Can the Internet help climate scientists? Not everyone thinks so.
"The Internet is a double-edged sword," Met Office scientist Peter Stott told a London courtroom last week. "There's a whole cacophony of voices on blogs, people with different opinions and people very motivated to dig around. But not in the 'big picture' details, frankly. That is not helpful to getting an overall balanced assessment."
Stott had just been asked whether widespread online participation in the UN's Intergovernmental Panel on Climate Change assessment-of-the-science process might improve it. The open source software development principle, that "given enough eyeballs, all bugs are shallow", came to mind.
The occasion was an Information Tribunal appeal brought by one-man information Inquisition David Holland. The retired Mancunian engineer's previous enquiries were seen by many as the catalyst for the famous "Climategate" email leaks.
"My interest in this was never to do with climate. I'm trained as an engineer, and I know the scientific method," Holland told El Reg in 2011, when he had sought access to large amounts of information from the British climate-science establishment - and was denied. Holland's FOI requests set off a catastrophic sequence of prevarication and obstruction by the responding scientists, which ultimately appears to have triggered the Climategate leaks and massive discomfort for all the researchers involved.
Now it's the turn of Peter Stott of the Met Office to come under Holland's microscope.
I actually felt a bit of human sympathy for Stott; you can bet he would have rather been somewhere else, and it transpires that Holland didn't actually want him there at all. Holland had wanted to cross-examine the head of the UK delegation to the IPCC, a Department of Environment and Climate Change official called David Warrilow, head of climate science and international evidence.
The procedural questions under the spotlight are Warrilow's bailiwick, not Stott's, but Holland was refused his man. Stott, we learned, had been pressganged into appearing by the Met Office's lawyers. Stott also had to defend his and allied organisations' refusal to disclose material on a basis - as we shall see - that's highly questionable. No intelligent person should have to waste his own time, or anyone else's time, defending the indefensible.
And the mere presence of a Met scientist is a bit of a red herring, as it's really the IPCC that is on trial; the case for the defence is being organised by the Treasury solicitor, paid for by you.
Judge Anisa Dhanji was not impressed by the defence's refusal to find someone so very germane to the case to stand up to cross-examination, and demanded that a written statement by Warrilow be included in the record.
So. Here we all were. Why was this happening, exactly?
The case for transparency
The IPCC is the United Nations organisation's process for providing climate advice to policy makers. Every few years it updates this advice, which takes the form of three gigantic reports: one assessing the physical basis (called Working Group 1, or WG1); another considering the impacts (WG2); and the third the mitigation options (WG3).
None of the groups does any original science. They're supposed to write fair summaries of the state of the science - although you'll notice that WG3 already plays with loaded dice: it is about "mitigation", not "adaptation" nor "economic costs" or even "low carbon technological innovation". Each of these groups writes three drafts in a rolling process before they're signed off as official UN policy.
What Holland is seeking the "zero order draft" – aka Draft No. 1 - of Working Group 1. He couldn't care less what's in it, but wants to establish the principle that citizens can see it. As it happens, much of this material is already all over the web. But as Judge Dhanji pointed out, that's by-the-by.
Holland justified his request on the twofold basis that the WG1 zero draft must, and should, be publicly available.
The "must" is the statutory obligation of the UK as a signatory to the Aarhus Convention, or to give it its full title, the "UNECE Convention on Access to Information, Public Participation in Decision-making and Access to Justice in Environmental Matters" and by signing up to EC directive 2003/4/EC (PDF) the UK has obliged all of its environmental regulations to be consistent with the Convention. (Not every EU country is a signatory, but the UK is: perhaps a case of civil servants' Euro-enthusiasm coming back to bite them.)
The Met Office argued back that the ZOD is most precious, and if ZOD material were to be disclosed, counsel argued, there would be grave damage to the UK's "international relations". Like the Sneetches, British environment bureaucrats would be stranded on the beaches: shunned, ostracised, and cast into outer darkness. Nobody would want to play with British climate scientists any more.
Yet this particular reason for refusing ZOD is a relatively recent one. It came about in the aftermath of Climategate, and a succession of other climate-related -gates, after which the IPCC's processes came under much-expanded scrutiny. The Inter Academy Council, a panel of dons, made recommendations, amongst others.
The international-relations grounds for climate secrecy first surfaced in 2009 and was used as a reason to knock back FOI requests, from various people including Holland himself. Its origins appear to lie in a request from University of East Anglia climate scientist Tim Osborne in 2009 to IPCC WG1 co-chair Thomas Stocker, which you can read here. Osborne asked Stocker:
UEA must provide evidence to support its reasons for rejecting the original request. I've been asked if we can obtain a further statement about confidentiality ...
... it would be very supportive if someone who currently represents IPCC (or at least IPCC WG1) could indicate that this is also the view/position of the IPCC ...
There are four specific items that we would ideally like to have you view on:
(1) Does the IPCC WG1 expect authors to keep confidential the emails/correspondence/chapter text that they receive from fellow authors during the drafting process?
(2) Would there be an adverse effect on the IPCC WG1 if we were to break this confidentiality? (Note that we might be forced to break it *during* the drafting of the next report)
(3) Would there be an adverse effect on UEA's relationship with IPCC WG1?
(4) In providing views on items (1)-(3), are they your personal view or can we say that they represent IPCC WG1 position?
"Can we say that" is particularly revealing.
It's a common misconception - a sign of the media's deference to scientists, perhaps - that the IPCC consists of a properly appointed actual panel somewhere. As Stott cheerfully confirmed, beyond a small technical administrative support team called the TSU, there isn't really "an IPCC" at all. Self-selecting scientists kick off the assessment process, often gatekeeping material by their friends and colleagues, hard-green campaign groups etc. From the Second Draft stage of the review process the room is full of government officials - the international bureaucracy effectively takes over. The rules are set by the participants, making it up as they go along, Lord of the Flies-style. The Osborne-Stocker exchange illustrates a normal example of one scientist colluding with another in an attempt to prevent the public finding out how the process works.
Stott described in court the process which allows any member of the public to "self-certify" as an Expert Reviewer and join the process at the First Order Draft (ie, the second) stage. He pointed out that Comments generated at this stage were made public last time.
This is true, but bear in mind that the IPCC went to great lengths to avoid Draft Review Comments actually being read last time around - "making public" the comments by depositing an un-indexed paper archive in a library in Massachusetts. In the internet era, that's tantamount to "burying them in a hole in the ground", says Holland. Today, the Comments are indeed officially released on the actual interwebs - but too late to stop errors being spotted, Holland retorted. He cited Glaciergate as an avoidable error.
"On the internet, half of what I've asked for is already there," said Holland. "It's extremely boring stuff, and contains nothing confidential. I can see no reason for it not being released."
Holland's other argument, the "should", is an ethical one. He cited the former IPCC head John Houghton who wrote, "for assessments that inform public policy it is essential that things are seen to be done: the assessment should be completely open and transparent". He contrasted this with "indefinite confidentiality" implicit in refusing access to the ZOD.
Not to mention the small matter of compliance with Aarhus.
The IPCC's defence was twofold: the ZOD work needed to be kept from the public because the participants were learning the ropes, and because it offered a sanctity of private thought required for this vital education. Exposing the scientists to the public at this stage would be like mocking a Learner Driver for knocking over the traffic cones. This would be horribly cruel, defence argued.
"The Zero Order Draft is a training document we can take to a few people to tell us 'are we on the right lines here?' 'have we learned to write a document?' The lessons we learn we can then take on to produce the first First Order Draft to formal review," Stott said.
"We're learning how to write a report. All of us have to learn that."
"It's an interim internal document for ourselves to learn how to do this job ..."
Judge Dhanji wondered what the harm was?
"Well the issue can and does get misused and misrepresented as representative of the IPCC assessment. We've had that in the past with leaked drafts."
And again: "it's an internal document for our own internal use, we can expose our ignorance, expose our learning process, to learn how to write a document. That's the distinction between ZOD and the other drafts," he replied.
"Things that are said in the name of the IPCC during process may come back to haunt them in that people who disagree with the IPCC may scour material for inconsistency," he declared.
Defence counsel raised these concerns, which obliged Judge Dhanji to castigate him for coaching Stott.
"Too much leading," she warned.
Bear in mind that nothing in IPCC rules actually specifies that the ZOD stage is sacrosanct. Participants may prefer it, but rules is rules. Or not.
So Stott answered questions he couldn't, at one point offering his not-so-expert opinion on International Relations.
"It's a bit hypothetical, isn't this?" he replied at one point - surely the truest words spoken all day.
In contrast to the most recent climate-science-related FOI hearing I attended in October (Newbery vs the BBC), the judge was scrupulously exact with both parties, while the two lay judges actively engaged with very good questions, clearly intended to improve their understanding.
In Newbery's case, they asked no questions, and merely scowled. It later transpired that one of the lay judges was a former local government official whose huge redundancy payoff had only been exposed through FOI requests. The other was a former Labour councillor who had received money for campaigning on climate change - and whose views on "deniers" were well known.
It'll be interesting to see the Tribunal's decision in due course. ®