Our response Our response

Thu, 2 Sep 2010

University of East Anglia's Response

1. The University is indebted to Sir Muir Russell and his team for conducting a comprehensive, thoughtful and challenging Review into the allegations which have been made against the University's Climatic Research Unit (CRU) following the publication of a large number of emails and other material, which are believed to have been obtained illegally from a back-up server in CRU.

2. The Review was conducted over a period of seven months and looked for evidence of manipulation or suppression of data by scientists in CRU or of manipulation of the peer review process, and addressed issues relating to compliance with the Freedom of Information Act (FoIA) and the Environmental Information Regulations (EIR) for the release of data and correspondence, and the governance and security structures for CRU. 

3. The University welcomes Sir Muir's approach to the Review, which invited any party, evidently including those who had levelled allegations through the media and the "blogosphere", to make representations to it. The team was painstaking in its analysis and transparent in making all evidence it received, and the records of interviews conducted available on its website, other than where that would be defamatory or otherwise unlawful.

4. The University welcomes the findings that:

"On the specific allegations made against the behaviour of CRU scientists, we find that their rigour and honesty as scientists are not in doubt"

"We do not find that their behaviour has prejudiced the balance of advice given to policy makers. In particular, we did not find any evidence of behaviour that might undermine the conclusions of the IPCC assessments."

5. These findings reflect those of other reviews and inquiries conducted both prior to and subsequent to the publication of the Review.

5.1 The Commons Science and Technology Committee (in March 2010), following its inquiry, stated that: "Within our limited inquiry and the evidence we took, the scientific reputation of Professor Jones in CRU remains intact".http://www.publications.parliament.uk/pa/cm200910/cmselect/cmsctech/387/387i.pdf

5.2 Lord Oxburgh's Scientific Assessment Panel reporting in April 2010 similarly found that: "We saw no evidence of any deliberate scientific malpractice in any of the work of the Climatic Research Unit and had it been there we believe that it is likely that we would have detected it".http://https://www.uea.ac.uk/documents/3154295/7847337/SAP.pdf/a6f591fc-fc6e-4a70-9648-8b943d84782b.

5.3 More recently the US Environmental Protection Agency (EPA) published its "Notice of Denial of the Petitions to Reconsider the Endangerment and Cause or Contribute Findings for Greenhouse Gases". The EPA conducted a comprehensive review of a number of issues including the allegations which had arisen as a consequence of the publication of the emails and other material. The Denial of the Petitions and the accompanying volumes are a substantial body of evidence and careful analysis. The summary (Section A) of the Denial states inter alia "Petitioners … rely on an assumption of inaccuracy in the science … based on various statements and views expressed in some of the e-mail communications between scientists at the Climatic Research Unit (CRU) of the University of East Anglia … and several other scientists…EPA's careful examination of the e-mails … shows that the petitioners' claims are exaggerated, are often contradicted by other evidence, and are not a material or reliable basis to question the validity and credibility of the body of science … Inquiries from the UK House of Commons, Science and Technology Committee, the University of East Anglia, Oxburgh Panel, the Pennsylvania State University, and the University of East Anglia, Russell Panel, … have examined the issues and many of the same allegations brought forward by the petitioners … These inquiries are now complete. Their conclusions are in line with EPA's review and analysis of these same CRU e-mails. The inquiries have found no evidence of scientific misconduct or intentional data manipulation on the part of the climate researchers associated with the CRU e-mails … These inquiries lend further credence to EPA's conclusion that petitioners' claims that the CRU e-mails show the underlying science cannot or should not be trusted are exaggerated and unsupported". http://www.epa.gov/climatechange/endangerment/petitions/decision.html#I-A

6. The Review expresses a number of concerns and raises broader issues, but is critical of the University and CRU in aspects of its adherence to the spirit and intent of the FoIA or EIR. The University accepts the criticisms levelled by the Review and values the many recommendations of the report for improvements in the processes for dealing with FoIA and EIR requests and will seek to implement them fully. The University accepts that an apparent reluctance to provide access to data gave the impression that CRU was attempting to hide issues relating to its science. Clearly the conclusions of the Review are that there was nothing to hide, which underscores the point that the interests of CRU, the University and the dissemination of its research would have been best served by a more proactively helpful response to requests for information relating to data used for CRU's published analyses.

7. The University is collaborating in two significant activities arising from the issues raised in paragraph 6 above. Firstly, while the University is already undertaking a number of steps to improve engagement with FoIA/EIR, a senior team led by the Vice-Chancellor will be meeting with the Information Commissioner and his colleagues to review the University's processes and to seek guidance as to additional improvements which may be necessary. Secondly, CRU, in partnership with the Science and Technology Facilities Council E-Science Centre is embarking on a Joint Information Systems Committee funded project. This will examine how best to provide standardised access to processed climate data, linked both to raw observations and meaningful descriptions of intermediate processing. http://www.jisc.ac.uk/news/stories/2010/07/data.aspx

8. In the following paragraphs we paraphrase and summarise the key, more detailed, findings and recommendations of the Review and, where appropriate, add a comment in italics. References to the Review are shown as (page, paragraph).

Scientific integrity

9. The Review's further comments on the robustness and transparency of the research of CRU are welcome. 

Land Station Temperatures

9.1 CRU was not in a position to deny anybody access to temperature data. The team demonstrated that anyone could download station data, directly from primary sources, and construct a temperature trend analysis that agreed very well with that produced by CRU. This was an invaluable and instructive contribution to refuting allegations made against CRU that data was withheld as a barrier to challenge. (53,6.7)

9.2 The Review team demonstrated that its analysis of the temperature trend remained largely consistent regardless of stations selected and the use of adjusted or unadjusted data. CRU had not manipulated its selection of station data or its analysis to achieve a pre-determined outcome to show a rise in global temperatures. (53, 6.7)

Temperature Reconstruction from Tree Ring Analysis

9.3 There was no evidence that past temperature data as derived from tree ring proxies was misleading, nor was there evidence in IPCC AR4 of the exclusion of other temperature reconstructions that would show a different picture. The extent of the uncertainties surrounding such past temperature reconstructions were extensively covered in AR4, including the divergence of tree ring proxies from instrumental records in more recent times. (59, 21)

9.4 CRU did not withhold underlying raw data, having directed the single request for information to the owners of that data. (61,29)

Peer Review and Editorial Policy

9.5 There was no direct evidence of subversion of the peer review or editorial process. (68,18)

Misuse of the IPCC process

9.6 Allegations that in two specific cases there had been a misuse by CRU of the IPCC process, in presenting AR4 to the public and policy makers, could not be substantiated. CRU researchers were part of a large group of scientists taking responsibility for the AR4 text, and were not in a position to determine the content. (13, 26)

Disclosure of data and correspondence

10. The Review is critical of the handling of requests for disclosure of data and other material, and also makes a number of recommendations to the University.

Land Station Temperatures

10.1 CRU should have made available an unambiguous list of the stations used in the construction of its land temperature record at the time of publication. We accept that the University was slow in providing this data when requested. (51, 32)

Temperature Reconstruction from Tree Ring Analysis

10.2 The figure of trends in land temperatures supplied for the WMO report, whilst not misleading per se in splicing past temperature reconstructions and modern day temperature records, did not clearly describe the mechanisms used in the construction of the figure in the caption. The University would comment that the figure was an illustration for the cover of the report and additional explanation was supplied on the inside cover and in the text. (13, 23)

10.3 There was a delay in archiving tree ring data by its owners. The University will, as part of a wider protocol documenting the agreements over the use of data provided by others, promote the benefits of such data being archived and accessible. In some instances, however, this will not be achievable, particularly where the commercial interests of the owners come into play. (62, 38)


10.4 While the University had widely distributed initial guidance on the introduction of the FoIA/EIR regimes, there was evident confusion within CRU as to how these should be applied, for example, to data, codes and personal correspondence. There was insufficient priority given to motivating staff and continuing their education in this respect. Senior staff need to make clear their commitment to transparency and to resourcing the process.The University accepts that all staff from the top down must be better engaged with the FoIA/EIR regimes. The University has already begun a programme of further training with awareness raising for senior staff. The Vice-Chancellor has written to all staff to underline UEA's commitment to this. All new staff will receive a written statement concerning responsibilities under the FoIA/EIR together with annual updates. All staff with a particular role in the implementation of FoIA/EIR (recognising the general obligation of all colleagues) will attend annual workshops to update their knowledge. A programme specifically for staff closely involved with EIR will be mounted. The resources available to the FoIA team have been increased. (91, 25 and 91, 33)

10.5 There was a failure to recognise the extent to which more careful engagement with requesters would have been both appropriate and helpful to avoid fuelling the fire of suspicion. We fully accept this criticism and our various steps referred to in this response seek to address it. (91, 26)

10.6 There was a tendency to give unhelpful responses: failing to address the question asked or giving partial answers. There was extensive delay in providing details of those station identifiers which were not the subject of confidentiality agreements. Again, we accept the University could have performed much better in responding to these requests and steps are being taken to address this. (91, 27)

10.7 A number of emails appeared to incite deletion or evidence deletion of other emails, although there was no evidence of emails being deleted that were the subject of a request for disclosure. We accept this shows insufficient awareness of and focus on obligations under the FoIA/EIR, but we welcome the finding that there was no attempt to delete information with respect to a request already made. This confirms assurances already given to the Vice-Chancellor by colleagues in CRU that they had not deleted material which was the subject of a request. We have underlined that such action would have been one of the key elements necessary to constitute an offence under Section 77 of the FoIA and Section 19 of the EIR, the others being that information had actually been deleted, that it was deleted with the intention to avoid disclosure and that it was disclosable and not exempt information. Professor Jones has commented that, while emails are cleared out from time to time, this is to keep accounts manageable and within the allocated storage. (92, 28)

10.8 There is an imbalance of authority between the Information Policy and Compliance Manager (IPCM) and senior academic staff holding information which may be the subject of a request for disclosure. There is also a lack of constructive challenge in the appeals processes. The University has amended its protocols to allow the Director of Information Services and the Registrar to become involved at an early stage in the consideration of sensitive cases and for reviews of any decisions not to disclose information then to be undertaken by others at a senior level in the University. These changes are being formalised in a revised code of practice. (93, 29 and 93, 30)

10.9 There is a lack of understanding of the presence of long-duration back-ups of email which, had it been stronger, would have led to a greater challenge of assertions regarding the availability of material. We accept this, albeit, as is recognised in the Information Commissioner's guidance, retrieving such data may not always be a practical option. (93, 31)

10.10 There was a fundamental lack of engagement by the CRU team with their obligations under FoIA/EIR.CRU is now more clearly integrated within the management and administrative structures of the School of Environmental Sciences, and the Head of School will take greater responsibility for compliance with FOIA/EIR requirements. The University also has undertakings from the Director of Research (Professor Jones) of necessary improvements in this regard in the future. (93, 32)

Broader issues

11. The Review identifies a number of broader issues which are a valuable commentary on the process of scientific debate, both generally and in the particular instance of climate science:

11.1 Much of the challenge to CRU's work has not followed the conventional method of checking and seeking to falsify conclusions or offering alternative hypothesis for peer review and publication. Again, the Review has been invaluable in demonstrating that the great bulk of the temperature data used by CRU was already readily available and that there was no barrier to checking or seeking to offer alternative hypotheses compatible with the data. Attempts simply to taint the science with the content of email exchanges are not the appropriate way to probe or challenge the conclusions (15, 35 and 15, 36)

11.2 The scientific community must learn to communicate its work in ways that recognise the emergence of the blogosphere and non-traditional scientific dialogue. That this provides an opportunity for unmoderated comment, for challenge without inhibition and for highly personalised critiques of individuals and their work to be promulgated without hindrance is a point well taken. The University accepts this is a necessary but challenging task. (14, 31 and 15, 33)

11.3 The research community must establish very clearly the requirements of funders for the release of data and its archiving, and the associated costs. We agree, noting that these requirements should be proportionate to the likely wider value and importance of the data. (104, 36)

11.4 It is important for policy makers and lobbyists to understand the limits on what science can say and with what degree of confidence. Alternative viewpoints should be recognised in policy presentations, with a robust assessment of their validity, but challenges should always be rooted in science rather than in rhetoric. The University fully endorses this observation. Challenges to science should come through peer review publication substantiating the alternative; not through criticism of emails which, as the Review states, are rarely definitive evidence of what actually occurred. (14, 32)

11.5 While Peer Review is an essential part of the process of judging scientific work, it is not a guarantee of the validity of the individual pieces of research, and the significance of challenge to individual publication decisions should not be exaggerated. We agree with the Review that robust challenges to the publication of research which experts believe, in good faith, does not meet the standards required is commonplace and should not be dismissed as an attempt to "silence" critics. (15, 33)

Sector wide issues

12. The Review raises a number of issues which require a sector-wide debate, and engagement between the representative bodies and the ICO. The University strongly supports this approach and will seek to promote further consideration of these important issues through Universities UK, the Joint Information Systems Committee (JISC) and other bodies.

12.1 Raw data, meta-data and codes necessary to allow independent replication of results should be provided concurrent with peer reviewed publication. However, so far as preliminary analyses, drafts of scientific papers, plans for future research, peer reviews and communications with colleagues are concerned, the American approach, where these are exempt from disclosure, is one which is worthy of consideration. (94, 34)

12.2 CRU was the subject of an orchestrated campaign of FoIA/EIR requests, and while more positive engagement by CRU would have mitigated this, conceivably there are situations where such campaigns could recur and overwhelm any small research unit. The ICO is urged to provide guidance on how best to respond to such campaigns. (95, 34)

12.3 The ICO could produce further guidance as to how long it is reasonable to retain data without releasing it, pending full publication as part of a peer reviewed paper. We agree that this is an important concern. The many benefits of publication are set out elsewhere in this response. Nonetheless data sets, carefully assembled, may result in a number of publications for an individual, the very foundation on which a scientific reputation is built. For how long is it reasonable for an individual to have their intellectual investment protected? (95, 34)

12.4 There should be a standardised way of defining station data and meta-data, and for publishing a snapshot of the data used for each important publication. We will discuss with the WMO but this will not be a trivial undertaking. (53, 40)

12.5 The storage of important research data, and the associated meta-data which make that data useful, should be specified by those funding research and there should be a clear statement as to which data should be placed in the public domain and any constraints on the timing of its release. (104, 36)


13. The Review makes a number of recommendations to the University on risk management and on the storage and security of data.

13.1 The University was insufficiently alert to the implications of the external attitudes which existed towards the work of CRU and of the attention of external pressure groups, and mitigation measures should be put in place. Greater CRU security, a bias for openness and a properly resourced policy on data management and availability should have resulted. The University will undertake a Faculty-based risk assessment of all areas of the University's research; implement more centralised IT support to ensure appropriate security levels; and develop processes which ensure that senior management are informed of emerging problems in a timely fashion. The University will participate (with others) in projects to improve the storage of and access to research data both specifically in respect of climate data and more generally. (103, 33)

13.2 Universities should develop formal approaches to the training of researchers in basic software development methodologies and best practice. We shall consider the development of a programme of workshops for researchers in appropriate disciplines. (103, 34) 

13.3 There should be a formal approach to storage and archiving of meta-data where a university is hosting a unit of such international significance as CRU. We agree and have successfully bid for grant funding to support a project for our three principal data sets. It is anticipated that the results of this project will provide an exemplar for climate researchers, including those outside UEA. (103, 35)

13.4 At the point of publication of research, enough information should be available for others to reconstruct the process of analysis, including the source code. The University accepts this should be the case, unless valuable intellectual property or other commercial constraints are in play. (104, 37)

13.5 Where the University establishes a framework and standards in areas such as information systems but allows local interpretation, this should be subject to robust audit. The University is centralising the control of IT systems and intends to reduce the level of discretion of research groups and others for the control and management of IT. Adherence to overall policies will be part of the future programme of internal audit. (104, 38)


14. The University will update this document as the steps set out in the body of the document are progressed