I spent some time over at skepticalscience reading the post and discussions. There’s a whole bunch of interesting discussion about Greenland over there, a small amount seems relevant to Jason Box’s claim the results of FKM would be obsolete at the time of publication. (The irrelevant material include all the discussion of high surface temperatures, and the discussion of the truth of global warming. FKM is a reconstruction of melt data and does not deny the truth of global warming.)
In light of what I read at SS, today, I want to focus on these questions:
- What are the major results and claims in FKM2011?
- Are any of those claims already made obsolete based on data from 2010?
- Are any of those claims likely to be obsolete?
To identify the major claims, I will turn to the abstract:
A reconstruction of annual Greenland ice melt extent, 1784–2009
Oliver W. Frauenfeld,1 Paul C. Knappenberger,2 and Patrick J. Michaels3
Received 17 August 2010; revised 21 December 2010; accepted 31 January 2011; published 19 April 2011.The total extent of ice melt on the Greenland ice sheet has been increasing during the last three decades. The melt extent observed in 2007 in particular was the greatest on record according to several satelliteâ€derived records of total Greenland melt extent. Total annual observed melt extent across the Greenland ice sheet has been shown to be strongly related to summer temperature measurements from stations located along Greenland’s coast, as well as to variations in atmospheric circulation across the North Atlantic. We make use of these relationships along with historical temperature and circulation observations to develop a nearâ€continuous 226 year reconstructed history of annual Greenland melt extent dating from 2009 back into the late eighteenth century. We find that the recent period of highâ€melt extent is similar in magnitude but, thus far, shorter in duration, than a period of high melt lasting from the early 1920s through the early 1960s. The greatest melt extent over the last 2 1/4 centuries occurred in 2007; however, this value is not statistically significantly different from the reconstructed melt extent during 20 other melt seasons, primarily during 1923–1961.
Supplementing the abstract, I provide Figure 2 from FKM to which I added some lines which will facilitate discussion of the text in the abstract.
Now, let’s start parsing sentences. The first sentences everyone agrees on: ice melt has been increasing. The second sentence reads:
The melt extent observed in 2007 in particular was the greatest on record according to several satelliteâ€derived records of total Greenland melt extent.
FKM2011 estimated the melt by gathering together melt estimates from three groups, “XPGR data of Abdalati [2007], the SMD data of Mote [2007], and the impXPGR data of Fettweis et al. [2007] to produce a composite melt index representing the total annual melt extent across the Greenland ice sheet for 1979–2009 (Figure 2, blue line).” I added a horizontal blue line passing through the melt index for 2007. As you can see, it is the maximum in the figure presented by FKM2011, which includes data through 2009.
The following sentence reads:
Total annual observed melt extent across the Greenland ice sheet has been shown to be strongly related to summer temperature measurements from stations located along Greenland’s coast, as well as to variations in atmospheric circulation across the North Atlantic.
Everyone agrees that melt across the Greenland ice sheet is strongly correlated to summer temperatures and in particular, more ice melts when summer temperatures are high rather than low.
FKM continues:
We make use of these relationships along with historical temperature and circulation observations to develop a nearâ€continuous 226 year reconstructed history of annual Greenland melt extent dating from 2009 back into the late eighteenth century.
Based on the observation that more ice melts when summer temperatures are high and less ice melts when summer temperatures are low, FKM decided to estimate the ice melt back to roughly 1840 and where possible even further back. This reconstruction is accomplished by examining summer air temperatures and features of the atmospheric circulation that pre-date the satellite record.
Creating this reconstruction is the main contribution of the paper.
Mind you: the reason people would be interested in the reconstruction is to note how the current melt compares to past melt. So, one might wish to compare the 1979-2009 melt (blue dots) recorded by satellites and estimated by the reconstruction (open circles) to the melt rate during previous periods. This is why the abstract places the recent melt in context of the values indicated by the reconstruction:
We now arrive at the FKM results that are putting some people’s knickers in a twist:
We find that the recent period of highâ€melt extent is similar in magnitude but, thus far, shorter in duration, than a period of high melt lasting from the early 1920s through the early 1960s.
This sentences makes two claims: 1) shorter in duration, 2) similar in magnitude.
It’s not clear to me precisely how the authors determined the beginning and end points of high-melt and low-melt periods. Had I been asked, I might have asked them to clarify this. However, this statement seems clearly true. On the figure above, I defined the high melt periods as those during which the 10 year lagging average was greater than 0. These are indicated by bold fuschia lines at the bottom of the graph. Even accounting for the lag in the 10 year mean and padding the by an additional 10 years, the final bold fuschia line is shorter in duration than the one corresponding to the previous high melt period.
The claim “similar in magnitude” is a qualitative statement. However, it seems justified based on quantitative statements we could make based on the graph:
- the maximum value of the 10 year lagging mean is lower than that achieved during the previous high melt periods,
- the reconstructed value for the 2007 melt (open circle) is lower than the maximum reconstructed value recorded during the previous melt period and the measured value based on satellites falls well within the range of values during a previous rapid melt period.
That is: It’s similar to what we saw during the past high melt period.
Finally, we do get to a sentence that seems to really be bothering some:
The greatest melt extent over the last 2 1/4 centuries occurred in 2007; however, this value is not statistically significantly different from the reconstructed melt extent during 20 other melt seasons, primarily during 1923–1961.
What this sentences actually says is that 2007 is the highest value on the graph above. It is higher than the magnitude of all other satellite measurements (i.e. blue dots) and also all other reconstructed values (i.e. grey circles). Absolutely true.
Based on this someone might want to decree 2007 a record year for the ice melt index. After all, people are always wondering whether or not we can say something is a record with any confidence. So, can we say that 2007 was a record based on this reconstruction?
The text then goes on to note that this value does not fall outside the ±2σ error bars for the estimate of the melt index for 20 other melt seasons; this means given the uncertainty in the reconstruction, 2007 may not be a record. That is: any one of the other 20 years may well have had a higher melt index than 2007.
The fact that 2007 melt index falls within the uncertainty of our knowledge of 20 previous melt indeces can been seen by noticing that the horizontal blue line I drew falls inside the light grey traces showing the uncertainty in the reconstructed (i.e. estimated) melt during 20 other melt seasons.
The only thing that final sentences is telling us is that 2007 cannot be decreed a record melt year based on this reconstruction. That’s it.
So, are the results and claims of this paper obsolete based on 2010 data?
It seems to me the results and claims are not obsolete.
In the first place the reconstruction back to 1780 cannot be made obsolete by publication of 2010 data. No matter how rapid the 2010 melt, people will be able to refer to the reconstruction and compare future data to that reconstruction.
In the second place: It seems the 2010 melt data used in this paper are still unavailable. Neither Jason Box nor the guys at Skeptical Science have linked to any evidence that the melt data used in this paper are available. This means that the authors of FKM could not add the “blue dot” corresponding to 2010 to figure 2 above. They also could not discuss it in context of their reconstruction and still can’t. So, the paper is not “obsolete” in the sense of leaving out that bit of data.
However, Skeptical Science did find a paper published after the revised version of FKM was submitted which includes satellite melt data for Greenland. Had FKM been based on this particular groups melt data and rather than 3 others, then maybe FKM would be “obsolete” in the sense that the melt in 2010 might have greatly exceeded 2007 and so we would be able to decree 2010 a record, or change any other claim in the abstract of FKM.
So, let’s examine the paper by Tedesco et al (2011) which provides some insight into the melt rate in 2010 relative to 2007. Here is their graph showing the melt index anomaly:
At first glance I couldn’t tell if the standardized melt index in 2010 exceeded the value recorded in 2007. I drew a horizontal line through the 2007 standard melt index anomaly and it does seem to me the melt in 2010 was higher than in 2007. So, the melt in 2010 was, indeed high. But, is this enough to change FKM’s results or specific conclusions?
- Is knowledge of the 2010 melt index likely to drastically change the reconstruction back in time: No. That is, it seems to me the major results will not be obsolete.
- Let’s assume the relative melt indices for 2007 and 2010 reported by the three groups FLM used to create their index will agree with those are show in Tedesco. When that dot is added to FKM2011 figure 2, will we be able to state with confidence the 2010 melt index exceeded the melt index during the previous melt period? Probably not. That blue dot will represent almost the same melt index as for 2007. Given our uncertain knowledge of the melt index values in the past, the slightly higher melting in 2010 relative to 2007 will not be sufficiently high for people to say it set a record for the past 200 or so years.
- If we compute the 10 year running mean adding the estimated melt from 2010, will that value set a record? Nope.
- If we add the extra year to the duration of the melt period will that make the duration of the recent melt period as long as the previous one? Nope. Not even close.
So, whatever gripe people have with FKM2011, it seems to me the claim that it is “obsolete” ought not to be one of them. It doesn’t appear obsolete based on available melt data. After melt data from other groups come out we can examine whether any claim in FKM2010 became obsolete when melt data for 2010 became known. Based on data available now– but unavailable when FKM was submitted — likelihood Box’s claim will be correct seems pretty low. But who knows? When the time comes, we can look at that.
Precisely.
-Chip
“it does not deny the truth of global warming”
It is interesting that a paper irrelevant to the discussion of “the truth” of Global Warming would need to be characterized as having conformed to a dogmatic position about “the truth” of Global Warming.
Good thing we’re used to this kind of nonsense, or we might think think it odd.
Andrew
Chip–
I thought graphs showing how including 2010 melt data might help. It’s worth nothing that neither Box, Skeptical Science nor– now Eli Rabett– seem to actually want to reveal the current estimates for 2010 melt data, show how the estiamtes for 2010 melt data compare to 2007, discuss how incorporating melt data would affect your figure 2 or any specific conclusion in your paper.
Instead we are being shown figures indicating that Greenland’s 2010 summer was warm and and melt was higher than most other years, this will make a difference to your conclusions. Like 2007, Greenland’s summer was warm. Like 2007, Greenlands melt was high. But it doesn’t look like adding 2010 data — had they been available– would make any difference to your results or interpretation of those results.
This year is probably not as warm as last, so likely adding 2011 won’t make an difference. Come 2012, maybe it will. But that hardly seems like a reason to not publish the reconstruction that permits us to place current melt indices in context.
Lucia,
I think the questionable claim the results are ‘obsolete’ is simply an attempt to control the political narrative. If Greenland melt rates are not “much greater than ever, and accelerating rapidly”, then claims of meter-plus sea level rises by 2100 become much less likely; the most credible motive for immediate and drastic public action on CO2 is diminished.
.
All the apoplexia is just politics masquerading as science.
Lucia,
Do you think that had we been able to include the data for the 2010 melt season, and then concluded something like:
“The greatest melt extent over the last 2 1/4 centuries occurred in
20072010; however, this value is not statistically significantly different from the reconstructed melt extent during2015 other melt seasons, primarily during 1923–1961.then Box, Eli, SkepticalScience, et al. would be celebrating our paper?
For all the clamor, you think so, but somehow, I doubt it.
-Chip
Chip–
Uhh… no.
There is something even more mystifying:
Jason Box is a co-author of the Tedesco paper. That paper indicates that the melt index of 2010 barely edges out 2007. So:
* By the time Box wrote his blog post, if read your paper and refreshed his memory as to the results of the Tedesco paper, he should have known adding the 2010 data point would not change your results. Anyone who looked at the results in the Tedesco paper would see that that data will not change the results of your paper.
* At the time he refused to re-review your Dec submission, he must have known that adding the 2010 data would not change your results. It’s true the general public wouldn’t know– but presumably Box, a co-author would have known and it’s hard to believe he would have forgotten the results in a paper going through the review cycle.
* In August 2010, when Box insisted you incorporate the 2010 melt data in your paper, he must have known the 2010 melt index data were not yet available from the team he works with and likely knew the time table for publishing. He should have been perfectly well aware that third parties would not have access to melt index data to use in downstream analyses. His own team would not be publishing their data for months.
It’s all rather mystifying. It’s equally mystifying that commenters at Skeptical Science, Rabett Run etc. seem incapable of looking at what your paper actually says, looking at what the 2010 data really are and seeing that as far as we can tell, adding the 2010 data would not materially change the results of your paper.
Lucia,
You wrote:
But at least the responsible editor at JGR was capable!
-Chip
There was a paper 2 years ago in Geophisical Research Letters titled, “Arctic air temperature change amplification and the Atlantic Multidecadal Oscillation” by Petr Chylek,1 Chris K. Folland,2 Glen Lesins,3 Manvendra K. Dubey,4 and Muyin Wang5
It is quite rmarkable the similarity in curves between the figure you’ve pasted in above and figure 2 and 3 of the Chylek paper. Is there any reason to believe that the cycle of warming and cooling over 60-70 years will not continue?
lucia:
Nothing in the FKM paper would have been found objectionable if the troubling passage in question had been amended as follows:
Genuflection performed. Problem solved. “Nihil obstat” applied by The Consensus. And best of all, 2010 melt data availability no longer matters.
ha lucia, see if skepticalscience will accept your observations about Box?
steven mosher–
I wasn’t previously registered to comment there. I did. While waiting for the email I looked more carefully at Tedesco, got distracted and wrote this post. I’ll go see if their blog shows trackbacks.
Hmmm… not a good test. I typed a rhef= instead of a href= . I fixed the link. Presumably WP will send out a ping on the update.
Chip –
This is somewhat off-topic, in that it doesn’t touch upon the relevance of 2010 data to the conclusions of FKM2011. I can add nothing to what Lucia and yourself have written.
At Dr. Box’s site, the question is raised as to whether melt extent is a useful metric. In one comment, by Dr. Pelto I think, it’s suggested that melt volume (mass balance) is more important, which incorporates both melt extent and melt duration. And in Dr. Box’s latest, he suggests positive degree-days as being a good predictor of melt volume.
As the title of your article clearly indicates that the topic is melt extent, my question is whether any reviewers suggested during the review process that the article was not worth publishing because it focuses on a metric of little utility.
Also, perhaps you could comment on the relative value of melt extent as a metric. Thanks.
HaroldW,
Thanks for question.
Certainly (and as we note in our paper) surface ice melt is but one component of the more complete picture of the surface mass balance (SMB) of Greenland (which can pretty directly be tied into sea level change). But, of the various components, surface melt is probably the one most directly influenced by (short time scale) temperature (or incoming radiation) variability. Other components of SMB like total run-off include a precipitation influence and glacial changes are even less well understood (although certainly the temperature plays a significant role).
And since our longest-term direct observations include temperature and NAO (a circulation feature that we think thinks influences melt (in addition to temperature), through an albedo (i.e. radiation) effect (through winter snowfall variability)), it seemed that trying to reconstruct ice melt extent was the best way to go. And we are far from the first group to study surface melt extent in Greenland (as evidenced by the fact that our melt extent index is derived from a combination of melt extent calculations determined by three different research groups, and there are others out there as well).
All that being said, a paper has recently been drawn to my attention that was published in 2009 by Wake et al. (which Jason Box, interestingly, was a co-author). In that paper, a reconstruction of surface mass balance was performed back to 1866. Qualitatively, at least, their SMB reconstruction seem pretty similar in character to our melt index reconstruction, indicating that historically surface ice melt is pretty closely related to SMB variations—a relationship which we assumed in making our speculations about sea level rise in the conclusions of our paper.
I hope this answers your question!
-Chip
I agree with Steve F that the Box et al claims of ‘obsolescence’ are an attempt to control the political narrative. Isn’t this what government ministers always say about any research into the effects of their policies, ie they are based on old figures (which are the only ones available of course) and that the situation has changed in the last however long?
Strange then that Box et al were not critical of Steig’s Antarctica paper which took almost a year to reach publication and was obviously based on data much earlier than that.
Indeed one could say that they are, in a sense, effectively saying all scientific research is obsolescent since new data is always available:-)
A few comments.
Mass balance is a useless measure since we (including Pelto) have no idea what the “balance” is to start with. It is artificial estimate of “something like melting” but there is no “mass” measurements behind it. Pelto goes out and measures the melt flow off glaciers for a short time and makes a call on the amount of mass lost during the year. Nobody is measuring the actual mass in the Greenland glaciers.
Secondly, the Chylek paper and the AMO. At the peak of the melt season in July and August of 2010, the AMO was at the third highest in history (with 1937 and 1878 being higher). So, assuming there was a large melt in 2010, the relationship with the AMO found by Chylek continued into 2010. It has now gone down to close to Zero or a normal level in its 600 year history and should go a little lower in 2011 yet – a lagged response to the ENSO.
http://img718.imageshack.us/img718/1132/amoapril2011.png
http://img818.imageshack.us/img818/1042/weeklyamogulfstreamapr2.png
question:
“The greatest melt extent over the last 2 1/4 centuries occurred in 2007; however, this value is not statistically significantly different from the reconstructed melt extent during 20 other melt seasons, primarily during 1923–1961.”
How high would 2010 have to be to be statistically different from the reconstructed melt?
Steve Mosher–Eyeballing the ice melt extent difference graph from FKM it looks like a satellite melt index of about 2.8 would definitely be statistically significantly different from all reconstructed. That is: It would fall outside the uncertainty intervals of all past melt indices. I think everyone would call that a record.
I’d have to do a calculation using the binomial theorem to figure out probabilities to say more. Just up… need coffee. But I’ll do this later.
Ok… I’m sipping coffee. I’ll give an explanation, but I still won’t know the answer because to know the full answer, I need to squint at this.
It would really be better to have the numbers.
Assume we will know the 2012 melt index with certainty because we think satellites are perfect.
Suppose the 2012 melt index falls just at the upper end of ±95% sd for only 1 reconstructed (estimated) data point from the previous melt period. (Examining figure 2, it looks like this would be the melt index for 1929 or so. Let’s just say I’m squniting right and use 1929 as the date.)
We know that the true melt index for that 1 estimated point falls within a range. Owing to the definition of the sd intervals, there is a 97.5% likelihood that the 2012 melt index is higher than the one for 1929 and we’d pretty much call the record broken.
But now suppose the 2012 index falls inside the 95% uncertainty intervals for twodata points. Either of those data points have at least a 2.5% of being lower than the 2012 index, but the chance that both are lower could be as high as (1- 0.975^2)=0.049375. I’d suggest that we do use 1 sided tails for calling the record. If all I had to go on was my eyeballs and a graph, I’d noting 4.9%<5% I'd call this a record-- statistically. (Or, failing that, I'd say: "Chip! Give me the actual p values for the two years so I can do a better computation instead of simplifying! Someone who doesn't want to call a record would definitely want the underlying data because less simplified calculation would likely result in getting a higher probability this is not a record.)
If 2012 falls inside 3 uncertainty intervals, there will be at least a 7% probability it’s not a record. If 2012 falls inside 5 uncertainty intervals, there will be a greater than 10% probability it’s not a record. In statistics even using 1 tailed, no one would ever call this a proven record.
Because 2007 falls inside 20 uncertainty bounds, the probability it is not a record is over 40%. (And the 40% is computed using the simplification that 2007 is just at the 95% uncertainty interval for all 20– which it’s not. So, really, the probability it’s not a record might be quite a big higher. I’d need more detailed data from Chip to say.)
But pretty much: Until the observed ice melt index falls outside the 95% uncertainty intervals for fewer than 5 reconstructed points from the previous melt, you can’t say with confidence that the observed data point falls outside the range from the previous melt periods. When there are fewer than 5, a computation is required. When it’s outside the range of all past reconstructed temperatures, it’s definitely a record (based on information from the reconstruction.)
@Bill Illis
GRACE measures the mass using changes in gravity.
http://www.skepticalscience.com/print.php?r=85
Chip (#75578) –
Thanks for the response, it was very helpful.
Bill Illis (#75586)
“Nobody is measuring the actual mass in the Greenland glaciers.” The GRACE satellites provide an indirect measurement of mass, or more precisely, mass change. See e.g. http://www.agu.org/pubs/crossref/2009/2009GL040222.shtml
One major issue can be generalized to this:
For a proxy reconstruction of a climate record, what is the minimum acceptable practice for including recent data?
FKM2011 is an article that was submitted for review in mid-2010. It discusses data from the summer of 2010, but does not include them in its analyses and graphs. Its reconstruction runs from 1784 to 2009. For Dr. Box, FKM2011 falls well below any acceptable standard in this regard. Many mainstream climatologists and advocates seem to agree with him.
A related issue is at the heart of some of the “Hide the decline” controversies highlighted by the hacked UEA emails. Recent proxy records — treerings, in this case — were available, but in different ways were excluded from various analyses. Other graphical presentations were tided up by removing un-pretty post-1960 traces from figures.
“Recent” for FKM2011 = 2010 (still not finalized).
“Recent” for treerings = ~1960 through the 1990s and 2000s. Thirty to forty years.
Is sauce for one goose also sauce for thirty to forty ganders?
lucia (Comment #75605),
Does this assume the satellite data is perfectly accurate? If there is some uncertainty in the satellite data, would that change the analysis, or is the uncertainty in the satellite data already taken care of by the uncertainty of the reconstruction data? Seems to me it is, but I just want to be sure.
SteveF–
Yes. I’m assuming the satellite melt index is perfectly accurate. I doubt the uncertainty in the reconstruction fully accounts for the uncertainty in the inaccuracy in the satellite melt index for 2007.
I think that neglecting the uncertainty in the satellite melt index is’ likely an ok assumption relative to the inaccuracy in the reconstruction particularly in light of other assumptions I made. But it is true that if we want to account for absolutely all uncertainty down to the gnats-a** we would need to account for that.
If we were comparing 2007 to 2010 satellite records,then we would need to consider the uncertainty in the satellite measurements.
Chip sent me some data, so I’m going to fire up R and s_l_o_w_l_y code something up to find the probability. The estimate will rely on — get this —assumptions. (I’ll assume normal distribution etc.)
As I have tired of repeating myself, I’ll not comment on what I think of the paper or Dr. Box’s review other than to say I am grateful that the review and the graphs of ice melt were called to my attention. It tends to put things in perspective and in this case reaffirm some already held views. Details are less important than what is so obviously there from which anyone can easily draw their own conclusions.
AMac(#75609) –
A better comparison would be to Wake et al. 2009, which produced a surface mass balance history from 1866-2005. From its abstract: “[W]e compare annual SMB estimates from the period 1995–2005 to a similar period in the past (1923–33) where SMB was comparable, and conclude that the present-day changes are not exceptional within the last 140 years.”
Dr. Box apparently didn’t object strongly to omitting the very large melt year 2007 from this analysis, as he is one of the co-authors of this paper.
edit — fixed link
Mosher– I did the calculation to find the probability that 2007 was a record based on the uncertainty in the reconstruction. It’s 11.09692%. I used values Chip sent me. Will post later on….
Lucia, thanks for your usual careful analysis.
“Jason Box’s claim the results of FKM would be obsolete at the time of publication.”
On his blog (“selective science = pseudo-science”) Box claims that FKM was obsolete when it was submitted – in August 2010!
I wonder if he will join the discussion here to justify his claims.
And I wonder if he will post the comments I put on his blog:
Lucia,
perhaps we can ask Chip et al for their code and play around with 2010 values.
steve–
Chip sent me an excel spread sheet for figure 2. People don’t always use R (and they don’t have to.)
excel is ok. as long as that is the tool he used to create the chart and you can recreate it from the tool he sent
Does anybody know if Konrad Steffen is about to give an update?
His last report of Melt Extent was from 2007.
http://cires.colorado.edu/steffen/greenland/melt2007/index.html
steven–
I got he information I wanted. That’s my standard. 🙂
Hans– I don’t know. You’d have to ask Konrad Steffen.
You could start here or here or Chip could just plug the surface temperatures and NAO stuff into his model for a ballpark. Just sayin
Eli—
You using more words might help readers understand what the heck you think one could accomplish by “starting” at those two links. I’m aware of the information at those links.
Eli, you could just kiss me deleted by lucia
Mosh,
Careful there, I hear rabetts have cooties.
Just fleas, Zeke.
No links here to the Watts Magnum Opus. It has been in development, and much talked about, for years. It was finally linked to on Watts own web site with barely a word spoken, and disappeared off the front page soon after. I would have thought that his pride and joy would have deserved more.
bugs–
I’m not entirely sure what’s bothering you.
This may come as a surprise to you, but I don’t read Anthony’s blog every single day. Now that I looked, I notice Anthony posted yesterday. It seems to be rolling down at the natural rate of roll-down on blogs.
I’ll have to get the paper and read it. But you’ll note that I mentioned on another thread that I’ve been busy trying to figure out something of interest to me.
I suspect Zeke may have read Anthony’s paper since Zeke is interested in computation of the temperature series and will want to read what they specifically discovered. I’ll probably write a “congratulations post”.
The Knappenberger/Michaels mode of operation seems to be to hunt around for something that suits their expectations. Takes a while, but they find a statistical tid bit here and there. As time passes, their hunting expeditions come up with less and less.
I wondered what happened to the ol’ moshpit.
It was beginning to look like he was taking this “published author” thing a bit too seriously, but glad to see he’s back!
Re: John M (May 20 16:36), WTF
Haha I got deleted.
eli can kiss my hairy [snip by moshpit]
bugs (Comment #76093),
Say what? Try for once to get past the politics and address the technical issues. If you have something of technical substance to say, then I am certain you could shock many (if not all!) by making a substantive technical comment. I note that when Steig et al published their Antarctica reconstruction in Nature the focus of Id, McIntyre, O’Donnell, and others was on the erroneous mathematical treatment, not if Eric is/was a bad person. Maybe you could raise you sights to something of substance as well.
LOL. This is only one CA topic attacking Steig the man.
http://climateaudit.org/2009/08/14/steigs-mystery-man/
Did you actually read it bugs? – it’s really an analysis of an interesting pattern of events.
It’s about Steig the man, and not the science. There are several more topics in a similar vein. SteveF is asking for the debate to be lifted above the politics, such as McIntyre does. McIntyre’s main game is the politics.
bugs – quote from your link to Steve’s article:
“Hu McCulloch posted on Steig’s failure to allow for serial correlation in their confidence interval calculations (even though they had said that they had done so.) The recent Steig corrigendum made precisely the same correction – without acknowledging McCulloch.”
Sounds like it is about the science and process to me – got any quotes from that article that are about the “man”? Or do you just want to do political posts?