Recently, Geoffrey Boulton’s report and Nature editorial provided more pious language urging data archiving by hoarding scientists. As I mentioned in my initial comments on Boulton’s editorial, there have been many such pious pronouncements over the years without the slightest impact on, for example, the serial non-archiving couple of Lonnie Thompson and Ellen Mosley-Thompson, who, as it turns out, is an even worse offender than husband Lonnie, if such can be imagined. Their long career of non-archiving has flourished despite clear U.S. federal government policies dating back to 1991 which, on paper, require thorough data archiving by the climate community as a condition of receiving grants. Unfortunately, the U.S. climate funding bureaucracy has been thoroughly co-opted by the climate industry
and has failed to enforce regulations that, on paper, would require the Thompsons and others to archive data. Unfortunately, Boulton failed to do any assessment of why even apparently mandatory government policies have been insufficient to deter to serial non-archivers.
U.S. Policy on Climate Data
In an early Climate Audit post, I briefly reviewed the high-level policy under the U.S. Global Change Research Program, a policy that would be entirely satisfactory, if it were enforced. The policy statement stated:
Full and open sharing of the full suite of global data sets for all global change researchers is a fundamental objective. As data are made available, global change researchers should have full and open access to them without restrictions on research use….
For those programs in which selected principal investigators have initial periods of exclusive data use, data should be made openly available as soon as they become widely useful. In each case the funding agency should explicitly define the duration of any exclusive use period…Deciding when data become widely useful is the responsibility of the funding agency, which should explicitly define the periods of restricted access, if any. In the past, some Principal Investigators have retained data for indefinite periods and this has inhibited their widespread use. This practice should be eliminated
In 1997, they added that grant language should include explicit terms on how data was to be archived:
SUGGESTED DATA PRODUCT REQUIREMENT FOR GRANTS, COOPERATIVE AGREEMENTS, AND CONTRACTS. Describe the plan to make available the data products produced, whether from observations or analyses, which contribute significantly to the grant results. The data products will be made available to the grant without restriction and be accompanied by comprehensive metadata documentation adequate for specialists and non-specialists alike to be able to not only understand both how and where the data products were obtained but adequate for them to be used with confidence for generations. The data products and their metadata will be provided in a standard exchange format no later than the grant final report or the publication of the data product’s associated results, whichever comes first.
Although Boulton’s report made recommendations with similar objectives, Boulton did not review U.S. experience or why seemingly sufficient U.S. policy has been unequal to the (admittedly difficult) challenge of getting the Thompsons to archive data funded by the U.S. National Science Foundation.
While Lonnie Thompson has been a frequent example at Climate Audit of a serial non-archiver, it turns out that Ellen Mosley-Thompson is even worse. She has spent her entire career in the ice core business> According to her CV, she has led “nine expeditions to Antarctica and six to Greenland to retrieve ice cores”. However, a search of the NOAA paleo archive for data archived by Ellen Mosley-Thompson shows only one data set from Antarctica or Greenland associated with her. Lest this example be taken to mar her otherwise unblemished record of non-archiving, the data was published in 1981 while she was still junior and, according to its readme, it was transcribed by a third party and contributed in her name. I believe that it’s fair that she has not archived at NOAA (or, to my knowledge, elsewhere) any data from the “nine expeditions to Antarctica and six to Greenland”.
Mosley-Thompson has had an important leadership role in the U.S. PARCA program (Program for Regional Arctic Climate Assessment), which drilled 49 short and medium-length cores in Greeenland between 1995 and 1998. She was senior author of a summary article in 2001 – see here.
Despite the importance of d18O as a climate proxy and the promised benefits of the PARCA Program for Regional Arctic Climate Assessment, not a single d18O measurement from the PARCA program has been archived at the NOAA paleoclimate archive nor, to my knowledge and I’ve looked very carefully, elsewhere.
Squiggles for 6 of Mrs Lonnie’s Greenland cores (5 PARCA and one 1989 core) and 3 of her Antarctic cores (dating back to the early 1990s) were shown in a 2006 article. None of this data has been archived.
Nor, to my knowledge, has any other American scientist archived any Greenland d18O measurement for any year subsequent to the inception of the IPCC in 1990. I’ve looked closely at the major archives (NOAA Pangaea and NSIDC and the most recent archived d18O measurement from Greenland by a U.S. scientist appears to be from 1986 (see here), the most recent year archived from 1989 hole GRIP89-1.
The total failure of the PARCA program to archive a single d18O measurement is really quite remarkable.
Will an editorial from Boulton change the practices of the serial non-archivers at Ohio State? Not an ice core’s chance.
The only thing that would make hard cases like Lonnie Thompson and Ellen Mosley-Thompson pay attention – or for that matter the field as a whole – is to treat violations of U.S. federal data archiving policies as seriously as, say, an NCAA recruiting violation. In my opinion, the entire Ohio State ice core program should be put on probation by the NSF until they’ve archived their entire data set – not a small task given over 20 years of systematic non-compliance.
Postscript: Only one of the 10 Mosley-Thompson ice cores illustrated in Figure 1 of Mosley-Thompson et al 2006 here goes back to the Medieval period – a O18 series from Plateau Remote, Antarctica. A series that ought to be of considerable interest to multiproxy authors given the shortage of highly resolved series that include the medieval period. This series was not considered by Gergis. Nor was it included in the Neukom and Gergis list, although two other unarchived Mosley-Thompson series (Dyer Plateau, Siple Station) were considered. CA readers will probably not be surprised to learn that the unarchived and unconsidered Plateau Remote O18 series did not have a Hockey- Stick shape, as shown below.
Excerpt from Mosley-Thompson et al 2006 (Ann Glac) Figure 1
The only publication of this data, to my knowledge, was in a volume from a 1996 NATO workshop here (edited by Jones and Bradley.) This article raises a number of interesting questions about the apparent long-term cooling of d18O values at Plateau Remote over the Holocene. None of these caveats recur in the 2006 article, by which time articles by the Thompsons tend to be little more than global warming pamphlets.
Postscript 2 – Ellen Mosley-Thompson has an interesting cameo appearance in the Climategate emails. She was the EOS editor who rushed through the Mann et al 2003 EOS article on Soon and Baliunas 2003. The article took about 10 days from being commissioned to being accepted. They giggled among themselves when Willie Soon inquired about the peer review process. The character assassination of this article has not been fully analysed. In one despicable email, Tom Wigley acknowledged that Soon and Baliunas might have a point that 20th century precipitation was not unusual (a theme revisited in AR5 Zero and First Draft), writing to Mann and others (2003-06-06 682.):
Mike,
Well put! By chance SB03 may have got some of these precip things right, but we don’t want to give them any way to claim credit.
Mann, Wigley and others accomplished this by misrepresenting the actual content of Soon and Baliunas. The Mann article was rushed through by Mosley-Thompson just in time to be used against Soon and Baliunas in a Senate hearing in late July 2003 – the hearing at which von Storch announced his resignation as editor of Climate Research.
Update June 3: I’ve edited this post slightly. See here.
303 Comments
I write science grants on a regular basis
But not in the climate world — that I don’t do
The rule applies everywhere else: On its face is
Enough — it’s what government PMs pursue
For each Program Manager demands reports
Of the work that’s in progress and funded by grant
The data gets archived, with code of all sorts
Why don’t they enforce this for climate?
They can’t.
===|==============/ Keith DeHavelle
+1 Keith DeHavelle, nice irony and elegant control of metre. Form matters.
That’s Ms Lonnie to you,
If there’s 1600 meters on file
Of ice cores they’ve had for a while
But no data is stored
Then I am not on board:
Here a Ms is as good as a mile
===|==============/ Keith DeHavelle
This is one of the better ones 😉
Excellent!
The attack on Soon and Baliunas in EOS and the character assassination that followed is probably what drove Sallie Baliunas from the field. Ellen Mosley-Thompson also has that to answer for.
Willie Soon has persisted. Willie’s courage has been in the face of unremitting personal assaults.
Thank you Steve. Curiously I was writing about Lonnie just this morning. So, Mrs Lonnie was involved in allowing lies / misrepresentations of Soon and Baliunas to knock them out of the picture. That was a key moment of malfeasance IMHO. \\snip ot
I suppose that if I tortured a body of data I wouldn’t be too keen to release the body.
It would be very interesting (telling perhaps)to see what would be left, if all climate related papers which had not been independently verified, through lack of data and codes etc: were removed from the entire body of climate research.
My feeling is that this exercise would decimate,(reduce to one tenth), the total body of research.
Decimation would reduce the total by one tenth. The word comes from a Roman military punishment in which one in ten of the unit were executed.
“The word comes from a Roman military punishment”
Not really. That is colloquial usage, albeit from 2000 years ago. The word in its original Latin means one tenth. It was often used in the sense of a tithe, and can be found with this meaning in documents of the Roman Catholic Church. In the sense you describe, we could call it Roman military slang. It is generally used in English to mean “severely damage”. Some people like to criticize this usage, but basing the correction on ancient slang seems somewhat ironic.
Keeping on the pedantic theme, I find this article extremely interesting. Whichever way I look at it, I find it hard to make any defense for not producing supporting data and yet at the same time expect to have one’s work taken seriously. The concept of “honor” is rarely raised in science circles. Perhaps it should be shocking that it might be.
Oh, honorable scientist, how did you arrive at these conclusions?
Decimation was indeed a Roman punishment. In his Conquest of Gaul Ceasar recounts employing decimation on a Legion for cowardice in battle. It didn’t refer to giving every tenth legionnaire – chosen by lot – a spanking.
Its unlikely but possible that she gives the data to anyone who asks. Sure, this isn’t as effective as archiving but it would still represent full access to everyone. Have you approached her to ask for the data ?
And if she were hit by a bus?
I’m not defending her non-archiving. I’m trying to find out if (a) this is just weak data management from an otherwise excellent and cooperatative scientist or (b) a long-term attempt to keep outsiders away.
If its (a) then Steve’s snarky post is inappropriate. If its (b) then the post is well justified.
The terms of the grants that she operates under
Make clear that this data is not private plunder
The Thompsons have failed to provide when we ask
But that’s not the sole cause to take her to task
In ClimateGate files, as Steve noted above
Ms. E. Mosley-Thompson shows skeptics no love
And for Climate Crisis is all in the bag
(On hearing of Soon, said: “I wanted to gag”)
But Climategate II 1480 text shows
That within the Team ranks, some ice data flows
There are a few (non-Thompson) cores in that mail
But archiving? Publicly? They say “No!”
Fail.
===|==============/ Keith DeHavelle
I’m well aware of all of this. However, please explain why a snarky ad-hominem post would be justified if my scenario (a) is true ?
I’ve had limited contact with her. Early on (prior to CA), I asked Lonnie Thompson for data. She responded on his behalf, wondered who I was and then didn’t provide data. Subsequently, Climatic Change established a data policy and required them to provide the previously refused data (which was the first time that information from Dunde, Guliya or Dasuopu had been archvied.) On a later occasion, I asked her for a series that she had sent to another scientist and she sent it.
I had subsequent outright refusals from Lonnie Thompson and journals publishing him – there are numerous CA posts on this. PNAS refused to require him to provide an archive and accepted wilfully obtuse responses. Thompson hasn’t published Bona-Churchill data which has been the topic of CA posts. This means that he hasn’t archived it either.
Mosley-Thompson’s failure to archive has other effects even if she had an unblemished record of responses (which she doesn’y.) For example, the Plateau Remote data has not been included in multiproxy surveys. Had it been archived, it would have entered into the record.
Roger, if you are so keen to find out whether (a) or (b) is the case, why don’t you ask her for some of the data?
Paul
I’m not the one making an ad hominem post against her. Why should it be up to me to show she’s a good scientist ? The onus lies on the people making the attacks to show that their snarkiness and suspicions are justified.
Steve
Your post would have been far stronger if you had gone through the excercise of requesting the data first before writing today’s post.
There’s nothing wrong in principle with ad hominem attacks as long as all reasonable steps are made prior to publishing to ensure that such posts are fully justified. I don’t think you’ve done that here.
Even if she were to duitifully and cheerfully send the data when requested – it still isn’t archived.
These data are not personal property, and this method of operation has a limit on it – her lifetime. What happens to the data 100 years from now?
Why do you see this as an Ad Hominen attack? (Normally considered as a way of disparaging a logical argument by drawing attention to irrelevant information about the presenter) The post is a chronology of events and a criticism of behaviour that the author clearly considers mistaken. Please identify the thesis under discussion, presented by Thomson from which this post is meant to distract us?
Roger, there’s nothing “ad hominem” about a post that identifies facts about an investigator’s poor practice that are relevant to the topic at issue. The sole remedy to poor practice – if there is one – is to identify it and the bad actors who persist in it. In science classes in both high school and college EMT’s short comings in the data documentation department would have earned her a failing grade, if that element was part of the assignment – and it generally is. As a British friend expressed it, her failure is “rank bad practice.”
Roger:
you put the choices as
(a) this is just weak data management from an otherwise excellent and cooperatative scientist or (b) a long-term attempt to keep outsiders away.
Your fallacy is false dilemma
Your legal mistake is assuming facts not in evidence.
It could very well be weak data management from a shoddy scientist who is also attempting to keep outsiders away.
I watched a 2010 AGU keynote presentation given by Ellen. Lucky for me I was standing behind two
prominent and publically recognizable mavens of the AGU. Unlucky for them they didn’t see me.
I would say their opinion of her ( excellent?) did not match yours.
This is not about her. This is not about her stature in the business. This is about the best practices for an very important science. Paleo climatology provides the best estimates of climate sensitivity. The uncertainty in that figure is literally worth trillions of dollars. It is not best practices to leave that data un archived.
Since these studies are funded by the taxpayers, someone (Chris Horner, maybe)ought to bring a lawsuit against the U.S. National Science Foundation to compel it to, in turn, compel all these scientists to archive and release their data to the public; who, after all, have bought and paid for it.
The idea that a scientist need not release his data rests on a conceit: that scientists are honest. But science does not work because scientists are honest, but because their work can be checked. In experimental sciences one can repeat the experiment (and students do this in college labs every day). When repeating en experiment is not possible, the data should be made available. Drilling your own ice core is not possible. The fear of an error being found keeps experimentalists honest, and weeds out the fakers and incompetents (and there are some). But if data are not released in non-experimental fields, where is the feedback to find mistakes? Engineers have their work checked and rechecked so the wings don’t fall off or the building fall down. Should we trust them? Only if they allow us to check their work, which is why they allow it. Isn’t that simple?
I read the ’06 paper linked as Thompson, Thompson, Lin
One sentence jumped out due to an admission tucked within:
Does this not make it seem that they acknowledge the emergence
Of a parallel effect to Briffa’s famed Tree Ring Divergence?
It seems they’re almost saying that the last few decades’ scores
Don’t show as good predictors (clear from all the Greenland cores)
And these are decade midpoints, thus abstracted at least twice
With such results I’m not surprised the publishing’s on ice
It seems the line-up’s better if the temperature’s untinkered
But this they cannot see if ideologically blinkered
===|==============/ Keith DeHavelle
“Does this not make it seem that they acknowledge the emergence
Of a parallel effect to Briffa’s famed Tree Ring Divergence?”
If you look at enough O18 and D ice core series covering a few centuries to a millennium or two- and particularly those in proximate locations – and compare those results with the really long term series with stable isotopes, you might conclude that the stable isotope ratio methods can find large changes in temperatures that occur between the inter glacial and glacial maximum periods but have problems when changes are smaller – like the modern warming period.
I also believe that scientists who do not release their data to the public, like the Thompsons, and their funding agencies are fair game for speculation why they do not release it.
If the data were used in published paper but not released to the public I could speculate that the published data was cherry picked and release of all the data would confirm this action.
If the data were known to be measured but consequently not released to the public or used in a publication one could speculate that the specimens in the case of ice cores were not properly handled. This would be embarrassing to admit by both the handlers and those agencies funding the project. Another reason could be because the data did not fit some preconceived ideas about what it should look like.
What I find curious about this failure to make data public is that I would think, in effect, the funding agency owns the data or certainly must have rights to viewing it. Why would not that agency make the data public or force the scientist to make it public? Are there details that would be embarrassing to not only the scientists but the funding agency as well?
I suppose to a public viewed forgiving in these matters or naive, the agency and scientists could fall back on the excuse that the scientists doing the project have first rights to publication and then the data are released to public, but that the first publication is taking quite awhile to do and the funding agency is somehow alright with giving the scientists an unlimited gestation period. That excuse could, of course, be used to cover up the suppositions I have made above.
Which means the public’s right to know about data relating to the “greatest threat to humanity ever known” takes a back seat to scientific prestige. That really gives me a warm and fuzzy feeling.
NOT!
Plus the US federal policy in 1991 considered the problem of unlimited delays and established a policy against it.
Of course, some scientists are going to try to abuse the system.
That’s why I’ve tried to focus attention on the regulatory failures, though, for some reason, readers seem to prefer focusing on the obstructive scientists.
Roger Posted Jul 1, 2012 at 4:41 PM
You have the stick by the wrong end. Why on earth should anyone have to ask her for her data? What is this, the childhood game of “Mother May I”? I am among those who paid for her to go and collect the data, and paid her handsomely, and the terms of her getting the money included archiving the data I paid for. As a result, I’ll be damned if I’ll ask her for anything, the onus is on her, not us.
w.
Willis – I don’t have the wrong end of the stick at all. If the situation is someone who gives it out freely to everyone who asks then this is equivalent to archiving it albeit in a very imperfect way. That person would not merit the type of snarky post that Steve has written. Instead, a post emphasising the importance of archiving for reasons of efficiency and setting a standard would be appropriate.
If, however, the data are shared among a select group of scientists (or not shared at all) then Steve is absolutely right to adopt the tone he used in this post.
I’d like to know which of the above two scenarios describes this person’s approach to data sharing. I’m not prepared to believe the worst about anyone until I have evidence. I’m a great believer in the importance of archiving. I’m also a great believer in the importance of evidence-based thinking.
You write that you’ll “be damned if [you] ask her for anything”. This is a rather silly uncompromising attitude. What would you then do if the data were stored and you wanted to use them but the description of the data was ambiguous ? Refuse to send a mail since it violates your principles ?
Roger, maybe you’re right and Willis, who seems uncharacteristically tetchy today, is wrong, but there’s more to archiving than just giving out the data isn’t there. Shouldn’t she have archived the data she used, the data she rejected, here approach to choosing data etc. etc.? For years Jones of the CRU ya boohed everyone be pointin to the GHCN and saying the data are in there. They weren’t of course because he meant the manipulated data, he’d lost the raw data when they redecorated, or something. Now if he’d have archived it in a scientific manner…
Geronimo
I wholly she should have archived her data. Its bad practice not to have done this.
Geronimo here makes the point that I believe Roger totally missed and that is the proposition that we have to see all the data and not just that data used in a publication. That proposition has nothing to do with snark or defense of the Thompsons but rather a common sense statistical approach and a view of what hard science has to do to be considered hard.
How did I miss the point ? What part of “she should have archived her data” implies that I believe a subset of data should have been stored ?
Roger Posted Jul 2, 2012 at 4:44 AM
No, that’s not equivalent to archiving it in any way at all. An “archive” is somewhere that anyone can freely access the data. It does not depend on one person agreeing to release the data on a case by case basis, even if they might do that. It is a PUBLIC repository that anyone can access to examine all of the data.
In addition, you seem to have ignored the point that the terms under which she took my money to do the study REQUIRE that she archive the data, and this is not satisfied by her privately holding on the data and releasing it upon request, no matter how freely she might release it.
This is the 21st century, Roger, and the rules have changed. She is ignoring that, and I can imagine why. When people hide something, it’s usually because they have something to hide. The curious part to me is that you are struggling mightily to justify her actions, and I can’t imagine why. The old ways are no longer adequate. I don’t care if she gives it out freely, that’s not the point. An archive allows me free unfettered access at 2AM without any delay, I don’t have to email her and say pretty please can I see your data and then get a message that she’s off on some glacier drilling holes and will get back to me in a month … in the 21st century, that is totally inadequate.
Finally, archiving these days is so easy to do, requiring such little work, that I totally fail to see why you are acting as her apologist. Since she is required to archive the data by the terms of her grants, and since it is so easy to archive it … then why are you working so hard to come up with some bogus justification for her refusal to make her data public despite repeated requests? I don’t get it.
w.
Willis
If the data are available to anyone who wants them then its archived and available. You accuse me of ignoring your point yet I notice you build your case by ignoring mine.
I put forward the hypothesis that she would give it away to anyone who wants it. Sure, you can call this a “request” if you want but no more so than going to the library and reqesting the librarian to get you some reference books from the storage area. A scientist at a public university can easily act as an archivist. Now, whether she would give her data away freely is an unknown and it shouldn’t have been prior to today’s CA post.
BTW I’m not acting as an apologist for anyone. I just dislike ad honimem attacks unless they are properly through through. They tend to diminish the people making them far more than those who are nominally on the receiving end.
Finally, there is no need TO SHOUT. I’m not stupid. I just disagree with you (and Steve).
Roger said,
” … she would give it away to anyone who wants it.”
Until she decided not to, or she lost it or she died.
Roger, disliking “ad hominem” attacks is a good thing. The inability to differentiate between “ad hominem” and identifying poor practice is not. Willis’ reference to the “old days” is actually mistaken. Throughout the history of the US until we moved into the 1970s publicly funded data was also made available publicly. Bugaboos about security and increasing issues with “IP” and little, special, side deals that excluded bits of publicly-funded work have snow-balled into the mess we have now.
The funding agency originally enforced data availabilty. This is why for instance you can, minimal cost, acquire high resolution maps of the entire US. It why soil maps are actually free, along with all the necessary documentation to understand and use them. You should not have to ask EMT for a copy of her data. Indeed, you should be able to ask the agency or agencies who funded her junkets for it.
Whether she will supply it to you or not is irrelevant. Whether she is really a nice person with poor organizational skills or not is irrelevant. As long as she was using public money to do her work, she was quite literally a public employee and that “work product” was not “hers.”
So, in fact, it might not be a bad thing to start asking the funding agencies, rather than the publications, where the data is.
Roger
Even under this condition the data is not available to everybody. It is only available to those who have prior knowledge of the data’s existence.
If it is archived then it is available to everybody.
Roger
Why should we have to ask the author for data that is required by law to be published? That undermines the rule of law.
Also, if the data is not archived we don’t know what to ask for.
How much has she done about which we are unaware? How many inconvenient series are thrown (metaphorically) done the nearest bore hole.
Mooloo
I don’t follow the logic of your post.
You would ask for all the raw she took.
raw -> raw data
David L. Hagen
If she gives it away to anyone who demands it then how is that not respecting the law ?
<blockquote) … and the terms of her getting the money included archiving the data I paid for.
Did they? Were the data “deliverable?” Were measures for their retention and means for their distribution if requested a specific line item in their grant? In plainer words, were they contractually obligated to do this, or was it an ethereal “recommendation?”
The post above is factual
The wording underscored:
The duty was contractual.
And then, of course, ignored.
===|==============/ Keith DeHavelle
Keith,
Is the wording to which you refer compulsory? Is it ever a contract requirement for grants in this area? Was it a requirement of any of the Lonnies’ grants?
If so, then there appears to be a flagrant failure to perform all of the obligations undertaken by receiving the grant funding.
It looked to me that the practices which I agree should be contractual requirements are not and that these practices are only recommendations from on high, however ineffective that might be.
john ferguson
All were alarmed by our collective refusal to slow down our use of fossil fuels. One wryly summed up our behaviour as a “remarkable experiment”, a quip I pass on to Thompson as we leave his ice-core room. “He forgot one word,” Thompson says, ready as ever to add to the record what is missing. “It’s a remarkable uncontrolled experiment.”
National Geographic, Sept 2004, p74.
Don’t know if this is what you’re looking for, but I found the following, seems to be a number of short cores:
PARCA-1997-cores.XLS
PARCA-1998-cores.XLS
Steve: Interesting. Hadn’t seen this – this is from the web archive and not presently online. However, doesn’t have any d18O values.
Steve, there are indeed some del O18 records. Hit the first 3 bottom tabs to find some in 1997 cores.xls
That’s what I thought too, but it turns out those are acculumation rates estimated from del 18O, not the underlying values. However, this paper may the be paper that describes the data. From the abstract:
“While most of the final results, such as surface velocities and accumulation rates, have been published previously in various reports and papers, the raw data and basic details of the observations are not easily accessible. The purpose of this review is to present these data, including data hitherto unpublished in a single place, and to evaluate and review their quality for those who have used or plan to use these observations.”
Thank you mt, I see what you mean. It’s not so tidy a presentation and I apologise for jumping the gun in trying to be helpful. Geoff.
This woman got her PhD in 1979, long before current controversies were dreamed of, at a time when not all that many women were getting PhDs. /Whatever/ your view on her professional activities subsequently, the references to her here as ‘Mrs Lonnie’ are a pretty cheap shot.
Steve: Lonnie Thompson’s failure to archive has been a longstanding issue at Climate Audit. That she’s married to Lonnie is relevant to the point of the post. I might add that she responded on behalf of Lonnie to an inquiry from me.
Not optimal irony perhaps. Skimming it earlier I was reminded of Groucho Marx: “You’re a disgrace to our family name of Wagstaff, if such a thing is possible.”
None of which excuses or explains the references to her as “Mrs Lonnie”. It’s one thing to criticise an individual, or indeed a pair of individuals who happen to be married and are therefore inevitably very closely associated, for their activities in the professional sphere: but you have to stoop pretty low when you make the fact of their marriage a hook on which to hang your contempt for their science.
Re: Jon Grove (Jul 2 11:16),
“Contempt for their science”? I seem to have missed that. Rather, I find a well founded “name & shame” exercise for their behaviour with respect to archiving the data upon which their publications are based. Since both of them have been doing the non-archiving thing for many years I don’t find it problematic to point out the fact that they are married.
My comment was supposed to be to “Roger”, not bob edgar. Sorry!
Steve, it’s your blog of course, but I have to agree with Jon Grove that “Mrs. Lonnie” detracts from the point of your post.
If one wished to emphasize that they are not just colleagues but married, “Mrs. Lonnie Thompson” would be a more neutral mode of reference.
That said, Bob Edgar and Andrew Russell are quite right to jump on Jon Grove’s equation of your relentless criticism of the Thompsons’ scientific lapses with “contempt for their science”.
I would not have used the term myself, but it highlights Steve’s point that the Thompsons are apparently like-minded about archiving behaviour. Many “field” people are. They really exist to be out in the field, don’t like labs, computers, or analysis and writing. There are lab folks for that. They possibly are obstructionist about who acquires access to their data, especially since the format embarrassingly resembles an abandoned bird’s nest, and they don’t want to answer questions like, “isn’t there any metadata?” I also would not be astonished to discover that it really is publicly available at an undisclosed location.
I think that the funding agencies are the principle parties that should be responsible for ensuring the conditions of a grant are met, and should withhold a significant chunk of the money (25%?) until all obligations are fulfilled.
One reality of doing research is that data accumulates at geometric rates while office space stays constant. If I had not included all my electronic data in a printed form in my thesis, it would be long gone. I am really not surprised when I hear a researcher say, “I don’t have that data any more. I was looking for it myself six months ago. My office has moved four times since I collected it and things have vanished each time, never to return.”
RE Roger:
***Its unlikely but possible that she gives the data to anyone who asks. Sure, this isn’t as effective as archiving but it would still represent full access to everyone. Have you approached her to ask for the data ?***
Giving the data to one person who asks is not full access. If you had been following Climateaudit, you would know this has happened in the past by Jones and it was not shared.
If she is a real scientist she would follow the archiving guidelines given by the various publications.
Are you really suggesting that someone who would give freely the data to anyone who contacts her but who didn’t put in a public archive isn’t a “real scientist” ?
If she refused to give the data away then I may well agree with you. However, neither you nor I have any evidence that this is the case.
Are you really suggesting that someone who is required by the terms of her grant to publicly archive her data but who repeatedly refuses to do so is a real scientist?
All of this “isn’t a real scientist” stuff doesn’t help your case. Its simply sloganeering. I recall the “not a real scientist” slur used against Steve. It was daft in that context and its daft here. Its pure adhominem hyperbole.
The difference between making data freely available when someone requests it (and please disprove this before ranting**) and putting it in a public archive is miniscule. To then declare that this means she’s not a scientist is ridiculous.
** A somewhat pejorative term I know but IMO justified if one looks over your contributions to this thread.
The is a huge difference between archiving and making data available on request, even before considering the legal requirements.
Archiving freezes the data at a point in time. This is a critical step in establishing the “authentic” data for replication.
Making data available on request creates multiple versions of the same data, which may or may not be identical. Who can say which one is the correct version? Which one do we use for replication?
> An ad hominem attack is when you refute someone’s argument by attacking them, rather than the point they made.
We should not conflace an ad hominem **argument** with an ad hominem **attack**, which is self-describing. Calling a couple “non-archiving,” with all the connotation it entails, might very well be considered as a personal attack, ergo a slur.
Same goes to calling Thompson a prick. It might be true. But it is still what it is, and nothing else.
Roger,
If you have the courage of your convictions, why don’t YOU ask her for the data. And while you are at it, ask her why she refuses to archive that data.
Andrew
I’m surprised by your post. Isn’t it the job of the people making the slurs to find out all the facts before launching into ad hominem attacks ?
An uncritical attaboy attitude doesn’t help Steve.
Hi, Roger. You have called this an “ad hominem” attack quite a few times. I don’t think Steve’s post is ad hominem. An ad hominem attack is when you refute someone’s argument by attacking them, rather than the point they made. Calling Ellen M-T a serial non-archiver, however, IS the main point of the post. Steve is speaking directly to that with all the evidence he has accumulated.
Further, the ancillary points are not ad hominem, either. One could make the argument that Steve has implied she is biased and unscientific by pointing out, for example, her brief-but-significant appearance in climategate emails. And this is true – these facts do give that impression. But Steve does not draw that conclusion explicitly and he did this, I believe, by design. Steve basically said, “Here is some more evidence worth considering that may help us to understand Ellen M-T’s actions.” He DIDN’T say, “M-T is biased and unscientific for these reasons…” And he also didn’t say “M-T’s paper on so-and-so is wrong because she doesn’t archive her data.” That would have been an ad hominem attack.
As to the other point about archiving vs. sharing, Roger is essentially correct … his point is both valid AND insignificant. It is valid because, in the parlance of networking and computer science, when you go to a webpage, you are making a TCP/IP “request” for that page, one that is filled automatically. Roger is saying there is no real difference between fulfilling that request automatically vs. manually. It is tedious going back-and-forth like this on all the difference and similarities between automatic and manual fulfilling of data requests.
However, this (valid) point is also insignificant. It would have been significant if Steve’s post said, “L-T never shares her data because it isn’t archived.” But he didn’t say this. Instead he said “L-T never archives her data.” This is the main point – it’s even in the title! – and it seems to be true, thanks to the copious evidence provided.
Roger, I think if you want to argue that non-archiving “doesn’t matter” – which seems to be the thrust of the “archiving vs. sharing” debate – then you have to provide some evidence towards that end, rather than just the possibility of some evidence.
Roger
First: I’m not entirely sure which words you consider to be “slurs” or “ad hominem attacks”. No one has criticized the Thompson for merely being who they are. Nor has anyone said what they are doing is bad because they are the Thonpsons. People have criticized the Thompson’s for their actions which people think do not meet appropriate standards. This is not “ad hominem”.
Whatever or criticism someone makes they only need to discover facts that are relevant to the criticism. Whether the Ellen provides data privately is not relevant any criticism that she does not place data in a public archive. And no– it doesn’t become relevant merely because you make an argument that private sharing would be good enough for you personally. The fact that you might find private archiving acceptable doens’t magically transform it into public archiving.
You are, for some reason, going on and on and on about the hypothetical possibility that she might give data out to “anyone” if only they ask. And you seem to suggest this ought to make some sort of difference. Well… clearly, this would still not be a public archive — and so makes no difference to any criticism that she does not place her data in a public archive.
But maybe if she does hand out data privately, that might make a difference to what you claim is your view– which is that would be ok if she provided it privately.
But even though you tell us it makes a difference to your view– and suggest it ought to make a difference to ours, you seem to be unwilling to discover whether she does provide it privately. Maybe she does. Maybe she doesn’t.
Maybe if she does, you could persuade someone, somewhere that– while SteveMc is correct that she does not archive publicly– her failure to publicly archive isn’t so bad.
I think if you are going to keep yammering on about your theory of how people’s views should be changed based on her willingness to provide data privately, you ought to do the work to find out if she even does give out data privately.
As it happens: No matter what you discover, I’ll still think she ought to archive the sorts of stuff she collects publicly. I’ll think it’s fair of SteveMc to point out that she doesn’t. And as long as she doesn’t archive publicly, I think it’s perfectly fair to call her a “non-archiver” because that’s accurate. The label merely connotes “one who doesn’t archive” and I don’t consider a “private data trove” to be “an archive”.
As for your throwing around accusations that someone needs to do more work to support whatever “slur” you imagine is being throwing around: You may think it’s a “slur” to call someone a “non-archiver”. But if you do, I’m a bit mystified why you would consider the term to represent a slur when you seem to think it’s ok for a person to be a non-archiver. If it’s ok to be such a thing, the term to describes an accurate acceptable features must not be slur!
Just because a term is descriptive doesn’t make it a slur.
I too would like clarification on what Roger regards as a “slur”. At times, it seems to me that members of the climate community regard not archiving as a badge of honor. It is a matter of fact that Ellen Mosley-Thompson has a lengthy track record of not archiving data i.e. is a serial non-archiver. Contrary to Roger’s allegations, I examined the archives carefully before making these statements and, thus far no one has produced a shred of evidence to contradict any of my assertions.
With regard to postscript 2, it is worth spelling out that the team hatchet piece in EOS, in which they included data they knew to be wrong (see CA post ‘perpetuating rubbish’), was in fact Ellen M-T’s idea. See email 3323.txt,
Science proceeds by way of a conversation among those with something to say or with observations to report. Until the post WW2 era, this was pretty much a free enterprise business. Anyone could contribute to the conversation and they could say what they liked and make what claims they liked on what evidence they liked. In due course questions would be settled, at least for a while, as what became accepted as the most compelling arguments led to a single conclusion. What constituted a compelling argument depending on the logic, the apparent veracity and quality of the data, the perceived integrity of the data source, the reputation of the channel of publication, etc. — as judged by those engaged in the conversation.
Today, governments fund many scientists and thus seek to make rules about how scientists should proceed. They set goals, they establish codes of scientific ethics, they may demand data archiving, disclosure of interests, etc. Is this a good thing? Overall, I doubt it. It leads to Lysenkoism and Al Gore’s Nobel Prize.
One understands that auditing the work of others is not possible if they refuse to release their data. But so what? Those who will not release data must be judged on that. Journals that do not establish reasonable rules on data archiving must be judged on that. There are no certain criteria for establishing the truth. But letting government set the rules of the game guarantees that the “scientific consensus” will always be more questionable than it need be. We should care less about enforcement of government agency rules about data archiving in climate science and more about getting government out of climate science altogether.
You’ve described the conditions pertaining in what Michael Polanyi called “The Republic of Science,” where a spontaneous order can generate scientific progress because each investigator’s interest in making discoveries and getting cited internalizes incentives to pursue practices that facilitate discovery by others. But in a regime of external funding and policy-relevant science, the incentives are not properly internalized by the investigators, who will therefore tend to under-disclose, under-archive, under-criticize their funder’s preferred set of beliefs, and overstate findings conducive to to the policy preferences of their social reference group.
Yes. Well put. So Michael Polanyi, would have condemned any role of government in climate science, as I assume would his Nobel-Prize-winning son, John who wrote:
Faced with the admitted difficulty of managing the creative process, we are doubling our efforts to do so. Is this because science has failed to deliver, having given us nothing more than nuclear power, penicillin, space travel, genetic engineering, transistors, and superconductors? Or is it because governments everywhere regard as a reproach activities they cannot advantageously control? They felt that way about the marketplace for goods, but trillions of wasted dollars later, they have come to recognize the efficiency of this self-regulating system. Not so, however, with the marketplace for ideas.
And here’s another comment on the counter-productive effect of government funding of science from Terence Kealey, Vice-Chancellor of the University of Buckingham:
There is a central myth about British science and economic growth, and it goes like this: science breeds wealth, Britain is in economic decline, therefore Britain has not done enough science. Actually, it is easy to show that a key cause of Britain’s economic decline has been that the government has funded too much science…
So even though she is a published scientist who uses her own name in her professional life, to you she’s just somebody’s wife who doesn’t even get mentioned by her own name in your blog post title. Do you view all married women as appendages or even possessions of their husbands?
Problem is, if they don’t provide the means to replicate their work they ain’t really scientists. The data has become an appendage or even a possession, to coin a phrase. That’s the issue here. It’s completely gender neutral.
Problem is, if they don’t provide the means to replicate their work they ain’t really scientists. The data has become an appendage or even a possession, to coin a phrase. That’s the issue here. It’s completely gender neutral.
Kind of over the top. Even if their data were available, “replication” calls for more than re-analyzing old data. That can be part of replication, but it is not the most important part. What it offers is the means to conduct an autopsy of the analytical methods. To replicate her work would actually require either collecting new ice cores from the approximate same locations, or re-sampling the original cores, and running the entire suite of chemical analysis once again. After that, you can then reapply the mathematical methods or an alternative.
Asking for the “data” in this particular situation pretty much says, “I want to check your arithmetic.” The instant response is a red face and a p***** off state of mind. The fact that someone like Steve, who already has a track record for embarrassing effectively-amateur statistical operators like Mann, would only enhance the paranoia, unless of course it wasn’t really paranoia.
Not delight that someone is interested? And not the slightest trouble because all you have to do is point the person asking to a public URL.
That’s the norm for a real scientist. What you speak of is pathologically wrong. It really is.
Aren’t you overreacting somewhat here? In my view, Steve’s title seems to be a not unreasonable way of stressing that the partner of a serious offender of the archiving process is herself an offender of similar magnitude with a long track record of undisclosed data. Two peas in a pod…
As far as mentioning her own chosen professional last name, I counted ten such usages of such in the body of the post starting in the first paragraph as opposed to only three “Mrs Lonnie” mentions. I noticed nothing in the post that even hinted in any way of the existence of a subservient relationship between the two of them.
Louise, and those above, aren’t over-reacting. They’re deliberately changing the subject to Steve rather than Ellen Mosley-Thompson’s lack of archiving.
No-one is actually deeply offended by someone else being called “Mrs Lonnie” even – and I accept it is – in slightly bad taste. The purpose of the outrage is to deflect. Please don’t feed the Trolls.
I know a lot of people who would object to Steve’s description of this person as Mrs Lonnie and who have no dog in the climate fight.
What evidence do you have that Louise is a troll ? Why do you assume that people only want to talk the topic that you find most interesting ?
Mooloo: you kind of make my point for me here. I’m not changing the subject. The subject is the question of archiving. It is perfectly reasonable to point out the fact that EMT and LT are wife and husband, although I fail to understand how this makes any substantial difference to the matter in hand.
Is the suggestion, for instance, that Mosley-Thompson’s archiving practice would somehow be less vulnerable to criticism if she had no association with Thompson? Is it worse because she /is/ associated with him? Or does this fact actually make no difference at all to science and discussions about best practice? What do you think?
As for the choice of language, yes it does matter, because it is not neutral: you may as well just call her Lonnie’s little lady, because that is the implication. The form adopted three times in the blog is exactly the kind of usage that still crops up regularly when people want to help their audience along, with a wink and a nudge. It’s unnecessary, unpalatable — and troublingly /not/ superficial. Why not just make the case with facts that suit the argument? (My parents were researchers and collaborators of a departed generation, who worked at the same institution. Had anyone ever dared to refer to my mother in this way, she would have given them short shrift indeed — and quite rightly so. But no-one /would/ have dared do so in a public theatre: the technology has changed things, and not for the better in my opinion).
I do not regard Mosley-Thompson’s non-archiving as merely an “appendage” to Lonnie’s non-archiving. I believe that she is a serial non-archiver in her own right. I believe that my post clearly drew attention to her conduct as a serial non-archiver in her own right and not merely as an appendage.
If anything, I think that my post pointed out that she was Lonnie’s equal or superior in non-archiving.
I also think that it is relevant that they are a serial non-archiving couple and that it would be incorrect to discuss each non-archiver’s conduct in isolation.
Sure, they are married. There are ways of referring to that status that aren’t implicitly demeaning: and ways that /are/.
But I still don’t see what specific difference the marital connection makes to any argument about scientific practice. On the whole, I would have thought that the fact of this legitimate contractual relationship makes any shared behaviour you identify between this couple less interesting rather than more so. What exactly is it that we are supposed to /learn/ by dwelling on this marital partnership, above and beyond the very simple observations about archiving? Why is it so significant that it requires this level of emphasis?
Today’s New York Times article also refers to Ellen and Lonnie as a couple. I fail to understand why the NYT is permitted to observe this connection, while I’m supposed to overlook it.
I saw that article, front page and above the fold! All I can say is WOW! Steve, either you have the most amazing ability of anticipation and timing, or the NYTimes rushed out this article in reaction to you yourself. What a complement! The latter is not out of the question, the PR people at Oglivy or whomever who is running the show operate with 24 hour and similar short news cycle-type thinking, and may have rushed out the article having seen yours.
I personally enjoyed the NY Times article very much, having first read yours. But its ending, the Heyhoe-like corrupting of religious experience for purposes of the AGW industry, was just too corny. Or does the Times now think its readership is now susceptible to such strategies?
Roger Posted Jul 2, 2012 at 1:11 PM
Roger, why do you think they have policies requiring people to archive data?
The obvious reason is, if Mosley-Thompson is out of town, or not feeling well, or out on a glacier, the data is NOT AVAILABLE. I do wish you’d think about the difference between a public archive, and a private individual who lets data out on a one-by-one basis, and then only if they are in town, and not sick, and not working, and not in a bad mood, and have the time to reply, and are not mad at me because I called them “Mrs. Lonnie”, and are able to understand my language, and …
If you think these two situations are equal:
• on the one hand a public archive that I can access immediately at 2AM without having to be able to speak English, and
• on the other hand someone doling out data as and when they get a request (in English, or maybe the request has to be in Lithuanian because they don’t speak English) and when they are available and in the mood and not gone for three months drilling in Antarctica,
you truly don’t understand the concept of “public archive”. Your method worked fine for years, Roger, but we’ve moved way, way past that now.
w.
“we’ve moved way, way past that now.”
But I thought it was the complaint that we have not moved past that now.
the idea that there is some established formal scientific method to which everyone must strictly adhere or have their reputations publicly trashed, if they are not forced to withdraw from science altogether, seems totalitarian.
If you don’t think much of what X or Y have done, or the way the have or have not made there data available to you or anyone else, fine, make your criticism known. But the idea that there must be some kind of scientific police force, i.e., government compulsion, that forces scientists to behave in particular ways, will mean worse science, not better.
Science needs to remain a free enterprise and largely unregulated process. If government seeks to regulate what scientists can say, or the conditions on which they say it, we will suffer from the inevitable conflict of interest between government and science. Government is to do with the exercise of power and if it controls most of science as it does today, it will continue to favor science that supports its policy — as in the case of climate science — thereby corrupting and diverting the search for knowledge.
“the idea that there is some established formal scientific method to which everyone must strictly adhere or have their reputations publicly trashed, if they are not forced to withdraw from science altogether, seems totalitarian.”
CS, I too agree that one has to make judgments based on what a scientist or agency does regarding making their data public. I also do not think that those scientists and the community that might support them should be overly sensitive to those judgments that might, in effect, say or imply we do not hold your results very high or important because we do not have the data to test it or we say that it causes us to question your motivations for not providing it. Surely the IPCC would be in a position to make that judgment and the fact that they do not leads us to judge that body accordingly.
In an entirely private enterprise I would have great faith that the market place would reward those who can be trusted and punish those who cannot – at least eventually. Unfortunately in today’s world of government funding it is not the market that operates and if we as taxpayers fund an operation we should have a right to know the results. If a scientist wants to keep his data to himself let him go to private sources for it – and there you might or might not be able to maintain your prima donna status. What better way to establish and maintain a totalatarian regime than to fund only those scientists who provide the “correct” results and public curiousity be damned.
“if we as taxpayers fund an operation we should have a right to know the results”
I agree, but are governments ever motivated by consideration of citizens’ rights?
But I question the value of demanding that right, because it presupposes a government right to regulate science, which I do not accept.
The tradition has been that a scientist is a gentleman, or the female equivalent, and can be relied upon not to misrepresent the facts. This is analogous to the assumption that a parliamentarian never lies. Thus just a parliamentarian cannot be accused in parliament of lying, so traditionally, scientists were not asked to publish their raw data.
These are, in a way, absurd rules, but because of the general assumption of integrity among scientists, it means that the career of a scientist, like that of a parliamentarian, if caught in a lie is finished. This provides a powerful incentive for integrity in science.
cs, she must believe it’s golf.
Parliamentarians can’t be accused of lying in parliament? That’s interesting news. Congress critters sure as heck can be accused of lying in Congress. They do it all the time!
If you go by “tradition”, traditionally scientists were not funded by the government. Those who were funded had delivered whatever it was the people who funded them wanted. Those who might need to prove something to outsiders (like the Patent office) etc. delivered whatever it is the outsiders needed to make their decision.
Also, at least on task oriented research funded by the government many scientists were required to archive– though those archives were cumbersome to access. The stuff might consist of lab notebooks, punched cards, tapes etc. At national labs with quite a bit of stuff was archived in boxes stored in records vaults. At PNNL, I remember having to get someone to scan records to make sure they weren’t radioactive before sending them to the DOE records vault!
Now the taxpayer government funds lots of work and its reasonable for taxpayers to want the benefit of the work funded by the government. Requesting people to archive data as part of a deliverable is useful. It’s pretty easy even for university types to find an archive for data underlying a publication and — as written– NOAA requires this. But evidently, they don’t try very hard to make sure anyone actually does it.
People who want the data have a right to complain. And if NOAA doesn’t enforce it’s rule, public shaming of scientists who don’t voluntarily comply seems the only mechanism for the public to try to enforce appropriate behavior.
CS,
You may not realize this, but what you are advocating is “post-normal” science. The issue has nothing to do with coercion or “science police” and everything to do with best practices. Taking someone’s word for fact is an error of authority. It can’t really be avoided in science, since you really can’t go to the trouble of reinventing everything, but in a controversy, one side cannot simply demand every one on the other side to take their word for fact. That is not scientific practice. At best it is politics and at worst religion – viz calls for “crucifying” skeptics.
There is an immense literature of philosophy of science, but in the end, the effective distinction between science and religion or politics is productiveness. Science yields useful new knowledge or useful alternate understanding of old knowledge. That requires that each scientist addressing the same research question with the same methods be able to report whether the original results were reproducible. Science is not “free enterprise,” that is economics.
An inventor may be scientifically trained, but unless he wants his ideas public, he has to depend upon being able to replicate his own work. The confusion of science and economics is what has lead to the present IP political debacle, where increasingly we see public money applied to private gain when publicly funded research is sequestered, patented and used for private gain. This is not by any means limited to “climate science” either. There are outstanding examples in every major technical discipline that enjoys tossing “science” into the title.
Willis
When did I ever say that the approach I postulated was anything but imperfect ? You accuse me of missing the point and yet studiously manage to misrepresent everything I’ve written.
I looked through the Boulton report and found no trace of a discussion of computer codes. This may not have been in his original remit. But isn’t that also an important element of accessibility?
There was none despite the what seems to be awful problems with the code and no version control over the data , none of the ‘reviews ‘ covered this .
Amazing given how great the claims , that the poor quality of the basic tools used to support these claims was never questioned.
Some people are missing a serious point here. This data program that involved ice drilling was a data infrastructure type program. It was obtaining data that was believed to be widely important to many people, to be a base for other studies. Like the UAH satellite data. Or NOAH ocean records. Not some obscure study on a species of frog or something. Not data to be given out to the friends of the scientist. Certainly not to remain unarchived for 20 yrs. Why would the funds be given for such expensive work and then have it never see the light of day?
Oh, yes, we have mapped the sand bars that endanger shipping using gov money, but you can’t see the results for 30 yrs or maybe never. Yes, we have discovered the Higgs Boson (trust us).
In addition to promoting good scientific practice, data archiving serves another important but more mundane purpose. Funding for research through NSF and other granting agencies is always in short supply. Without access to data collected with prior support, new investigators would have to duplicate effort and reinvent wheels to gather the same data again multiple times. From the point of view of the granting agency this is an inefficient use of resources that could be put to better uses. NSF and other grantors therfore have a direct, pecuniary interest in data archiving. It is one reason why these agencies have adopted clearly stated policies that grant recipient must have a viable plan for archiving data. In the case of the Thompsons the question is, did such policies exist with the programs that provided their grant monies, and if so what specifically were those policies?
Roger originally said (July 1, 4:41PM):
“Its unlikely but possible that she gives the data to anyone who asks. Sure, this isn’t as effective as archiving but it would still represent full access to everyone. “
In writing these words, Roger created an “inverse-strawman” of a researcher who was violating the official rules, but perhaps not the full spirit. At the same time, Roger says it’s unlikely, and not as effective as the official procedure. However, in reading further, it appears that some here have mistakenly taken Roger’s imagined “informal compliance” as IN FACT being this researcher’s SOP. Where did that come from?
My postulated “informal compliance” may well even be formal compliance depending on how rules are interpreted.
Regarding the confusion between the argument being a postulate rather than a statement of fact, I would suggest that people bother to read the content of my posts before leaping to erroneous conclusions and writing thunderous replies.
I’ve read each of your posts Roger, and I can easily conclude two things, and one flows from the other. 1. You make false accusations against Steve for making “ad hominem” attacks, when that’s clearly what you’re trying to do to him.
snip
Is there anyway to find out how much money the NSF paid the Thompsons over the years to collect and utilize these cores?
CS Posted Jul 2, 2012 at 4:48 PM
I agree that we don’t need a scientific police force. However, we do need clear requirements for what we expect when we give people money to do scientific investigations. And we have such requrements … but they are not enforced. That’s the problem.
It’s not scientific cops, it’s just common sense to say if you are doing science on the public’s dime, you need to to disseminate that information as widely as possible. We can’t each go to Antarctica and mount a drilling operation, so we pay a few people to do it and ARCHIVE THEIR FINDINGS for the benefit of all the rest of the scientists.
Without that last part, why on earth should they be funded using public money?
w.
Amen!
“it’s just common sense to say if you are doing science on the public’s dime, you need to to disseminate that information as widely as possible.
Common sense, yes. But it ain’t government policy. As I’ve discussed above, governments control information very carefully. What is disseminated should conform with policy, not set policy.
“But it ain’t government policy.”
As Steve indicated above it is government policy. If you think Steve is wrong then please demonstrate your case with some evidence.
I think the following, from the Annals of Mathematics website, is the simple and obvious requirement for any research paper that depends upon data and computer programs:
“Computer-assisted proofs of exceptionally important mathematical theorems will be considered by the Annals.
The human part of the proof, which reduces the original mathematical problem to one tractable by the computer, will be refereed for correctness in the traditional manner. The computer part may not be checked line-by-line, but will be examined for the methods by which the authors have eliminated or minimized possible sources of error: (e.g., round-off error eliminated by interval arithmetic, programming error minimized by transparent surveyable code and consistency checks, computer error minimized by redundant calculations, etc. [Surveyable means that an interested person can readily check that the code is essentially operating as claimed]).
We will print the human part of the paper in an issue of the Annals. The authors will provide the computer code, documentation necessary to understand it, and the computer output, all of which will be maintained on the Annals of Mathematics website online.”
Really, is that so hard?
Roger Posted Jul 2, 2012 at 11:48 PM
Real science is not a slogan, Roger. It is a way of looking at reality. Real science is distinguished from other ways of investigating the world inter alia by the idea of transparency. I make some claims, and if I am a scientist, I make all of my data and code public so that people can freely investigate to find out what is wrong with my claims. That transparency allows them to see each and every bit of the data, and everything I’ve done to the data, so that they can poke holes in it. If they can’t poke any holes, my claims are tentatively taken as being true.
However, if I don’t put my data out there for people to investigate, I am not engaged in science. Instead, I am making unfalsifiable claims, which we generally call “anecdote” or “religion” or “faith” or “belief”. Transparency is at the core of science, and science does not and cannot work without it.
The difference between somebody keeping the data under their own personal control, and having available in a public archive, is huge. For starters, as I pointed out above, if you don’t speak the person’s language it’s kinda hard to ask for it. The same is true when Ms. Mosley-Thompson is off the grid in Antarctica. The fact that you don’t see that is odd. And what if Ms. M-T dies while on an expedition? Minuscule difference? Not. At that point, the data could be lost forever, hardly a minuscule issue when I helped pay for the data and it is not duplicable.
Heck, what if she just loses the data? We’ve seen that happen more than once in climate science, from Phil Jones on down the line. So no, a person privately holding data is not the same as a public data archive, for a host of reasons.
More to the point, however, is that she is not keeping her word. When she got the grants she agreed to the terms, among which were archiving the data. She has not archived it.
Finally, for all we know, she is altering the data as we speak … I am certainly not saying that she is doing that, I’m pointing to another reason why archiving is important—it prevents post hoc “adjustments” to the data. I greatly doubt that she is altering the data, but there is nothing to stop her from doing so … and given the historical actions of other AGW stars when they were not being watched, that’s not a good thing.
Look, if you don’t follow the rules of scientist, you are just peddling your personal opinion. She has refused to archive her data despite taking the money. That makes her a wholesaler of anecdotes, and a teller of uncheckable tales … but it doesn’t make her a scientist.
Oh, please, don’t mistake your opinion for the truth. Sure, you disagree with me, but that doesn’t mean either that you are right, or that my contributions are worthless. That’s just your ego patting itself on the back before the votes are even cast, much less counted.
Roger
As per the terms of her contract she is supposed to publicly archive data. She did not do it and in the absence of publicly archived data her work is not scientific. It is a breach of contract.
And no amount of hand waving apologies by you on her behalf can convince anyone in this aspect. Stop digging when you’re already in a hole.
Willis
If someone writes IMO as I do then it implies that they are certainly not mistaking their opinion for truth; they are, in fact, qualifying their statements. Ironically, your accusation describes your *own* posts which are not qualified at all (hence the description of rants, especially given your use of upper case).
I stand by what I wrote. If someone has a policy of giving data to anyone who says “give me the data” then its difficult to describe them as being “not a scientist” or not following the scientific method. It may be an imperfect method of data storage but is certainly not worthy of the type of ad hominem nonsense witnessed on this site. A well reasoned and well researched ad hominem attack can be very effective. This was a poor ad hominem post from Steve.
Its interesting that Steve frequently compliments (or at least moderates his criticism) of climate scientists who send him the data. By and large the posters follow his lead then. This time he attacks without even asking what this person’s data policy is or whether or not the data are available to all, albeit stored in an imperfect way. Once again the posters follow his lead and don’t ask whether or not such a position is consistent or fair.
The posters don’t do Steve a service by acting as attaboys.
In simple English, she’s required to archive the data. Period. No ifs and buts. How difficult is that to understand?
Roger
Actually, in a comment above SteveMc describes his interactions with her.
But since you think it’s important for us all to collect more data on Dr. Lonnie’s data availability practices, I think you should collect it. (No. I don’t think assigning the task to willis or SteveMc is appropriate– I think you should do it.
So, I suggest you write Dr. Lonnie and ask her for a significant amount of data that underlies a figure that appears in a paper she published. Presumably, you fall under the category of “all”. And you should be able to write something brief like
Dear Dr. X
I have been reading your interesting paper entitled ‘XYZ’ and I’m particularly interested in the data underlying figures (list) and tables (list). I would like access to all data required to create those figures and tables.
Thank you,
Roger
Or write whatever you think is suitable. Maybe you’ll have to go back and forth a few times to figure out what precisely she has.
If and when you get the data, post the data somewhere and let people inspect the data to see if you were given what you requested.
If you get a refusal, let us know and tell us the grounds for the refusal as soon as you get it. If you don’t get any data within a week, let us know and then report back at monthly intervals a few times.
If you get no answer, resend the email once a week until you get an answer.
Based on your comments, I think this experiment is necessary, and clearly you are the one who should do it.
And if she gets hit by a bus…?
Roger, you said “This time he attacks without even asking what this person’s data policy is or whether or not the data are available to all, albeit stored in an imperfect way.”
What you are missing is the fact that Steve is “attacking” her for not archiving the data. That’s the whole point of this post, and it seems to be true. He isn’t, as you say, attacking her for not *sharing* data.
Lucia is correct: if you want to make the point that sharing data manually is just as good as archiving, then you need to make that point with evidence rather than speculation. Otherwise, your point doesn’t speak to the focus of this post – non-archiving data – at all.
Roger,
I stand by what I wrote. If someone has a policy of giving data to anyone who says “give me the data” then its difficult to describe them as being “not a scientist” or not following the scientific method. It may be an imperfect method of data storage but is certainly not worthy of the type of ad hominem nonsense witnessed on this site. A well reasoned and well researched ad hominem attack can be very effective. This was a poor ad hominem post from Steve.
There two issues that you are confounding. One, upon which I agree with you, is that if a scientist freely makes their data available upon request, you cannot accuse them of failing an obligation of best scientific practice.
The second, which is Willis point, is that if, as condition of receiving public money, you are required to to archive the data in a publicly available fashion, then your personal office or den and availability on your own personal schedule will not satisfy the requirement, and wouldn’t have even in the “old days” that Willis mentions. It happens to be why we have institutions like the Library of Congress, the Smithsonian, and such, places where critical information is available. Again, this not, contrary to Willis, “unscientific.” It does fail to meet the obligation imposed by the grant. There is absolutely nothing “ad hominem” about identifying such a failure. It does not call the results of her work into question, simply her failure to meet her obligations to other taxpayers (US, Canadian, or whoever paid the tab for her – and her husband, Mr. Ellen).
What new readers to Climate Audit (Roger above perhaps) might not realise is that over the years Steve has become a very good judge of character. (Honestly, you should have seen him as a newbie way back in the begining trying to communicate with Dr. Mann et al)
Those of us who have read right from the start has seen this skill honed to a fine edge.
You will know them by their fruits, (or their lack there of). Certainly Steve does, moreso than most other people.
Steve can be surprisingly deferential, polite and kind to those who are polite and helpful, or even just polite. Personally, for whatever it might be worth, I trust his judgement.
Jeff
I’ve also been reading CA since it first came online and have a great deal of respect for Steve.
Regarding this case, he has had little to no dealing with this person so its difficult to attach any value to his character judgement. If anything, I’d speculate that his antipathy towards her is as much due to dealings with her husband than anything she may have done.
And the “Mrs Lonnie, Serial Non-Archiver” is more than just a title, it paints a picture that worth many thousands of words to those familiar with the style.
“If someone has a policy of giving data to anyone who says “give me the data” then its difficult to describe them as being “not a scientist” or not following the scientific method ….. This time he attacks without even asking what this person’s data policy is or whether or not the data are available to all, albeit stored in an imperfect way.”
Roger, 2.04 AM
If a researcher were to accept funding, from a body that imposes its own data policy as a condition of that funding, what relevance does the researcher’s own data policy have in that context ?
Cassio
Where does it specify in the grant conditions that a researcher who freely gives all the data to anyone would be violating the rules ?
Mooloo
“What is it about climate science that has people argue fiercely in public that black is white?”
I could ask what it is about this blog that requires the faithful to always agree with everything Steve writes.
Since you think I’m arguing black is white then the simplest way of resolving this is for you to quote all the relevant rules that you think this person may be breaking and then point out unambiguously how she is doing this.
**I could ask what it is about this blog that requires the faithful to always agree with everything Steve writes. **
Maybe it has something to do with the fact that his writings are well researched and usually on the money.
And nobody is required to agree with him. Your comments would have long been deleted at certain blogs.
Roger, you are an exemplary credit to the upside down world that climate science resides in! By my observation of what I have seen over the past 5 years, serial obfuscation, serial publication manipulation, serial non archiving, serial cherry picking, serial non use of professional statisticians, serial non public debating of the science, serial projecting of personal misdeeds, are all standard practices in “climate science”.
I think your being unfair , serial inability to meet the standards expected of an undergraduate should be added to the list , we don’t won’t to sell them short do we.
Thank you for the kind remarks but I’m not a climate scientist. Ironically, that’s the only insult you give which stings.
However, there are quite a few specific offences you’ve associated me with: “serial obfuscation, serial publication manipulation, serial non archiving, serial cherry picking, serial non use of professional statisticians, serial non public debating of the science, serial projecting of personal misdeeds, are all standard practices in “climate science”.
Would you care to back up any single one of these ad hominem attacks with respect to *anything* I’ve said. And please, don’t invent a position I don’t hold. Lots of people already have and its meaningless.
Its ironic (or maybe not) that a response to complaints about ad hominem attacks which haven’t been properly researched should lead to more ad hominem remarks.
Loaded question. In fact, the faithful are not required to always agree with everything Steve writes; sometimes they disagree right here in comments.
BTW: Have you requested data from Ellen yet? I think it’s important you do that so we can all learn how she responds to “anyone”. It’s especially important to your position –not so much to anyone else’s. That’s why you out to make the request.
Yes, I’ve been wondering about this, myself.
Let’s see … this post went up on Canada Day (July 1) which is when Roger embarked on this particular “campaign” with the first of many comments in which he seemed far more intent on proving (by dint of sheer repetition) that Steve was wrong to comment on the Thompsons’ notable failures (compounded by the failure of the Journals) to comply with requirements that their data be archived.
I’m left with the impression that in Roger’s view, there is no difference between a “personal request” for data (with which the author may or may not choose to comply) and making it publicly available via archiving in compliance with the policy – albeit evidently not the enforced practice – of the journal(s).
And it’s now the Glorious Fourth (Happy Independence Day to our southern neighbours … OK, OK, spellchecker, in deference to the occasion, “neighbors”), and Roger doesn’t appear to have done anything at all towards supporting his “argument” – except to keep repeating it, and cast unwarranted aspersions on those whom he cannot seem to browbeat into agreeing with him!
A lot of sound and fury from Roger, signifying, well … very little, I suppose!
hro001
I find it interesting that I’m credited with “sound and fury”. I suggest you objectively look over the posts to me and from me to see who gives the “sound and fury”. Can you perhaps provide some quotes from to illustrate the “sound and fury” tag ?
Its also interesting how you misrepresent my position not least since I was very clear that I see little difference between a researcher freely giving his/her data to anyone who wants it (not selectively giving it, this I’ve never stated although I’m sure it is comfortable for some to think it is my position) and having it in a formal archive. Re-read any of the repetitions you complained about if you’re unsure. Better still, provide a quote to show you’re right – this pesky demand for evidence again. Those who are the recipients of ad hominem attacks (me in this case) rather like to see the evidence of our crimes.
The position on data-sharing that I outlined above is in fact that held by Dr Leif Svalgaard for a recent publication. I’m sure most of you are aware of his studies. His work has led many to question CAGW assertions. He is feted, with some justification, in sceptic circles. You can find in this link to a WUWT article his refusal to archive all of his data but instead give it out on request:
http://wattsupwiththat.com/2010/05/29/leif-svalgaard-on-the-experience-of-peer-review/ .
How many of you will take the opportunity to vehemently slur him as not being a “real scientist” ? Or will you soften your stance because you’re sympathetic to him ? Perhaps you’ll find excuses and invoke get-outs in policies etc.
For those who are tempted I suggest you avoid dismissing this as trolling. Its not. Its at the heart of much of what I’ve been arguing. Ad hominem attacks are a nasty business and best to be avoided unless necessary. Join in when you’re really sure you know what you’re doing and you’ve persuaded yourself with all the evidence that its right to do so and that you’re not allowing prejudice to cloud your judgement.
snip
Steve -no need to divert into a discussion of the Svalgaard situation. I’m not familiar with the case. I would encourage him to comply with the standards advocated here.
Snip if you want Steve but I think the Svalgaard case is important. A central theme of this post is an ad hominem attack on someone. As I’ve mentioned a few times there is nothing wrong with this as long as the necessary research is done justifying the attack. I think it wasn’t in this case.
A lot of people piled in and made a lot of ad hominem remarks about Mrs Lonnie. Even if she would give away her data freely this was derided as being wholly unacceptable. In this situation she wouldn’t have been a “real scientist”.
Are these attacks rationally motivated on the basis of scientific practice or because she’s part of the “enemy” ? This is a relevant point. In one instance its mobbing, in another its fine. Maybe its a bit of both. Setting up an equivalent example of someone in “the good guy’s” camp following the policy I posited for Mrs Lonnie (and providing solid evidence that this is the case) is a good way to illustrate how a discussion and attitudes are motivated. Ultimately, the aim must always be to abandon prejudice and follow fundamental principles and logic. Dealing with a situation in which someone of whom one approves has done something you would condemn in others is a good way to find out how much we are guided by our prejudices.
As I mentioned, snip away all you want Steve, but I think the above points are valid ones, especially so when you write your “name and shame” posts.
Steve: I snip and delete a lot of piling-on comments. I try to be consistent in standards. I’ve spoken out against Cuccinnelli and the Heartland billboard. I have limited time and energy and haven’t read Svagaard’s work nor followed the dispute. My issues with Thompson have been going on for nearly nine years though.
Roger, you have stated:
“The difference between making data freely available when someone requests it (and please disprove this before ranting**) and putting it in a public archive is miniscule. To then declare that this means she’s not a scientist is ridiculous.”
What, in your opinion, are the miniscule differences?
What are the scientific advantages of data-on-request over a public archive?
Roger
No it’s not. You may need to learn what “ad hominem” actually means.
Providing data privately is not the same as putting it in a public archive. You can call this observation “derision”, but it is simply a fact.
You are the one who seems to want to wave away the accurate observation that she has not archived publicly. You are the one who wants to characterize criticism of her non-archiving as “at attach”.
She hasn’t archived. This is a simple fact. It can be observed even if someone somewhere might consider her “the enemy”.
First: the link you provided is to the top of an article– and frankly, though I read the article and searhed on “Leif Salvagard says” I found nothing at that link to provide details of whatever it is you might claim he did or did not do.
If Salvgard hasn’t archived materials supporting his publication and those materials are unique to him, then people can observe that and criticize him and should. If his work is funded by NOAA, then by not providing underlying data, he too might be in violation of their policy. Whether or not he is might depend on whether he is the originator of the data and whether the data has been archived by the originators. Is his work published by NOAA? Does he operate a telescope or a satellite? Is he the originator of the data?
Those are rhetorical by the way: Because no matter what the answer to those question it would not change any of the following:
1) Steve’s post is not an ad hominem attack. It is a discussion of behavior and criticims of behavior.
2) The post is about whether or not Dr. Ellen archived.
3) It appears she did not.
4) People who think that’s bad are criticizing her for her behavior.
Your persistent attempts to defend her by changing to subject to something else– and then claiming your subject changes have something to do with the post– are ridiculous!
Re: lucia (Jul 5 19:24),
So I went and read the PDF document at WUWT and come to the conclusion that Dr. Svalgaard’s position is difficult to disagree with. Here is the relevant text, first the comment from the reviewer and then the reply. I have bolded the parts that I find compelling, YMMV.
Further down in the document:
Finally one more exchange witht the reviewer:
I would like to repeat Lucia’s observation above, however:
I apologize for the length of this post but I felt the actual text was a compelling refutation of the assertion made by Roger earlier in the thread.
Thanks Bob, that was a very interesting post. I wonder who the reviewer was with the strange double standards.
Perhaps I may make a remark. This whole issue is moot [and a bit silly]. The data is in the public domain [the World Data Centers] which is where we got it from in the first place. As with any archive of enough data there are several issues with faulty data [e.g. from November 31st]. We outline some of those issues in our paper http://www.leif.org/research/2007JA012437.pdf section A4.2 and the actions we have taken to correct those the best we could. Unfortunately there exists no mechanism to re-inject corrected data into the World Data Centers, so other researchers cannot readily take advantage of [or reproduce our findings exactly] our corrected data. The data is of modest [but still unwieldy] size (3 Gigabytes) and held in a standard format that can be read by all without special software. We have put the data in NASA’s Virtual Magnetospheric Observatory http://science.nasa.gov/heliophysics/heliophysics-data-centers/virtual-magnetospheric-observatory-vmo/ which unfortunately seems to be defunct. Having foreseen such development the data is [and has been] also available on my website [open for all – no request needed, although I will also honor specific requests for subsets of the data] http://www.leif.org/research/data
Regarding the ad hominem attack. This post picks out one researcher among many and attacks in a very snarky tone.
However, Mrs Lonnie’s approach is consistent with practice in the field. Why must she be singled out to be “named and shamed” ? Why personalise what is a field-wide issue ?
Furthermore, the implication of the criticism is scientific misconduct (“not a real scientist” was one of the criticisms made here). This is a very serious charge. Reputation assailed, damage done. Inevitable question mark over her science. However, when it comes to the key question of availability of the data in question no attempt was made to find out if she would freely give it away. Given previous posts I doubt Steve would have written the post the same way had she given away her data. Reputation not assailed, damage not done (perhaps). I’ve no idea what she would have done. However, this is irrelevant to the central point that one collects as much evidence as possible before making an attack since reputations are at stake. This good practice wasn’t followed this time.
Steve: I did not say that Ellen’s non-archiving amounted to ‘scientific misconduct”, which as defined in most academic codes of conduct would not include non-archiving. Indeed, as I’ve observed, non-archiving is almost a badge of honor in the discipline. Complaints to journals about the Thompsons have accomplished little in the past. As to your allegation that I did insufficient research into the Thompson’s archiving prior to making the post, I deny that totally. Neither you nor anyone else have pointed to a single factual error in the post. I carefully examined all archives before making the statements in the post and, to the best of my knowledge, the statements are correct. As to your allegation that I’ve had insufficient correspondence with journals about the Thompsons, I have pages and pages of correspondence, some of which I’ve published at CA but much of which I haven’t.
Roger, you have very transparently created a straw woman for the purposes of avoiding the discussion of the real issues that develop out of non-archiving. The Thompson’s are merely excellent examples of those who have either not archived data or had to be prodded to do it. Do us all a favor and either move on or discuss the issues at hand.
Roger
Criticizing an individual researcher is not an “ad hominem attack”. Adopting a “snarky tone” is also not an “ad hominem attack”. An ad hominem argument is “an attempt to negate the truth of a claim by pointing out a negative characteristic or belief of the person supporting it”. No one is trying to suggest that the reason we should believe failure to archive is bad because the Thompson’s are bad people. Rather: 1) failure to archive is bad because failure to archive is bad and 2) These specific people are failing to archive in various and sundry ways.
Steve didn’t say failure to archive makes the Thompsons “not real scientists”.
But it turns out that some people think it — evidently willis may. While you may disagree with those people (as I do) those people are permitted to express their opinion. Moreover, expressing the opinion that failing to archive when it is a formal requirement under the grant makes the person “not a real scientists” is not an “adhomimen attack”. And get a grip: Willis expressing that opinion is not tantamount to accusing anyone of scientific misconduct.
Geehh.. it turns out some people think poorly of scientists who do not archive. They say so. But you know what? It’s the scientist’s failure to archive that is damaging the scientists reputation.
First: You can keep calling this a “key” question all you like. But the fact is: it is not a “key” question. It is utterly irrelevant to the finding that she does not deposit material in public archives.
But this answer does seem to be important to your position about judging her. But you won’t make any attempt has been made to get the answer to this question which only you think is a “key” one.
If you truly think it matters whether Thompson supplies data privately, you should stop acting like a knuckle head and go ask her to supply you data privately. That answer will be irrelevant to SteveMc’s post and most other people’s points. Despite that, many might be interested to learn whether in addition to not archiving data she also refuses to hand out data privately. Let us know what you find.
Roger –
Could you please go look up the definition of ad hominem? Then read it, and post a link to the source you consulted. You’ve been told many times that you’re using it wrong (you are) but continue to steadfastly ignore it.
If we can’t get past this very, very basic definition, I have to think that, in addition to repeatedly trying to change the subject, you’re being willfully obtuse and not discussing/arguing in good faith. Thanks.
> One can see too the public appeal of any ad hominem attack. Personal criticism fuels much of our public debate – certainly it is diet on which the mainstream media has grown fat. It holds within it the possibility of a hostile exchange – if not, then of provoking one – and so inevitably they are given much prominence and presented with much fanfare and drama; all of which is to the detriment of real and meaningful discussion.
To repeat the point made above (twice, where the moderator kept the one wrong placed — let’s not wonder why), an ad hominem attack should not be conflated with an ad hominem argument.
w
I will also note that not **one** of Roger’s comment got through the RSS feed.
Again, let’s not wonder why.
check your feed. I get them just fine.
Oh, “let’s not wonder why” says the clone
In a rather transparent-snark tone
With his false accusations
(I get Roger-rations)
Perhaps you’re not getting your own.
Since the comments go out on the spot
And don’t wait for “approval” or “not”
And since we get the feed
(Yes, including your screed)
Your complaint, like your argument:
Shot.
===|==============/ Keith DeHavelle
You need to remind roger that
1. there is a difference between an ad hom argument and an ad hom attack
2. describing someones behavior is not an attack on their person.
3. calling a murderer a murderer is not a slur. It’s a fact.
So then why did you post a link to the definition of an ad hominem argument?
They may be using the phrase “ad hominem attack”, but there is nothing on that page which distinguishes that from the standard definition of an “ad hominem argument”.
Roger> “Regarding the ad hominem attack. This post picks out one researcher among many and attacks in a very snarky tone.
However, Mrs Lonnie’s approach is consistent with practice in the field. Why must she be singled out to be “named and shamed” ? Why personalise what is a field-wide issue ?”
Dear lord, how monumentally ignorant are you? WHEN did you start reading Climate Audit?
Steve McIntyre for YEARS has been “naming and shaming” those who refuse to archive their data. Are you ignorant of the Yamal posts? Of the CRU posts? Of the Hockey Stick? Of all the other postings over the years where the host of this blog has documented how he politely asked for data and been rudely and nastily refused?
The “consistent with practice” you seem to think is no big deal is in fact a VERY big deal. It is the documenting of deliberate violations of the principles of the Scientific Method that has made Steve McIntyre such a powerful voice against the CAGW band wagon. The simple fact is that those who refuse to archive their data and methods, who refuse to allow independent verification of their “studies”, are NOT scientists. And to you that is “personalizing” an “issue”? Disgusting.
Are you ignorant of the simplest facts of your “consistent with practice” hidden data policies? Where when they are discovered (like Mann’s “CENSORED” ftp directory) or forced out (like Briffa’s Yamal data by a Royal Society publication) – that the raw data is cherry picked, then massaged with phony-baloney statistical methods, or just literally turned upside down. Phrases like ‘short-centered PCA’, ‘Yamal’, and ‘Upside Down Tijlander’ are infamous among those who have been given an honest look behind the “climate science” curtain by Steve McIntyre.
So let me repeat my first question? How monumentally ignorant are you?
Steve – I never wrote that you thought Ellen’s non-archiving amounts to “scientific misconduct”. I said that that there was an implication from the post. This is the case regardless of any intention. Here, I don’t imply a deliberate implication left hanging in the air. The response of the posters (“not a real scientist” etc.) suggests I’m right in my assessment of the implication. Regarding the climate scientists wearing a refusal to share as a badge of honour, the scientific field is broader than climate science and on many occasions in her professional career she will have to deal with non-climate scientists. Your posts do matter even if it may be convenient for some to dismiss them.
I stand by the accusation of you not having done enough research although here my wording was admittedly poor. This CA post can be read as an attack on a scientist (look at the comments). From previous posts on CA I think your position and attitude to her would have been softened had she given all the data in question freely to you (or anyone) upon sending a mail. I think you should have excluded this hypothesis before writing your post not least since it would have been a trivial exercise.
Ultimately, regardless of intention, this post has called into question the scientific output of a scientist because of her apparent failings. Just how serious these failings are wasn’t fully explored.
I can imagine a conversation between Ellen and her faculty Dean. The Dean comments she has been personally attacked in the internet on a noted sceptic blog – which by now is begrudgingly known for often hitting the target – an unethical attitude to data sharing. Ellen’s response -“he didn’t even ask me for my data” is a pretty good one.
Roger It still remains in your power to make such a request to prove your point , when will you do this ?
I write again – hopefully this post will end up where it should end up.
KnR
My point has been consistently misrepresented here. With reference to what I’ve written (not what you think I’ve written) what do you think my point is and how would it be proven/disproven should I contact Mrs Lonnie ?
It might
I’ve answered your post. Please respond to mine.
You claimed she would provide the data if asked and therefore there is no need to correctly archive it .
One, you failed to support this claim in any way but with ‘becasue I said so ‘ your can prove it by doing the asking are you going to ?
Two, the reasons why there need to correctly archive data have been explained to a number of times , you just ignored them . Unfortunately hand-waving does not make these problems go away , if she losses them their gone , if she dies there gone , if your computer gets corrupted there gone etc
Once again given that great claims are being made and great changes are demeaned its only fair that a fairly low level of data management , so low a students would be reacquired to do it , should be expected as a minim.
Doing good science involves far more than the exciting bits of filed work and press releases , it requires all the boring paper work and data management work too.
Roger said:
“I can imagine a conversation between Ellen and her faculty Dean. The Dean comments she has been personally attacked in the internet on a noted sceptic blog – which by now is begrudgingly known for often hitting the target – an unethical attitude to data sharing. Ellen’s response -”he didn’t even ask me for my data” is a pretty good one.”
You’re not going to get away with this line, which you keep repeating, by trying to wear us all out. Thanks to global coverage, while some are sleeping, others are reading. 🙂
For the umpteenth time, it’s not about data sharing, it’s about data archiving, and they are not the same thing at all, for reasons that patient posters above have spelled out in language that even a physicist should be able to understand. And, Steve has not accused Ellen Mosely-Thompson of “an unethical attitude to data sharing” anywhere that I can detect. He has, correctly, accused her of failing to archive her data, which nobody has been able to demonstrate to be incorrect.
Johanna
Apologies if you receive two replies to this. My browser hung when I submitted an earlier reply.
I don’t quite know where to begin. First, you misrepresent my position. Read carefully my response to Steve and *don’t* cherry pick quotes out of an appropriate context.
Second, don’t take the comments section of this site too seriously. Regarding me writing when (some) others sleep, you post reads like “how dare Roger, in a different time zone, write something to which the rest of us object when we’re asleep. He should do it when we’re logged on so we can tell him he’s stupid?”. Daft.
Finally, in spite of the name calling I’ve received (eg troll) all I’ve done is put forward my objection to this post. Outside of the CA poster world I doubt it would be remotely controversial.
I should add (one more time..) that my motivation for commenting on this thread is that I don’t like the idea of attacking people too much. Attacks too often diminish the attacker rather than the recipient for many reasons, not least that one questions the motives of the attacker. Furthermore, regardless of the careful nature of any given attack, the message is invariably distorted as it winds it way in the wider world. In other words, any attacker ought to be ultra-careful and examine all facts prior to attacking.
This has been my position from the start – I’m amazed that this remotely controversial. That said, I shouldn’t really be surprised since my own stance has been misrepresented here (“Roger means this” – no I don’t! “Roger thinks that”, nonsense!). It all pretty much illustrates why I’m right to be concerned about this CA-post. People read what they want to read. In many cases even a hint of criticism is enough for them to assume the worst.
Having read through the comments I’m not really any the wiser what your point really is. You’ve repeatedly described SMc’s piece as an ad-hom, when (as several others have pointed out) it isn’t. You’ve repeatedly equated giving data to anyone who asks for it with archiving, when (again, as others have pointed out) it isn’t; you have agreed, however, that she should have archived her data. As far as I can tell, there are no factual inaccuracies in the piece. You complain of being misrepresented, yet, in addition to your repeated ‘ad-hom’ claims, you were the one who said the piece implies scientific misconduct. I don’t actually see anything controversial in your views on the potential dangers in ‘attacking’ people, but am perplexed by your reaction to this piece.
DaveS
You’ve repeatedly described SMc’s piece as an ad-hom, when (as several others have pointed out) it isn’t.
As I’ve explained I think it can be legitimately described as adhominem. This post picks out one person for engaging a common practice within a field and “names and shames” in a very snarky way. Others have disagree this is adhominem – that’s not the same thing as “pointing out its wrong”, its called disagreeing. My position is that neutral observer would be hard pressed to not see something adhominem in this (by the common definition of adhominem as a personal attack, link was provided). You and others disagree. So what.
“You’ve repeatedly equated giving data to anyone who asks for it with archiving, when (again, as others have pointed out) it isn’t.”
No I haven’t. Where have I equated the two approaches ? Read my posts here and quote examples. But please *don’t* cherry pick. I’ve written quite a bit on this.
“As far as I can tell, there are no factual inaccuracies in the piece.”
Where did I ever write there were factual inaccuracies ?
“You were the one who said the piece implies scientific misconduct.” Yes, it does have an implication of scientific misconduct. This is my worry. See, for example, the “not a real scientist” slurs against her in the comments (not from me btw). Clearly I was right, others are reading it that way.
I’ll rather characterize your reference to my work as ad-hom or at the very least gross misrepresentation. E.g. my “refusal to archive data”. Upthread I have explained what the situation is. Study that carefully and apologize.
Meant for ‘Roger’
Leif,
thanks for the thorough reply. Roger’s attempt to liken your situation to Ellen or Lonnie Thompson seems singularly misplaced. It’s also interesting that your referees took notice of your data situation, whereas referees of the Thompsons have obviously paid no attention whatever to archiving.
Regards, Steve
An important point: in many fields there are no ‘established’ public, formal archives. In my case, the World Data Centers only accept data from original data providers [not from me who just corrected obvious errors]. The NASA VMO repository has gone away, so what is one to do? At least back in 2007, JGR charged money for ‘eternal’ storage. My current solution [free access to all on my website] seems to me to be reasonable. I wish that everybody would just do similarly, to the extent it is practicable [with petabytes that may be hard]. ‘Condemning’ me for my ‘attitude’ without doing their homework reflects badly on the people who do this.
“‘Condemning’ me for my ‘attitude’ without doing their homework reflects badly on the people who do this.”
Amazing that Roger comes here shouting about not doing your homework and then provides a real, and not imagined, example of that. Once again I very much doubt Roger’s sincerity in these matters. He started with an improbable excuse for Ellen Thompson that she could willingly and freely provide all her data to anyone requesting it, but would not put that data at a site that was specifically set up to do that, namely the NOAA repository. That proposition is insulting to those of us who are aware of the opportunities to archive data in her case and actually is insulting to Ellen Thompson for making it appear she is blithely unaware archiving opportunities and would for some unfathomable reason want to dole it out piecemeal. Actually she has archived as lead author at NOAA one ice core data set, Greenland Camp Century particle data, and has 8 other ice core data sets archived at NOAA that she coauthored with L. Thompson as the lead author.
The data in question in this thread is that from the numerous expeditions of E. Thompson to Greenland and Antarctica which is different than that data noted above. In my email to E. Thompson I asked if there were any extenuating circumstances that might prevent her from archiving these data. I also asked if she would make that data available to me or other individuals. I have not received a reply from her. In numerous email exchanges with climate scientists in the past, I have had all of them reply to me. In one case the reply noted that a team was going to reply to me, and when they did not, a second request to the team leader produced a reply. The only non reply from a climate science related query was to the IPCC concerning making public the documentation of how the participants determined the probability that their group placed on evidence noted in the IPCC publication.
This is posted as a “reply” to Steve’s post of July 7, 2012 at 4:18pm to see if the “reply” function is working now. Unravelling who said what to whom in this thread has often been quite difficult. I was initially misled into wrongly attributing a quote several times.
Doctor Svalgaard:
Having read virtually all of the above thread, I suspect that Roger is the apologizing type…
we shall see. somehow I have the feeling he/she is not
Four years ago I thought CAGW was real enough, as I believed that climate scientists were honestly studying this issue. Sadly, I came to realize that this was not the case. Honest science requires openess, eg, data archiving. Time after time the failure to release data was a cover up for cherry picking.(I think this is Rogers real goal right now) I am now sickened by the field of “climate science”. I don’t think it has any useful purpose other than to perpetuate bad science. IMO, all funding should be stopped.
No Ed-B just the funding for non archived studies. That is all any reasonable skeptic would look for.
Roger,
Did you do your homework?
Did you see that Steve had already requested data from the Thompsens?
Did you see what all the details were regarding Leif?
Looks like no to both.
Beware of anonymous people throwing darts. Those of us who are publically known
operate by different standards than folks like you.
Re: Roger (Jul 7 08:38),
Roger, you are factually incorrect in claiming that this post is Ad Hom. Your interpretation of “personal attack” is incorrect, as defined on the very site you referenced.
I am not disagreeing with you. I am pointing out that the sources you used disagree with you.
You claim this post was a “personal attack.” According to the site you reference, that would be
Please show where Steve failed to provide evidence, and instead made an abusive remark.
In fact, please show where Steve made an abusive remark even with evidence.
It’s not there.
The post was not a personal attack.
The post was not an ad hominem attack.
In fact, by your definition, YOU are the one making ad hominem attacks according to the “personal attack” definition:
– You have provided no factual evidence for your claims regarding Steve’s post.
– Instead, you have substituted an ongoing series of non-factual posts that amount to abuse… pounding away until you think you’ll make a mark.
Time to apologize, time to reflect on your own actions here, time to learn how to do what Steve does so well:
– Investigate carefully
– Learn the facts
– Post carefully yet boldly, pointing out with evidence where “they” got it wrong
– Gracefully point out when “they” make corrections and apologies
– Apologize and correct your own work when your errors are noted.
Looking forward to a change of attitude!
“I should add (one more time..) that my motivation for commenting on this thread is that I don’t like the idea of attacking people too much.”
And yet you come here and attack people because they make valid criticisms of non-archiving scientists. Steve has used two of the worst offenders to illustrate his concern. That’s how arguments are made; you illustrate the problem.
You made your point days ago. Your snarky attacks on Steve have been noted. I’m interested to see if, when Steve publishes his next audit, you have any think of scientificvalue to contribute.
My guess is you won’t.
Roger
‘I should add (one more time..) that my motivation for commenting on this thread is that I don’t like the idea of attacking people too much. Attacks too often diminish the attacker rather than the recipient for many reasons, not least that one questions the motives of the attacker. Furthermore, regardless of the careful nature of any given attack, the message is invariably distorted as it winds it way in the wider world. In other words, any attacker ought to be ultra-careful and examine all facts prior to attacking.”
but you engaged exactly in that kindof behavior from the cozy protection of anonymity
Apologies – I misread your quote about others reading.
However, in true climate science fashion, the details of the argument didn’t anyway matter in this case since the conclusion, “daft”, still very much holds. As does the statement “don’t take the comments section of this site too seriously”.
Roger, you make this accusation without having accurate knowledge of what I’ve done, not all of which is covered in CA posts.
Lucia – yours is a bizarre question. Where I come from its the responsibility of those making public slurs to make sure they have the full facts before attacking.
Roger–
Who has made a public slur? SteveMc discussed the tendency of these people to not archive publicly, and what SteveMc says is true.
You then seem to wish to change the topic of SteveMc’s post to whether or not they would be willing to hand out data privately, and seem to want us to their willingness to hand out data privately makes some difference. I say this is a subject change because whether these people hand out data privately is utterly irrelevant to anything SteveMc has accused them off. So he would hardly be expected to check this fact
This bit of information isn’t relevant to some other persons argument– it is only relevant to your argument. And you seem to be slurring and attacking SteveMc for not getting this information.
And where I come from you should be providing the data to support your argument. And you should not be slurring or attaching SteveMc for not obtaining data that is only relevant to a point you want to make.
I think it is totally “not bizarre” of me to suggest that you request data to discover whether these people are willing to hand out data privately.
Are you going to try to get this data? Or are you going to find every excuse to continue to slur and attack others for not getting information that is irrelevant to their arguments but required to support yours? (I suspect the answer is you are not going to try to get this data.)
Roger Lucia’s question is not bizarre.
You’ve made a claim about Ellen that she would share the data with anyone who asks.
We know this to be false. At least I know it to be false as do a few others.
You can prove the falsity of your own claim to yourself ( do an experiment)
by asking for the data yourself.
If you get it merely post it as proof. If you dont get it, then you can join the rest
of us who know that she neither submits data to archives nor supplies it to anyone who asks.
It’s not a slur to note that somebody is a lousy scientist. Most people can only dream of being a lousy scientist.
“Cassio
Where does it specify in the grant conditions that a researcher who freely gives all the data to anyone would be violating the rules ?”
Roger 3.37 AM
Perhaps you would care to answer my question before posing one of your own.
Oh, why bother ! I apologise to Steve, and to all readers, for having taken the space to put a question to Roger, who is clearly not interested in answering other peoples’ questions, but only those that – much more conveniently – he asks himself.
Casio
An answer to your question is that the researcher should follow the policies. End of story. I’d already made clear what I thought this person should have done. I’m surprised you think that I’ve been in anyway evasive.
Now, please answer my question.
Perhaps Roger would be helped by a Josh cartoon of Ms Mosley-Thompson with the caption “ceci n’est pas une archive”.
Roger would prefer to be helped by a rational evidence-based counter argument.
Dictionary definition of archives – “a place where public records or other historical documents are kept.”
Note that word “public”. Is that evidence based enough for you?
So if she supplies all data in her archive to everyone who asks then how is that then not making it publicly available ?
You are just trolling. Archiving means archiving. Period.
Trolling, no. I’ve never trolled in my internet life although I understand its convenient for others to think so. Even if I would suddenly troll it would be far more ethical than attacking somebody I didn’t know and not making the slightest effort (trivial – all a mudslinger need do is send a mail) to find out how justified my attack was.
Roger–
First, there is a difference. Archiving means it will be available if she gets hit by a car and dies. Archiving means that people don’t have to hunt down her email address. Archiving means people can be reasonable sure it is static. They can reference the archive and so on.
But anyway: As I said before, if you think it’s important for us to know if she will had out data you should contact her, get the data and prove to the world that she shares willingly with anyone who asked. I suggest you ask for each and every one of her papers.
On October 19, 2003, I wrote to her, cc Lonnie Thompson:
She responded:
I replied:
She replied on Oct 20:
This inquiry to Mosley-Thompson was a result of a comment that I had posted on a Yahoo blog about non-archiving of data used in Mann and Jones 2003:
Timo Hameranta circulated this comment to a lengthy email list. I responded to him that I would ask the authors criticized to archive their data and did so, sending the above email to Mosley-Thompson among others. This got notice in the Climategate emails and is the second time that I’m mentioned. ( CG2 1566. 2003-10-20). Mann wrote Jones and Briffa, cc Bradley, Hughes, Diaz, Rutherford (this is 2 weeks before publication of MM2003):
At 11:14 19/10/2003 -0400, Michael E. Mann wrote:
Jones replied:
ANother point in this respect is that authors have complained about being “pestered” for data. Or that one request simply leads to another. The climate community has by and large closed ranks in solidarity with the idea that requests for data were “distracting” scientists from their “real” work. I asked authors to archive their data as a solution to this supposed imposition,
It is not “publicly” available because 1) you have to address her privately to acquire a copy, and 2) suppose she falls down a crevasse, has a car accident, or develops amnesia? Who will know what data you are requesting? If you think such examples are absurd you really need to study the history of science. That availability without having to go through any specific individual is the difference between privately available and publicly available.
CG2 1566 is here. After that mentioning McIntyre became a bit of a habit.
1. Steve has already asked for the data.
2. handing data out to whoever asks for it runs a huge risk unless you have a
method for insuring that you always hand out the same data. One reason
why we put things in archives is to insure that all parties get the same data.
If you rely on a scientist to hand out a copy multiple times, then the
scientists ability as a good data custodian becomes the issue, rather than the science.
I cannot count
the number of times people have sent me different versions of the “same” data.
They were good scientists but utterly untrained data custodians and document
control specialists. Scientists should not control data or manage it unless they
are trained in the discipline.
Long story on Lonnie in the NYT today: http://www.nytimes.com/2012/07/03/science/earth/lonnie-thompson-climate-scientist-battles-time.html
Excerpt …
To some climate scientists, the Thompson ice core record became the most convincing piece of evidence that the rapid planetary warming now going on was a result of a rise in greenhouse gases caused by human activity.
“The reason Lonnie’s stuff is so powerful is that it’s so simple,” said Daniel P. Schrag, a geochemist at Harvard and director of its Center for the Environment.
“His evidence dismisses the idea that this is some sort of 300-year or 500-year cycle, which is what the skeptics and the deniers want to say. You say: ‘No, because Lonnie’s ice didn’t melt then. It’s melting now, but it didn’t melt then.’ ”
Re: Eddy (Jul 3 05:10),
Yes, but
And on the topic of this post:
Accepting the premise that he is nearing the end of his life (that is the main point of the article) one can wonder if he will ever publish the data he has collected.
bob
Accepting the premise that he is nearing the end of his life (that is the main point of the article) one can wonder if he will ever publish the data he has collected.
Yup. You’d think that he might want to make sure everything is published, to preserve his legacy. Or if he doesn’t care enough, that his associates might.
“Lonnie’s ice” – apparently he owns the ice too. not just the data.
Just to be clear, I don’t care less about this lady. She may well be a terrible scientist or another Einstein.
Its also clear that the majority of posters know little of her. And yet they are still willing to condemn her on the basis on one person’s leading post.
Nobody has bothered to clarify if she is (a) a dedicated researcher with an imperfect but open-to-all approach to data sharing or (b) if she has adopted an unacceptable proprietary approach to her data. Some have tried to pretend that it doesn’t matter. Sorry, it does, and you know it.
Similarly, even though I’ve asked. nobody has proven that she has unambiguously violated the rules. There seems to be an assumption that this is the case. Well, when it comes to ad hominem attacks on people I don’t know I prefer the assumption to be the other way round. Mud sticks. Its fine to throw it but only when you’ve made an effort to get hold of the facts. In this case, it would have taken a couple of short mails to ascertain the facts and find out if scenario (a) or (b) is correct. And yet, bizarrely enough, I’ve been told to send the mails, not the mud-slingers.
Roger, I’m not sure I’m following your arguments. If Mrs Thompson has written a paper containing any of the data and had it published in a scientific journal, then that raw data, along with the methodology for choice of data, any manipulations of data and the software used to find the signal from the noise in the data should be a matter of public record. Are you saying it is and the posters on this thread haven’t checked?
It may be, I haven’t checked, but if someone will give me the email address of Mrs. Thompson I’ll ask her for her data and check out her response. Come to think of it you could do that yourself and then we’d all know for certain.
If she is bothered about having her science checked then we are entitled to believe she’s done something to be ashamed of until the opposite is proved.
This is where we fundamentally differ. I think you are certainly entitled to believe she has done something to be ashamed. However, prior to slurring in public one has a responsibility to make some minimum checks or at least ascertain if those checks have been made.
By all means send off a mail (I’m sure if you google her name you’ll get it quickly). She could tell you to get lost but it wouldn’t logically affect my argument.
Seems simple to me. All government granting agencies, including the NSF, should cease giving grants to all “scientists” until prior works funded by the public are demonstrably 100% archived.
Problem solved!
“Similarly, even though I’ve asked. nobody has proven that she has unambiguously violated the rules. There seems to be an assumption that this is the case. Well, when it comes to ad hominem attacks on people I don’t know I prefer the assumption to be the other way round. Mud sticks. Its fine to throw it but only when you’ve made an effort to get hold of the facts. In this case, it would have taken a couple of short mails to ascertain the facts and find out if scenario (a) or (b) is correct. And yet, bizarrely enough, I’ve been told to send the mails, not the mud-slingers.”
In your seeming haste to defend, Roger, you may not be aware of this fact, but NOAA has a site that contains a large number of publicly available temperature proxies, including ice cores. The point is not whether we can provide wiggle room for those that do not choose to make their data available to the public, and data that was collected on the public’s nickel, but rather pointing out to the public that some scientists have not made their data public – and data that was collected many years before present time.
The fact that you attempt to sidetrack the discussion by references to ad hominem attacks and how Thompson might not make her data public and still be compliant with someone’s interpretation of the rules, only makes me judge your response as not taking seriously the need to make public these data and especially those publicly funded.
You have not really bothered with the issue of making all the data public to guard against requestors using only parts of the data and not reporting the entire set. There is also the issue of the perception of a good old boy network in climate science that shares data within the network and jealously guards against outside analyses. That perception suffers greatly from this withholding of data from the public- for whatever reasons.
Kenneth
I’m not defending anyone although I see that its convenient for folk to believe that. I’m asking you all to prove that she is a “proprietary scientist”. The proof I’m looking for is simple to get but nobody seems to think it worth their while to get it. Reminds me of climate science: “we’ve got enough proxies for our conclusions, don’t need to get more thanks, they might anyway complicate the tidy picture”.
To publicly slur someone requires evidence and not supposition. You all write with such certainty. I’m a scientist myself and find there is little certainty without prior investigation.
As for sidetracking the discussion, this is just a blatantly ridiculous line. This post is fundamentally an attack on someone. The faithful have all joined in. To request that one of you actually bothered to seek out some facts about the person you’re attacking in order to justify it is not sidetracking anything. Its at the heart of it.
Roger, if would come off your collective description and view of the people posting here you might be able to engage in an intelligent discussion. My point and I think the point of many posting here is that is that the Thompsons amongst other climate scientists have been delinquent in their efforts to archive. Issues addressed and points made are that (1) agencies are also to blame for not enforcing the rules, (2) not making all the data public but doling it out piecemeal can encourage cherry picking the data with the public not the wiser, (3) the public has a right to data that was collected with their funds and (4) organizations like the IPCC could do much more in considering only data that is publically available.
To continue harping on the lack of proof of the Thompson dole (which is not even the point – it is archiving) and attempting to classify the posters here as part of an amen choir is a very silly and transparent way of avoiding an intelligent discussion of the issues at hand.
**Similarly, even though I’ve asked. nobody has proven that she has unambiguously violated the rules. **
a) It seem quite clear from the postings at this site that archiving is required and was not done. So what proof do you require?
b) Where in the publications requirements do your questions a) and b) appear? Seem like you just made them up to suit your own interests. And I do not buy your statement that you do not care about this lady.
If its all clear then please point out exactly what the requirements are and how she has violated them.
See Willis below, But I will bet that you will ignore him.
As I noted in another comment, NSF officers have not, as a matter of course, implemented overarching data archiving policies. Regulators are too often cheerleaders for the climate industry and, in the process, appear to have neglected their regulatory obligations, both in contract writing and enforcement.
The science is settled. Why is there even an argument about archived vs hidden data.
ROger Posted Jul 3, 2012 at 10:11 AM
Roger, it’s impossible to prove a negative, that she has not archived her data. So what you are asking for is impossible, and if you don’t know that, you should.
However, neither you nor anyone else has located any such archive, despite searching for it.
Given that her data is not archived, here are the rules she is breaking:
Note that they say nothing about your bogus claim that simply holding onto the data privately is OK. They speak specifically and repeatedly about DATA ARCHIVES. Ms. Mosley-Thompson keeping the data on her PC, no matter how willing she might be to share it, is not an archive, Roger. It does not fit the criteria given in paragraphs 2 or 3. Her privately held data storage doesn’t have “procedures for data acquisition”. It doesn’t have “aids for locating and obtaining the data”. It doesn’t have “a clearinghouse … to prevent the purging and loss of important data sets”. We paid her to collect this information, and it has never been seen. Is it lost? Is it altered? Is it purged? We have no idea because her data is NOT IN A DATA ARCHIVE, which is a public repository with all of those safeguards.
Why is this so hard for you to understand?
They also require “full and open access”, which means that anyone can just go and get the data from a Lithuanian researcher without having to speak Lithuanian, and without having to wait until Mrs. Lonnie gets back to town after yet another of her taxpayer-funded vacations.
Please note that the “explicitly designed archive” for paleoclimate data already exists, and all she has to do is place her data there as is required by her grant.
w.
Willis,
unfortunately, as I’ve said on numerous occasions, NSF grant terms haven’t implemented the overarching federal policy.
I’ve drawn attention to the failure of regulators to implement archiving policies. In the case at hand, the acquiescence and neglect of regulators may have resulted in grants that do not specifically implement federal policy.
Readers are often too quick to blame scientists for capitalizing on regulatory failures while insufficiently blaming the negligent regulators.
I agree with you whole heatedly Mr. McIntyre. The issue should be regulatory failure on the part of our politicians and that is where the pressure must be applied for change!
Willis,
How do you derive compulsion from “should” the strongest word contained in the quotations you’ve provided? It would be nice to do something you “should” do but it cannot be the basis of a contractual obligation – and besides, what make you think this was your money? I thought it was everyone’s, or at least those of us who pay for this shooting match.
I note that a certain response is conspicuous by its absence. Calling —- calling—.
I’m surprised not to see any discussion about today’s New York Times front page story on Lonnie Thompson. I for one did not realize how ill he was and that he needed a heart transplant to survive. It is a very flattering piece, but the following quote should resonate with CA readers: “While Dr. Thompson has defended his interpretations on these points, he does have some regrets. One is that those years of frenzied drilling led him to fall behind in publishing his data, so some of the evidence he has gathered is not yet available to the broader scientific community.”
One logical solution to Thompson’s problem would be for them to simply archive the raw data and let other scientists analyze the data, if Thompson’s health is causing delay.
The difference between field operators and lab rats is inherent in Thompson’s remarks. I know more than one professor whose head would immediately go into a Futurama-style jar, if such things existed, because they simply never have caught up their lab work and archiving with their field work, even with agencies hounding them for results. Even important chunks of the hard data, rock samples and such such are still in paper bags where they were placed when collected in the field decades ago.
If you can’t follow standard procedures don’t then publish in the first place after all if you can’t show the data you have not finished the job. Illness would not be accepted as excuse for student making claims but not supplying the data that supports them , they would be expected to write only when well enough to do both , why should it be OK for professional not to .
In this post, I attempted to draw attention to Ellen Mosley-Thompson’s non-archiving record in her own right. I do not think that the language of the post either explicitly or implicitly implied that her non-archiving was merely an “appendage” to Lonnie;s non-archiving. Nonetheless, some readers have objected that referring to her as Mrs Lonnie diminished her non-archiving accomplishments. I have accordingly removed the term “Mrs Lonnie” from the post and the title.
Reasonable people can disagree as to whether Lonnie and Ellen are serial non-archiving couple or whether each is a lone serial non-archiver operating independently within the same household.
Probably referring to Lonnie as “Mr. Ellen” occasionally will help.
This post is also documenting another form of cherry picking: how do we know that the unarchived data from the many ice cores by Lonnie and Ellen are not examples of data that show something inconvenient about past history (MWP global, no hockey stick, or maybe ice cores are lousy thermometers, who knows?). Just the same as only reporting the subgroups that responded “properly” to a new drug.
Roger seems certain that Steve has not bothered to check whether Ellen has archived her data. He says in the head post that he searched diligently. Knowing Steve, if he says he was diligent, he was. If Roger can find the data in a public place, I am sure an apology would be in order, but I’m not holding my breath. Usually, in a paper, if the data has been archived the authors point to it. In the case of these 2 authors the papers do NOT point to archives and the data can not be found in the several public data repositories.
I count 9 data sets in her name at the NOAA Paleoclimatology Ice Cores Data Sets page (the first and only place I looked). That was the fruit of a 5 minute search, starting with a simple google search. I haven’t looked at these in any detail: but I only had 5 minutes to spare and it’s late. Then there are the other two sets noted by someone else. I dare say given a little more effort one could pull up a few more cases. I’ve never pursued Ice core data before: I’m all thumbs. Did I just get lucky?
Steve: Please read carefully and please do not assume that I am unfamiliar with these data sets. Mosley-Thompson led expeditions to Antarctica and Greenland. If you look at the data sets, only one is from Antarctica-Greenland – the early series that I described. NONE of the data sets from expeditions led by Mosley-Thompson to Antarctica/Greenland is archived. The datasets that you refer to are from Lonnie’s tropical cores – where Lonnie has made abysmally incomplete archives. Most of these were unarchived until an earlier campaign of mine resulted in these meager summaries.
There seem to be a couple more at the IRI/LDEO Climate Data Library (no time to see whether there’s an overlap). There appear to be some password protected PARCA ice-core sets at ucmerced (via the school of engineering pages). There’s a link to the 1997-8 cores on the PARCA pages at NSIDC. There’s a LT and EMT Kili core at ncdc.noaa.gov, and another set (one of the ones listed earlier I think, didn’t keep a note such is my eagerness to go to bed) at the Goddard GCMD site. Maybe this is all the same stuff repeated over and over. Maybe someone should check if they haven’t already and they have another 5 minutes.
Steve: Kilimanjaro is not in Antarctica or Greenland and was not an Mosley-Thompson expedition. I’ve parsed the PARCA pages at NSIDC and have been unable to locate O18 values. Maybe I missed O18 values, but I dont think so.
I looked at the obvious googles that you’re discussing. The PARCA information at umerced does not have O18 values. Again, in your rush to try to contradict me, please pay attention to what I’m talking about: Ellen’s expeditions to Antarctica and Greenland. Please look at the data to see if it has O18 values for relevant cores.
Not in a rush to try to contradict you: just in a rush to hit the sack, and interested to stumble on a few bits and pieces so quickly. (Best perhaps to button one’s lip at at 1am…). Still, archived data is archived data isn’t it?
Just stumbled upon the University of Copenhagen page for the Niels Bohr Institute, Centre for Ice and Climate — which I suppose must represent a model for the sort of thing you’d like to see? Very impressive. http://www.iceandclimate.nbi.ku.dk/data/
Steve: There’s nothing wrong with the NOAA website as a permanent archive. On the contrary, it’s excellent. The Denmark data set archives very summary data. In my opinion, all these people should archive their entire sample data sets. In mineral exploration projects, every sample is placed in company archives.
Back to the REASON for archiving data – to allow others to duplicate and verify your results. That implies sufficient and accessible data, e.g. ALL data and not HIDDEN data. I kinda think this isn’t the case with the Lonnie “doneagain” shuffle.
What happens to company archives if the company go bust?
Steve: My point was that it’s easy to record every sample and that there are no logistical obstacles to Thompson doing so. When companies go bust, information can get lost. One hopes that Thompson won’t lose his data.
I completely agree with the principle that data needs to be made readily available as soon as possible: and I suspect that emerging generations of scientists will have quite different views on this from their forebears (not least because it’s clear that students and young internet users now /expect/ data and publications in which they’re interested to be available online instantly, as a matter of course). There’s clearly a cultural shift taking place at the moment which is uncomfortable for some (not only academics) used to different norms.
I hope you don’t mind me pressing you on this point about corporate archiving to which you’ve referred (and about which I’m completely ignorant — like many others I suspect)?
I’m not sure I really see the difference between a private company archiving its own data, under its own auspices, and an individual scientist or research group ‘archiving’ the data they’ve generated privately, to make available on request.
I believe that in some contexts in the US (Alaska?) there are state-run facilities for private companies — including to archive data (e.g. mineralogical) in perpetuity. I don’t know how widely such arrangements are used. Presumably you’d see it as sensible for private scientists (is that better than ‘citizens’?), academics like Lonnie and Ellen, and corporate scientists all to meet the same standards, and to archive data in contexts where it doesn’t matter if they are ‘hit by a bus’ or the institutional equivalent? Or am I wrong?
Steve: my initial point was a narrow technical one. Lonnie Thompson keeps pointing to his very small archive at NOAA (provided grudgingly) and says that he’s archived his data. However, each core has about 3000 samples on which 6-10 measurements are taken per sample. If his data is as important as we are told – and given the difficulty of getting the data – every damn measurement should be archived. That should be his legacy. If he argues that that is an impossible standard, that’s why mineral exploration programs are relevant. Even small programs generate similar amounts of data and companies routinely place all their sample data into a company database. There’s nothing impractical about it.
Obviously company databases are proprietary. Under the terms of some exploration licences, companies may be required to archive data in ministry archives. In Ontario, there are exploration archives going back a century. Companies have some discretion on what they archive. In some cases, when companies fold, people will contribute orphan data to the ministry archive rather than see it lost, but it is optional. The Thompsons are funded by NSF and their results are being applied in the development of public policy. So there are many considerations in favor of them being required to archive data that do not apply to private mineral exploration.
In my opinion, the most compelling reason for the Thompsons to make a comprehensive archive is that this should be their legacy. They should embrace the concept, not begrudge it.
The man himself, in the New York Times, July 2:
The argument “it is not archived until everyone has access even if the author is hit by a bus” is not theoretical. An amazing number of tree ring scientists have sadly met their demise at an early age… and before they ever archived their work.
Not only that: with data sets stored in a myriad of boxes around a lab, when the lab moves from one building to another, lots of data can get lost.
These are very real, very painful challenges.
Re: Jeff Norman (Jul 3 20:25),
+1
Just my opinion:
If there is a breach of reuirements, the way to go would be to notify the institution of the breach and ask for a receipt and follow ups. If a public institution then does not investigate and enforce their requirements, it may put them in a difficult legal position.
From the NYT article at http://www.nytimes.com/2012/07/03/science/earth/lonnie-thompson-climate-scientist-battles-time.html?pagewanted=3&_r=1&ref=general&src=me :
I wonder if Ellen Mosley-Thompson has similar regrets about her data.
Even though she may not have been a co-author on the original publications of some of Lonnie’s cores, she did co-author summaries based on them like CC03 and PNAS06, and therefore has a personal obligation to see to it that the data used there gets fully archived.
In the recent discussion of Steig’s Siple Dome data, I ran across an archived file of data from Siple Station (a very different location, despite the name) that was co-authored by Ellen M-T, so perhaps she has archived some of her own data. Unfortunately I couldn’t relocate it. It may have been at WAISCORES rather than NCDC.
I find it interesting and curious that the journal articles I have read that were coauthored by L. Thompson and about his ice core findings make no major claims like some other articles by other climate scientists do about temperature reconstructions and the unprecedented modern warming. I was sensitive to these claims since after viewing some of the ice cores I was wondering what claims would or could be made.
When you hear comments from him and reported in the MSM one would wonder where these claims, no matter how vague and general, whence come. I think it might be this consensus way of thinking that I have a difficult time understanding.
Perhaps the most disciplined enforcement of hold-out-verification data blinding?
My hypothesis: no set contains unprecedented behavior in industrial times.
these type of statements make SM post/points relevant & needing resolution –
“The reason Lonnie’s stuff is so powerful is that it’s so simple,” said Daniel P. Schrag, a geochemist at Harvard and director of its Center for the Environment.
“His evidence dismisses the idea that this is some sort of 300-year or 500-year cycle, which is what the skeptics and the deniers want to say. You say: ‘No, because Lonnie’s ice didn’t melt then. It’s melting now, but it didn’t melt then.’ ”
Yes, this is a bit off topic here, but this statement caught my eye as well. I guess they must have searched for ice that might have melted in the past, but couldn’t find any. 😉
There seems to be a tacit assumption here that either all global ice melts simultaneously or none of it melts. On a trip to Alaska several years ago, I was in a heavily forested valley which the guide described as being covered with glacier ice 500 years ago. I couldn’t see that ice either.
It would seem MacIntyre is being economical with the truth.
I emailed the Thompson’s yesterday:
I’m assuming the archive is at OSU. Could you confirm please?”
I’ve just received an email from Lonnie Thompson stating:
So it would seem the data is archived, it is available and the Thompson’s have even offered to provide more data to MacIntyre.
Anon Posted Jul 5, 2012 at 1:16 AM
Not as economical as Lonnie and you are with the LINKS to the claimed datasets, Anon …
I also note that he says he released “individual sample data” to Steve … but if his data truly were archived in the Paleoclimate data base, why would he have to release anything?
Finally, I also note that he made the release of other data conditional on whether Steve “published a new interpretation”.
Again, Anon, if his data truly were archived as he claims … then what is it that he is so darned proud of withholding from Steve?
So it seems you are being economical with your critical thought processes. If his data were archived as it should be, he couldn’t withhold it. You really should think before uncapping your electronic pen and dishing out accusations … I can see why you want to remain anonymous.
w.
I accept “being economical with the truth” was a little off and apologise.
Have you checked the archives and can verify the archive does not exist?
Anon
You’ve reported they do exist. Surely you have visited them to confirm. Please just provide the links so we can see them.
Anon
Posted Jul 5, 2012 at 3:25 AM
Thank you sir, you are indeed a gentleman.
Since neither you nor he have identified which data he refused to give to Steve, how could I possibly do that?
w.
I went to the NOAA Paleoclimate database and searched for the Plateau Remote Antarctic ice core data and I was unable to find it.
“I went to the NOAA Paleoclimate database and searched for the Plateau Remote Antarctic ice core data and I was unable to find it.”
No, they have it listed at the NOAA Ice Core Gateway (listed between Newall Glacier and Shirase Drainage Basin), it’s just not available in a single click as many aren’t. Try contacting them.
http://www.ncdc.noaa.gov/paleo/icecore/current.html
Steve: You’re wrong. It would be better if you didn’t present information as facts when you don’t know. There’s nothing wrong with not knowing – just don’t pretend to know. The Plateau Remote data is not at the NOAA archive. They listed Plateau Remote as a site where an ice core has been drilled, but they don’t have the data.
Plateau Remote is listed at http://www.ncdc.noaa.gov/paleo/icecore/current.html , but has no link to data.
Not archived.
Just because there isn’t a direct link does not mean the data isn’t archived.
Steve: In this case, it does. I say this not just because there is no link, but because I have carefully parsed the entire NOAA FTP site and can affirm that the Plateau Remote data isn’t there. Look at ftp://ftp.ncdc.noaa.gov/pub/data/paleo/icecore/antarctica/
Putting a name in a list is not archiving.
Read the document on NOAA policy linked at the bottom of the NOAA page whose URL you posted previously. The first paragraph states:
If there is no “direct” link (which does exist for many other data sets), then either NOAA is not following its own policy or the data is not archived.
What exactly is an indirect link?
Steve, all of the names in the list that are black (not hyperlinked) are missing from the FTP page you link to as well. Are they all the fault of the Thompsons, too? Why would you even expect it to be on the FTP if it isn’t hyperlinked at the Gateway page?
Casey
D57
Dolleman Is.
DSS (Dome Summit South)
Dumont D’Urville
Dyer Plateau
James Ross Is. (Dallinger Dome)
Mizuho (Watanabe et al., 1978)
Plateau Remote
Shirase Drainage Basin
South Pole (Jouzel et al., 1983)
Windless Bight
Steve: Within this list, the Dyer Plateau and Plateau Remote are Ellen sites. I observed that ELlen Mosley-Thompson hadn’t archived data from expeditions that she led to Greenland and Antarctica – including the relatively PARCA cores.
I didn’t say anything about cores that they had nothing to do with. WHy would you make such an absurd argument? Some of the sites in this list are quite early e.g. the Mizuho site.
I parsed the FTP site because I’m thorough not because I “expected” data to be there that wasn’t linked at the gateway. I generally use the FTP site directly anyway.
It’s true that Lonnie T has had a recent frenzy of archiving in response to Steve’s persistence at ftp://ftp.ncdc.noaa.gov/pub/data/paleo/icecore/ :
Dunde was just added a month ago, on 6/4/12. Puruogangri was added 8/24/11, and Dasuopu, Guliya and Quelccaya were all added to in the last 18 months.
Please ask him for the URL to Bona Churchill and let us know what you learn: https://climateaudit.org/2007/11/12/gleanings-on-bona-churchill/ . I would, but he hasn’t responded to my e-mails since I called him out on the authorship of the series identified as “Dr. Thompson’s Thermometer” in AIT.
In any event, this post is about Ellen M-T’s cores, not Lonnie’s. Please e-mail her as well.
RE Ellen’s Plateau Remote data, J Bowers writes
Then what’s the URL and why isn’t it linked? The burden is on Ellen M-T to contact them if they didn’t link her data correctly, not on us.
Your question to Lonnie Thompson was irrelevant to this particular post.
Lonnie Thompson has a very inadequate archive at WDCP for his tropical cores. No data from Dunde, GUliya or Dasuopu had been archived until I complained to one of his journals, which required him (grudgingly) to archive data, which he can now point to. After many years of criticism, he’s added to the WDCP archive in the past year and even in the past few months. As Thompson says, he’s also provided grey versions of data to researchers, some of which have been archived in the third party collations. Unfortunately, the various versions for e.g. Dunde are inconsistent creating a total dog’s breakfast. This can and should be resolved by a comprehensive archive of all sample data – EVERY single measurement.
In my opinion, Thompson should not view the provision of such an archive as an imposition, but as his legacy. If he did what I suggest, he’d be able to look back at his career as being more complete than it is at present.
But back to my post. It was not about Lonnie, but about Ellen Mosley-Thompson and her serial non-archiving of Antarctic and Greenland ice cores. No one has contradicted my assertions about her. Lonnie’s assertion about Kilimanjaro is completely irrelevant to Ellen Mosley-Thompson’s serial non-archiving.
I re-iterate that Thompson’s own archive is completely inadequate relative to the supposed importance of his ice cores. I’m not saying that there is nothing archived for his tropical cores – but that the archive has been inadequate. Ellen’s is non-existent.
The question was to Lonnie and Ellen
I assumed one spoke for the other as you did
Anon —
Thanks for e-mailing the Thompson’s about their data.
However, I find Lonnie’s reference in his reply to “our data” to be a little ambiguous: Even if the e-mail was equally “TO” both of them, he may not have noticed this and may have interpreted it as referring just to his own controversial data, much of which he has indeed created NCDC archives for in the last 18 months. “Our data” may then be referring to himself and his coauthors on these specific cores, rather than to himself and Ellen.
In order to resolve this uncertainty, could you possibly now e-mail Ellen M-T directly (with a CC to Lonnie and a copy of Lonnie’s reply), asking her for the whereabouts of her Greenland and Antarctic data archives? You might specifically mention Plateau Remote.
Thanks!
Anon stated:
I challenge Lonnie Thompson to produce such an email to me from him or his associates. To the best of my knowledge and recollection, it doesn’t exist and Thompson’s claim to have sent such an email to me is untrue. I’ve searched my emails for the relevant period and have no record of receiving an email from Lonnie Thompson or any of his associates with the supposed notation. (To the best of my knowledge, Thompson has never responded to or acknowledged a single email, though Ellen Mosley-Thompson has responded to a couple. Nor did I ever correspond directly with either Lonnie or Ellen in connection with Kilimanjaro. )
I did receive Kilimanjaro sample data in 2005 from the editors of Science without any such notation. I had asked the editors of Science to require Thompson to provide detailed sample information on Dunde, Guliya and Dasuopu, all of which were used in multiproxy studies. In October 2005, they sent me sample information for two Kilimanjaro cores, which was obviously not responsive to my issues with Dunde particularly. I placed this information online at the time. While the Kilimanjaro information was interesting, it did not resolve the continuing issues with the Himalaya cores. I continue to press Science for this data but got nowhere.
At no point did the editors of Science ever suggest that the availability of further data would be linked to my publication of a re-interpretation of Kilimanjaro dates in academic literature. This would hardly be germane to their policies. Especially when I was seeking data for different cores in the first place. Nor was the Kilimanjaro data left unused. As a reviewer of AR4, I contested their reliance on Thompson’s dating of Kilimanjaro. The IPCC authors accepted this comment – one of the few that they accepted – the responder saying that he understood the problem that I had raised.
I have no idea why Thompson believes that he sent me an email with “the notation that if he published a new interpretation on those data in a peer reviewed journal that I would send him the next data set”. I have no record or recollection of ever receiving such an email and do not believe Thompson’s claim.
Interesting, hopefully they will respond
It will not matter whether they respond or not. It will not matter whether Thompson fabricated this exchange or not. It will not matter whether everything they claim can be shown to be untrue. Instead, they will get more hagiography, more glowing profiles in periodicals. More tales of heroics on mountain tops. Thompson, Mann, Ehrlich, have achieved media status that gives them automatic impunity.
One other point on the Kilimanjaro samples. I did three detailed blog posts taking a look at Kilimanjaro sample data:
In the last post, I questioned Thompson’s dating of the Kilimanjaro ice cap.
In my review of AR4 I commented on Kilimanjaro. The First Draft stated:
As a reviewer, I stated:
The AR4 author responded:
The statement was removed in the Second Draft and in the final report.
It is completely possible that he did in fact send data to some person named MacIntyre. Who could possibly know other than the sender and the receiver?
It seems to me that consistently spelling someone’s name incorrectly is a sign of disrespect and a form of ad hominem. What do you think anom?
Geronimo
Thread hijacking ? Its rather strange that calling for simple evidence gathering prior to making a public ad hominem attack should elicit should a response.
Regarding your points.
(1) She should put it into a formal archive. Where did I ever argue that she shouldn’t ? I also said there was de facto little difference in data availability between her giving it away freely and putting in an archive although giving it away freely is a far more imperfect way of doing things.
(2) If data are given away freely then anyone can get this anyway to do whatever audits they want. Data include raw data, processed data – its all data (as I mentioned earlier on in this thread).
(3) This is a strange point since it violates the argument I made in point (2). You said you wanted to keep to logic, so please respond to what I wrote, not what you would have preferred me to have written. Somebody could also manipulate any data easily enough and then put it into a public archive after several years. I’m not sure how the public archive argument protects in anyway protects against that.
Now, I await yours and others condemnation of Dr Svalgaard, the noted CAGW sceptic. I provided the proof of his attitude to data sharing. Its a click away (look at the report he provided) and is provided by a reliable source (himself through WUWT).
Roger:
Thread hijacking ? Its rather strange that calling for simple evidence gathering prior to making a public ad hominem attack should elicit should a response.
Regarding your points.
(1) She should put it into a formal archive. Where did I ever argue that she shouldn’t ? I also said there was de facto little difference in data availability between her giving it away freely and putting in an archive although giving it away freely is a far more imperfect way of doing things.
#########
The defacto difference between her giving it away freely upon request and her posting
a copy would be clear to anyone who managed data.
If you ask Ellen for A and I ask ellen for A, and she sends us a file. Do you
know that you have recieved the same file I have. If you think this is not
a problem, see Nic Lewis’s latest difficulty.
If ellen places A online, then we both know we are accessing the same thing.
(2) If data are given away freely then anyone can get this anyway to do whatever audits they want. Data include raw data, processed data – its all data (as I mentioned earlier on in this thread).
See above. We post data and code for a reason. That reason: YOUR METHOD is a known
source of errors. There is a reason why industry has moved from passing files
between people to posting information in a shared storage area. Passing around
individual files is broken. Been there done that, learned some tough lessons.
You have not learned these lessons.
(3) This is a strange point since it violates the argument I made in point (2). You said you wanted to keep to logic, so please respond to what I wrote, not what you would have preferred me to have written. Somebody could also manipulate any data easily enough and then put it into a public archive after several years. I’m not sure how the public archive argument protects in anyway protects against that.
Now, I await yours and others condemnation of Dr Svalgaard, the noted CAGW sceptic. I provided the proof of his attitude to data sharing. Its a click away (look at the report he provided) and is provided by a reliable source (himself through WUWT).
Dr Svalgaard is a personal friend. I condemn the practice of not sharing data regardless
of who practices it. It is not best practices. Still, he’s a lovely man and his wife
is charming. I would no more condemn him as a man than I would condemn Ellen as a woman.
I’m happy to condemn the practices of both.
“I also said there was de facto little difference in data availability between her giving it away freely and putting in an archive although giving it away freely is a far more imperfect way of doing things.”
Roger, the very obvious point here is that if the data are archived we need not test the de facto little difference. Has E. Thompson stated unequivocally that she will provide her data to anyone who requests it and can you show us where she might have made that statement? If you can then I will immediately make such a request to her and if she makes the data available to me I’ll make the data available to the public. Better yet, I could simply ask her to provide the data to NOAA – does that make more sense to you? If she did not fulfill my request within a reasonable period of time what would your opinion on the matter be then. Also if I test the system we would need to continue this conversation after it was tested. How best could that happen?
I’m well aware of your position. I’ve absolutely no idea if she has stated she’ll share her data or even if she would if she agreed to do this. The point is before attacking her she should have been asked for all her data. A trivial exercise. However, nobody bothered to exclude the hypothesis of a diligent researcher with an open-to-all attitude about her own work but with an imperfect archiving policy. Pity, before launching public ad hominem attacks its the least one can do. It help gets the full picture.
Now, I note that you say nothing about Dr Svalgaard who has stated that he’ll share on request. However, as has been repeatedly argued to me this is an awful way to do science. Can I expect a full condemnation of his practice ?
I’m unfamiliar with the Svalgaard case. In my opinion, authors should archive all data as used together with turnkey scripts at the time of publication. If Svalgaard hasn’t done this, he is hardly alone but I would certainly encourage him to comply with the standards advocated here.
I agree although it will take time to effect a sea change in attitudes, which is what’s required. However, there are changes afoot. Where I work we are now reminded to make data public, in particular for climate science and other areas influencing public policy. Much of this is likely down to your efforts as much the technology making it easier. There aren’t too many others making the case; data access is an unfashionable cause to push.
Roger
Posted Jul 5, 2012 at 11:12 AM
A few points of note.
First, an “ad hominem” attack is attacking the person rather than attacking their scientific ideas. It is done to divert the issue from one of the actual scientific claims that they are making to the person themselves. For example, “Your ideas are wrong because you don’t have a PhD” is an ad hominem attack that I get all the time.
But Ellen M-T’s scientific ideas form no part of this discussion, so an “ad hominem” attack in this case is not even theoretically possible. The question under discussion is whether she has archived her data or not, so to say that she has not done so, as Steve has done, is totally on point and is not “ad hominem” in the slightest.
Regarding that actual subject, her archiving practices, you claim that Steve has not done “the necessary research”.
You seem to have missed the part where Steve says:
That’s called “the necessary research”, Roger. You seem to think that asking her directly would make things clearer. But as the example upthread of “Anon” asking Lonnie if he has archived his data clearly shows, asking Ellen if she archived her work is a fool’s errand. One-half of the non-archiving couple are more than happy to flat-out lie about the question of whether their work is archived, so only a fool would trust the other half of the couple. Lonnie clearly claimed to Anon that his work was archived … but it’s not.
Finally, both Mosley-Thompson and Thompson-Mosley are quite free to defend their archiving record right here. If I were being attacked for my archiving, I’d be the first one to show up here or anywhere to defend my own practices … as would you, I assume, as you seem to be an honest man.
As a result, for people with common sense the fact that they have not defended their archiving record, either here or elsewhere, forms part of the “necessary research”.
But heck, Roger, if asking her if she archived her data is so gosh-darn important in your world, if you think it should be part of the “necessary research” … then why are you wasting your time here discussing this with me? Why haven’t you asked her, so you can get over this idea that she has archived her data?
Most of us know she hasn’t done so, because we’ve looked. You see, that’s part of the point of a public archive—you can find a scientist’s data without having to ask them where it is. That’s why NOAA has the paleoclimate archive. But if you think she’s archived it, go ahead and ask her … I’ll wait, I’m a patient man …
w.
PS—I note that you keep reiterating your claim that a private individual holding on to data and being willing to share it is the same as archiving.
Your statement is like me saying “As a basketball player, I’m equivalent to Shaquille O’Neal, albeit in a very imperfect way.” Either I’m a basketball player who is actually equivalent to Shaq or I’m not (actually I’m not, in case you were wondering) … and similarly, either private ownership plus a willingness to share is actually equivalent to public archiving or it’s not. (It’s not, in case you were wondering.)
I and others have pointed out three big reasons why her holding on to the data privately is not equivalent to archiving in the slightest, even if she shares it freely:
1. What if she is hit by a bus?
2. What if (a la Phil Jones and a number of others) she simply loses the data?
3. How would we know if she decides to secretly alter the data, or it is accidentally corrupted?
Those are non-issues if the data is archived, but are huge issues if it is not. There are other reasons, including accessibility when she is not accessible, and language difficulties.
You have not responded, as far as I know, to a single one of these issues. All you have done is reiterate your claim that data residing on Ellen’s PC is the equivalent of data archived at the NOAA Paleoclimate Archive … so how about pointing out how Ms. M-T privately holding the data gets around those large problems?
“Pity, before launching public ad hominem attacks its the least one can do. It help gets the full picture.”
I have emailed Ellen Thompson concerning making her data from her Antarctica and Greenland expeditions public and asked if there were any extenuating circumstance that prevented her from making it public. That action does not change my judgment that publicly funded work of this nature should be archived. I also find it very difficult to believe that Thompson will at this time concede in making it public by a simple prompting from me or anyone else. Pressure from her peers might make this happen but I do not see that occurring any time soon.
But you, Roger, seem not much interested in the subject at hand which is archiving data but rather in attempting to make your references to slurring Thompson and ad hominem attacks stick by repetition. Archiving of climate science data is not so much a matter of personal choices but rather of the science community deciding what gets taken seriously and what does not.
If you are interested in archiving and what could be done to encourage it for the climate science community, you should put away your low opinion of the people posting here and attempt to enter into an objective discussion. Otherwise I suggest you move on, since you can only express your opinion so many ways and your allotment has been filled.
Your kind of posting is nothing new here. I have seen others do the same and with one thing mainly in common: they never discuss the issue but get side-tracked in the personalities of posters here.
By the way, there was a puff piece on Lonnie Thompson on the front page of the New York Times earlier this week. While there was brief mention at the very end about his critics, the main purpose of the piece seemed to be merely providing an excuse to include “Global Warming” in a story headline at a time of high temperatures on the east coast.
Lucia, you definitely don’t need my approval but IMHO you definitely have Roger pegged accurately.
One thing I noticed about both Thompsons is that both of them got their Masters and PhD degrees from Ohio State and have spent every minute of their subsequent academic careers at Ohio State. You can see lists of their degrees and links to their CVs here:
http://bprc.osu.edu/Icecore/GroupP.html#lonniethompson
Most universities will not hire their own PhDs because doing so can lead to an inbreeding/nepotism type of problem. Those that do will only hire their own after they have gone elsewhere and distinguished themselves first.
When I was a graduate student in economics at Ohio State, there were zero Ohio State PhDs among the tenure-track faculty in the econ department. And when we were finishing up our degrees and going on the job market, we didn’t bother applying to the department for a position. We knew we would not even be considered.
Here’s a copy of the letter I am sending to the provost at Ohio State and also the Dean of the college of arts & sciences.
July 6, 2012
Dr. James McCown
Dr. Joseph Alutto
Executive Vice-President and Provost
The Ohio State University
Office of Academic Affairs
203 Bricker Hall
190 N. Oval Mall
Columbus, OH 43210
Dear Provost Alutto:
I am an alumnus of Ohio State, having earned my PhD in economics there in 1997. The reason why I chose Ohio State for my graduate studies was because of its excellent reputation as a research institution. Unfortunately, it has come to my attention that two of the faculty members at my alma mater are engaging in academic misconduct that is also in violation of federal and state law. Therefore, I am writing in the hope that you will persuade these professors to act in a manner that upholds the reputation of the Ohio State University.
As you are probably aware, Professor Lonnie Thompson of the School of Earth Sciences and Professor Ellen Mosley-Thompson of the Department of Geography have undertaken significant research involving ice core paleoclimatology. Much of this research has been financed by the US taxpayers with National Science Foundation grants. The Thompsons’ research has potential implications for public policy in both the state of Ohio and the US federal government. Thus, it is important that other researchers in climate science be given access to the data that the Thompsons are using, in order to verify the accuracy of their results and ensure their methods are valid.
Both Professor Lonnie Thompson and Professor Ellen Mosley-Thompson have refused repeated requests that they share their data, or place it on an archive accessible on the internet. Mr Steve McIntyre of Canada, Professor Huston McCulloch of Ohio State’s economics department, and several other people have requested access to the data, only to have their requests ignored or refused.
The Thompsons are required by law to make the data from research projects funded by the National Science Foundation available to other researchers. I ask that you advise them to comply with the law. And if they fail to do so, I ask they be placed under disciplinary action by the university.
Sincerely,
Dr. James McCown
PhD, The Ohio State University, 1997
Bob etc
Please, no excuses for Dr Svalgaard (after all – none were afforded to Mrs Lonnie).
Where are Dr Svalgaard’s data ? I’m an interested member of the public who wants to verify his conclusions. My motives are irrelevent. How do I access the data ? Surely, I don’t have to contact him, not with today’s technologies etc. I’m not interested in his position or motivations for his data sharing policy (such a discussion was forbidden for the Mrs Lonnie case). I’m interested in getting hold of the data for a published study without having to ask the scientist first. This is a question of basic scientific ethics. It has been argued that a scientist who does not make all his/her data publicly available is not “a real scientist”. This is essentially the argument put forward by a lot of people this thread.
It is straightforward to store data and make a turnkey approach to algorithms especially with today’s technologies. Steve has demonstrated this.
No excuses, no evasions. How do I download *all* the data (raw data, metadata etc) from Dr Svalgaard’s published study ? Where has it been publicly sitting for the past 2 years ?
You spent quite some time on your post. It would be much simpler if you just showed me how I’m able to reproduce Dr Svalgaard’s work without contacting him first.
I agree. Throw Dr. Svalgaard under the bus if he has not made data public. So do you agree that Lonnie and Ellen should be thrown under the bus too?
Then we’ll see who responds to the pressure…
Roger, what surprises me from the Svalgaard exchange is that the referees asked that the data be archived – something that hasn’t happened to the Thompsons.
If the Svalgaard issue pertains to issues pertaining to 3 GB, I don’t want to debate this. I haven’t criticized authors for not archiving huge databases (though the definition of huge is changing); I’ve restricted my criticisms to “small” datasets such as the Thompsons. As I’ve said before, I think that Svalgaard should adhere to the same expectations as those set out for the Thompsons.
Hi Steve,
I rather suspect that the reviewer was simply trying to make life difficult for Svalgaard (the phrase: “going to town on him” springs to mind in this context). However, with anonymity one can never be sure of anything.
Granted there is subjectivity in size but I don’t think 3GB is in anyway a large database when it comes to storage and access. Maybe 10 years ago perhaps. I also doubt the posters would consider it to be a legitimate obstable. I recall a previous discussion on CA when lots of posters were criticising particle physics experiments for their data management policies. They wouldn’t countenance the argument that the data sets (~100 million GB for data alone, not counting simulation) + the GRID data management infrastructure couldn’t be easily made available for public use.
Bob Edgar’s post raises an important issue about the consequences of rules that are enforced on a rather arbitrary basis. You can give a pass to that with which you agree and nail to the wall (even when there are extenuating circumstances) that which you do not. I see that not only in organizations and agencies but in responses to FOIA requests.
Ideally, key ice core data should be centrally archived at the NCDC FTP site at ftp://ftp.ncdc.noaa.gov/pub/data/paleo/icecore/ .
However, data might be archived elsewhere. Most academics now have university-provided websites where they can upload their data less formally. Thus, Eric Steig recently posted Siple Dome dD data on his U. Wash. website, which although not ideal is better than nothing.
For their part, the Thompsons control the Byrd Polar Research Center’s website, and indeed there is an icecore data archive page there, at http://bprc.osu.edu/Icecore/dataandimages.html .
Although this page does have a link for Ellen M-T’s Greenland (PARCA) data, the link is dead when I tried it,so this doesn’t count as even an informal archive.
It also has a link for Lonnie’s 2002 Bona-Churchill data, (see https://climateaudit.org/2007/11/12/gleanings-on-bona-churchill/ ) but all it leads to is a statement that preliminary results will be presented at the 2004 AGU meetings.
That’s hilarious, Hu, and just so typical. Dead links to supposed archives, and claims that the data will be presented in 2004 …
w.
Hu/Willis,
I’ve exhausted all my tricks trying to find that PARCA link. Even Wayback doesn’t have it.
http://wayback.archive.org/web/*/http://geog.sbs.ohio-state.edu/courses/parca/
Even the reference to it at NSIDC is DOA
http://nsidc.org/data/parca/data.html#parca-17
Its a total memory hole. The machine was removed from the net and appears to have had a block on it to prevent robots crawling it and archiving pages. There’s nary a trace of it anywhere.
Re: Anthony Watts (Jul 7 01:29),
Anthony, the address in that link, http://geog-www.sbs.ohio-state.edu/courses/parca/cores.html, is likely to be or have been on the Ohio State Intranet. Anyone here with access to the OH Intranet?
jeez —
I’m at Ohio State, and in SBS no less, but have never seen a domain name that begins like geog-www. Perhaps it’s an internal geography department domain. The fact that the folder is /courses/ suggests that this was uploaded for a specific course, and perhaps only geography department members have access. (Ellen M-T is in geography).
A newish university-wide system called CARMEN allows materials to be uploaded for access only by students enrolled in a specific course, but perhaps this was an older geography-specific system.
I’m at home now and the link doesn’t work from my laptop. I’ll try it from my office computer Monday.
FYI – This link:
http://www.sbs.ohio-state.edu/courses/parca/cores.html
(remove the ‘geog’), returns a 404 error, which implies that the server is there, but the address is no longer on the site.
With the ‘geog’ the server itself is not found.
While the requirements re archiving of data is the focus of this post, the processing of the data is of equal importance.
My simplistic mind says that in a statistics/data-heavy field like climatology all raw data should be archived AND presented as is.
Post-rocessing presentation then can be done, along with methodologies clearly laid out. Code then should be presented for perusal, to see if it actually does what the methodology says is supposed to be done in the number crunching.
Without before and after graphs, how is anyone supposed to critique the study?
It is NOT for funding agencies/groups to dictate methodology, but I assume, Steve, that such things are covered elsewhere in the rules?
Steve Garcia
Apologies if this case has been cited before. I came across it and it made me think of Climate Audit.
http://docs.justia.com/cases/federal/district-courts/pennsylvania/pawdce/2:2009cv01480/94664/26/
“On November 11, 2009, Plaintiff brought suit against the United States Environmental Protection Agency (“EPA”), United States Department of Health and Human Services (“DHHS”), the Centers for Disease Control and Prevention (“CDC”), National Institutes of Health (“NIH”), and the National Institute of Environmental Health Sciences (“NIEHS”), seeking an order of Court requiring Defendants to disclose certain scientific records and research data. According to the Complaint, [*2] Ms. Pohl requested this information from Defendants under the Freedom of Information Act (“FOIA”) in August 2007. The data related to grants which the EPA and the DHHS, through its component agencies, the CDC, NIEHS and NIH, had awarded to Bruce Lanphear, a researcher at the Center for the Study of Prevalent Neurotoxicants in Children at the Children’s Hospital Medical Center in Cincinnati, Ohio. Based on his research, Dr. Lanphear had concluded that children exposed to lead particles in the air, even at very low levels, suffer a decrease in IQ levels. (See Doc, No. 10 at 1.) This research led, in turn, to a Final Rule by the EPA dated May 1, 2008, adopting a new ambient air quality standard which cut by 90% the amount of lead permitted in the air.”
*****
“It appears to the Court that Plaintiff is establishing two separate scenarios in the proposed amended complaint. First, the Government Defendants violated FOIA by failing to provide the Requested Data despite the fact that the refusal was not based on one of the nine exceptions. This failure can be remedied under FOIA through an order of Court requiring them to produce the data. Alternatively, the Government Defendants violated the Shelby Amendment 3 by allowing the Research Defendants to withhold the Requested Data despite the provisions of Circular A-110 which state that if a Federal agency has granted an award under which data were produced which the Federal Government uses in “developing an agency-action that has the force and effect of law,” the awarding agency “shall request, and the recipient shall provide,” the research data so they can be made available to the public through FOIA. This is a separate violation because the Government Defendants failed to take “discrete agency action” they were “required to take” under the Shelby Amendment and Circular [*13] A-110, regardless of the interplay with FOIA, See Norton v. S. Utah Wilderness Alliance, 542 U.S. 55, 64, 124 S. Ct. 2373, 159 L. Ed. 2d 137 (2004) (the APA provides relief for an agency’s failure to take action it is legally required to take.) Actions brought to enforce the Shelby Amendment or Circular A-110 can be remedied through application of the APA or the mandamus statute. See Salt Inst. v. Thompson, 345 F. Supp. 2d 589, 601-602 (E.D. Va. 2004) (where two underlying prerequisites are met – the agency must have taken “final agency action” and the action is not committed to agency discretion by law – it is presumed that APA judicial review applies.) At this point in the litigation, Plaintiff will be allowed to proceed with her alternative theories of relief.”
This case was ultimately settled, with the requested data being provided.
Another interesting case
http://www.hhs.gov/dab/decisions/dab1232.html
In a plagiarism case, lack of primary data supports an infeence of plagiarism.
DR. C. DAVID BRIDGES, DAB No. 1232 (1991)
“In determining to recommend to the Debarring Official that the
Respondent be debarred, I considered the strong public interest in
protecting the integrity of the scientific research process, and in
preventing the potential for abuses of federal funds. The record
indicated that Dr. Bridges’ disregard for accepted standards of conduct
affected his actions throughout 1986, when the research in question was
conducted, and continued to affect Dr. Bridges with regard to his
failure to maintain and make available adequate primary data related to
the research he conducted. I further conclude that there are no factors
that would mitigate against a conclusion that debarment is warranted.
The Respondent’s disregard for accepted standards of conduct and his
efforts to deceive others about the conduct of his experiments reflects
an overall lack of integrity, care, and judgment.”
See Also
“ORI also suggested in post-hearing briefing that a previous decision had set a documentation retention standard for research data. Contrary to ORI’s suggestion, that case did not find scientific misconduct based on lack of documentation, but was a plagiarism case which merely referenced a regulatory standard for record retention under federal grants as support for an inference drawn from the lack of any original primary data for published work. Dr. David C. Bridges, DAB 1232 at 83-87 (1991). We also note that the retention requirement is a three-year requirement, and there is no basis in the record here for concluding that Dr. Imanishi-Kari had failed to meet it, to the extent it applied. See 45 C.F.R. 74.21 (1984-86). Similarly, ORI’s reliance on a 1989 publication as establishing standards for research conducted before that time is misplaced.” DAB No. 1582
The earlier case cited referenced the SHelby Amendment. Discussion on the Shelby Amendment from a law review article
“The Freedom of Information Act (FOIA) allows for public access to a variety of governmental records upon request. Federal agencies have promulgated regulations to implement the FOIA, including the Department of Health and Human Services (and hence National Institutes of Health [NIH] and its institutes). Concerns have been raised from across the disciplinary spectrum about the possibility that individuals or organizations could utilize the FOIA process to compel involuntary disclosure of research data and other related information from investigators who are funded by NIH and its institutes. Although compelled disclosure is always a possibility depending on the circumstances (see discussion of researcher privilege above), use of FOIA procedures to accomplish such disclosure for federally funded research information appears to only be a remote concern at this point.
The Freedom of Information Act was passed and signed in 1966 with much fanfare and commendable goals. It established a basis for public inspection of nonsensitive governmental records” to offer more transparency in how the federal government operates–arguably a foundation of any representative democracy. The Act focused on information and records directly in the possession or control of governmental agencies. However, over time, the focus expanded to access to information that was in the peripheral control of agencies. Nevertheless, a seminal decision by the U.S. Supreme Court in 1980 established that a federal agency is not required to disclose research data collected under activities funded by the agency if the agency does not have actual possession of the data. It would seem that this case would have ended concern about researchers being compelled to disclose data under the FOIA. However, politics intervened.
In the 1990s, the U.S. Environmental Protection Agency (EPA) updated its clean air standards based at least partially on a longitudinal health and mortality study conducted at Harvard University. When Congressional opponents of the new standards requested the raw data generated by the study (which EPA and HHS–the study funder–did not have), Harvard refused to disclose the data citing confidentiality concerns. Senator Richard Shelby (R-Alabama) then inserted language into a Bill that directed the Congressional Office of Management and Budget (OMB) to amend OMB Circular A-110, which governs the administration of grants to universities, hospitals, and other nonprofit organizations, to require that all data produced by federally funded research be made available to the public under the framework of the FOIA.
There was widespread concern over the possible implications of the “Shelby Amendment” for researchers. These concerns were tempered when the OMB finally published its proposed changes to Circular A-110, and offered access to much fewer kinds of information than the Shelby Amendment implied. In the final rule, OMB only required disclosure for data used in published studies that are utilized by a federal agency to create law or policy. The OMB rule, while quite narrow in scope compared to earlier fears, has not allayed all concerns among researchers. However, while judicial discussion of the FOIA in this context has occurred, the courts have yet to apply the provisions in Circular A-110. Thus, although Congress has discussed the involuntary disclosure of research information under the FOIA and even included it in legislation and rules, it does not appear to be a prominent concern at this point. The case law (or lack of it) is telling and when the issue is ultimately litigated, the courts will still likely use a balancing test–weighing the public benefits of disclosure against the private concerns of individual confidentiality.”
17 Psych. Pub. Pol. and L. 333, 346
Two of the main perpetrators of the “tropical glacier mass is a proxy for tropospheric temperature” (as opposed to tropical moisture) meme come under scrutiny. The NYT rushes to their defense. Ah, the world is an amazing place.
Thanks for the comments.
Ad hominem is perfectly well described as a term denoting personal attacks
http://en.wiktionary.org/wiki/ad_hominem .
I’m afraid I do see this is a personal attack. One person has been singled out for critism for conduct which is commonplace in a field (and in a highly snarky way). IMO any neutral observer would be hard pressed to say that there was nothing ad hominem in this CA post.
On another note, I enjoyed reading the posts about Dr Svalgaard. As predicted, excuses, evasions and none of the language reserve for Mrs Lonnie. S. Mosher came closest with a condemnation and a “not best practice” statement. However, this was followed with a “still, he’s a lovely man and his wife is charming” statement. A nice (and highly appropriate) bit of human warmth. Still, none of the hyperbole of “not a real scientist” here.
Re: Roger (Jul 7 03:23),
Roger, in what way is Steve’s post a “personal attack”? Note that the site you reference defines the term. And the definition is consistent with the real definition for ad hominem.
Sorry, you are just blowing troll-smoke here.
Steve’s post is factual. Steve is highly respected worldwide as being one of the least emotional/snarky commenters in the field. He receives more abuse than you can imagine, yet continues to respond on an even keel, and with devastating command of the facts.
Give it up. And please stop this never-ending push to water down the need for archiving, as well as trying to blur the focus of this post.
You want to go after 100% of everyone who fails to archive, all at once? Fine. Go for it. Or… provide enough resources so Steve can lead such an effort.
Roger, if you can’t bring yourself to do either one, then your perspective is completely unrealistic. Steve is one person, doing way more than one person’s job in whistleblowing on this and other important issues in climate science.
KnR
My point has been consistently misrepresented here. With reference to what I’ve written (not what you think I’ve written) what do you think my point is and how would it be proven/disproven should I contact Mrs Lonnie ?
Roger: it is not polite to completely take over someone else’s blog like you are doing here.
As someone who has hunted down this type of data and requested it, I can say that the response to requests, even if one is not infamous like our host, is less than 50% because a) the author is deceased, b) the author does not feel like sending data or even replying, c) the email has changed and the author can not be tracked down. This is totally unprofessional. Archiving was invented for a reason.
One of the persistent problems documented here and elsewhere is inconsistent data. Data that does not match the figures in papers, truncated data, data only part of which is reported, cherry picking. There is a real problem that needs addressing by allowing others access to the data. These data in the case of sediment cores or ice cores cost millions to obtain, not a few thousand, and are a world-wide scientific treasure. They should not be hoarded for decades by absent minded professors.
“I don’t like the idea of attacking people too much”
Why start commenting on this now? Climate Science has been waging an attack on “skeptics” for decades.
Andrew
> So then why did you post a link to the definition of an ad hominem argument?
Because what mattered to me was the **effect** the authors were describing, issuing ubiquitious criticisms that target persons, using these criticisms in a public debate to fuel or provoke an hostile exchange, promoting (in the technical sense) much fanfare and drama, at the expense of real and meaningful discussion.
As auditors would say, sound familiar?
These authors were not concerned by other discourses than (political) debates, and so the basic definition they are using have no currency for our current op-ed.
Let’s recall the title of this op-ed: **Lonnie and Ellen, Serial Non Archivers**.
If auditors can’t see how this might very well amount to an ad hominem attack on at least three counts, there’s not much incentive to play the dictionary game. Dictionary games should be kept in the parsomatic genial mecca anyway.
Claiming that this is a purely descriptive title is often seen as a justification for the use of the D word.
Trying to portray CA a formal platform to discuss archivistics issues lacks honor.
Contradicting Roger’s point that way only reinforces it; gatekeepers should let go.
Naming and shaming should be owned, more so when one is a serial namer and shamer.
That last epithet was purely descriptive, of course.
PS: On the other hand, the two technical questions are interesting, considering we’re dealing with archiving issues.
Myles ALlen recently commented here, introducing the term “name and shame” as a tactic against non-archiving climate scientists:
I regard these battles over data as a total waste of time. Data being relied upon by IPCC should be archived. Nor do climate scientists do their field any good by not archiving data or by descending into CRU-type FOI battles.
The “effect” you were looking for was to create a link to a known fallacious argumentative technique. However, since the key element of the ad hominem appellation is that the disparagement be “unrelated to the discussion at hand”, such a link is spurious. Criticisms of poor scientific behavior are indeed personal, but that behavior is the point of the post. If these criticisms become ubiquitous(!), then it is only because the issue has blatantly continued over a prolonged period of time and is unlikely to be improved by polite discourse with people who refuse to discuss the topic outside of their insular tightly-knit circle.
The rest of your comment is ridiculously silly.
“Dictionary game”? When you are completely off-base on the meaning of terms and phrases, it becomes a “game”? From my experience reading previous comments you have written, it seems that playing “word games” has constituted your sole modus operandi.
“Lacks honor”? Give me a break. We do indeed discuss science on this blog – including archiving issues. Somehow, you just don’t seem to become involved in the meat of those discussions. Now, why would that be the case…
“Contradicting Roger’s point”? Which one would that be? Roger has contributed many opinions, but nothing with any factual basis. His accusations of Steve not having done his homework on the issue were completely false. He has even stooped to arguing that the blog post will corrupt the minds of weak-willed people: “People read what they want to read. In many cases even a hint of criticism is enough for them to assume the worst.” Fie on Steve for such ruthless irresponsibility!
test. my previous comment did not post.
So, I’ll try again:
The issue is moot.
All data are in public domain [the vast majority in World Data Centers] which is where we got them from in the first place. As we remark in one of our papers http://www.leif.org/research/2007JA012437.pdf section A4.2:
“A4.2. Quality Control of WDC Data
[54] A large database invariably contains errors of many types: timing, calibration, sign, transcription, omission, and misunderstanding. Information about the data (metadata) is sorely lacking, especially for older data in the classical WDC data format. Fortunately, the data themselves can often be analyzed to bring to light many errors and allow correction of much of the data. Unfortunately, it is difficult to propagate such corrections back into the publicly available databases.”
We have corrected [as best we could] obvious errors [e.g. data from November 31st and the like], and as we say:
“[56] A few other examples further illustrate problems of data quality. The data for VQS is given in local time rather than UT, and the data for the series POT-SED-NGK has the 7th through the 10th day of every January for many early years designated as ‘‘missing,’’ although the data is present in the yearbooks and in older versions of the electronic data. Apparently, some recent ‘‘cleanup’’ was attempted with unintended consequences. For years from 1900 through 1907, the WDCs at times have ‘‘Y2K’’ problems where there is confusion about 2000 –2007 and 1900– 1907, as the century has no unique designation within the ‘‘WDC Exchange Format.’’ Attempts by the WDCs to rectify these various problems have often led to introduction of other problems or to loss of data, and there is no standard
procedure for feedback from researchers to the WDCs”
We deposited the whole dataset in NASA’s Virtual Magnetospheric Observatory http://science.nasa.gov/heliophysics/heliophysics-data-centers/virtual-magnetospheric-observatory-vmo/ which unfortunately now seems defunct. Having foreseen such a development we also keep the whole dataset on my website at http://www.leif.org/research/data in a simple text format which can be read by anybody without special software. This data is available to all [I look forward to see the paper by ‘Roger’ where he reproduces our results from the openly available data] and we’ll gladly make available subsets by request [which we have done in a number of cases] if the user is only interested in a part of the data.
Sometimes our effort of data control bears fruit: http://www.ann-geophys.net/29/283/2011/angeo-29-283-2011.pdf
“The first indication of something unexpected in the early Eskdalemuir hourly values was from correspondence with
Leif Svalgaard in 2004 where he presented ratios of the newly defined Inter-Hour Variability indices for the UK observatories…”
The various ad-hom over here at Climate Audit directed at me personally are mild compared with what I normally am exposed to.
There is a simple system used in the engineering design induxtry. Especially in the design of life hazard systems/structures.
It is called a “retention”, it is a significant percentage of the contract(grant) value and is only released upon the supply(archive) of all data, manuals, calculations, references and third party reviews.
Works extremely well!
What a wonderful, effective filter CA is!
Inevitably dues have to be paid:-
Due process and due diligence.
To “McIntyre et al” I tug my forelock!
Ouch!
Steve – I referred consistently to Mrs Lonnie not to Lonnie.
The old correspondence you post on another thread may well refute something I didn’t allege.
This is all getting sterile so I shan’t post again on this. I would advise the faithful not to read too much into this (no paranoia here from me – they’ve already ascribed a number of positions to me which I don’t hold). I stand by everything I’ve written, with the exception of the posts about Leif Svalgaard which, while factually correct, gave an incomplete picture. I’m confident that, outside the world of CA-posters (who themselves represent a tiny subset of CA-readers), my complaints would not be considered remotely controversial.
Re: Roger (Jul 9 04:13),
Your complaints aren’t controversial here either.
It’s quite simple: you came here with a chip on your shoulder, and made non-factual complaints. You’ve been proven wrong, based on evidence and sources. The facts trump any opinion or attempt at stirring up controversy. (Bottom line: contrary to your assertions, Steve did do his homework; even freely-shared data is not equivalent to archiving; you’re the one making personal attacks, not Steve.)
And now you appear to be imagining that your false accusations can be blithely sidestepped by “standing by” what you wrote (even though proven false) while insinuating that people here are simply the “faithful.”
Roger, I’m sad to say that I can’t see that you’ve added anything helpful or thoughtful, let alone accurate, to the discourse. Maybe you’ll do better next time.
Roger, what you wrote about me is not ‘factually correct’. I never ‘refused’ to archive data, nor do I have an ‘attitude’ towards non-archival. On the contrary, I’m fighting for archival: “We are working on a resolution through IAGA (EWC is the Chair of the IAGA Interdivisional Commission on History) for the next IUGG Scientific Assembly to address the archival and quality control of the extremely valuable early (18th and 19th century) geomagnetic data”.
It is one thing for you to have made mistakes. It is a lot worse not to admit it. Man up and apologize.
Leif, the irony is rich. Roger does toward you what he had falsely accused Steve of doing toward Ellen Mosley-Thompson. Now Roger seems unable or unwilling to acknowledge and apologize. I thought that Roger had so much concern for unjustified personal attacks, but that concern has proved highly selective.
Lots of comments on Thomson non arching over at WUWT but it seems they have archived stuff with NOAA, Any comments on current situation with data archiving or non archiving by Thomson?
Steve: I am completely familiar with the NOAA archive. The archive is incomplete. Pointing to it is simply a trick.
9 Trackbacks
[…] https://climateaudit.org/2012/07/01/mrs-lonnie-serial-non-archiver/ […]
[…] (click to enlarge, source) […]
[…] on Climate Audit, there’s an interesting thread going on about Lonnie Thompson and his wife Ellen, who Steve […]
[…] Lonnie and Ellen, A Serial Non-Archiving Couple […]
[…] https://climateaudit.org/2012/07/01/lonnie-and-ellen-serial-non-archivers/#comment-341083 […]
[…] integral of the scientist’s contribution. Lonnie could make a huge and lasting impact by archiving his ice core data. Share this:EmailTwitterFacebookDiggStumbleUponPrintRedditLike this:LikeBe the first to like this. […]
[…] […]
[…] the Ohio State University , taken with a grain of salt since Dr. Thompson and his wife Ellen are serial non archivers of ice core data (even when asked for it), which prevents others from checking his […]
[…] this data prompted some discussion of the Thompsons as serial non-archivers, referring to my post here. It is worth commending Thompson for prompt archiving of the present data, but that does not refute […]