In today’s post, I’ll follow the affair through Keenan’s complaint to SUNY, which is documented in the Climategate emails and at Keenan’s website.
Keenan, unconvinced by Wang’s statement about the existence of the station meta-data histories in the face of the contrary statement in the contemporary technical report NDP039, filed a materials complaint with Nature in July 2007, seeking evidence of the station histories said to have been consulted in Jones et al 1990. Nature rejected the request on the basis that it was unreasonable to expect climate scientists to retain unique historical information and were untroubled by the issue of whether the station histories had actually been available to the authors in the first place:
given the 17 years that have elapsed since publication, and that these histories were (in 1990) apparently only available as hard copies from an institute that has long since moved location, the authors’ inability to supply you with the requested materials, while regrettable, is excusable. We wish you every success in tracking down these old documents.
In early August, Keenan filed a misconduct complaint against Wang at SUNY, placing the complaint at his website. This was followed up almost immediately by his submission of an article to E&E about the events for an issue being edited by Benny Peiser.
On August 29, Peiser sent Keenan’s article to Jones for peer review. Jones immediately forwarded the article to Wang, Karl, Mann and Trenberth, resulting in a flurry of correspondence, including considerable discussion about whether Keenan had violated a supposed “confidentiality agreement” with SUNY – an issue that later became important in SUNY withholding report information from Keenan:
Trenberth, who had previously suggested that Jones label critics as “lazy and incompetent” and that he “cast aspersions on their motives and throw in some counter rhetoric”, now suggested a not uncommon tactic in climate science to Jones:
“You could try a patronizing approach of over explaining the difficulties.
Wang replied to Jones from China the same day (Aug 29) that he had forwarded the article to SUNY officials, who told him that Keenan had violated confidentiality:
and i have forwarded the file to the vp research and she wrote back to me that keenan has violetted the confidentiality, as i have told her in the very beginning 812. 1188508827.txt
Jones asked Wang (Aug 30) for clarification about Keenan’s supposed breach of confidentiality:
Can you clarify what you mean by violated confidentiality? I presume you mean that Keenan agreed to do nothing on the issue until the SUNY process has run its course. I presume this will conclude sometime this autumn 1188508827.txt
Wang replied (Aug 30):
the confidentiality means that keenan needs to keep the “inquery” confidential during the process of sunya “inquery”.
Jones passed the news about the SUNY confidentiality agreement to Mann and Treberth (Aug 30):
He got a reply to say that Keenan has now violated the confidentiality agreement related to the allegation. 1188478901.txt
Mann forwarded the article to Gavin Schmidt and they jointly reverted to Jones (Aug 30):
First, if there are factual errors (other than the fraud allegation) it is very important that you point them out now. If not, Keenan could later allege that he made the claims in good faith, as he provided you an opportunity to respond and you did now. Secondly, we think you need to also focus on the legal implications. In particular, you should mention that the publisher of a libel is also liable for damages – that might make Sonja B-C be a little wary. Of course, if it does get published, maybe the resulting settlement would shut down E&E and Benny and Sonja all together! We can only hope, anyway. So maybe in an odd way its actually win-win for us, not them. Lets see how this plays out… RealClimate is of course always available to you as an outlet, if it seems an appropriate venue. But we should be careful not to jump the gun here.
Later on Aug 30, Jones reverted to Wang, telling him that “libel is quite easy to prove in the UK” and wondering whether he should inform Peiser that Keenan had “broken his agreement” with SUNY:
1. Libel is quite easy to prove in the UK as you’re not a public figure. Perhaps when you’re back you ought to consider taking some legal advice from SUNY. Assuming the paper is published that is.
2. More important. I think I should send a short email to the editor Peiser and inform him that Keenan has broken his agreement with SUNY over this issue. If I don’t, they could say I had the chance and didn’t. Can you check with SUNY whether the folks there think I should? I just don’t want to do anything that later could be construed as the wrong thing now. I could also point out some factual errors.
Wang replied to Jones (cc Karl) (incorrectly) speculating that Keenan was “in a panic” because the IPCC report was “coming out soon”, suggesting how Jones should respond to the E&E review. This was a wild misunderstanding of motives – however, it shows the all-too-prevalent mentality in the field that sought to reduce any sort of inquiry or dispute to IPCC current events. He added:
I have also asked SUNYA’s opinion about what you should do within the SUNYA framework. But be careful that you do not know much about SUNYA action.
On August 31, Jones expanded the circle to include Tom Wigley, telling Wigley that Keenan was about to be notified by SUNY of his violation of a confidentiality agreement:
Keenan’s about to be told by SUNY that submitting this has violated a confidentiality agreement he entered into with SUNY when he sent the complaint. WCW has nothing to worry about, but it still unsettling! 813. 1188557698.txt
Wigley, who had been director of CRU in 1990, expressed concern that the claims to have examined station history were untrue and that this was a real issue – even if it later proved not to make a difference – advice that Jones ignored.
Seems to me that Keenan has a valid point. The statements in the papers that he quotes seem to be incorrect statements, and that someone (WCW at the very least) must have known at the time that they were incorrect. Whether or not this makes a difference is not the issue here.
In fact, as at August 30, Keenan had not heard back from SUNY, let alone entered into a confidentiality agreement with them. On August 31, Keenan heard back from SUNY for the first time. Contrary to Jones’ expectation, the SUNY letter did not notify Keenan that he had “violated a confidentiality agreement” nor make any reference to Keenan’s actions to that date (with which we now they were aware). University official Lynn Videlka sent Keenan a letter advising Keenan that the university had opened an inquiry on his allegation, including a copy of the university procedures on misconduct complaints, one of which was the following:
Responsibilities: The complainant is responsible for making allegations in good faith, maintaining confidentiality, and cooperating fully with an inquiry and/or investigation.
The covering letter drew Keenan’s attention to the requirement of confidentiality.
The attached letter informs you of the status of the University at Albany’s response to your earlier communication. It also apprises you of mandatory policy and procedural requirements including the requirement of confidentiality.
Please contact me if you have any questions about this matter.
Around Sep 4, as Mann and Schmidt had suggested, Jones submitted a short review of Keenan’s E&E submission. This time, Jones stated Zeng, not a coauthor of Jones et al 1990, had selected the stations, re-iterating Wang’s point that Zeng had had access to the station histories when the article was written:
Two networks (one of 60 and another of 205) were developed around 1990. The 60-station network contained data for 12 meteorological variables and information on the station histories, but the 205-station network contained mean temperatures and precipitation totals only, without station histories. This was because of a lack of resources at the time. The 42-station pairs used in the two 1990 papers were selected by Professor Zeng (who was a co-author on Wang et al., 1990) from the 60 and 205 station networks. In making her decision she did have access to the station histories and the site population values.
In his response, Keenan re-iterated one more time the explicit statement from NDP039 that “station histories are not currently available”:
In particular, the 1991 report (and the 1997 revision) explicitly states that for 49 of the stations claimed to be studied by Jones et al. and Wang et al. “station histories are not currently available”.
In his reply to Keenan’s response, Jones responded that NDP039 (Tao et al 1991) did not “explicitly” address the 49 stations in Jones et al 1990 – a criticism that was irrelevant since the 49 stations had already been determined to be within the 205-station network of NDP039. Jones re-iterated his statement that Zeng had the station histories, explaining that they lacked resources in 1989-90 to digitize the station histories;
Attached is Tao et al (1991). Nowhere in it does it explicitly state for 49 of the stations claimed to be studied by Jones et al. and Wang et al. are ‘station histories not currently available. It says this for the 205. I’m attaching Tao et al.. It is a scanned pdf, so the find/search facility won’t work. Zeng had the station histories for the 84 sites we used. They didn’t have adequate resources in the 1989-90 period to digitise everything. Keenan has been told this.
In this case, Jones did not control the journal and Keenan’s article was published.
Jones contacted Mann and Gavin Schmidt again in mid September 2007. Mann said that Wang needed to sue or threaten to sue and that he had a lawyer who would do so pro bono:
Wei Chyung needs to sue them, or at the least threaten a lawsuit. If he doesn’t, this will set a dangerous new precedent. I could put him in touch w/ an leading attorney who would do this pro bono. Of course, this has to be done quickly. The threat of a lawsuit alone my prevent them from publishing this paper, so time is of the essence.
Jones also forwarded the article to UEA officials. Michael McGarvie questioned the propriety of Keenan publishing an article while the SUNY inquiry was underway and thought that FOI officer Palmer should be contacted about the FOI claims in the article. Jones also notified new Head of School Jacquie Burgess, who commiserated that this almost amounted to “harassment”,
At SUNY, the preliminary “inquiry” stage commenced. It interviewed Keenan by telephone see here on Dec 7, 2007. It issued its report on Feb 18, 2008 (see Climategate documents here, with a redacted version being sent to Keenan on Feb 20. It observed that Wang told them that the station logs “existed in paper form” and had been used by his “colleague” (Zeng) to select stations for use in the paper, that the Committee should contact Zeng to confirm the existence of the logs in station form and that “established procedures” had been used to adjust for some of the station moves e.g. altitude even though this wasn’t mentioned in the publications. The Inquiry Committee’s conclusion was that, without evidence about the paper station logs, it had no alternative other than to recommend an investigation.
After careful and thorough review of the evidence, the Committee has concluded that without the written input from the Respondent’s colleague, we cannot determine the accuracy of the Respondent’s explanation for the station selection. While the respondent maintains that there may be additional evidence available that could allow for a clear and final decision to be rendered, for this Inquiry Committee to examine this case further or in more depth would be beyond its charge. Consequently in the absence of any such available evidence regarding the paper station logs, there is simply no way for us as an Inquiry Committee ot conclude that there is not sufficient evidence to warrant further investigation.
On Feb 22, Wang prepared a submission to the investigation, with an appendix said to have come from Zeng (it turned up in the Climategate documents and was never sent to Keenan for comment.) Wang said that the station histories had not been included in the DOE report because of the “huge effort” required to digitize the station histories:
While the station history was included in the 1991 DOE report for the 60‐station network, they were not included in the 205‐station network (published in 1993 DOE report which Professor Zeng is also a co‐author) due partly to its not being a requirement under the US‐China agreement, and partly to the huge effort coupled with inadequate resources (manpower and hardware) to digitize them.
That digitization of meta-data for the stations was not a “huge” commitment of digital resources, even in 1990, can be seen by the station histories for the 60-station network, published as Appendix B of NDP039.
Wang said once again that the information had been available at the time, but was “no longer available due to several office moves”:
the 49‐stations (most rural stations) are based on her [Zeng’s] recollection (together with checking against the present‐day station location), simply because the original station history manuscripts (archived at IAP) and her detailed notes were no longer available due to several office moves over the almost 19‐years time span.
Zeng’s recollection of the details of station moves in the appendix was oddly precise despite the passage of 18 years. Her recollection as documented in the Wang submission was also frequently inaccurate in cases where her recollection could be checked against the documentation preserved in NDP039.
The next time that the curtain is drawn on this is on May 23, 2008, when SUNY notified Keenan that the investigation had concluded that there was no misconduct, giving Keenan 14 days to add to the record, while refusing to give Keenan a copy of the report that he was asked to comment on.
After careful review of the evidence and thoughtful deliberation, the Investigation Committee finds no evidence of the alleged fabrication of results and nothing that rises to the level of research misconduct having been committed by Dr Wang.
As the institutional official responsible for this case, I have accepted the Committee’s findings and the Report. You have fourteen (14) calendar days from the date of this letter to provide any comments to add to the report for the record.
On June 4, 2008, SUNY “explained” to Keenan that he
did not receive a copy of the Investigation report because the report did not include portions addressing your role and opinions in the investigation phase.
Two days later (June 6, 2008), Keenan objected, describing the request for comments without access to the report as “Kafkaesque”. Keenan also objected to procedural breaches by the inquiry, including the university’s failure to notify him that an investigation had commenced. Keenan also noted the investigation’s breach of their obligation to interview “all individuals involved in making the allegation”:
The investigation process will include, but not necessarily be limited to, examination of pertinent research data and written materials, interviews with all individuals involved either in making the allegation or against whom the allegation is made, and statements from or interviews with other individuals who might have information regarding the allegation.
On June 25, 2008, without replying to Keenan, the University president sent his final decision to Wang (not sent to Keenan until August 11, 2008)
After studying the report of the Investigation Committee assessing an allegation of data fabrication levied against you and upon further weighing the recommendation of the Vice President for Research, I concur that there is no evidence whatsoever that you have committed data fabrication or any research misconduct with respect to this allegation.
On August 11, 2008 (six weeks later) , SUNY notified Keenan of the SUNY final determination, warning Keenan:
The University’s misconduct policies and the Office of Research Integrity regulations preclude discussion of any information pertaining to this case with others who were not directly involved in the investigation.
There matters rested for a while. The following year, on March 18, 2009, Keenan sent an FOI request to SUNY for the reports of the Inquiry and Investigation Committees. The university turned down the FOI request on April 14, 2009 as an “unwarranted invasion of personal privacy”.
The affair was mentioned from time to time in misconduct literature. In May 2009, Aubrey Blumensohn made a fairly detailed report, strongly criticizing procedures at SUNY:
In the absence of any explanation to the contrary, it seems that the methodology for station selection as described in these two publications was false or at best grossly misleading.
Wang maintains that hard copy records do exist detailing the location of stations selected by himself outwith the published methodology. However the refusal to clarify “method” is inappropriate and a form of misconduct in and of itself. It does not lend credence to Wang’s assertion that fraud did not take place. It would also be necessary to see records of stations that were not selected, in order to confirm that selection was indeed random, and only “on the basis of station history”.
The University at Albany is in a difficult position.
If the University received such records as part of the supposed misconduct investigation, then they could easily resolve the problem by making them available to the scientific community and to readers. If the University does not have such records then they have been complicit in misconduct and in coverup of misconduct. If the University at Albany does have such records, but such records are not in accordance with the stated methodology of the publications, then the University has more serious difficulties.
“Investigations” of scientific misconduct should themselves align with the usual principles of scientific discourse (open discussion, honesty, transparency of method, public disclosure of evidence, open public analysis and public discussion and reasoning underlying any conclusion). This was not the case at the University at Albany. When you see universities reluctant to investigate things properly, it provides reasonable evidence that they really don’t want to investigate things properly.
Peiser distributed Blumensohn’s article on CCNET, reaching Wigley’s attention, who sent a lengthy and thoughtful email on May 4, 2009 to Jones and Santer (their reply is not in the Climategate emails)”
Do you know where this stands? The key things from the Peiser items are …
“Wang had been claiming the existence of such exonerating documents for nearly a year, but he has not been able to produce them. Additionally, there was a report published in 1991 (with a second version in 1997) explicitly stating that no such documents exist. Moreover, the report was published as part of the Department of Energy Carbon Dioxide Research Program, and Wang was the Chief Scientist of that program.”
“Wang had a co-worker in Britain. In Britain, the Freedom of Information Act requires that data from publicly-funded research be made available. I was able to get the data by requiring Wang�s co-worker to release it, under British law. It was only then that I was able to confirm that Wang had committed fraud.”
You are the co-worker, so you must have done something like provide Keenan with the DOE report that shows that there are no station records for 49 of the 84 stations. I presume Keenan therefore thinks that it was not possible to select stations on the basis of …
“… station histories: selected stations have relatively few, if any, changes in instrumentation, location, or observation times”
[THIS IS ITEM “X”]
Of course, if the only stations used were ones from the 35 stations that *did* have station histories, then all could be OK. However, if some of the stations used were from the remaining 49, then the above selection method could not have been applied (but see below) – unless there are other “hard copy” station history data not in the DOE report (but in China) that were used. From what Wang has said, if what he says is true, the second possibility appears to be the case.
What is the answer here?
The next puzzle is why Wei-Chyung didn’t make the hard copy information available. Either it does not exist, or he thought it was too much trouble to access and copy. My guess is that it does not exist — if it did then why was it not in the DOE report? In support of this, it seems that there are other papers from 1991 and 1997 that show that the data do not exist. What are these papers? Do they really show this?
Now my views. (1) I have always thought W-C W was a rather sloppy scientist. I therefore would not be surprised if he screwed up here. But ITEM X is in both the W-C W and Jones et al. papers — so where does it come from first? Were you taking W-C W on trust?
(2) It also seems to me that the University at Albany has screwed up. To accept a complaint from Keenan and not refer directly to the complaint and the complainant in its report really is asking for trouble.
(3) At the very start it seems this could have been easily dispatched. ITEM X really should have been …
“Where possible, stations were chosen on the basis of station histories and/or local knowledge: selected stations have relatively few, if any, changes in instrumentation, location, or observation times”
Of course the real get out is the final “or”. A station could be selected if either it had relatively few “changes in instrumentation” OR “changes in location” OR “changes in observation times”. Not all three, simply any one of the three. One could argue about the science here — it would be better to have all three — but this is not what the statement says.
Why, why, why did you and W-C W not simply say this right at the start? Perhaps it’s not too late?
I realise that Keenan is just a trouble maker and out to waste time, so I apologize for continuing to waste your time on this, Phil. However, I *am* concerned because all this happened under my watch as Director of CRU and, although this is unlikely, the buck eventually should stop with me.
P.S. I am copying this to Ben. Seeing other peoples’ troubles might make him happier about his own parallel experiences.