the CMIP5 archive, the models do show trends of sea ice loss during the 1920s/1940s consistent with the warming trend during that time-period. I don’t see this in the sea ice records that date that far back (i.e. Chapman and Walsh), but that’s in large part because the ice cover has observational gaps and has been filled in with climatology. These trends rival some of what we saw in the 1990s, but not what we’ve seen the last decade. I know these are model results so some here may disregard them all together.