The climate scientists at the centre of a media storm over emails released on the internet were disorganised but did not fudge their results, an independent inquiry into the affair reported today.
The inquiry, the second of three set up in the wake of the controversy, found "absolutely no evidence of any impropriety whatsoever", according to Lord Oxburgh, who led the investigation.
Instead, Oxburgh said, many of the criticisms and assertions of scientific misconduct were likely made by people "who do not like the implications of some the conclusions" reached by the climate experts.
Oxburgh said: "Whatever was said in the emails, the basic science seems to have been done fairly and properly."
The review gave scientific processes at CRU "a clean bill of health" but did raise some issues of concern. Record-keeping was patchy, it said, while the scientists did not use the best possible statistical techniques to analyse their data.
David Hand, a statistician at Imperial College London, who sat on the enquiry panel, said the CRU scientists had been naive over their use of statistics, but there was no evidence that the better techniques would have produced different results. Poor record-keeping was common among scientists, Oxburgh said, while the CRU experts could not have anticipated the future public interest in what had been an "unfashionable" area of science for much of their careers.
Oxburgh said the scientific papers contained the necessary caveats and expressions of uncertainty where required. But he criticised the way these caveats were often stripped away when such research was presented by other bodies, such as the media, government agencies and the IPCC.
Oxburgh singled out a graph of global temperature used in a 1999 report for the World Meteorological Association, which spliced three different data sets, as an "unfortunate representation of a very complex piece of science". The graph was prepared by CRU experts, and was the subject of the infamous email from Jones in which he described how he had used a "trick" to "hide the decline". Jones said the relevant error ranges were included in the WMO document.
At a press conference to launch the review's findings, Hand re-ignited a long-standing row about a high-profile study published in 1998 by scientists led by Michael Mann at Penn State University, US. The paper featured an emblematic graph known as the "hockey-stick" that showed temperature rise in the twentieth century was unprecedented in recent history. Hand said the study gave him an "uneasy feeling" because it used "inappropriate statistical tools". The hockey-stick effect was genuine, Hand said, but the 1998 paper exaggerated it. He praised Steve McIntyre, a Canadian climate blogger who led much of the criticism of the CRU scientists, for identifying the problem.
Oxburgh said sustained requests to CRU scientists for data and computer codes from McIntyre and others could have amounted to a campaign of harassment, and that the affair left several unresolved questions about how Freedom of Information laws should be applied in an academic context.
The report also said it was "unfortunate" that the UK government had introduced widely copied policies to charge for environmental data sets, such as those used by the CRU scientists and requested by critics. The move impeded the flow of data between researchers, it added.