It is difficult to imagine a more bizarre academic dispute. Where exactly are 42 weather monitoring stations in remote parts of rural China?
But the argument over the weather stations, and how it affects an important set of data on global warming, has led to accusations of scientific fraud and may yet result in a significant revision of a scientific paper that is still cited by the UN's top climate science body.
It also further calls into question the integrity of the scientist at the centre of the scandal over hacked climate emails, the director of the University of East Anglia's Climatic Research Unit (CRU), Dr Phil Jones. The emails suggest that he helped to cover up flaws in temperature data from China that underpinned his research on the strength of recent global warming.
The Guardian has learned that crucial data obtained by American scientists from Chinese collaborators cannot be verified because documents containing them no longer exist. And what data is available suggests that the findings are fundamentally flawed.
Jones and his Chinese-American colleague Wei-Chyung Wang, of the University at Albany in New York, are being accused of scientific fraud by an independent British researcher over the contents of a research paper back in 1990.
That paper, which was published in the prestigious journal Nature, claimed to answer an important question in climate change science: how much of the warming seen in recent decades is due to the local effects of spreading cities, rather than global warming?
It is well-known that the concrete, bricks and asphalt of urban areas absorb more heat than the countryside. They result in cities being warmer than the countryside, especially at night.
So the question is whether rising mercury is simply a result of thermometers once in the countryside gradually finding themselves in expanding urban areas.
The pair, with four fellow researchers, concluded that the urban influence was negligible. Some of their most compelling evidence came from a study of temperature data from eastern China, a region urbanising fast even then.
The paper became a key reference source for the conclusions of succeeding reports of the Intergovernmental Panel on Climate Change – including a chapter in the 2007 one co-authored by Jones. It said that globally "the urbanisation influence … is, at most, an order of magnitude less than the warming seen on a century timescale". In other words, it is tiny.
But many climate sceptics did not believe the claim. They were convinced that the urban effect was much bigger, even though it might not change the overall story of global warming too much. After all, two-thirds of the planet is covered by ocean, and the oceans are warming, too.
But when Jones turned down requests from them to reveal details about the location of the 84 Chinese weather stations used in the study, arguing that it would be "unduly burdensome", they concluded that he was covering up the error.
And when, in 2007, Jones finally released what location data he had, British amateur climate analyst and former City banker Doug Keenan accused Jones and Wang of fraud.
He pointed out that the data showed that 49 of the Chinese meteorological stations had no histories of their location or other details. These mysterious stations included 40 of the 42 rural stations. Of the rest, 18 had certainly been moved during the study period, perhaps invalidating their data.
Keenan told the Guardian: "The worst case was a station that moved five times over a distance of 41 kilometres"; hence, for those stations, the claim made in the paper that "there were 'few if any changes' to locations is a fabrication". He demanded that Jones retract his claims about the Chinese data.
The emails, which first emerged online in November last year following a hack of the university's computer systems that is being investigated by police, reveal that Jones was hurt, angry and uncertain about the allegations. "It is all malicious … I seem to be a marked man now," he wrote in April 2007.
Another email from him said: "My problem is I don't know the best course of action … I know I'm on the right side and honest, but I seem to be telling myself this more often recently!"
An American colleague, and frequent contributor to the leaked emails, Dr Mike Mann at Pennsylvania State University, advised him: "This crowd of charlatans … look for one little thing they can say is wrong, and thus generalise that the science is entirely compromised. The last thing you want to do is help them by feeding the fire. Best thing is to ignore them completely."
Another colleague, Kevin Trenberth at the National Centre for Atmospheric Research in Boulder, Colorado, urged a fightback. "The response should try to somehow label these guys and [sic] lazy and incompetent and unable to do the huge amount of work it takes to construct such a database."
In August 2007, Keenan submitted a formal complaint about Wang to Wang's employers. The university launched an inquiry. Reporting in May 2008, it found "no evidence of the alleged fabrication of results" and exonerated him. But it did not publish its detailed findings, and refused to give a copy to Keenan.
By then, Keenan had published his charges in Energy & Environment, a peer-reviewed journal edited by a Hull University geographer, Dr Sonja Boehmer-Christiansen.
The paper was largely ignored at the time, but Guardian investigations of the hacked emails now reveal that there was concern among Jones's colleagues about Wang's missing data – and the apparent efforts by Jones and Wang over several years to cover this up.
Those concerns were most cogently expressed to Jones by his ex-boss, and former head of the CRU, Dr Tom Wigley. In August 2007, Wigley warned Jones by email: "It seems to me that Keenan has a valid point. The statements in the papers that he quotes seem to be incorrect statements, and that someone (W-C W at the very least) must have known at the time that they were incorrect."
Wigley was concerned partly because he had been director of the CRU when the original paper was published in 1990. As he told Jones later, in 2009: "The buck should eventually stop with me."
Wigley put to Jones the allegations made by the sceptics. "Wang had been claiming the existence of such exonerating documents for nearly a year, but he has not been able to produce them. Additionally, there was a report published in 1991 (with a second version in 1997) explicitly stating that no such documents exist."
This is believed to be a report from the US department of energy, which obtained the original Chinese temperature data.
Wang's defence to the university inquiry says that he had got the Chinese temperature data from a Chinese colleague, although she is not an author on the 1990 Nature paper.
Wang's defence explains that the colleague had lost her notes on many station locations during a series of office moves. Nonetheless, "based on her recollections", she could provide information on 41 of the 49 stations.
In all, that meant that no fewer than 51 of the 84 stations had been moved during the 30-year study period, 25 had not moved, and eight she could not recollect.
Wang, however, maintained to the university that the 1990 paper's claim that "few, if any" stations had moved was true. The inquiry apparently agreed.
Wigley, in his May 2009 email to Jones, said of Wang: "I have always thought W-C W was a rather sloppy scientist. I would …not be surprised if he screwed up here … Were you taking W-C W on trust? Why, why, why did you and W-C W not simply say this right at the start? Perhaps it's not too late." There is no evidence of any doubts being raised over Wang's previous work.
Jones told the Guardian he was not able to comment on the allegations. Wang said: "I have been exonerated by my university on all the charges. When we started on the paper we had all the station location details in order to identify our network, but we cannot find them any more. Some of the location changes were probably only a few metres, and where they were more we corrected for them."
The story has a startling postscript. In 2008, Jones prepared a paper for the Journal of Geophysical Research re-examining temperatures in eastern China. It found that, far from being negligible, the urban heat phenomenon was responsible for 40% of the warming seen in eastern China between 1951 and 2004.
This does not flatly contradict Jones's 1990 paper. The timeframe for the new analysis is different. But it raises serious new questions about one of the most widely referenced papers on global warming, and about the IPCC's reliance on its conclusions.
It is important to keep this in perspective, however. This dramatic revision of the estimated impact of urbanisation on temperatures in China does not change the global picture of temperature trends. There is plenty of evidence of global warming, not least from oceans far from urban influences. A review of recent studies published online in December by David Parker of the Met Office concludes that, even allowing for Jones's new data, "global near-surface temperature trends have not been greatly affected by urban warming trends."
Keenan accepts that his allegations do not on their own change the global picture. But he told the Guardian: "My interest in all this arises from concern about research integrity, rather than about global warming per se. Jones knew there were serious problems with the Chinese research, yet continued to rely upon the research in his work, including allowing it to be cited in the IPCC report."
From sceptic Doug Keenan to Dr Wei-Chyung Wang and Prof Phil Jones – 20 April 2007
"I ask you to retract your GRL paper, in full, and to retract the claims made in Nature about the Chinese data. If you do not do so, I intend to publicly submit an allegation of research misconduct to your university at Albany."
From Jones to Dr Kevin Trenberth
"I seem to be the marked man now !"
From Prof Michael Mann to Jones
"This is all too predictable. This crowd of charlatans is always looking for one thing they can harp on, where people w/ little knowledge of the facts might be able to be convinced that there is a controversy. They can't take on the whole of the science, so they look for one little thing they can say is wrong, and thus generalise that the science is entirely compromised."
From Trenberth to Jones and Mann – 21 April 2007
"I am sure you know that this is not about the science. It is an attack to "undermine the science in some way. In that regard I don't think you can ignore it all … the response should try to somehow label these guys lazy and incompetent and unable to do the huge amount of work it takes to construct such a database."
From Prof Tom Wigley to Jones – 4 May 2009
"I have always thought W-C W [Wang] was a rather sloppy scientist. I therefore would not be surprised if he screwed up here … Why, why, why did you and W-C W not simply say this right at the start? Perhaps it's not too late? I realise that Keenan is just a troublemaker and out to waste time, so I apologize for continuing to waste your time on this, Phil. However, I *am* concerned because all this happened under my watch as director of CRU and, although this is unlikely, the buck eventually should stop with me."
• This article was amended on 2 February 2010. One sentence in the original read: "Of the rest, 18 had certainly been moved during the story period, perhaps invalidating their data." The word story has been corrected.