In a blow to the Real Climate “hockey team” one team member’s paper, Steig et al Nature, Jan 22, 2009 (seen at left) has been shown lacking. Once appropriate statistical procedures were applied, the real data spoke clearly, and it was done in a peer reviewed paper by skeptics. Jeff Condon of the Air Vent writes via email that he and co-authors, Ryan O’Donnell, Nicholas Lewis, and Steve McIntyre have succeeded in getting a paper accepted into the prestigious Journal of Climate and asked me to re-post the notice here.
The review process was difficult, with one reviewer getting difficult on submitted comments [and subsequent rebuttal comments from authors ] that became longer than the submitted paper, 88 pages, 10 times the length of the paper they submitted! I commend them for their patience in wading through such formidable bloviation. Anyone want to bet that reviewer was a “team” member?
As WUWT covered in the past, these authors have demonstrated clearly that the warming is mostly in the Antarctic Peninsula. Steig et al’s Mannian PCA math methods had smeared that warming over most of the entire continent, creating a false impression.
WUWT visitors may want to read this primer which explains how this happens. But most importantly, have a look at the side by side comparison maps below. Congratulations to Jeff, Ryan, Nick, and Steve! – Anthony
Jeff writes:
After ten months of reviews and rewrites we have successfully published an improved version of Steig et al. 2009. While we cannot publish the paper here, we can discuss the detail. Personally I’ve never seen so much work put into a single paper as Ryan did and it’s wonderful to see it come to a successful conclusion. This is the initial post on the subject, in the coming weeks there will be more to follow.
Guest post by lead author Ryan O’Donnel.
——–
DOING IT OURSELVES. . . a tongue-in-cheek reference to the RC post here:
Improved methods for PCA-based reconstructions: case study using the Steig et al. (2009) Antarctic temperature reconstruction
(Accepted 11/30/10, Journal of Climate)
Ryan O’Donnell Nicholas Lewis Steve McIntyre Jeff Condon
| Abstract
A detailed analysis is presented of a recently published Antarctic temperature reconstruction that combines satellite and ground information using a regularized expectation-maximization algorithm. Though the general reconstruction concept has merit, it is susceptible to spurious results for both temperature trends and patterns. The deficiencies include: (a) improper calibration of satellite data; (b) improper determination of spatial structure during infilling; and (c) suboptimal determination of regularization parameters, particularly with respect to satellite principal component retention. We propose two methods to resolve these issues. One utilizes temporal relationships between the satellite and ground data; the other combines ground data with only the spatial component of the satellite data. Both improved methods yield similar results that disagree with the previous method in several aspects. Rather than finding warming concentrated in West Antarctica, we find warming over the period of 1957-2006 to be concentrated in the Peninsula (≈0.35oC decade-1). We also show average trends for the continent, East Antarctica, and West Antarctica that are half or less than that found using the unimproved method. Notably, though we find warming in West Antarctica to be smaller in magnitude, we find that statistically significant warming extends at least as far as Marie Byrd Land. We also find differences in the seasonal patterns of temperature change, with winter and fall showing the largest differences and spring and summer showing negligible differences outside of the Peninsula.
Copyright © 2010 American Meteorological Association (early online release to be available on or around Dec. 7th) |
Temperature trend Deg C/Decade Click to enlarge.
Some of you remember that we intended to submit the analysis of the Steig Antarctic reconstruction for publication. That was quite some time ago . . . and then you heard nothing. We did, indeed, submit a paper to Journal of Climate in February. The review process unfortunately took longer than expected, primarily due to one reviewer in particular. The total number of pages dedicated by that reviewer alone – and our subsequent responses – was 88 single-spaced pages, or more than 10 times the length of the paper. Another contributor to the length of time from submission to acceptance was a hardware upgrade to the AMS servers that went horribly wrong, heaping a load of extra work on the Journal of Climate editorial staff.
With that being said, I am quite satisfied that the review process was fair and equitable, although I do believe excessive deference was paid to this one particular reviewer at the beginning of the process. While the other two reviews were positive (and contained many good suggestions for improvement of the manuscript), the other review was quite negative. As the situation progressed, however, the editor at Journal of Climate – Dr. Anthony Broccoli – added a fourth reviewer to obtain another opinion, which was also positive. My feeling is that Dr. Broccoli did a commendable job of sorting through a series of lengthy reviews and replies in order to ensure that the decision made was the correct one.
The results in the paper are generally similar to the in-process analysis that was posted at CA and here prior to the submission. Overall, we find that the Steig reconstruction overestimated the continental trends and underestimated the Peninsula – though our analysis found that the trend in West Antarctica was, indeed, statistically significant. I would hope that our paper is not seen as a repudiation of Steig’s results, but rather as an improvement.
In my opinion, the Steig reconstruction was quite clever, and the general concept was sound. A few of the choices made during implementation were incorrect; a few were suboptimal. Importantly, if those are corrected, some of the results change. Also importantly, some do not. Hopefully some of the cautions outlined in our paper are incorporated into other, future work. Time will tell!
Lastly, I’ll give a shout out to other folks whose comments helped shape the paper by their comments and analysis. In particular, Roman, Hu, and Carrick . . . thanks!
Congrats! For as long as I’ve been following this on the web, for one from the “jesters” to be published… this feels like a real breakthrough.
PS. You do realize, if he doesn’t find a way to nit-pick this paper to death (treatment they never seem to give to the “Team” work), Gavin is going to take credit for pushing you guys to do this. Just wait!
The introduction article is very enlightening. http://wattsupwiththat.com/2009/02/28/steigs-antarctic-heartburn/
I think Ryan is being far too polite (I guess intentionally, so as not to be confrontational and get this important paper published.).
Without repeating the details shown in the above link , Steig 09 was an awefully sloppy piece of work. The most favourable way to view it they got the results they wanted with 3PCs and did not go any further, though it takes a load of credulity to not to think they tried using more.
Almost doubling the period of the real data by projecting backwards in time as part of a calibration control is almost humorous.
Seeing the net result was that there was not any warming on the Arctic peninsula when this was where all the data was and clearly showed warming … Well why bother publishing known falsehoods if you are not out to deceive?
So hat’s off the authors of this paper for some credible analysis that actually ties up with reality. Congratulations for your perseverance and the fruit it has borne.
I hope this signals the beginning of an acceptance of real science into peer reviewed journals. Long over due.
Hmm I cant find the Arctic peninsula , I must have meant Antarctic peninsula 😉
John Whitman says:
December 2, 2010 at 7:46 am
“Ryan, Nic, Steve & Jeff….Next step is inclusion in AR5. : ) John”
With the current state of affairs at CoP16, will there be an AR5?
londo says:
December 2, 2010 at 4:42 am
Any sort of statistical detection/estimation process suffers the same problem. It’s not so much the method itself as application of the method. PCA is not some magical algorithm, in fact, it is quite common in a variety of fields. Indeed, solutions for PCA include eigenvalue decomposition, singular value decomposition, and even more recent developments such as the PAST algorithm which is used in communications applications.
Mark
While not the best reference for detail regarding PCA, “Independent Component Analysis” by Hyvarinen, Karhunen, and Oja contains a brief overview of PCA as a pre-cursor to implementing ICA. They specifically mention some of the above, in particular, use of the PAST algorithm described by Yang in 1995.
Mike says:
December 1, 2010 at 8:29 pm
Very interesting. I have no idea who is right. But, both papers conclude Antarctica is warming.
Here’s a hint. It could be because…
http://notrickszone.com/2010/11/10/a-wind-in-antarctica/
toto says:
December 2, 2010 at 1:29 am
Your link; I loved this comment…
John Philip says:
1 June 2009 at 11:21 AM
Thanks for a very clear explanation of the choice of 3 PCs. It exposes this (link to WUWT) as the [poorly edited and ungrammatical] nonsense that it undoubtably (sic) is …
Complaining about grammar on WUWT and he can’t spell!
To everyone saying “rebuts” conflicts with Ryan O’s call to not “repudiate” Steig, get a dictionary, because the two terms do not have the same meaning. A “repudiation” means that the paper has no merit, that all conclusions are wrong. A “rebuttal” means a counter argument or proposal, showing an alternative viewpoint. Anthony’s choice of terms is completely correct and proper.
Here are the definitions from Dictionary.com:
Rebut –verb (used with object)
1. to refute by evidence or argument.
2. to oppose by contrary proof.
–verb (used without object)
3. to provide some evidence or argument that refutes or opposes.
Repudiate –verb (used with object), -at·ed, -at·ing.
1. to reject as having no authority or binding force: to repudiate a claim.
2. to cast off or disown: to repudiate a son.
3. to reject with disapproval or condemnation: to repudiate a new doctrine.
4. to reject with denial: to repudiate a charge as untrue.
See! Not synonyms, and the proper use was chosen by Anthony.
Well done. I will read the paper once it becomes available online.
I am surprised that no one has commented further on Roy Martin’s piece about the volcanism in the Antarctic and the Peninsular in particular. That the Antarctic Peninsular area is warmer than other nearby regions is well known and, from the link provided and elsewhere had been warming in recent times.
The role of volcanic activity is significant. The crescent shaped Deception Island was mentioned and with a photograph as being a partly submerged caldera. Visiting tour ships operators encourage passengers to swim in the warm water of the ‘bay’ or inlet, as my wife reported from her trip there in December 2005. The position of the warm area of water does move about from time to time.
It is a puzzle that Nature publishes papers which are contrary to known facts about regions of the world without due diligence and challenge by the so-called “peer” reviewers.
Roy at 1:10 am:
‘It would be interesting if an expert on the sociology of science (such people do exist) were to investigate the peer review process for papers on climate change and compared sceptical papers with those upholding the “consensus” view to see if double standards applied.’
I think a number of cases are accumulating, and the analysis might best be done with detailed case studies rather than a necessarily more superficial quantitative analysis. One thinks of my colleague Garth Paltridge who described at CA his experience in trying to get published that looked at 50 years of radiosonde data looking in vain for evidence that water vapour had increased with temperature (the key positive feedback mechanism in the models), only to be dismissed by one reviewer (effectively) for trying to undermine the consensus. Add the Climategate e-mails on trying to keep such research out of the journals (and out of the IPCC reports) and there is the basis for a good paper there.
It all fits with the views of philosopher of science Paul Feyerabend that scientists behave like a cartel or club and defend their paradigms by fair means or foul. It might not matter much with some science, but in this case they are demanding society invests billions on the strength of it — and for this reason we have every right to insist on the openness and repeated attempts at falsification Karl Popper recommended. (It was not accidental that his philosophy of science resonated with his political philosophy in The Open Society and Its Enemies and The Poverty of Historicism).
I covered a lot of this in my (pre-Climategate) book Science and Public Policy: The Virtuous Corruption of Virtual Environmental Science, which has now been reduced in price to $40 (down from $110). For ordering details (and some amusing discussion with an environmental activist) see:
http://rogerpielkejr.blogspot.com/2010/11/aynsley-kellows-science-and-public.html#comments
In Canada, what the Globe and Mail wrote back then about Steig et al. 2009:
“Antarctica not as chilly as thought
MARTIN MITTELSTAEDT
Globe and Mail Update
January 21, 2009 at 1:02 PM EST
For years, Antarctica has been considered the one part of the world that global warming has left untouched.
Scientists thought the icy continent at the bottom of the world was experiencing stable temperatures or even cooling slightly. This was considered an unalloyed good thing because it meant Antarctica’s massive ice sheets weren’t in danger of melting, unlike the rapidly disappearing sea ice around the Arctic.
But now it appears that the view on the stability of Antarctica’s climate may have been wrong.
A new study, using the most sophisticated analysis of satellite and weather station temperatures to date, concludes that Antarctica has actually grown warmer at a pace similar to the rest of the southern hemisphere – about 0.5 degrees Celsius over the past half century.”
Not laughing already?
”We can’t be quite as comfortable in thinking, ‘Oh, at least Antarctica … is safe because it’s getting cold,’ because it’s unfortunately not getting colder, but getting warmer like everywhere else,” said Drew Shindell, an atmospheric scientist at the NASA Goddard Institute for Space Studies and one of the researchers who conducted the study.
Although the research, appearing in Thursday’s edition of Nature, does not blame global warming for the rise, it says the trend ”is difficult to explain” without pointing to the rising concentrations of greenhouse gases in the atmosphere related to human activity.”
The Globe’s usual suspect innuendos…
“Getting a fix on the trend in Antarctic temperatures has been one of the most important issues facing climate scientists. The continent contains so much ice that any substantial warming threatens to raise sea levels, although it would likely take centuries to swamp coastal cities. The smaller of the two Antarctic ice sheets would raise oceans by about 6 metres if it melted entirely.”
Fearmongering…
”It’s a long-term threat, but given the magnitude of the threat, I would find it worrying” that temperatures are rising, Mr. Shindell said.
Antarctica, as the only formerly uninhabited continent, does not have a long history of climate data and accurate records from weather stations have only been available since 1957. Most of the stations are located along the coastline, and do not give much of a clue about what has been happening to temperatures in the vast interior. Satellite data have been available only since the 1970s.
The weather stations indicate that the Antarctic Peninsula, the long finger-like projection that points to South America, has grown warmer, but many other coastal stations have had slight cooling, leading to a belief that most of the continent was not being affected by global warming. Scientists also have found that the depletion of the Antarctic’s ozone has contributed to cooling since it developed in the 1970s.
Based on these observations, the authoritative Intergovernmental Panel on Climate Change, the UN scientific group that monitors global warming, concluded in 2007 that the continent had experienced ”little change” during the last half of the 20th century, except for the peninsula where warming was ”pronounced.”
But Mr. Shindell said that if the new study had been available earlier, it is likely the IPCC would not have been so sanguine about the state of Antarctica’s climate.
The IPCC said it was confident that six of the world’s seven continents were getting warmer, Mr. Shindell said, adding that ”I think that conclusion will now be changed to all seven.”
Sound: trumpets of apocalypse…
Because Antarctica has had only about a half century of weather station data, the researchers were unable to rule out the possibility that the raising temperatures they calculated were related to some normal, natural fluctuation in its weather patterns.
The research was conducted by a team that included scientists from a number of U.S. universities and government laboratories.
The group made their warming estimates by comparing data from both satellites and ground weather stations.
They found that temperatures from the stations and the satellites corresponded, a finding that allowing them to infer temperatures in the vast areas of the continent that don’t have weather monitors.
They used a similar approach to deduce temperature trends throughout the continent between 1957 and when satellite data became available.”
All this gearing up for Copenhagen…
Oh and the Globe and Mail, Canada’s national newspaper, took over two weeks to acknowledge climategate in their webedition…
Notice how quickly the Globe article was published… january 21
Let’s see how fast the O’Donnell et al. 2010 paper will be picked up by the Globe at publication…
I do think we need more research into the cause of the antarctic peninsula warming.
Is it due primarily to adiabatic warning from wind descending from the interior plateau or due to volcanic heat? The concurrent ocean warming makes one wonder whether the ocean or the winds are the culprit.
I like Ed Caryle’s thesis of heating due to descending air masses. We know that mid latitude deserts around the world are due to descending air masses associated with Hadley cells which originate as ascending air masses or evaporative cooling at the ITCZ to the south and are completed with surface trade winds from ne to sw. Descending air releases heat and lowers relative humidity resulting in reduced cloud cover and associated solar heating.
The question I have for Ed is can such a heat pump produce enough joules to actually increase OHC adjacent to western antarctica and the Peninsula?
Don’t look now, but it is MY opinion that the warming is largely confined to the Palmer Peninsula.
========================
Well done Steve. Nice job…
dbleader61 says:
December 2, 2010 at 11:33 am
With the current state of affairs at CoP16, will there be an AR5?
————-
dbleader61,
Indeed, no AR5 (or at the least a significantly postponed one) would be entertaining indeed.
John
The importance of this rebuttal or whatever you want to call it becomes even more obvious when you re-visit the Guardian’s take on the Steig et al. article when it was published – “Research ‘kills off’ climate sceptic argument by showing average temperature across the continent has risen over the last 50 years.”
http://www.guardian.co.uk/environment/2009/jan/21/global-warming-antarctica
Presumably various other media outlets had similar reports.
In response to
Mike says:
December 1, 2010 at 8:29 pm
“Very interesting. I have no idea who is right. But, both papers conclude Antarctica is warming.”
Actually I think this says when you correct the most obvious errors, such as selection bias and infilling errors, than most of the warming vanishes. It does not yet address the question of whether the method has other errors or even works at all.
sHx says: December 1, 2010 at 11:49 pm
Unless Anthony Watts is setting up the likes of Joe Romm and Gavin Schmidt for a ‘gotcha’ moment, a correction to the title of the post is warranted.
To qualify to be in their group, he would need to use the word repugnant.
Richcar 1225 you may want to update your understanding of circulation…
The warming in the peninsula was well known before Steig et al. 2009 (see Leroux various books, the latest being Dynamic Analysis of Weather and Climate, 2ed. Springer 2010) and can easily be explained through atmospheric circulation dynamics as warm air moist advections are channelized by relief as a result of stronger catabatic anticyclones coming out of Antarctica. In fact the same dynamic can be noticed along the Chilean coast where temperatures have been cold at sea level -dense cold air- while temperatures have risen in the mountains where warm air is advected against relief as the strong mobile anticyclone progresses north. It is also these warm air advections on the WA peninsula that bring record snowfalls there. Dr. Steig should at least know that depressions affecting the area are deeper and their frequency has been increasing, hardly a sign of global warming!
Hi Ryan,
I’m hoping that you’re still reading comments and might answer the couple of questions that I’d had, and then consider a request…
I’d asked if there was any idea of why the larger cooling area exists within the continent, if there are topological features or something else that might account for it?
I’d also asked if your improvements could be applied to how Arctic temps are being calculated – and my apologies, because I have no idea if some of the same Steig methods are even being used for Arctic calcs….
Oops, I just thought of a third question – are the Steig calcs/methods being used by NASA GISS HadCrut, etc? In other words, does the Steig data/calcs/methods affect the ‘official global temps?’
Then finally the request – if it’s correct that the majority of the warming has occurred decades ago as a few others have posted, could you favor us with at least the years that warmed along with how much (deg C/decade or year), and when that warming turned to cooling along with how much cooling from then to present day? I understand totally if that’s not possible because of publication requirements, but am hoping you’re not constrained in this regard and might be willing to post this info.
Thanks so much, and again, many kudos on your paper and it’s publication!!!
[FWIW: Apostrophe removed. “Kudos” is both the singular and the plural form. ~dbs, mod.]
re moderator note: FWIW: Apostrophe removed. “Kudos” is both the singular and the plural form. ~dbs, mod.]
Thanks dbs – spelling has never been my forte, that’s for sure. I’ve no idea why I automatically put the apostrophe in, but hopefully I’ll remember in the future.
[Reply: lots of people make that misteak. I was just performing a public service. ~dbs]
I haven’t seen a misteak in AGES dbs – when I was a little kid, I had a ginormous eraser – like 6 inch x 3 inch x 1 inch, with MISTEAK printed across it. Thanks for the grin!
Ryan and his co-authors deserve the highest accolades for not only for fixing the lapses in Steig’s Antarctic reconstruction, but for getting their paper past a hostile reviewer. I suspect that the latter was accomplished because they concentrated on methodological issues, rather than confronting the more substantive issue of data manufacture.
PC methods work most reliably in the context of multiple (overlapping) data sets that may be somehow flawed, but in the aggregate contain ample information about the signal components. And derived EOF’s can be very powerful in compactly identifying complex modes of oscillation where such exist in a naturally defined space (e.g., a basin of water). But neither of these conditions applies to Antarctic temperatures, where only a handful of widely scattered station records cover more than 50 years. Aside from the recent satellite data, most of the other stations have significant gaps in their time series and average only ~30 yrs in length. Thus any EOF reconstruction of a 50-yr gridded record relies largely on data that is not a product of measurement, however flawed, but of manufacture by the algorithms employed.
Despite the beguilingly detailed maps provided, no one with any sense of the dimensions–temporal and spatial–over which terrestrial temperatures are strongly correlated would take any such “recontructions” to the bank as physical reality. Whether Antarctica cooled or warmed on the whole in the recent past thus remains an open question. It is further complicated by the confounding mathematical effects of natural multidecadal oscillations upon multidecadal regressional estimates of trend.