If the beleaguered climate scientists of the University of East Anglia have a weakness, it is in their statistics 鈥� yet their conclusions that the planet is warming stands on solid ground.
That鈥檚 the conclusion of the third independent inquiry into 鈥�climategate鈥� 鈥� the fallout from last November鈥檚 release of hundreds of emails from the Climatic Research Unit (CRU) at the university, which is located in Norwich, UK.
Crucially for the scientists who work at the CRU, the inquiry 鈥� led by Ron Oxburgh, a former chair of the UK House of Lords science and technology select committee 鈥� cleared them of any professional misconduct.
Advertisement
鈥淲e found absolutely no evidence of any impropriety whatsoever,鈥� said Oxburgh at a press briefing in London today.
鈥淲e are absolutely satisfied that these people were doing their job fairly. I don鈥檛 think they even minded what the outcome [of their research] was, as long as it was as close to truth as possible,鈥� he continued.
Career decision
That will have been a comfort for the CRU team. As Oxburgh put it, they were 鈥渦nlikely to have worked as scientists again鈥� had the inquiry panel found them guilty of professional misconduct.
The panel found that the statistical tools that CRU scientists employed were not always the most cutting-edge, or most appropriate. 鈥淲e cannot help remarking that it is very surprising that research in an area that depends so heavily on statistical methods has not been carried out in close collaboration with professional statisticians,鈥� reads the inquiry鈥檚 conclusions.
However, 鈥渋t is not clear that better methods would have produced significantly different results,鈥� the panel adds.
Messy data
Oxburgh was commissioned by the university to review the professional conduct of the CRU scientists. The UK Royal Society recommended his appointment and that of the panel of scientists which assisted him.
, president of the UK Royal Statistical Society and a member of Oxburgh鈥檚 panel, said the work of climate scientists is a 鈥減articularly challenging statistics exercise because the data are incredibly messy鈥�.
Climate scientists must gather temperature data from disparate sources. Over the course of decades, temperature probes may have to be moved 鈥� because of the growth of cities, for example 鈥� and temperatures are not always measured in the same way at different stations around the globe.
Hockey stick
He said the strongest example he had found of imperfect statistics in the work of the CRU and collaborators elsewhere was the iconic 鈥渉ockey stick鈥� graph, produced by of Pennsylvania State University in University Park.
The graph shows how temperatures have changed over the past 1000 years (see graphic, right).
Hand pointed out that the statistical tool Mann used to integrate temperature data from a number of difference sources 鈥� including tree-ring data and actual thermometer readings 鈥� produced an 鈥渆xaggerated鈥� rise in temperatures over the 20th century, relative to pre-industrial temperatures.
That point was initially made by climate sceptic and independent mathematician . The upwards incline on later versions of the graph has been corrected to be shorter and less exaggerated (for the full story of the hockey stick controversy, see Climate: The great hockey stick debate, and Climate myths: The 鈥榟ockey stick鈥� graph has been proven wrong).
Hand said he was 鈥渋mpressed鈥� by McIntyre鈥檚 statistical work. But whereas McIntyre claims that Mann鈥檚 methods have 鈥渃reated鈥� the hockey stick from data that does not contain it, Hand agrees with Mann: he too says that the hockey stick 鈥� showing an above-average rise in temperatures during the 20th century 鈥� is there. The upward incline is just shorter than Mann鈥檚 original graphic suggests. 鈥淢ore like a field-hockey stick than an ice-hockey stick,鈥� he told New 杏吧原创.
Two other inquiries, by British members of parliament and Pennsylvania State University, have also cleared the scientists of misconduct. Two further inquiries, one led by former British civil servant Muir Russell and the other conducted by British police, are still under way.
