by Gavin and Mike
The long-awaited and surprisingly thorough Muir Russell report (readable online version) was released this morning. We’ve had a brief read through of the report, but a thorough analysis of this and the supplemental information on the web site will have to wait for a day or so.
The main issue is that they conclude that the rigour and honesty of the CRU scientists is not in doubt. For anyone who knows Phil Jones and his colleagues this comes as no surprise, and we are very pleased to have this proclaimed so vigorously. Secondly, they conclude that none of the emails cast doubt on the integrity and conclusions of the IPCC, again, something we have been saying since the beginning. They also conclude as we did that there was no ‘corruption’ of the peer-review process. Interestingly, they independently analysed the public domain temperature data themselves to ascertain whether the could validate the CRU record. They managed this in two days, somewhat undermining claims that the CRU temperature data was somehow manipulated inappropriately. (Note that this exercise has been undertaken by a number of people since November – all of which show that the CRU results are robust).
All in all, none of the various accusations and insinuations that have been floating around the blogosphere have been sustained. (See some of the early media coverage of the report).
However, there are two issues that have come up that deserve some comment. The first are the evolving practices of data presentation and access, and the second is the issue of how to handle Freedom of Information (FOI) requests.
As climate science has moved away from single researcher/single study/single site analyses towards synthesis across multiple studies, across the globe and involving more and more researchers, practices that were appropriate at one time don’t necessarily scale up to the new environment. Data requests dealt with on an ad-hoc and informal basis work fine if only a couple of people are interested, but more formal and automated procedures are needed when the data sets grow and many more people are involved (see the PCMDI/CMIP3 archive of model results for instance). Given too, the obsession in certain quarters with irrelevant details of smoothing techniques and end-point padding in decade-old papers, it is clear that the more information that is put out as supplementary material to the creation of high-profile figures, the better off we will be. Examples of this for figures in IPCC AR4 already exist, but it will be helpful for IPCC to adopt this practice more generally. Historically, this hasn’t been done – mainly because no-one thought it particularly interesting (most smoothing methods produce very similar results for instance), particularly for figures that weren’t for publication in the technical literature.
One example of this was the cover art on a WMO 1999 report which, until last November, was completely obscure (we are not aware of any mention of this report or this figure before November in any blogospheric discussion, ever). Nonetheless, in the way of these things, this figure is now described as ‘an icon’ in the Muir Russell report (one of their very few mistakes, how can something be an icon if no-one has ever seen it?). In retrospect (and as we stated last year) we agree with the Muir Russell report that the caption and description of the figure could indeed have been clearer, particularly with regard to the way proxy and instrumental data sources were spliced into a single curve, without indicating which was which. The WMO cover figure appears (at least to our knowledge) to be the only instance where that was done. Moving forward, nonetheless, it is advisable that scientists be as clear as possible about what sorts of procedures have gone into the preparation of a figure. But retrospective applications of evolving standards are neither fair nor useful.
With respect to the continuing barrage of FOI requests (which are predominantly for personal communications rather than for data), we can attest from personal experience how disconcerting these can be at first. Since there are no limits on what can be asked for (though there are many limits on what will be delivered), scientists presented with these requests often find them personally invasive and inappropriate. Institutions that do not have much experience with these kinds of requests, and who are not aware of what their employees do that is, and is not, covered by the legislation, are often not much help in sorting out how to respond. This can certainly be improved, as can the awareness of the community of what is recoverable using these procedures. While it is not relevant to the legislation, nor to what can be released, the obvious bad faith of many of the requesters indicates that actual information about the functioning of public bodies is not the primary goal in making these requests. However, it would be a terrible mistake for scientists to retreat from the public discussion on climate science because of these attempts at intimidation.
We will post on more specific aspects of the report, and perhaps the legacy of the whole affair over the next few days…