Another UK climate data withholding scandal is emerging

As many WUWT readers know, Steve McIntyre’s tireless quest to get the raw data that makes up the gridded Hadley Climate Research Unit HadCRUT dataset has been fraught with delays, FOI denials, and obvious obfuscation. In some cases the “dog ate my homework” is the excuse. The UK Register has an excellent summary of the issue.

A similar issue has been brewing in parallel over tree ring data in the UK. Doug Keenan tells us the story of getting the “ring around” for over 2 years trying to obtain what many would consider a simple and non controversial data request. – Anthony

Guest Post by Doug Keenan

Queen’s University Belfast is a public body in the United Kingdom. As such, it is required to make certain information available under the UK Freedom of Information Act. The university holds some information about tree rings (which is important in climate studies and in archaeology). Following discusses my attempt to obtain that information, using the Act.

When a tree is cut, you can often see many concentric rings. Typically, there is one ring for each year during which the tree grew. Some rings will be thick: those indicate years in which the environment was good for the tree. Other rings will be thin: those indicate the opposite.

Scientists study tree rings for two main purposes. One purpose is to learn something about what the climate was like many years ago. For instance, if many trees in a region had thick rings in some particular years, then climatic conditions in those years were presumably good (e.g. warm and with lots of rain); tree rings have been used in this way to learn about the climate centuries ago. The other purpose in studying tree rings is to date artefacts found in archaeological contexts; for an example, see here.

Tree-ring data from Northern Ireland

One of the world’s leading centers for tree-ring work is at Queen’s University Belfast (QUB), in Northern Ireland. The tree-ring data that QUB has gathered is valuable for studying the global climate during the past 7000 years: for a brief explanation of this, see here.

Most of the tree-ring data held by QUB was gathered decades ago; yet it has never been published. There is a standard place on the internet to publish such data: the International Tree-Ring Data Bank (ITRDB), which currently holds tree-ring data from over 1500 sites around the world. QUB refuses to publish or otherwise release most of its data, though. So I have tried to obtain the data by applying under the UK Freedom of Information Act (FoI Act).

I have submitted three separate requests for the data. Each request described the data in a different way, in an attempt to avoid nit-picking objections. All three requests were for the data in electronic form, e.g. placed on the internet or sent as an e-mail attachment. The first request was submitted in April 2007.

QUB refused the first request in May 2007. I appealed the refusal to a Pro-Vice-Chancellor of QUB, who rejected the appeal. The primary reason that the Pro-Vice-Chancellor gave for rejection was that some of the data was in paper form and had not been converted to electronic form. The Pro-Vice-Chancellor additionally claimed that after data was converted to electronic form, “It is then uploaded to the International Tree Ring Data Base”. There might indeed be some small portion of the data that is not in electronic form. My request, though, was for a copy of the data that is in electronic form. So, is all data that is in electronic form available at the ITRDB, as the Pro-Vice-Chancellor claimed?

QUB has in the past published the results of various analyses of its tree-ring data (most notably its claim to have sequences of overlapping tree rings extending back in time many millennia). In doing the analyses, the sequences of tree-ring data are analyzed statistically, and the statistical computations are done by computer. This is well known, and moreover has been stated by QUB’s former head tree-ring researcher, Michael G.L. Baillie, in several his publications. (Indeed, Baillie and his colleague Jon R. Pilcher, also at QUB, wrote a widely-used computer program for tree-ring matching, CROS.) Obviously the data that was used for those computations is in electronic form—and it has not been uploaded to the ITRDB. Thus the claim by the Pro-Vice-Chancellor is untrue.

The Pro-Vice-Chancellor further claimed that to organize the data in “the very precise categories which [I] have specified” [in my request] would entail a vast amount of work. My request, though, was merely for the tree-ring data that had been obtained and used by the university; that hardly seems like precise categorization. Moreover, I later submitted a second request for “the data about tree rings that has been obtained by [QUB] and that is held in electronic form by the university”. That request was also refused. And a third request that was very similar to the second was refused. All three requests were refused in whole, even though the university is required to make partial fulfillment when that is practicable.

The UK Information Commissioner’s Office

After half a year of trying to obtain the information from QUB, I appealed to the Information Commissioner’s Office (ICO). The ICO is charged with ensuring that the FoI Act is enforced. My appeal to the ICO was submitted on 24 October 2007. The ICO notified me that an officer had been assigned to begin investigating my case on 14 October 2008. Such a long delay is clearly incompatible with effective working of the Act.

The ICO then contacted QUB, asking for further information. QUB then admitted that almost all the data was stored in electronic form. Thus QUB implicitly admitted that its prior claims were untruthful.

QUB now asserted, however, that the data was on 150 separate disks and that it would take 100 hours to copy those disks. (These were floppy disks—the type that slide into computers and, prior to the internet, were commonly used to carry electronic data.) It takes only a minute or two to copy a floppy disk, however; so the claim of 100 hours to copy 150 floppy disks is an unrealistic exaggeration.

QUB also said that it considered photocopying a printed version of the data, but that this would take over 1800 hours. As noted above, all my requests were for data that is in electronic form; moreover, I have repeated this point in subsequent correspondences with QUB. The statement from QUB about photocopying is thus not relevant.

On 22 December 2008, the ICO sent me a letter rejecting my appeal, on the grounds that the time needed by QUB would exceed an “appropriate limit” (as stipulated in the FoI Act). The ICO had accepted QUB’s explanation for refusing to release the data without question, and without discussing the explanation with me. I telephoned the ICO to raise some objections. To each objection that I raised, the ICO case officer gave the same reply: “I’m satisfied with their [QUB’s] explanation”.

I also offered to visit QUB with the case officer, to demonstrate how quickly the data could be copied (e.g. from floppy disks), and to copy the data myself. This seemed particularly appropriate because the officer had told me when she started on the case that she would visit QUB as a standard part of investigation, yet she had not made such a visit. The officer, though, declined my offer, again saying that she was satisfied with QUB’s explanation.

There is a mechanism to appeal an ICO decision, to a tribunal. I told the case officer that I wanted to do so. The officer replied that, in order to file an appeal, I would need a formal Decision Notice from the ICO. I requested a Decision Notice. The officer then informed me that the ICO would send a Notice, but that, because they were busy, it would take about two years to do so.

Environmental Information Regulations

I discussed the above with a colleague, David Holland. Holland said that my request should not have been processed under the FoI Act. His reasoning was that the information I was requesting was about the environment: environmental information is exempt from the FoI Act and requests for such information should instead be processed under the Environmental Information Regulations (EIR). He pointed out that the tree-ring data clearly fits the definition of “environmental information” given in the EIR. It also clearly fits the common (dictionary) definition.

I had been aware that the EIR existed, but had assumed that the EIR was essentially the same as the FoI Act. After the discussion with Holland, though, I checked and found that there is one major difference between the EIR and the FoI Act: under the EIR, there is no limit on the amount of time that a public institution requires to process a request. In other words, even if QUB’s original claim that some of the data was only available on paper were true, or even if QUB’s revised claim that copying data from disks would take 100 hours were true, that would still not be a valid reason for refusing to supply the information.

I am not an expert in how to apply the EIR or FoI Act, though. So I telephoned the ICO headquarters to ask for guidance. There I spoke with a Customer Service Advisor, Mike Chamberlain. Chamberlain told me the following: that the information seemed obviously environmental; that there was no limit on processing time that could be used to refuse a request for environmental information; that I could freely visit a site where environmental information was held in order to examine the information; and that it was the duty of the public authority (i.e. QUB) to determine whether the EIR or the FoI Act was applicable. Chamberlain also confirmed everything that he told me with someone more senior at the ICO.

It is regrettable that I had not realized the above earlier. My initial request to QUB, in April 2007, had stated the following.

It might be that this request is exempt from the FOIAct, because the data being requested is environmental information. If you believe that to be so, process my request under the Environmental Information Regulations.

QUB, however, had not processed my application correctly. I should have caught that.

There is another issue. I had described the information to the ICO case officer by telephone and also by e-mail (on 24 November 2008). Hence the case officer must have known that the information was environmental, and thus exempt under the FoI Act and only requestable under the EIR. Why did the ICO not act on that? On 29 January 2009, I e-mailed the case officer, citing the above-quoted statement from my request to QUB and saying “I would like to know the reasoning that led to my request being processed under the Freedom of Information Act, instead of EIR”. Initially, there was no reply.

The EIR was enacted pursuant to the Aarhus Convention, an international treaty on environmental information that the UK promoted, signed, and ratified. Failure to implement the EIR would constitute a failure by the UK to adhere to the Convention. So, a few weeks after e-mailing my question to the ICO, and with no reply, I contacted the Aarhus Convention Secretariat (ACS), at the United Nations in Geneva. The ACS has a mechanism whereby individuals can file a complaint against a country for breaching the Convention. I had an initial discussion with the ACS about this. That turned out to be unnecessary though. The Assistant Information Commissioner for Northern Ireland contacted me, on 10 March 2009: he was now handling my case and, moreover, he had visited QUB and seen some of the data.

On 22 April 2009, I received a telephone call from the Assistant Information Commissioner for Northern Ireland. The Assistant Commissioner said that he was preparing a Decision Notice for the case, and he made it clear that the Notice would hold that the data should be released under the EIR. The next I heard anything was on 13 July 2009, when it was announced that the Assistant Commissioner had been suspended. On 13 August 2009, I telephoned the ICO: I was told that a new officer would be assigned to the case within the next few days and that a draft Notice, which had been written by the Assistant Commissioner, was in the signatory process. I am presently awaiting further word.

Another example—Gothenburg University

I have previously been involved with obtaining tree-ring data from another institution: Gothenburg University, in Sweden. Sweden has a law that is similar to the UK’s Freedom of Information Act (the Swedish law is the Principle of Public Access). In 2004, Swedish courts ruled that the law applied to research data held by universities. In a famous case known as the “Gillberg affair”, a researcher at Gothenburg University refused to obey the law. As a result, both the researcher and the rector of the university were convicted of criminal malfeasance. (The researcher received a suspended sentence and a fine; the rector received a fine.)

Gothenburg University does substantial tree-ring research. On 10 April 2007, I requested their tree-ring data. The university’s lead tree-ring researcher repeatedly resisted, claiming that it would take weeks of his time, and that he was too busy to do it. On 22 April 2008, I sent a letter to the (new) rector of Gothenburg University, saying that if the data was not supplied, I would file complaints with both the Court and the Parliamentary Ombudsmen of Sweden. The next day, all the data was submitted to the ITRDB.

What transpired with Gothenburg University exemplifies the importance of laws on Freedom of Information for tree-ring data.

Motivations for withholding data

Some people have asked why QUB does not want to release the data. In fact, most tree-ring laboratories do not make their data available: it is not just QUB and Gothenburg that have been reluctant. The reason for this was elucidated by Peter M. Brown, in April 2007. At the time, Brown was president of the Tree-Ring Society, which is the main international organization for tree-ring researchers. Following is an excerpt (the full explanation is here).

… they ARE my data. Funding agencies pay me for my expertise, my imagination, and my insights to be able to make some advance in our understanding of how nature works, not for raw data sets. … It is the understanding and inferences supplied by the scientist that funding agencies are interested in, not her or his raw data.

In other words, even if the research and the researcher’s salary are fully paid for by the public—as is the case at QUB—the researcher still regards the data as his or her personal property.

There are only a few tree-ring laboratories where attitudes are different. One example is the University of St Andrews, in the UK. Almost all tree-ring data held by St Andrews is freely available in the ITRDB.

It is notable that QUB continues to withhold its data even though, in 2009, the tree-ring laboratory at QUB was effectively closed. The closure was primarily due to the lab lacking funds, which presumably resulted from having almost no research publications (i.e. the lab had not been producing anything; so funding agencies declined to support it). The dearth of publications occurred even though the lab has some extremely valuable data on what is arguably the world’s most important scientific topic—global warming (as outlined here). This problem arises because the QUB researchers do not have expertise to analyze the data themselves and they do not want to share their data with other researchers who do.