In my last post, I observed that Ellen Mosley-Thompson’s archiving record was even worse than that that of her husband, Lonnie Thompson, whose failure to adequately archive his ice core measurements has long been a subject of criticism at Climate Audit. In particular, I observed that Ellen had archived nothing from the 15 expeditions to Greenland and Antarctica that, according to her CV, she had led.
This post has occasioned fresh commentary on Lonnie Thompson’s archiving, which I will review in today’s post. In particular, I will assess Thompson’s statement in an email to a CA reader stating:
…our ice core data are archived at the World Data Center NOAA Paleoclimate data base in Boulder Colorado…
Despite Thompson’s claim, no data whatever is archived for many ice cores. For other cores, my issue is that the Thompson archive is completely inadequate, as I’ll discuss below. I remain mystified by Thompson’s intransigence in establishing a comprehensive and meticulous archive of his measurement data, as, in my opinion, he should regard the establishment of such an archive as an essential part of his scientific legacy and his #1 priority given his age and health.
Cores With No Archive Whatever
As noted in my recent post, according to her CV, Ellen has led “nine expeditions to Antarctica and six to Greenland to retrieve ice cores”. Antarctic sites include Plateau Remote, Dyer Plateau and Siple Station; Greenland sites include those in the PARCA program e.g. GITS, D2, D3, Raven, Tunu. Despite Lonnie Thompson’s claim that “our ice core data are archived at the World Data Center NOAA Paleoclimate”, no data from any of these Ellen-led expeditions has been archived at the NOAA Paleo website.
In my earlier post, I noted that one Greenland data set associated with Ellen had been archived, but pointed out that this came from a much earlier expedition that she had not led and that the data had been transcribed by third parties. For further clarification of this, in 1966, the first long ice core was drilled in Greenland at Camp Century (1387.4 m to bedrock). Lonnie Thompson studied dust concentrations in this core as part of his 1977 thesis entitled Microparticles, Ice Sheets and Climate (Ohio State University Institute of Polar Studies Report 64). These results were published in an academic journal in 1981 by Lonnie and his wife, Ellen Mosley-Thomspon as Temporal variability of microparticle properties in polar ice sheets (Journal of Volcanology and Geothermal Research 11). In 1990, this data was transcribed at WDC/NSIDC and eventually archived at NOAA (with the most recent deposit dated in 2007). This is the only archived Greenland/Antarctica associated with either Thompson and, as I stated in the earlier post, does not come from one of the 15 expeditions to Antarctica/Greenland led by Ellen Mosley-Thompson.
Subsequent to this early consideration of dust in the Camp Century, Greenland ice core, Lonnie’s career, as is well known, has been devoted to high-altitude tropical ice cores. In the late 1980s and 1990s, Thompson drilled a series of tropical ice cores: Quelccaya 1984, Dunde 1987, Guliya 1992, Huascaran 1993, Sajama 1997, Dasuopu 1997, Kilimanjaro 2000 and Puruogangri 2000. Each expedition was punctuated with a short article in Science, none of which, in my opinion, can really be considered as a comprehensive technical report on the ice cores.
Since 2000, Thompson has conducted expeditions to Bona-Churchill (two cores 2002 – NSF award here), Quelccaya and Coropuna (four cores 2003 – NSF award here), Puncak Jaya, New Guinea and Nevado Hualcán, Peru (four cores 2007 – see NSF award here and most recently Alto dell’Ortles, Italy (see NSF award here). See Thompson’s NSF award history
here.
Nothing has been archived at NOAA from any of these expeditions despite Thompson’s assertion that “our ice core data are archived at the World Data Center NOAA”.
NSF funding for Puncak Jaya, Hualcan and Ortles has not expired (it continues in each case to March 2013), but NSF funding has expired for the other three programs (Bona-Churchill – May 2006; Quelccaya/Coropuna – June 2007; Naimona’nyi – March 2010). Nor has there been any publication of the Bona-Churchill ice core, a lacuna noted at CA for a number of years – see here.
Cores with Very Defective Archives
Let me now turn to the archival situation for the early cores, where Thompson has uniformaly failed to provide anything like a comprehensive or definitive archive of measurement data, but has (in some cases, grudgingly) provided digital versions of a single figure in an article, which Thompson has then attempted to pass off as an adequate archive.
Thompson’s NSF award for Quelccaya/Coropuna contains a relevant description of the major component of an adequate archive (though, needless to say, Thompson has not established such an archive for these cores):
This award will help obtain and analyze four ice cores to bedrock from the Quelccaya and Coropuna ice caps in the Peruvian Andes. These new cores, covering the last 10,000 years will provide a more robust and detailed history of Pacific sea surface temperature variability than has been possible with previous cores. Approximately 6,000 samples will be obtained from each core for stable isotopes of oxygen and deuterium, insoluble dust concentration and size distribution, and soluble aerosol chemistry analyses. These and other proxy measurements will be used to reconstruct a regional temperature and precipitation history and produce a high-resolution record of El Nino Southern Oscillation (ENSO) variations
This is precisely in accordance with the sort of definitive archive that I’ve advocated for a number of years. Each Thompson ice core has several thousand samples, on which various measurements are typically taken: O18, dust, chemistry, etc. If Thompson’s ice cores as important as represented, then they deserve nothing less than a definitive archive i.e. the measurements for all 6000 or so samples for each core. These are not “large” archives. I presume that much of the data is already in digital form somewhere at Ohio State.
Other information that is required for a definitive archive is detailed information on layer thickness (including photographs of ice core) used to date the core. The dating of rapidly thinning ice cores is subject to a variety of uncertainties. Subsequent investigators should be able to re-examine Thompson’s dating and, if they disagree, they should have the information that enables them to reassess the isotope history. (An example of such a re-consideration is Vinther’s recent reassessment of the 1970s and 1980s vintage Agassiz cores in light of information from more recent Greenland drilling.)
Instead of this sort of definitive archive, Thompson’s archive at NOAA for Dunde, Guliya and Dasuopu were, until a few months ago, nothing more than decadal O18 chronologies for the period 1000-1990 with no detailed sample information, no ancillary dust, chemistry, nothing on deeper portions of the core.
Aside from mere craftsmanship, there is an important additional reason why Thompson should be expected to archive all measurement data for these cores. In multiproxy studies, the raw measurement data (by sample) is combined with the dating assigned to the samples to produce (for example) isotope “chronologies” (to borrow the term used by dendros). As discussed at CA on numerous occasions – see, for example, here, Thompson has published or distributed inconsistent chronologies (see the figure below.) Some of the inconsistencies are hard to understand. The Dunde version of Thompson et al (PNAS 2006) is inconsistent with the version in Yao et al (Ann Glac 2006), of which Thompson was coauthor. The Dunde version of Thompson et al (PNAS 2006) is consistent with the version used in Mann et al (1998), but not with the intermediate Thompson et al (Clim Chg 2004).
Dunde Versions – see here.
Correspondence with Thompson and Journals
Until 2003, Thompson had archived nothing from any of his Himalaya cores (Dunde, Guliya, Dasuopu), all of which have been important in the paleoclimate discussion. In October 2003, shortly prior to publication of our 2003 article on MBH, I requested data on these cores from the Thompsons, but got nowhere.
In 2004, I was asked to review a response by Mann et al to our 2003 article. In that capacity, I asked to see supporting data for their submission (data that they had refused to supply to me late in 2003.) Stephen Schneider said that, in his 28 years as editor, no reviewer had ever requested supporting data. I replied that times change and that I wanted to see the supporting data. Schneider said that this would require approval by their editorial board. I asked that this be obtained. This incident is documented in 2004 Climategate emails. As a result, Climatic Change adopted a policy requiring authors to supply supporting data, but not code. I then insisted that Mann supply the requested supporting data; Mann appears to have withdrawn the submission rather than comply with the new data policy.
In March 2004, I requested supporting data for Thompson’s 2003 article under the new policy, a request that was supported by Climatic Change. However, instead of providing a comprehensive and definitive archive, Thompson simply archived decadal O18 chronologies for Dunde core 3, Guliya and Dasuopu for the top part of the cores from AD1000 to 1990.
Thompson has subsequently pointed to this absurdly inadequate archive as evidence that he has “archived his data”. However, for the various reasons set out above, archiving the published decadal O18 histories for these sites is only one small component of a definitive archive.
Over the years, I’ve attempted to persuade both Thompson himself and the various journals that a definitive archive is both required and long overdue. I’ve collated this correspondence here. In 2004 and 2005, I tried to persuade Climatic Change. In 2005 and 2006, I tried to persuade Science to require Thompson to provide a definitive archive of cores that Thompson had published in Science. This resulted in them supplying sample information for two Kilimanjaro cores that were interesting (see contemporary CA discussion here) but not relevant to my request for sample data from Dunde, Guliya and Dasuopu. I then had a lengthy correspondence with Science re-iterating my request for details on these three cores; the editors said that they corresponded with Thompson on the matter, but the requests were stymied and ultimately the Science editors stopped answering without ever answering or resolving the matter.
In 2007, I tried to obtain measurement data for the three cores once again with both the editor of PNAS and then Ralph Cicerone, the president of NAS. Once again, the correspondence went nowhere. I asked for an archive of sample data, carefully explaining the difference between sample data and a decadal chronology. PNAS responded that Thompson had told them that he had already provided what I asked for and that I should simply go to the NOAA site – a site that I was obviously well aware of and which I had distinguished in my initial response. And which obviously didn;t contain the sample data that I had asked for. Both Thompson’s unresponsive answer and PNAS’ seemingly wilful obtuseness to the problem (think Gavin Schmidt or William Connolley on upside-down Tiljander) are all too characteristic of the climate community and have contributed to the present adverse atmosphere.
Rather than enforce journal policies, Cicerone told me to write Thompson, who had refused for years to provide this data. So I wrote to Thompson one more time, this time copying Cicerone, Gerald North, Brooks Hanson of Science. Once again, Thompson ignored my email.
In comments on the prior thread, CA reader Roger has criticized me for being insufficiently diligent in trying to obtain measurement data from Lonnie Thompson. I believe that the attached correspondence completely refutes his criticism.
As I observed above, it remains a source of great puzzlement to me why Lonnie Thompson does not regard the establishment of definitive archives of his data as an integral part of his scientific legacy. At this point, it is evident that it will be difficult, if not impossible, for him to carry out the definitive statistical analyses of his data. Over the past 10 years, he has fallen further and further behind. The logical solution to his quandary is simply to provide a comprehensive and definitive archive and thereby let any interested scientist carry out the analyses that Thompson himself will either never carry out or not carry out for many years. Nor in my opinion should Thompson, at this age and stage of his career, spend his time haring off to yet another drill site. No vice president of exploration in a mineral exploration company would personally feel obliged to sit on a drill rig. Surely there are other people capable of supervising the next drill program. Thompson should let them do their job.
Climate Audit 
138 Comments
Outstandingly logical – thank-you for continuing to pursue this subject matter. Sorry, but I could not open the “attached correspondence” – got an error message.
Link should be fixed now,
There’s an answer, but one you’ll abhor:
That the archive exists for each core
And is centrally stored
But the Team that’s On Board
Are the ones that this data is for
So no one chases LT, that’s plain
(And the versions are hard to explain)
But his legacy/fame
Is the use of his name
In support of the “burning world” flame
===|==============/ Keith DeHavelle
Reblogged this on Climate Ponderings.
Simply incredible. Steve, you deserve immense credit for perservering amidst so much instransigence and worse. What would be said of researchers in medical or life sciences who were so obstructive about allowing review of their data? Without commenting upon motives let’s simply say the behaviors are appalling and unscientific.
How can this dismal situation be allowed to persist for so many years? Sporadic provisions of “grey” data privately to favored researchers is wholly unacceptable. The Thompsons (and all similar researchers) should either be required to archive comprehensive data or else have their “research” off-limits to scientific citation.
Ditto.
Here is my theory. -snip – , he is well aware, from the hockey stick wars, that archiving the data would open it to intense scrutiny, and the drmatic conclusions on which his current legacy is based might be undone. Looking back, I’ll bet he realizes how sloppy some of the analysis was done and fears this will be the case. Best course then is to ride this out and continue to put his trust on the media-climate academic complex.
This scandal involves much more than the Thompsons, of course. For instance, Ohio State U. should be pressed to employ a couple of grad students to round up and archive all data that is in digital form already. How hard can that be?? The Byrd Center at Ohio State should be pressed (maybe by alumni plus the NSF) to get all data archived. Their reputation and the scientific legacy of ALL concerned deserves to be severely criticized if this squalid situation persists.
“Byrd Polar Research Center” is the full name:
http://bprc.osu.edu/
Lonnie Thompson is scheduled to give the keynote address at a conference they are sponsoring on “behavioral change” for a “sustainable” earth. I am truly glad his health is allowing him to resume a speaking schedule. Maybe he can be encouraged to consider some “behavioral change” for more responsible, REPLICABLE, and sustainable scientific practices.
Such a cavalier attitude toward the data. How does science progress without adeqate data? Such an attitude more readily supports that this is really about the politics and the message, not science. If it is really about the science of and derived from lengthy ice cores, how does the science move forward? And it is not as if much of the Thompson’s and other climate scientists search for, gathering, and preservation of such important data is not largely, publicly funded!
I don’t recall ever noting on Climate Audit whether the Thompson’s or other Climate research grant requests contain a request for additional monies specifically for the archiving and preservation of data collected.
I’m reminded of the scholars who originally got possession of the Dead Sea Scrolls. They kept them locked up for decades without publishing much of anything. Eventually, the rest of the world’s scholars rose up and demanded that they hand over the scrolls.
Sadly, in this case the ‘other scholars’ take the ‘there but for the grace of God’ attitude and keep their mouths shut.
Me too. The NY Times editorialized that episode as the “academic scandal of the 20th century”.
However, it was not the world’s scholars who rose up to release the data. Instead, it was Hershel Shanks, the layman editor of the popular Biblical Archaeology Review, who led the charge. He eventually found a backup copy that was out of the scholars’ control and published it, presenting them with a fait accompli.
Perhaps the NYT will one day call unarchived paleoclimate data the “academic scandal of the 21st century”? (Just dreaming!) 🙂
Hu and Steve, that thread on the Dunde data is remarkable. To all, anyone who (like me) has arrived here post-2008 really ought to read that thread and Hu’s comment:
(was anything ever resolved about inconsistent and changing data publication on Dunde? seems not?)
ahh, if I may link an update (for newbies like me) for a CA thread with Hu’s subsequent article that continued on this topic:
Good work, Hu and Steve, shame that the recalcitrance and non-archiving continues to this day….
See also https://climateaudit.org/2012/03/04/mann-on-irreproducible-results-in-thompson-pnas-2006/ . It turns out that the basic problem is that the 10-year averages in the 2000-year index in Thompson’s Figure 6 are not the averages of the 5-year averages of what purports to be the same index (back only 400 years) in his Figure 5.
Re Dunde, Guliya, etc, I suspect that some of the multiplicity of versions may arise from sites that have more than one core. One researcher may have gotten core 1, another core 2, a third the average of the two d18O readings, and a fourth the average of the two z-scores. Archiving all the data would help clear problems like this up.
Steve: that could explain a difference between Thompson and the Chinese. I doubt that it’s the reason for the difference between Thompson’s Dunde versions (though it is possible.) My own guess is that it pertains to the dating of the core – which is not nearly as much of a sure thing as people think. If you squint at the inconsistent Guliya versions, it looks like there’s two-century dating difference in the medieval period between versions, but that the “topology” of the curve looks similar, but is compressed/dilated in the versions.
Re: Skiphil (Jul 8 18:11),
Thompson took three cores at Dunde, two from Quelccaya, but apparently only one from Guliya.
Actually it was even more convoluted than that. The secretive scroll scholars made the capital mistake of circulating a concordance of some of the unpublished scrolls, in a digital age.
Two scroll researchers at Hebrew Union College, in Cincinnati, Ben Zion Wacholder and Martin Abegg logged all of that into a data base and used brute force programming to reconstruct the original scrolls from the concordance. Which they then published. That let the enscrolled cat out of the bag.
A month later, with the secret broken, William Moffett of the Huntington Library in Pasadena announced that microfilm copies of the scrolls stored in the library would be published.
That was the end of the stranglehold of the scholars. A contemporary LA Times story is here.
Norman Golb’s book about the controversy, “Who Wrote the Dead Sea Scrolls” is very good and fully conveys the pervasive tensions, egos, jealousies, and intellectual snipings that were so, well, atypical of quiet scholarly discourse. 🙂
Didn’t know about the reverse-engineering from the concordance Pat. That’s a book I have to read. Thanks.
I’ve read your correspondence
Your patience is commendable
It should be almost gone since
They think this stuff expendable
Your careful, skilled pursuit
Is much appreciated
They take a lot of loot
For work that’s overrated
The data might support
Their (forgone?) grim conclusion
But, willing to distort,
They’re fine with some confusion
It’s very sad that science
Has failed from non-policing
This most-useful appliance
They’ve jammed. They’re not releasing.
===|==============/ Keith DeHavelle
By chance, were samples of the ice cores retained from the expeditions that have missing data?
In my youth we used to go back to our retains all the time. If the ICE core retains are still around a representative sample could be reevaluated……..
If the raw data is not available, then can(should?)an independent evaluation of the actual retained cores be undertaken?
The problem here relates
To ice core loss with time
The ice just sublimates
And gets replaced with rime
While current atmosphere
Brings in contamination
The document linked here
Provides an explanation
===|==============/ Keith DeHavelle
Thanks for the link Mr. DeHavelle [and of course the verse], a great one hour diverticula.
Kakatoa —
A lot of their ice is in cold storage on campus in case someone wants to re-analyze it. See http://bprc.osu.edu/Icecore/facilities.html .
Of course, it be a lot cheaper (not to mention generate a lot less CO2) to just archive the existing measurements and pitch the ice.
Steve: Arrgh, how can you say that. Yes, the measurements should be archived, but one of the reasons for keeping the ice is that investigators in the future may have better analytical tools; the ice should be available to them.
strongly agree Steve.
even better, saw it into 2 or 4 & distribute to other sites/freezers to ensure against unforeseen events.
any idea if this has been done already, seems prudent?
Evening Hu and Steve,
We had to go back to our retains once as a new attribute was identified that was felt to be important for final system performance. This new material attribute was characterized with some care using a new analytical procedure on recent receipts of the critical material that we happened to have retains for (10+ years worth of them). Working in a regulated industry, we ended up going back to a representative sample of the 10 years worth of retains to evaluate them for the new attribute.
Low and behold we found out the new attribute, thought to be critical to system performance, had the same distribution and average response throughout the 10 years of retains. Nothing like a bit of retrospective validation to keep everyone happy.
Steve, just read your correspondence file.
If you have to go through this kind of process with all the authors in climate science, you must have some hidden trick to manage staying so calm and respectful.
I would just have gone crazy, i don’t know how you do it.
Lots and lots of squash…
The data must be lost.
Is that an observation or a suggestion?
I can think of 18,381,790 (2012 inflation adjusted) reasons why neither the Thompson’s nor OSU are interested in archiving data and making it available.
“I can think of 18,381,790 (2012 inflation adjusted) reasons why neither the Thompson’s nor OSU are interested in archiving data and making it available”.
Thanks. I was about to fire up the old Texas Instruments but will no longer have to do that.
The funding sponsor NSF(National Science Foundation) was started in 1950. Their operating budget for 2012 is 7.03 billion(according to their web site) so let’s face reality; 18 mill is chump change in the grand scheme of things.
True in the grand scheme of things but in the local college “dynamics” where the Byrd Polar Research Center runs on ~”66m / year, you don’t want to rock the “NSF Grant” boat!
Oops Make that ~$6m/year (http://www.dispatch.com/content/stories/science/2010/04/04/50-years-at-the.html)
Rick, I think the point is not what the funding amounts (whatever it may be) represents to NSF’s budget, but rather what it means to the Byrd Polar Research Center and the Thompsons. They are the ones who *could* see reasons not to rock the boat or call attention to any problems in the work (I’m NOT saying that has influenced them in this unduly, I have no idea). Millions of dollars may be insignificant “chump change” to large US govt bureaucracies but not to one smallish scientific research center like BPRC. Still, plenty of people display resistance to change or criticism quite aside from funding issues.
You’d think the Roger’s of this world would learn a lesson.
First he tries to divert the subject to steve. then to steve’s readers. Then to Leif.
In the end we get
1. More people asking Ellen for her data.
2. a detailed accounting of the record
3. an apology owed to Leif.
Hey Roger. ever here the phrase, when youre in a hole stop digging.
I sent an email to Mosley-Thompson as follows:
Steve, another great letter. The point about the legacy is for me unarguable. But do you ever feel like mentioning the context: that the findings of climate science, so called, are being used to justify major interventions in policy? Lonnie Thompson is coming to the end of his days. Isn’t that when one thinks not just of science but the wider implications of one’s life work? And don’t those wider factors scream out that the data must be archived, in full?
“Lonnie Thompson is coming to the end of his days.”
Of course, we are all coming to the end of our days, but a heart transplant as witnessed by many successful cases can put off that end for sometime.
Richard Drake, excellent point about the context, and I would like to underline a two-pronged argument about the context. Not only are *immense* policy decisions being urged based (in part) upon this kind of data, but also the scientific claims themselves are unusually large and ambitious. They claim to rule out, more or less, objections saying that current climate change is not unprecedented or frightening or dangerous to humans!
That is, the Thompsons and others argue a scientific case which is exceptionally broad and bold based upon these data sets. I note that they do make various caveats at times, but they are in effect arguing “ice core and other paleo records show there is current climate change unprecedented in several millenia at least!” As the saying goes, “extraordinary claims require extraordinary evidence.” It is the very ambition of the paleoclimate field to say so much about such important stretches of earth’s climate history over centuries and millenia (well a tiny stretch for the planet, but exceptionally important for humans) that should require higher and not lower data archiving and scientific practices in these paleo fields.
i.e., it is no adequate counter for them to point to other fields of science which may not have good data archiving — their fields is making extraordinary claims, important claims, the archival means are available, no more excuses!
p.s. I am also starting to read up on the mid-Holocene and possibilities that, say, 5,000 – 7,000 years ago the “climate” may have been warmer than today. IF that might be true and IF the natural causes are not well understood then that is a huge area of importance that is being neglected (potentially more important than arguments about the MWP). It seems clear to me that the reliability, sensitivity, and comprehensiveness of ice core data may prove of enormous importance to whether or not comparisons should be made between the Holocene Optimum and today’s climate.
Ken: agreed, my wording was loose. The NYT assumed something like this as context, that Thompson was thinking such things over after a good number of years in the field already. I wasn’t trying to make predictions.
Skiphil: Helpful indeed to distinguish the two aspects. Both point to archiving as essential. Thanks.
I share experts’ suspicions which may relate to why Lonnie hasn’t wanted to archive results, rooted in questionable techniques that auditing would flush out.
I think this situation
Is clear for all to see:
The data publication
Would end his legacy
For once it is exposed
To yours and kindred eyes:
“Catastrophe”? Disposed!
Unprecedented? … not well supported.
===|==============/ Keith DeHavelle
Although I think your verse
Deserves our approbation
I just cannot coerce
Your prior publication
Into a form that scans
And I suspect you pulled the trigger
On an version that demands
That you finish? … complete.
Steve, something OT (but maybe not):
I don’t know if you have followed the controversy on Science publishing an egregious article on “arsenic life” which got a big press conference with NASA in Dec. 2010, but I’m thinking that (1) there are some fascinating comparisons and contrasts with how they handle criticisms and data issues for the Thompsons et al, and (2) Science might be more receptive now to criticism of their policies and practices (or maybe they’ll just circle the wagons again, but at least they know right now that they screwed up badly on the arsenic life paper).
A couple of links with more links in case you are curious:
As this post goes live, so too go live two Science papers refuting the heavily criticized “arsenic life” paper published in the journal in 2010.
“Despite refutation, Science arsenic life paper deserves retraction, scientist argues”
conclusion of article at 2nd link (my emphasis):
“The article itself is, however, only a small part of the story. As we’ll see in future essays, the case provides an illustration of the abysmal failure of scientific peer reviewers, scientific journals, government and academic institutions, the media and numerous individuals to do their jobs with competence and integrity.”
One thought is that Science (or others) might need to be more receptive at present to arguments that their peer review and data practices have not been exactly exemplary! They need to consider the series of Thompson articles which they published over the years.
I realize the details and fields are quite different, but the common threads are (1) publishing splashy cover articles which (2) slipped in with poor peer review and inadequate analysis of data and methods.
My issue with academic journal articles on topics like a Thompson ice core is that academic journals, especially Science and Nature, assign importance to “originality”. Thus, each Thompson article in Science is structured differently. None of the figures are consistent nor is the data presented systematically.
In mining exploration, technical geological reports are highly structured. Every property report has more or less the same sections in the same order.
In my opinion, if this field is to advance, they would be better off to ensure that there are structured technical reports on each important site. A promotional/advertisement article in Nature/Science is fine, but they aren’t structured technical reports.
Shall we all agree to refer to people like these in future as “data deniers”? Or would “data denialists” sound better?
I rather like “Declining Data Non-Archivers”, myself; then we could refer to them as “DDNA” – not to be confused, of course, with DNA … then again, perhaps DDNA is indicative of the presence of an aberrant gene 😉
The Ellen and Lonnie Thompson paper linked below and above by SteveM in the previous thread might enlighten some suppositions for reasons for the hesitancy of the authors not to archive their data. The analysis in the linked paper paper is directed by the authors to the purpose of determining whether the ice core O18 ratios are sufficiently indicative of the instrumental period temperatures to be measured going further back in time in order to test whether the current warming is unprecedented on a millennial scale.
Most of the series that are presented in the paper show little or no warming trend over the 20th century and the authors choose to use before and during the 20th century comparisons to measure modern warming. That is a curious choice since something like 1900-1950 to 1950-onward would have thought to have been a better comparison if one were looking for an AGW signal. The paper then goes on to cite other papers and analysis that would appear to open up the issue of interpretation of the O18 ratio records to special conditions at individual coring sites.
Could the authors apparent equivocating on these matters of what the O18 measurements represent explain their hesitancy to archive? Would the agencies controlling the purse strings for these projects be less interested in seeing the records archived because of this equivocation?
I guess the authors might alone be able to address these suppositions.
Click to access Annals%202006%20EMT%20et%20al.pdf
As a non-scientist(but as a trial judge for 25 years and therefore rather accustomed to weeding out motive), my hunch, based in part on the fact that he has published so many different versions of the same data,is that he has long ago lost control of his data, its organization and its whereabouts, and cannot therefore collect it sufficiently for archiving, even if he wanted to. Probably the only solution is for the university to hire some graduate students to wade through everything and see if they can make a sensible collection of the data and archive it. No one here will hold their breath waiting for that to happen, however.
Hm, cock-up theory. I like it. Probably more to it than just that, tho’.
Maybe there is a simple reason why.
The ice core data does not give the “right” answer.
No Hockey Stick 😦
Don, re: “right” answers
I’m starting to poke around the web for info about the “Holocene Optimum” and it seems that the Thompsons’ ice core data may be crucial to dismissing claims that we don’t know enough or that astronomical data account entirely for climate changes 5,000 to 7,000 years ago. For the public, the NOAA paleo site just blithely dismisses any idea that there is a problem for their argument:
NOAA paleo site dismisses importance of Holocene Optimum
oops I lost a negative in there, I mean that the Thompsons et al and NOAA, Real Climate, et al want to dismiss any claim that astronomical data may NOT account for all significant warming 5,000 to 7,000 years ago. Of course, the burden of proof should be on THEM to show the reliability of the Thomspsons’ ice core data etc. to proving that any Holocene Optimum is purely an astronomical phenomenon.
IF there is a Holocene Optimum beyond what is driven solely by precession of the earth’s rotational axis, THEN that may be a serious problem for CAGW alarmism. That is what the Thompsons et al help to rule out, and that exclusion of hypotheses is an extraordinary claim which requires extraordinary evidence.
The Holocene Optimum is indeed interesting. There’s an interesting article yesterday in Nature CLimate Change by Esper and others (Wilson, Zorita, Timonen) pointing out that orbital changes should be noticeable in 2000-year proxy reconstructions. I have draft notes on this and will get to it.
Steve, have you been following the inline responses to that paper at realclimate? Interesting! http://www.realclimate.org/index.php/archives/2012/07/tree-rings-and-climate-some-recent-developments/
snip – let’s leave the Holocene Optimum for another day
Kenneth,
Are you suggesting another divergence problem?
James
Consensus Climate Scientist Handbook
page 12:
“If pressure to publicly archive data becomes unbearable, announce that the data has been lost.”
“Holocene Optimum”
Jan 1, 2007 at 2:33 PM
It is interesting that we now seem to have the absent-minded professor and the source of cutting edge knowledge that could reshape the world embodied in the same person. This is the stuff of books and movies during the early sci-fi era, but it ended about 50 years ago.
Look, we all get that some people are hopeless at administration but brilliant in their field – or vice versa. But, in this day and age, with millions of dollars of grant money involved, and archiving never cheaper or easier – surely they could pay a competent grad. student or retiree $50 grand a year to do it for them?
I would have thought that it would not be difficult to find someone who would relish the task, as well as the income, which is a drop in the bucket in the total funding.
That said, it is human nature to want to move on, which is why when issuing contracts for research work in a previous life I always included a line item for things like archiving and evaluation, with the understanding that payment was subject to delivery. The corollary was that non-fulfilment was a black mark against future bids.
It is a combination of structural problems and human nature. Relying on people doing the right thing in the absence of incentives and disincentives is always a doubtful strategy. When we are talking about public funds, and policy related research, it is Panglossian to the n to imagine that the outcome will not include slipshod, mistaken and even dishonest outcomes.
For the type of data from an ice core, I don’t think even $50k/yr would be needed. For some large field projects on CO2 enrichment with which I am familiar they do indeed have a data person to put it all together and archive it.
There is in fact a bias in funding agencies and by reviewers in favor of field work and against analyzing data already collected (“why didn’t you analyze it on the original grant?”)
The records may be a nightmare, though. Do they even have all the data or have they kept only what suits the narrative? I have to admit I had not begun to grasp how breathtakingly bad this whole Thompsonx2 data situation is until I just now started to go through past CA threads:
https://climateaudit.org/category/proxies/thompson/
It is truly shocking (to naive ppl like me anyway) what these ppl are getting away with.
Well, harry-read-me had the same problems. But, as Steve has said, it needs to be a condition of funding that data is properly archived at the time. And, like poor Harry’s, the lot of a retrospective archiver is not a happy one. But, we have two choices: throw our hands up and say it’s all too hard; or try to untangle the mess and see what is retrievable.
If we choose Option One, then we are announcing that science is now based on somebody’s say-so, not evidence – or that we will ignore all those papers as baseless.
It is worth reminding readers that the famous 1940s study of fruit flies and monogamy (Bateman) which allegedly demonstrated that males have an evolutionary bias towards promiscuity and females towards monogamy – cited thousands of times since – was never replicated until recently and found to be wanting, to put it mildly. See Gowaty of UCLA’s work for details.
The point is, even when it is easily possible to replicate work, it does not always happen, so that junk science can persist for decades. FGS, breeding fruit flies is the easiest and cheapest kind of critter research there is – which is why there is so much of it. Yet, no-one ever bothered to check the results of this ‘landmark study’, which cost peanuts, for more than 60 years. But, at least when someone decided to do it, it was possible. How can anyone check the results of the Thompsons’ (or anyone else’s) ‘landmark studies’ without the data?
Without that, they may as well have sat in an apartment in New York and written their ‘findings’ in a personal diary. I visited the Great Barrier Reef a while back, and have photos and diary notes to prove it. Absent data and records, however, anything else I might say about the GBR is anecdotal at best.
This sounds like a very common academic situation. Young academic scientists go out on fieldwork and collect data. It’s easy to collect data, hard to analyze it. It’s easy to get grants for glamorous fieldwork in exotic locales, hard to get grants for dry boring analysis of data collected long ago. So they build a surplus of unanalyzed data to sustain them when they are no longer young. But with all that goes on in academia, they never get around to it. Remember the closing scene from Raiders of the Lost Ark? All the old scientists have field data moldering in notebooks, unanalyzed and slowly being forgotten.
This data’s got its traders:
Earth-shaking allegations
Supported by citations thrown their way
Not like the scene from Raiders
This hiding situation’s
More like the ship in Independence Day
The public mob (per Schneider)
Is “panicky and stupid”
So statements can be far away from facts
Each data archive hider
Feels rightly instituted
But person’s smart … and now they can’t relax
===|==============/ Keith DeHavelle
It ain’t science if it can’t be replicated ⇒
If data and methods are not archived it can’t be replicated ∴
It ain’t science if it ain’t archived.
What does that do to corporate scientists producing proprietary data that is not readily available?
The incentive of a corporate scientist is to produce something that works. The “replication” of the science takes the form of repeated testing in real-world conditions. The incentives of an academic scientist are 1) novelty, 2) publications, and 3)awards, honorary degrees, etc. These incentives are not necessarily bad, but they can be satisfied by hasty work that is never replicated or proven in real-world tests. It’s not just climate scientists; I see the same thing in the chemical industry, and before in the medical device industry. Academics can do their job without the same degree of rigor that is demanded in industry.
But, if you want to have have vast influence on the lives of billions of people a higher degree of rigor is demanded. The Thompsons of the world don’t get this.
I completely agree that if you want to change public policy, you’ve got to show your working out. But I’m very worried by some of the prejudicial political and anti-scientific positions that seem to be energised by observations to this effect.
I’m not nearly as confident as you that the market exerts a cleansing influence on science: and not because of my politics. The scepticism (i.e. uncritical cynicism) that stands on political dogma is just another kind of faith. If one is going to be jaundiced about the workings of professional (i.e. salaried) science, I think it safer to be even-handed. One could invoke all sorts of negative tendencies that have effected or are likely to effect corporate science enterprises of different sorts — which would lead to generalisations just as half-true as those you’ve made about academic science. (And this comment is at risk of being snipped anyway.) In the end, the value of the science has nothing whatsoever to do with scientists or their circumstances.
The cultural distinction between academic science and corporate science seems to me to be imperfect. Where I work, the connections between academia and the private sector are growing increasingly complex and the distinctions more hazy — which will no doubt generate all kinds of interesting problems (e.g. as regards proprietary data) in the future. A number of academics I know own highly competitive companies which enable them to profit from their research. The bottom line for me is that it’s self-evidently a bit silly to assert that ‘honest’ or high-level science is now more likely to reside in private corporations than in universities.
I agree with you. My own critique tends to be quite narrow. Over the years, I’ve found the angriness of many commenters to be quite frustrating. I delete numerous such posts, under blog policies discouraging “venting”, “piling on” etc. I categorically don’t agree with the broadbrush angriness. Yes, I’m frustrated with the Team and with the wider community of climate scientists, but I’m not angry about many things that readers get worked up about.
Jon Grove –
Not sure if you are still attending to this thread or not, but if so…
I don’t understand what Steve sees in my words to suggest that I am angry. Nor can I see how what I said is, according to you, “anti-science”. Nor do I understand what politics has to do with my comment. All I am saying is that a medical device, for example, undergoes a tremendous amount of testing and intense scrutiny of the data, by both the company that produces it, and the FDA. This makes perfect sense because so many lives are riding on the integrity of the data. 99% of scientific papers that are accepted for publication do not undergo even a cursory review of the data, methods, or results. Peer-review does not fulfill that function and it is not supposed to. Peer-review establishes whether the work falls into generally accepted norms and is plausible.
There is no accusation of mis-conduct in any of the academic culture as I am describing it. But if the peer-reviewed papers are supposed to form the basis of far-reaching policies one would think they should be subject to at least the degree of scrutiny that a medical device is, for example.
Look, I strongly believe in archiving, but it’s ridiculous to say that something isn’t “science” if it isn’t archived. It may be or it may not be.
There’s a price for going a bridge too far with this sort of criticism, as it enables people to challenge this sort of overstatement and avoid the issues that were originally raised.
Steve,
You are right.
If the data is lost or hidden or simply unavailable, it may be science, but I think we can agree it is bad science at best. The pattern emerging from major branches of science regarding poorly documented misrepresented or incorrect results seems to be much more pervasive than just climate.
Steve: once again, I urge readers not to over-generalize. Science as an activity is highly successful. While I criticize Thompson’s data archiving, he has gone to great efforts to provide a permanent archive of his ice core and readers must always keep this in mind.
“This sounds like a very common academic situation. Young academic scientists go out on fieldwork and collect data. It’s easy to collect data, hard to analyze it. It’s easy to get grants for glamorous fieldwork in exotic locales, hard to get grants for dry boring analysis of data collected long ago.”
It does not make sense to me that after doing extensive chemical, particle and accumulation analysis of the collected ice that a naturally curious scientist would not be very interested in doing the remainder of the analysis which would not be all that difficult. I think either the project leaders were unable to make sense of what O18 represented in the instrumental to centennial time period and thus attempted to further measurements. Or that they decided to do some further analysis back in time and continued to not to be able to make sense of it. How much the advocacy position of the project leaders could have influenced such thinking I would leave to the judgment of individuals.
I suppose a bigger and more perplexing issue is why the funding agency did not pursue measurements with the project leaders and whether their judgment was biased in these decisions.
By the way collecting ice core data from these remote regions is not an easy task – it can be life threatening.
I seem to remember reading very recently that the Wellcome Foundation were now stipulating archiving data and metadata. Perhaps even more good practices. Can anybody confirm?
You people aren’t nearly cynical enough. Are there any Thompson-free papers that claim to have used ANY of the underlying data? Has anyone ever actually seen it?
As long as well-established, tenured professors keep getting grants to fund their research, they can (and will) continue to run their research programs as they see fit. That’s an unfortunate aspect of academic freedom. If the Thompsons are happy considering their publications, talks, Congressional testimony, IPCC authorship, and numerous awards their “legacy” – rather than their archived data – that’s what they will do so. Sorting through 30 years of data (some of which could be on legacy software and systems like tape or floppy drives) isn’t likely to be a pleasant job and would probably require full-time help.
Given “academic freedom”, the only parties that can put an end to such irresponsible behavior are governmental granting agencies and journal editors. Grant applications could require researchers to list data archived in connection with previous grants. New grants could require data archive plans (including release dates) and updates on progress. New grants could be withheld until data archiving for older grants had been completed. Accepted publications could be withheld until data and programs have been archived.
This comment and a few others are on target, I suspect. My impression is that they spend much of their energy and funding etc on the glamorous job of collecting and storing the precious ice cores. I think there is little time, money and energy spent obtaining real data from the ice cores once retrieved.
I have firsthand experience with this as I visited the ice core repository (being an OSU graduate in Geology gave me access) at TheOSU in 1999 and had a tour of the facility with my 11-year old daughter ( as part of her middle school science project). The facility was impressive, as far as the storage of the ice cores was concerned. On the other hand, the facility had very little room for data/core analysis ( one small lab with very limited equipment). I was told there was only one study going on at the time, involving melting small samples of certain cores and analyzing for a couple specific parameters, by a grad student. They also stated that they had funding request for future improvements to the lab, such as the purchase of a mass spectrometer. My impression is that they were collecting samples but there was very little analysis of the core samples once retained. It may be that much of the (non-archived) real data analysis that has been used for paleo studies was collected very early in their careers. I also fear that much of this data is essentially lost, i.e. will never be archived properly.
Thank you Steve for this very important “Name and Shame” campaign.
The core of the problem is that they’ve put the archiving of their data on ice.
In Archeology the saying is ‘not publishing your field data is equivalent to a bulldozer going over your site’ as the result is the same. Just printing a few pages in scientific publication with your most spectacular interpretations is good for fundraising but does not advance Science unless you make all your under laying proof – the field data – easily publicly available; the Internet is available since the last 3 decades for that, no excuses accepted and also no ‘Legacy’ till then.
Mr. McIntyre
There was a paper written in March 2011 by Joseph P. Simmons, Leif D. Nelson, and Uri Simonsohn entitled
“False-Positive Psychology: Undisclosed Flexibility in Data Collection and Analysis Allows Presenting Anything as Significant”
In case you have not seen this paper, I bring it to your attention. It is a short paper. Here is the link to the paper.
https://opimweb.wharton.upenn.edu/files/?whdmsaction=public%3Amain.file&fileID=4106
Their investigation was in the realm of psychology research, but after reading the paper, I was struck that it is entirely possible that much of the same behavior uncovered by the authors may not be restricted to psychology research, but be wide spread in the scientific community.
From their abstract:
“In this article, we accomplish two things. First, we show that despite empirical psychologists’ nominal endorsement of a low rate of false-positive findings (.05), flexibility in data collection, analysis, and reporting dramatically increases actual false-positive rates. In many cases, a researcher is more likely to falsely find evidence that an effect exists than to correctly find evidence that it does not. We present computer simulations and a pair of actual experiments that demonstrate how unacceptably easy it is to accumulate (and report) statistically significant evidence for a false hypothesis. Second, we suggest a simple, low-cost, and straightforwardly effective disclosure-based solution to this problem. The solution involves six concrete requirements for authors and four guidelines for reviewers, all of which impose a minimal burden on the publication process.”
Essentially what they show, is that because researchers do not stick to the original sampling plans, they actually migrate through many many analyses until they find and report one that “works”, and so the supposed 5% false positive risk is actually much greater.
The six concrete requirements that they say will eliminate this problem is as follows:
“Requirements for authors
1. Authors must decide the rule for terminating data collection
before data collection begins and report this rule in the article.
2. Authors must collect at least 20 observations per cell or else
provide a compelling cost-of-data-collection justification.
3. Authors must list all variables collected in a study.
4. Authors must report all experimental conditions, including
failed manipulations.
5. If observations are eliminated, authors must also report what
the statistical results are if those observations are included.
6. If an analysis includes a covariate, authors must report the
statistical results of the analysis without the covariate.
Guidelines for reviewers
1. Reviewers should ensure that authors follow the requirements.
2. Reviewers should be more tolerant of imperfections in results.
3. Reviewers should require authors to demonstrate that their
results do not hinge on arbitrary analytic decisions.
4. If justifications of data collection or analysis are not compelling,
reviewers should require the authors to conduct an
exact replication.”
Re: John Phillips (Jul 9 21:02),
This has been mentioned in comments on several previous posts. Plus, one of the recent top-line posts is focused on a followup to this article.
Oh gosh. Sorry. That’ll teach me to just parachute in on a blog that I haven’t been following for awhile without doing some back-reading first.
But I’m glad to hear the paper has been discussed!!
Here is another one for you, back from August 2005…
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC1182327/?tool=pubmed
And another:
http://www.plosmedicine.org/article/info:doi/10.1371/journal.pmed.0020124
It is not and never was just climate science.
Those concerned about physical core loss can see at least some core preserved in late 2006, material from National Geographic Magazine and Ohio State University, unable to find text author for acknowledgement.
The quote above the pic Geoff posted was “Some of the ice we have here is already gone from the mountains”. All the more reason to ensure the data is fully archived. Expensive field work was done, work that in some cases cannot be done again and all the information obtained with be lost without proper archiving. Let’s hope Dr Thompson will preserve his legacy.
Sera,
Thanks for the link to another very good paper, this time in the medical field on false findings. One has to wonder sometimes about the constantly changing “findings” of medical studies. “Coffee is good for you.” “No wait, coffee gives you pancreatic cancer.” “No wait, coffee make you live longer.” “No wait ….”
A study from University of Wisconsin reported recently on WUWT highlights the long-known fact that ice core oxygen isotope ratios reflect temperatures where the clouds formed, not where the precipitation fell. Thus changes in regional circulation may be at least as important as local temperature in determining isotope ratios.
Take a look at
Oxygen Isotope Changes In Tropical Ice, Quelccaya, Peru by P. M. Grootes, M. Stuiver, L. G. Thompson, and E. Mosley-Thompson, Journal of Geophysical Research, Vol. 94, No. D1, pp. 1187-1194, 1989
and you will see that the Thompsons were well aware of this problem a long time ago. This might be part of the reason they keep the data close; they could feel the data will be misread by anyone not equipped to interpret them in conjunction with regional circulation models. Indeed, careful study might reveal serious obstacles to interpreting ice cores as temperature proxies. The data are indeed precious, but perhaps not for the reasons advertised.
I am certainly not trying to excuse the non-archiving, just trying to give it some context. Know your adversary.
Steve: the interpretation of O18 in the tropics is a longstanding mainstream issue – – see Vuille’s criticism and many others. An issue long before I got interested in the field.
Bruce, I have been a consistent questioner of the problem with O18 ratio measurements you pose here and I have also supposed that making “sense” of relationship of isotope ratios to temperature may be a reason for the Thompsons not to make public their data.
From a regional perspective I do not think the exact location of the condensation site would be critical to teasing a global or regional temperature trend out of the data. I suspect the problem could be more one of the data not showing the consensus view of modern warming. As with all proxies I am not saying that the proxy is responding to temperature and in this case the proxy might be wrong.
Also something not discussed much at these blogs is that storage and handling these ice core samples can be critical to interpreting the cores with respect to temperature. Some of these cores were transported out of remote areas and in fact some cores were handled and analyzed at the site by melting down sections as I recall reading. The earlier cores could have been on the learning curve for handling and storing.
Regardless of details as noted above, the state of the original ice cores and any measurements taken from them should be made public. The project leaders could and should always explain any extenuating circumstances.
@ W.W. Wyagart. 4:14pm July 9.
Very impressive three lines of logic. Simple but the best things often are.
I “knew” how important archiving is but you have sharpened things for me.
Thanks,
Ed.
Sorry about the typo on your name, Mr Wygart.
Don’t know if it has been mentioned somewhere in this blog, but regarding the prestigious PNAS I remember reading (in another context) something that the PNAS has a slightly non-peer-review process of submission. What it comes down to is that a NAS member can push through an article for publication, if he is a co-author.
And my personal observation: both Science and Nature have been tabloids for at least two decades…
I can sympathize somewhat with the Thompsons plight. I used to do similar work, but for industry. (rock cores, not ice). I absolutely hated archiving data. If it was somehow possible to have neglected this responsibility and galavanted off to my next field project instead, I surely would have. Of course we had project managers who would never have tolerated us shirking our responsibilities.
“I remain mystified by Thompson’s intransigence in establishing a comprehensive and meticulous archive of his measurement data, as, in my opinion, he should regard the establishment of such an archive as an essential part of his scientific legacy and his #1 priority given his age and health.”
Exactly. I could make the case that providing quality raw data is more often likely to build such a legacy than the analysis of the moment. Sure there are exceptions. . . but they are exceptions.
In financial terms, it’s a very valuable “hedge”.
The real “raw data” is the ice cores. Measurements are subject to replication. The cores are not.
At this point the objective enquirer has to start weighing different hypotheses.
Possibility A is that the information exists but for psychological reasons the researchers have failed to comply with the conditions of their grants and archived it. They have also refused to answer requests for data for psychological reasons.
Possibility B is that the data does not exist.
This looks interesting
http://www.nature.com/nclimate/journal/vaop/ncurrent/full/nclimate1589.html
Solar insolation changes, resulting from long-term oscillations of orbital configurations1, are an important driver of Holocene climate2, 3. The forcing is substantial over the past 2,000 years, up to four times as large as the 1.6 W m−2 net anthropogenic forcing since 1750 (ref. 4), but the trend varies considerably over time, space and with season5. Using numerous high-latitude proxy records, slow orbital changes have recently been shown6 to gradually force boreal summer temperature cooling over the common era. Here, we present new evidence based on maximum latewood density data from northern Scandinavia, indicating that this cooling trend was stronger (−0.31 °C per 1,000 years, ±0.03 °C) than previously reported, and demonstrate that this signature is missing in published tree-ring proxy records.
I do hope that the data has been archived!
Steve: this is an interesting article on a central Climate Audit issue. I plan to post on it. Their data archive for this article is better than usual though more of a C+/B- than an A.
johanna:
The essence of the thing is that we are clearly expecting people to devote time and resources to something they are not being paid to do.
The problem, then is to get this requirement into their contracts, anything else is a cosmetic quibble.
I might add, that I find it astonishing that we object to someone not doing something for nothing when we otherwise would object to such an expectation as “socialist.”
It might also be reasonable to suppose that the grantees looked closely at the requirements of their grant and assigned values to the work elements which best (in their opinion) used the funds for what was needed- AND they didn’t do anything which wasn’t on the menu. That’s consulting 101.
I still favor the idea of getting into legislation a proscription from basing legislation or regulation on any “science” not having publicly available archives of all the data, software, etc.
it would be tough luck for the righteous guys who didn’t archive because they weren’t getting paid to do it, but it would certainly get the grant writers on board.
Steve: As I’ve observed over and over, there is ample authority for NSF to require archiving in the terms of their grants. Indeed, there’s an expectation that they do so. I’ve tried over and over to emphasize the role of the bureaucrats though hardly anyone pays attention to this.
/strong>
And Steve,
You seem about as worldly as they come. Do you expect people to devote time and resources to something they aren’t getting paid to do – of course in their professional pursuits?
In my opinion, it’s part of the job. I do not agree that scientists aren’t “getting paid” to archive data. They’re getting paid as scientists and, if archiving is part of the job, then they’re getting paid to do it as they are any other part of the job. Nor do I view the exercise as particularly onerous. By and large, scientists in question have probably spent far more time and effort in refusing to archive data than it would have taken in the first place. Further, archiving saves time – you don’t have to think about the version that you used. You’ve placed it in a permanent location where it’s not subject to the vagaries of your personal computer and backups.
More than likely, they wrote a grant that effectively promised the moon on a pittance, probably with accurate well-estimated field costs and vague handwaving to support the portion of the grant that dealt with analysis, write up, and archiving. It is part of the job, but many are caught on a treadmill where each project sees them farther and farther behind the curve for completion.
J ferguson: I agree with Don totally. I don’t know whether the Thompson contracts included funds for data mangement/archiving or not. If not, they should have. The journals require it, even if they don’t enforce it. But that is really beside the point. The Thompsons have spent millions of dollars collecting samples from which their data is derived. If it is not/never archived, it is lost to science, pure and simple. All that money, all that work, has or will be gone down the drain.
Additionally, you overlook the fact that Thompson has himself repeatedly said, albeit incorrectly, that it has been archived. So funding for management/archiving does not seem to be have been an issue with him.
It seems to me that if archiving is a requirement of the grant and they accept the terms of the grant, then they were paid to archive.
jferguson, perhaps I should have expressed myself more clearly. All standard contracts included things like archiving and evaluation, if required, but I made a point of separating them out and putting a dollar amount against them. It is much better than arguing after the event about an item in a laundry list of other items with no specific payment attached. It also has a psychological effect, IMO – it becomes an item to be ticked off by both the researcher and the commissioner in determining if the contract has been completed.
It’s good practice in drawing up any contract to separate out non-negotiable items and put an amount against them (making sure that the requirement is clearly defined). The laundry list parts of a contract are for things that are difficult to quantify in advance and/or are likely to require negotiation/revision during the term of the contract.
@ j ferguson “The essence of the thing is that we are clearly expecting people to devote time and resources to something they are not being paid to do.”
I have to disagree. As academics they are expected to produce “output”- usually in the form of peer-reviewed papers. If the journals that they submit to require supporting data they are failing in their academic duty not too.
More is the pity that said journals all too often do not stick to their own rules, allowing the sort of malfeasance we are commenting on.
If journals did stick to the rules, some academics would be struggling to publish in high impact journals and their grants would dry up, as would their careers.
In some cases this would not be a loss to science.
@ Don Keiller
And yet, so we’re told, many of these same people are quite willing to “devote time and resources” to the production of the IPCC’s assessment reports which they are also, so we’re told, “not being paid to do”.
Aye, there’s the rub! I wonder who set the precedent for “not stick[ing] to their own rules”. The IPCC or the journals on whose publications the (unpaid) writers of assessment reports depend … well, on which they appear to depend at least 2/3 of the time 😉
@hro001
@johanna
What we are discussing seems to be Ponzi science. The guys who do this stuff go from grant to grant, paper to paper without divulging the entire basis for their work. Because of that, it cannot be replicated or audited. Unless error in the method can be discovered which doesn’t depend on possession of the data, errors which do depend on access to the data will not be discovered and follow-on papers may be written based on defective premises.
One would think that the discipline would worry about this, but apparently not. I asked Judith Curry (on her blog) what happened when one of her PhD candidates set off to explore an avenue with questionable antecedents based on those antecedents or even worse, completed a dissertation premised on dubious science. She never answered, maybe because nothing like that ever happens in her shop.
Or, as she seems to worry about, trust is the issue and possibly because of the burden of keeping up, she has to depend on the work of people she respects to be reliable sans audit.
I’m much impressed that johanna put what we used to call “hammer” line items in her contracts putting a dollar value on delivery of the archived data.
This is exactly the sort of thing I was driving at. Not having the archive a line item with a value is like expecting Donna’s teenager to keep his room tidy.
And while we’re at it, does anyone have a grasp of how much of the IPCC inferences, or whatever they are, are based on papers sans archives? Or maybe if they use grey literature, they don’t expect an archive.
In an earlier incarnation, I consulted with an outfit that developed and manufactured biomedical devices. Their archiving was beyond belief and included sketches made on napkins in restaurants with wine spilled on them. They, and I suspect the FDA, wanted very much to be able to discover where the train went off the tracks, probably to make sure that train never left that track again.
It seems to me that since several of our countries, UK and Oz, in particular, are being forced by legislation to make significant sacrifices, the requirement that legislation can only rely on completely archived, auditable science – without exception.
oops: that last sentence should suggest that there be a legislative AND clear limitation that no other legislation or regulation can be based on unarchived science.
The trail of correspondence with the Thompsons and journal editors trying to get the data archived is remarkable. Related, Roger seems to have gone missing. I hope he’s feeling ok.
A quibble, Steve – an aging VP of exploration shouldn’t be running a drill rig. But s/he should get out to the site and look at the work. Recall the chief geologist of Bre-X failed to do that, whereas it did not take an outside consultants long to spot problems (appearance of gold in cores, contradictory statements about water sampling – it did help that one consultant could get by in the local language, etc.)
I once worked for a fool who hurt his reputation in the company because he would not go north of Edmonton. Beats me why, he was in the navy though I don’t know if he sailed on a ship. Though he may have been avoiding dealing with equipment problems up north, that’s probably what some of the operational people in Edmonton thought.)
OTOH I went all the way north, almost to Eureka, several days tour at least twice.
And I erred on the side of sending my staff north when there was a problem. (At least two of them got weathered in overnight, the silver lining is that really brought home the vagaries of flying in the High Arctic. At least one of them got up with the crew to run the 727’s APU every few hours to heat the airplane up.)
On one trip the airplane that took us to the High Arctic was carrying some geologists/geophysicists from TX, to a drilling rig location. They were excited at the ice ridges, and the rock outcrops on one of the islands near the Beaufort Sea (quite low, and not a great deal of snow). Me, I was more excited about the mountains on Axel Heidberg Island.
Environmentalists often don’t actually go look at something.
Steve: I discussed Bre-X in some early posts. Your facts are wrong. The Bre-X VP Exploration, John Felderhof, did visit the site. There’s also a difference between visiting the site and manning the rigs. A VP exploration will visit the site of a major drilling program but wont man the rig.
Not part of their job? Are you friggin serious?
All the published papers which refer to unarchived data, and which don’t adequately describe how the published results were derived from the data that is archived, will eventually be totally worthless and as such, discredited. I say we start throwing them out right now, and avoid the rush.
Science colleges can work with their own departments or university libraries; governmental groups and NGOs can damn well include funding for archival servers and the proper web tools and the maintainence thereof, as PART OF THEIR FUNDING, or they get NONE!
I’m just an engineer, not an academic, but I don’t understand how not achiving the data makes anyone’s life easier. How do these people verify their own results? How do they explain any discrepancies later pointed out in their results without a repeatable process starting with baseline data? Let alone be able to compare result sets from various locations and/or time periods. Is this an issue across academia or is it a specialty of climate scientists?
Geoff. I understand your bewilderment. In my very modest life as a geo-researcher, my time went as follows: a) collect data, b) organise data into some kind of formal matrix, c) interpret data.
I found it impossible to get from a) to c) without going through b), but maybe it’s a personal failure of my professionlism.
@Mickey I had the temerity to email Professor Phil Jones and say that papers that relied on data that was not pubicly available were “hearsay” and should be withdrawn and any conclusions/policy decisions should also be reconsidered.
Needless to say the response (behind doors, but revealed by Climategate) was quite hostile.
Needless to say nobody in authority gave a damn.
That is the real problem.
Don,
Bingo! If we are not seeing archiving, it’s because scientific ethics are insufficient reason. Grants (contracts, really) don’t require it, the journals do not insist, IPCC will base its recommendations on your stuff, sans archives, and therefore unauditable.
If the EPA were legislatively forbidden to base regulation on un-archived science, the grant writers might start including the requirement as Steve has pointed out they are enabled to do, and this problem might clean itself up.
Don —
I like the “hearsay” designation for articles, however peer reviewed, that rely on unarchived data. The journals can’t be expected to withdraw them, but still this is a good criterion for dismissing them.
A “Hearsay Paper” meme could be very powerful and effective at getting sloppy scientists to change their behavior and to begin archiving.
The problem of good archiving is not trivial, and I’m not minimizing the amount of work it requires. But seriously, if data collection and analysis is your job, you need to be good at this part of it.
And if we (the taxpaying ‘we’) are expected to make public policy decisions based on small statistical anomolies of dubious studies into complicated physical systems, the LEAST we can expect is that every paper’s charts and graphs should be reproducible by follow on scientists. The very least…
Has Roger given up? More important, this and related posts do underline the enormous amount of work Steve McI has done trying to clean up this dirty corner of climate research.
STEVE, FROM THIS POST ABOVE:
Hu McCulloch
Posted Jul 9, 2012 at 1:41 PM | Permalink
Re Dunde, Guliya, etc, I suspect that some of the multiplicity of versions may arise from sites that have more than one core. One researcher may have gotten core 1, another core 2, a third the average of the two d18O readings, and a fourth the average of the two z-scores. Archiving all the data would help clear problems like this up.
Steve: that could explain a difference between Thompson and the Chinese. I doubt that it’s the reason for the difference between Thompson’s Dunde versions (though it is possible.) My own guess is that it pertains to the dating of the core – which is not nearly as much of a sure thing as people think. If you squint at the inconsistent Guliya versions, it looks like there’s two-century dating difference in the medieval period between versions, but that the “topology” of the curve looks similar, but is compressed/dilated in the versions.
STEVE, REGARDING DATING OF ICE CORES, I was wondering, from your years in the Mining Industry, and I imagine friendships with quite a few geologists, I would be most grateful if you could comment, please: Some years ago after the EPICA (European Project for Ice Coring in Antarctica) issued results in a journal article (sorry cannot recall details of original publication) about that ~”800,000 year” ice core, bits about Saharan dust layers clearly visible in the cores, etc., I started an e-mail dialog with a Dr. Eric Wolf from the UK (one of the authors to whom correspondence should be addressed . . .). I posed the question to him wondering if the supervolcano eruption of Yellowstone (~630,000 years ago) might have left some sort of additional “marker” in their ice core. He pooh-poohed the points I raised and completely dismissed my comments having concluded that nothing in the atmosphere could have made it to the South Pole. For such a eruption, which expelled something on the order of 1000 cubic kilometers of matter, I cannot imagine nothing making its way around the entire globe, can you? He never responded to my last e-mail, so I gave up. Any thoughts on fine dust, elemental signature (e.g., elevated acid content, etc.) being transported globally by the atmosphere?
Thanks in advance, and looking forward to your reply.
Regards,
Bob
Steve: the geological formations that I’ve encountered in my business career are all from very very deep time – nearly all Precambrian. I have no geologic insight on this issue. I know that ice core-ologists do pay attention to tephra.
I was hoping that Roger would chime in…. I’ll be waiting.
@Duster
“And another: …
It is not and never was just climate science.”
Thank you Duster.
You’re quite welcome. Depending on the discipline, you can track these concerns far back in history. Francis Bacon in fact rants on quite at length about what we would call “irreproducible results,” “affirmation of the consequent” errors, investigators finding what they expect, etc. He advanced the experimental method as a means of increasing the objectivity of science – if it is reproducible, it is more likely to be useful knowledge. Since he was at his peak at the end of Elizabeth the First’s reign and the beginning of James the First’s reigns (late 16th and early 17th centuries), you can see that it has a long tradition. “Post normal” scientists have been known to toss around “Baconian empiricism” like it was a hand grenade, a devastating critque of the science of someone who disagreed with them.
The Bona Churchill Ice Core Record
Urmann, D. 2009. Decadal Scale Climate Variability During The Last Millennium As Recorded By The Bona Churchill And Quelccaya Ice Cores. Columbus, Ohio. Dissertation, The Ohio State University, Department of Geology. 255 pp.
This dissertation is available here as a pdf.
However, I can’t seem to find the actual data from Bona Churchill anywhere in the document. 😉
P. 56 graphs Bona Chruchill Core 1 (BC1) d18O versus inferred age for 535-1994 AD. It shows a Modern Tepid Period of roughly 800-1994, similar to its Medieval Tepid Period of roughly 900-1200 AD. These are separated by an LIA Warm Period, roughly 1300-1800. The “coldest” period was roughly 535-800 AD, with a pretty steady secular warming from 534 to 1300.
Although there is no data evident, the graph could probably be digitalized fairly accurately. Is there any new activity at NCDC?
I can’t find a full graph of d18O versus depth. Pp. 36, 47, and 48 have 75-81m, 146-164m, and 234-241m only. p. 82 goes down to 15 m only and compares this to BC2 and 4 shorter cores. (The total depth was 460m, which he estimates at roughly 265 AD, though he is only confident of the post-534 period). I have no reason to doubt his dating, but it would be appropriate to start with the actual data, and only then to show it versus inferred date.
Happy New Year to All!
Typo, for the “Modern Tepid Period” I think you meant to type dates of 1800 to 1994
Skiphil — Thanks, 1800 to 1994.
P. 131 has a clearer graph of d18O versus date. It only goes back to 800 AD, but then p. 56 indicates that the dating before 1000AD is much less certain than afterwards, so perhaps not much is lot by not going back to 535AD. The p. 131 graph is 21-year averages, but that’s still better than nothing.
The Acknowledgements section makes no mention of funding, by NSF or otherwise. I would have thought Urmann would have been an NSF-funded RA, on a Thompson PI grant.
Maybe he forgot to put that information in the thesis. However, he did put a lot of work into writing something that lengthy. I’m still wondering why he decided to post a reference to the thesis in this somewhat older thread.
Regarding the digitization of the graph, I would suggest that this would only give a somewhat limited view of the features therein. There did not appear to be anything new at NCDC when I checked there yesterday.
Hope your new year turns out to be good.
I hadn’t noticed that it was Urmann himself who wrote the Dec. 30 comment! That was certainly considerate and professional of him. This thread is as good as any, since it refers back to the 2007 Gleanings on Bona Churchill post, and is the most recent on Thompson data.
The line on the p. 56 graph is rather thick, so you are right that it probably does hide a lot of detail.
Chapter 3 of Urmann’s dissertation contains a lot of discussion of Quelccaya and its calibration to climate indicators. Quelccaya was a key series in the MBH99 HS graph (one of the top 3 contributors to its correlation to temperature), as well as in the Thompson et al CC03 and PNAS06 ice core indices.
For some reason, Quelccaya correlates strongly with global (or NH) temperature, while nearby Sajama and distant Bona Churchill do not. It’s of course an important statistical issue how to use the fact that some cores correlate closely with T while others don’t, without succumbing to overfitting. I had a new insight into this last fall that I hope to write up soon. (It’s probably been known for a long time, but still isn’t well enough known.)
A few years back I sent Steve a photocopy of Urmann’s 2004 MS disseratation, but it had no where near the detail of the 2009 PhD dissertation.
A question from this non-technical reader (i.e., me): I thought it interesting that the Urmann thesis states (with caveats) that the bottom of the BC1 ice core is thought to date from A.D. 265. Does that suggest or imply that the location was ice-free for some period of time before that date? Or is the ice dating situation murkier than that? Just seems of interest in the context of various current claims about “unprecedented warming” nowadays, etc.
I’m sure Steve has a much more informed opinion on this, but he’s taking a well-deserved vacation, so I’ll put in my 2 bits:
My guess (without reading the model) is that this (very approximate) number just means that any ice older than that has oozed out from under the overburden, so that there is no reason to think it was ice-free before this time. However, finding organic material at the bottom would place a bound on the potential age of the ice field.
Every ice field has finite depth, even if it is infinitely old. If it were perfectly balanced in the saddle, so that the flow was equally outward in both directions at all depths, the bottom of the field would then be infinitely old, but the annual layers would have zero thickness. However, no ice field is perfectly balanced, so I’m guessing this is how he can get a finite age at the bottom.
Maybe Dr. Urmann can help us?
Latest from the Thompsons et al.
2 Trackbacks
[…] https://climateaudit.org/2012/07/08/lonnie-thompsons-legacy/ […]
[…] did a fairly thorough review of Thompson’s non-archiving as of July 2012 here. Nick Stokes at WUWT claimed that my posts were refuted by his being able to locate Thompson data at […]