Foreword: I had originally planned to post a story on this, but Steven Goddard of the UK Register sends word that he has already done a comparison. It mirrors much of what I would have written. There is a clear discrepancy between the two data sources. What is unclear is the cause. Is it differing measurement and tabulation methods? Or, is it some post measurement adjustment being applied. With a 30 percent difference, it would seem that the public would have difficulty determining which dataset is the truly representative one.
UPDATE: The questions have been answered, see correction below – Anthony
Arctic ice refuses to melt as ordered
Published Friday 15th August 2008 10:02 GMT – source story is here
Just a few weeks ago, predictions of Arctic ice collapse were buzzing all over the internet. Some scientists were predicting that the “North Pole may be ice-free for first time this summer”. Others predicted that the entire “polar ice cap would disappear this summer”.
The Arctic melt season is nearly done for this year. The sun is now very low above the horizon and will set for the winter at the North Pole in five weeks. And none of these dire predictions have come to pass. Yet there is, however, something odd going on with the ice data.
The National Snow and Ice Data Center (NSIDC) in Boulder, Colorado released an alarming graph on August 11, showing that Arctic ice was rapidly disappearing, back towards last year’s record minimum. Their data shows Arctic sea ice extent only 10 per cent greater than this date in 2007, and the second lowest on record. Here’s a smaller version of the graph:
The National Snow and Ice Data Center (NSIDC)’s troublesome ice graphThe problem is that this graph does not appear to be correct. Other data sources show Arctic ice having made a nice recovery this summer. NASA Marshall Space Flight Center data shows 2008 ice nearly identical to 2002, 2005 and 2006. Maps of Arctic ice extent are readily available from several sources, including the University of Illinois, which keeps a daily archive for the last 30 years. A comparison of these maps (derived from NSIDC data) below shows that Arctic ice extent was 30 per cent greater on August 11, 2008 than it was on the August 12, 2007. (2008 is a leap year, so the dates are offset by one.)
Ice at the Arctic: 2007 and 2008 snapshotsThe video below highlights the differences between those two dates. As you can see, ice has grown in nearly every direction since last summer – with a large increase in the area north of Siberia. Also note that the area around the Northwest Passage (west of Greenland) has seen a significant increase in ice. Some of the islands in the Canadian Archipelago are surrounded by more ice than they were during the summer of 1980.
The 30 per cent increase was calculated by counting pixels which contain colors representing ice. This is a conservative calculation, because of the map projection used. As the ice expands away from the pole, each new pixel represents a larger area – so the net effect is that the calculated 30 per cent increase is actually on the low side.
So how did NSIDC calculate a 10 per cent increase over 2007? Their graph appears to disagree with the maps by a factor of three (10 per cent vs. 30 per cent) – hardly a trivial discrepancy.
What melts the Arctic?
The Arctic did not experience the meltdowns forecast by NSIDC and the Norwegian Polar Year Secretariat. It didn’t even come close. Additionally, some current graphs and press releases from NSIDC seem less than conservative. There appears to be a consistent pattern of overstatement related to Arctic ice loss.
We know that Arctic summer ice extent is largely determined by variable oceanic and atmospheric currents such as the Arctic Oscillation. NASA claimed last summer that “not all the large changes seen in Arctic climate in recent years are a result of long-term trends associated with global warming”. The media tendency to knee-jerkingly blame everything on “global warming” makes for an easy story – but it is not based on solid science. ®
Bootnote
And what of the Antarctic? Down south, ice extent is well ahead of the recent average. Why isn’t NSIDC making similarly high-profile press releases about the increase in Antarctic ice over the last 30 years?
The author, Steven Goddard, is not affiliated directly or indirectly with any energy industry, nor does he have any current affiliation with any university.
NOTE OF CORRECTION FROM STEVEN GODDARD:
The senior editor at the Register has added a footnote to the article with
excerpts from Dr. Meier’s letter, and a short explanation of why my analysis
was incorrect.
To expound further – after a lot of examination of UIUC maps, I discovered
that while their 2008 maps appear golden, their 2007 maps do not agree well
with either NSIDC maps or NASA satellite imagery. NSIDC does not archive
their maps, but I found one map from August 19, 2007. I overlaid the NSIDC
map on top of the UIUC map from the same date. As you can see below, the
NSIDC ice map (white) shows considerably greater extent than the UIUC maps
(colors.) The UIUC ice sits back much further from the Canadian coast than
does the NSIDC ice. The land lines up perfectly between the maps, so it
appears possible that the UIUC ice is mapped using a different projection
than their land projection.
Click for larger image
Because the 2007 UIUC maps show less area, the increase in 2008 appears
greater. This is the crux of the problem. I am convinced that the NSIDC
data is correct and that my analysis is flawed. The technique is
theoretically correct, but the output is never better than the raw data.
Prior to writing the article, I had done quite a bit of comparison of UIUC
vs. NSIDC vs. NASA for this year. The hole in my methodology was not
performing the same analysis for last year. (The fact that NSIDC doesn’t
archive their maps of course contributed to the difficulty of that
exercise.)
My apologies to Dr. Meiers and Dr. Serreze, and NSIDC. Their analysis,
graphs and conclusions were all absolutely correct. Arctic ice is indeed
melting nearly as fast as last year, and this is indeed troubling.
– Steven Goddard
Thank you Dr. Walt Meier, for your input. I’ve emailed Mr. Goddard directly to make sure he’s seen your response.
Would you be willing to answer questions about the NSIDC data and the presentations of it on this forum?
Thank you for your consideration. – Anthony Watts
Walt,
Thanks very much for your response. I just received and responded to your E-mail as well. I do understand that you have worked diligently on your methodology and that it has been throughly tested.
I understand the NSIDC definition of “extent” as being the area of ocean with >= 15% sea ice concentration. As I explained in the E-mail, when I measure that area in the UIUC maps (marked as concentration >=15%,) I see a 30% increase over last year.
Please advise why the UIUC concentration maps show a significantly larger increase in the amount of >=15% concentration ice, than the NSIDC graph shows.
If you or Mark can answer that question directly, technically and convincingly – I will be absolutely delighted to publicly retract and apologize.
http://igloo.atmos.uiuc.edu/cgi-bin/test/print.sh?fm=08&fd=12&fy=2007&sm=08&sd=11&sy=2008
The UIUC maps show a clear 30% increase in extent.
Regardless of any real or perceived inadequacies with the methodology, there should be closer agreement between the UIUC maps and the NSIDC graph. Even with any rounding errors due to pixels, or distortions in the UIUC projection, the difference is much too large. Indeed, the UIUC projection should tend to underestimate larger area relative to smaller ones, because lower latitude pixels get underrepresented on that map projection.
Basically, I am asking you to account mathematically why counting pixels on UIUC maps shows a 30% increase, and explain what is wrong with that methodology.
Also, do you still believe that these estimates will be realized?
http://nsidc.org/images/arcticseaicenews/200805_Figure4.png
And please comment on this widely published story.
“Mark Serreze, a scientist with the National Snow and Ice Data Center in Boulder, Colorado, told Agence France-Presse in an interview on June 27, 2008, that by the end of the 2008 summer, there might, for the first time in human history, briefly be no ice at the North Pole.”
http://en.wikipedia.org/wiki/North_Pole
This appears to be at odds with an earlier interview from several years ago.
Dr. Mark Serreze – There’s been open water at the pole before.
http://query.nytimes.com/gst/fullpage.html?res=9F00E3DD1E31F93AA1575BC0A9669C8B63
Once again, thank you for getting involved in the discussion.
Dr. Meier,
To speed the discussion up, a bit more on the type of explanation I am looking for wrt the ice extent dispute. Something like this –
1. “Pixel counting UIUC maps doesn’t work, because the map is distorted in a way which makes the area around 80N latitude too large. The distortion can be defined mathematically as follows ………”
or
2. “Pixel counting UIUC maps doesn’t work because the resolution per pixel is too low. The grid cell size is smaller than one pixel, so a region of four contiguous pixels may appear to represent more extent than what is measured on the higher resolution grid. This can be defined mathematically as follows……….. The reason we believe our grid size to be the correct one, rather than the one pixel grid you are using in the article, is because …….”
Thanks again. It is very late and I need some sleep. I look forward to seeing your response.
Steven Goddard
“Please advise why the UIUC concentration maps show a significantly larger increase in the amount of >=15% concentration ice, than the NSIDC graph shows.”
If the UIUC maps show a larger increase, where is the difference? Where do the NSIDC maps not show ice and the UIUC maps do show ice?
Dr. Meier, this is apples and oranges.
The chart you refer to appears to be sea ice concentration. It does not depict ocean area with sea ice as does the NSIDC chart and the UIUC maps under consideration, and the numbers are off by about two million sqKm.
The UIUC maps appear to show a recognizable and significant difference in percentage between the same time last year and now that does not correlate at all well with the NSIDC graph. That is my concern, and what I understand Mr. Goddard’s concern to be as well. Simply put (hint) the NSIDC chart tells me there is about 10% more ice than last year, and the UIUC maps tell me that there is about 30% more ice than last year.
Thank you, Dr Meier, for your input.
However, Mr Goddard has posed some relevant questions.
As they say, the ball is now in your court.
Earlier I complained I was having some trouble matching the UIUC map colours to the included scale. Closer examination showed that the 100% ice cover on the map isn’t well represented on the scale. But using the sum of absolute differences of RGB values as a metric and a threshhold of 50 I can get what looks like a reasonable match with the ice extent on the UIUC maps. I would have to play around a bit more before being totally happy with this.
But with this threshhold I get a 30.5% increase in pixel count for 20080811 over 20070812, in support of Steven Goddard’s original claim.
Somebody asked why I limited my ice area calculations to north of 63 degrees. The answer is that it was a lazy way of filtering the (probably) spurious concentrations near the coasts of Hudson Bay and the Great Lakes. I’ve actually looked at it both ways, and it only made 2 tenths of a percent or so difference, which isn’t significant when we’re arguing whether the true difference is 14 percent or 30 percent. Especially since it was a 0.2 percent increase.
No, Smokey, the ball is not in the NSIDC court. Rather, the match is over. Steven Goddard refuses to look at the data, and those scientists who have looked at the data, have determined that Goddard’s measurement of 30% increase in sea ice extent, from August 11, 2007 to August 11, 2008, is incorrect.
Dr. Meiers made the point quite clear:
“8. An image is simply a way to convey data; it is not data itself. Therefor it is not proper to do analysis on the image. You need to use the data.
9. The gridded data, when analyzed, must account for the projection in terms of the area of the grid cells. You have to sum the ice, weighted by the correct area for each grid cell. NSIDC uses a polar stereographic projection with a true latitude of 70 N. Other than at 70 N there will be distortion that needs to be corrected for, as NSIDC does.
10. NSIDC freely distributes all the data, tools to work with the data, and the grid cell area files. So anyone can do their own analysis.”
People on this thread did the analysis and confirm Mr. Goddard’s conclusion is incorrect…
Why doesn’t Mr. Goddard use the actual data? What does he have against the actual reported data?
Yesterday, I already suggested to Mr. Anthony Watts that he needs to publish a correction, but he embargoed my comment. I will attempt to make the comment again.
REPLY: PaulK Your “request for correction” was one that you wrote with specifics about language and placement. That’s not how it works. It’s not your blog, you don’t have the priviledge to dictate terms, and you don’t get to tell me what to write. The story is written by owned by Mr. Goddard, and since then he has offered to print a correction if his questions are satisfied, I’m going to respect his right to choose what he wants to do and when.
Right now we have some interesting questions in play, and I want to see how they are answered.
Steven Goddard has created a graph of his difference calculations. The values appear to (approximately)
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19
24 23 28 41 19 23 20 27 40 31 32 28 32 45 42 41 26 30 21
Can I be an optimist and hope that the latest figure of 21 percent represents a convergence towards my calculations?
Glenn
“The UIUC maps appear to show a recognizable and significant difference in percentage between the same time last year and now that does not correlate at all well with the NSIDC graph.”
The alleged difference in the UIUC maps doesn’t correlate well with the UIUC graph either, as Walt Meier pointed out. We again get back to the issue that everything is in agreement except for Mr. Goddard’s numbers.
Images certainly are data. They contain location, intensity and color information.
Why do we send satellites into space to take images of the earth? Isn’t ice data derived from “brightness” images? And at the other end of the spectrum, our understanding of microscopic objects is almost exclusively derived from images.
Image processing is a huge field, and the base technique of almost any image processing algorithm includes counting pixels of different colors and intensities.
The assertion that “images aren’t data” is simply not accurate. Many fields of science and engineering rely very heavily on digitized image data.
Patrick Henry… In some cases, you are correct, but not in this case. The image Goddard is using, was generated from a database, not directly recorded.
The key mistake Goddard and dipole are making, if I understand Steven Talbot and Jeff correctly, is that there is no way to show the surface of the globe on a flat screen as a continuous image. So the image people have been eyeballing, MUST have some errors in it. So where are the errors?
Here I will attempt an explanation… (and please correct me if I get this wrong, Steve or Jeff). It is very difficult to illustrate a curved surface on a flat piece of paper, or a flat computer screen. So “projections” are used for maps, drawings or illustrations depicting the surface of the earth. Steven Goddard assumed that the projection used on Cryosphere showed the area close to the North Pole as accurate to scale, and then that this projection would not represent areas correctly at lower latitudes.
But in reality the scientists at NSIDC were one or two steps ahead of Mr. Goddard. They used a projection accurate at 70 deg N, not at the pole (90N), as Mr. Goddard assumed. The scientists used this projection, because historically, the latitudes around 70 deg N, is roughly mid-way between summer mins and winter maxs in ice extent. (No one expected the massive ice melts nearer the Pole , that we saw last year, or this year, for another 50-100 years.) The scientists picked a projection that worked best for measuring sea ice extent, in a non-warming world. Where is 70N? It is about halfway down the West Coast of Greenland.
https://www.cia.gov/library/publications/the-world-factbook/geos/gl.html
If Cryosphere (UIUC) uses the same projection in the image Mr. Goddard is working with, that the NSIDC uses (and they should), then the image that you and I and Mr. Goddard are looking at, is not accurate at latitudes higher than 70N, and the higher the latitude, the bigger the inaccuracy. This isn’t because someone made an error; this is because you can’t show a three dimensional object in two dimensions without some error.
Imagine you are looking at the North Pole, looking down at a globe, and you are trying to draw that image on a flat sheet of paper (making a projection). If the area shown on the flat image is accurate down around the mid-west coast of Greenland, then a square inch of area on the globe at that latitude will occupy a square inch on the flat image. But the globe has more surface area (square inches) than the flat image. So the area, and the number of square inches above the 70N latitude, on the flat image, will be LESS than the actual area on the globe. So the flat image you and I see, shows less area in the Arctic Basin, than the actual area on the globe. If that area is covered by ice, and included in the ice extent, then Goddard ended up measuring fewer pixels in the Arctic Basin, than he should, because he is using a flat image to measure his data. And the guys eyeballing it, are making the same mistake. Since the YOY differences in sea extent are at lower latitudes, then the difference in areas there seem bigger, and then this difference is divided by the total extent of the sea ice to get the percentage change.
Lets try it with some approximate numbers: By getting a lower count of pixels at the high latitude, Mr. Goddard is dividing the difference of around 750k sq KM by less than the actual 5500k sq KM (which would calculate to 13.6% increase), he instead divides by a lower number of pixels equivalent to 2500 sq KM and gets a 30% increase. But his analysis is wrong, as are the eyeball estimates. The actual NSIDC data show that the YOY extent was about 13.5% different on August 11.
OK, that is how I understand the mistake. The key issue here, is that taking measurements from a flat image representation of our planet, will always end up in error, UNLESS you adjust for the projection error. Mr. Goddard should have known that, and he should have heeded Steven Talbot and Jeff, when they pointed out the projection error.
REPLY: That is a good explanation, and I appreciate the detail. Another question that needs to be answered is how the satellite derived data is “photographed”. As I understand it, it isn’t a single snapshot from above the pole, but a series of successive orbit slices that are patched together. Thus, how do we get from a curved rectangular view “slice” from the satellite persepctive to the raw data and to the imagery? What sort of projections errors or uninetentional distortions are introduced and are they relevant?
While there may be errors in the analysis presented here, how would the public know which presentation from which group better reflects the reality of sea ice? Problems like this one we are discussing arise when data presentations for public cconsumption, perhaps uninetntionally, show differences that can’t be easily resolved by the layman.
Assuming the description of the projection used in the UIUC images as the view from 10,000 miles above the pole, I have corrected my pixel count (21:39:11) to take account of the fact that pixels further from the pole represent a slightly larger area. This increases the extent figure by about 3%, and the increase of 2008 over 2007 goes up slightly to 31.1%.
Hope I got my trigonometry right.
Paul K (00:05:31)
The key mistake Goddard and dipole are making…
Originally I was just counting pixels. If I claimed further significance for the results it was unintentional. Where’s the ‘mistake’?
You are claiming that the interpretation of pixel count as ice extent is invalid because of spacial distortion. Steven Goddard described the projection used on the UIUC maps, and in my last post I incorporated this into my calculation. It didn’t seem to materially affect the outcome.
If you, or anyone else, can demonstrate I’m using the wrong projection I can have another go.
Hmm. I think my projection formula needs fixing. I am correcting for tilt but not change in distance from the observer. I will try again. Of course if this is the wrong projection I am wasting my time.
Has UIUC been emailed to see what they say on the topic? I get the feeling NSIDC are a bit unhappy with this story being broke before the author got in touch with them first so may not be that forthcoming? Or am I reading it wrong?
It does beg the question though that if the data is right, and having read all this I still think it is, then why are the “popular” graphs at Cryosphere seemingly so far out of kilter? A nice visiual should enable you to compare one year to the next without having to count pixels or apply algorithms or actually just junk it and go back to the data, because then it is useless.
Maybe these are questions more for UIUC than NSIDC, so I would like to know if anyone has had any input from them?
Talking of images, the direct NW passage is either nearly open or pretty well closed depending on what you look at.
http://iup.physik.uni-bremen.de:8084/amsr/arctic_AMSRE_nic.png
http://www.seaice.dk/iwicos/latest/amsr.n.ice.20080819.gif
http://ice-glaces.ec.gc.ca/Ice_Can/CMMBCTCA.gif
http://manati.orbit.nesdis.noaa.gov/ice_image21/D08232.NHEIMSK.GIF
There’s not too much consistency there to me to be honest. Are all these just pretty to look at in the main?
Paul K:
Is that the same as, “The science is settled”? Or maybe, “Consensus”?
Let’s cut to the chase here: explain this.
I think the distance might be 10000km; not 10000 miles. I did an estimate based on the latitudes of regions I can make out just on the perimeter. 10000km would mean the perimeter is roughly latitude 23.
“Another question that needs to be answered is how the satellite derived data is “photographed”.”
That information is publicly available for the NSIDC data from the NSIDC web site:
http://nsidc.org/data/docs/daac/nsidc0051_gsfc_seaice.gd.html
and info for the original data used by Cryosphere Today (Cryosphere Today does not clearly acknowledge their data source) is available from the NCEP web site:
http://polar.ncep.noaa.gov/seaice/support/ssmi.about.html
Images are data, but they’re not necessarily good data, and they do not contain location and intensity information. They contain an x,y location in image space and values for red, green, and blue. If you know the projection (and we don’t know the projection of the Cryosphere Today images) you can calculate location, but it’s not intrinsically in the image. Similarly, the RGB values don’t translate directly to intensity: they merely represent the relative strengths of three different sub-pixels on a computer monitor. Additionally, the Cryosphere Today images use lossy compression (even the PNGs contain compression artifacts) so they can’t be used for precise quantitative analysis. (Contrary to an assertion earlier in this thread, JPEG noise can not be removed.)
REPLY: I’ve been working with computer graphics since 1983, and my company produces computer graphics as seen on the side panel to the right. Thus I’m qualified to speak to the concerns of image formats and compression since I have used them both since their inception and produce them daily.
PNG is a 32 lossless compression image format, there are no artifacts, just like with the 8 bit GIF compression, there are no artifacts. The run length encoding schemes simply packetize pixels that are the same into tokens. It is totally lossless and without any compression or noise artifacts to convert the packets of RLE data to pixels upon display.
http://en.wikipedia.org/wiki/Portable_Network_Graphics
So there is no error of any kind in reading PNG pixels and translating it’s RGB component back to the numeric value of the original data as defined by the color key.
Unlike PNG, JPEG is not perfect, and much depends on the original compression setting, known as the Q factor. See the sample photographs in this article to see what I mean: http://en.wikipedia.org/wiki/JPEG
JPEG’s work because humans can resolve detail better than they can resolve color. NTSC and PAL color television use the same trick. They preserve the detail (luminance) while mostly compressing the color (chrominance). The weighting equation assigns the most compression to reds and blues, while preserving green, which we resolve better.
As for removing JPEG noise from it’s lossy scheme. I’ve done it and have a program that does that that is commercially available. Granted, depending on the compression factors applied to luminance and chrominance originally there may not be a 100% perfect restoration, but noise can be removed. Most of the error is usually in color space. JPEG does not appreciably alter the shape of the photo or graphic, thus in this exercise, while there may be a margin of error for recovery, Nonetheless for a comparison of two images to determine differences in pixels, it can be done with a small margin of error. I estimate about 2% or less for these images. To prove that to yourself, create a couple of images that have varying graphical color elements of known pixel area, then save as PNG and as JPEG, then run a pixel count on each. You may find that while the colors count varies in JPEG, the luminance count varies little, and perhaps not at all depending on the compression setting used.
But the Cryosphere today images, such as these here: http://igloo.atmos.uiuc.edu/cgi-bin/test/print.sh?fm=08&fd=15&fy=2007&sm=08&sd=15&sy=2008
are PNG and totally lossless. So I would expect accurate pixel counts between them.
That doesn’t address the larger questions about earth’s spherical geometry and what distortions may exist, but the approach of pixel counting/comparing two images can be expected to give repeatable results with a margin of error for JPEG and exact results for PNG – Anthony
Mr Goddard,
I don’t know whether or not Dr Meier will respond further to your questions, but I shall make my own observations as to whether or not you have anything to correct in your original article. Quotations are, of course, from your original.
1. “Yet there is, however, something odd going on with the ice data.”
You have no evidence that there is anything odd going on with the data. Your remaining case is that the visualisation of that data suggests the sea ice extent is greater than is the case. As has been made clear, NSIDC data is freely available to you. You have made no reference to it. IMV you should withdraw your implication and clarify that your concern is with visualisation exaggerating the differences in ice extent.
2. “Their data shows Arctic sea ice extent only 10 per cent greater than this date in 2007”
This figure has been corrected, and you should correct the statement.
3. “The problem is that this graph does not appear to be correct.”
You make the presumption that the graph is not a correct representation (and imply, IMV, that NSIDC data is thus not correct). Why did you not or do you not consider the possibility that the imaging is “not a correct representation”? It has been made clear to you that NSIDC data is accessible for examination. Your prejudiced presumption should be corrected.
4. “Other data sources show Arctic ice having made a nice recovery this summer. “
As Dr Meier has made clear (and others before him), “An image is simply a way to convey data; it is not data itself.” Your description of “other data sources” is highly misleading, suggesting that the issue is own of conflicting data rather than (apparently) divergent impressions of how that data is conveyed. This should be clarified, if you wish not to mislead your readers.
5. “Additionally, some current graphs and press releases from NSIDC seem less than conservative. There appears to be a consistent pattern of overstatement related to Arctic ice loss.“
This is a serious allegation, and one that is entirely unsupported by your ‘analysis’. On the contrary, your remaining issue with the image rendition of the data amounts to a potential concern that the image is overstating the data in its visualised indication of sea ice extent. Your allegation is thus unfounded and potentially defamatory.
Whether or not you gain an explanation that satisfies you of the visualisation apparently exaggerating differences in ice extent, the above points should be corrected if you are to show responsible concern for not misleading your readers.
PaulK,
I have described in some detail the distortion in the UIUC projection, but obviously not enough. In that projection, there is nothing magical about 70 degrees or any other latitude – other than 90 degrees. The oft quoted 70 degree number relates to the projection being used by NSIDC, not UIUC.
Starting with the ideal case of a viewpoint at infinity (no parallax) – at the pole, the image is undistorted. As you move away from the pole, the distortion increases. At the equator, the distortion would be infinite. In this case, the distortion would be = 1 / sine(latitude)
Looking at the UIUC maps, the actual viewpoint is from a height less than infinity, so the latitude where the distortion becomes infinite is at a higher latitude than 0 – maybe ten or twenty degrees. That isn’t hugely important because there is no ice anywhere near that latitude.
In that type of projection, a smaller circle centered around the pole (i.e 2007) is less distorted than a larger circle (i.e. 2008.) If the calculations were corrected to adjust for this, the larger circle (2008) would be adjusted upwards more than the smaller circle (2007.) This would have the effect of making the percentage gain in 2008 larger than what I reported. I did not make any adjustments to the raw pixel count, and that is why I described the methodology as conservative. The actual 2008 percentage gain is somewhat larger than I reported.
If there is a flaw in the methodology, it is not at that point. It would have to be tied to an inaccuracy in the UIUC map or my understanding of that map. This may well be the case, and I am keen to understand where the error is.
If Walt or Mark can explain to me where the error is, I will most certainly retract and apologize. There appears to be a discrepancy between the UIUC maps and the NSIDC calculations. I’m certain that the NSIDC calculations were done carefully and diligently.
UPDATE:
Cryosphere Today has changed their color scheme presentation very recently.
See the new color scheme here:
http://arctic.atmos.uiuc.edu/cryosphere/
It appears the archived images are unchanged and remain the same color scheme. I can’t find any announcement for why this has been done, if somebody knows of one, pass it along please.
Anthony,
Thanks for your follow up on the image color issue.
In the case of “ice extent” calculations, the image color precision is not important. All pixels >= 15% are treated equally in extent calculations.
If the calculation was for “ice area” using the images would likely introduce more error, and jpeg would probably not be the best format. I haven’t and wouldn’t attempt to do a calculation of sea ice area, using the images.