From The Sydney Morning Herald, some questions over the accuracy of Australian temperature data:
LONDON: Australian weather records for an international database on climate change were a "bloody mess", riddled with entry errors, duplication and inaccuracies, leaked British computer files reveal.
The Herald found the criticism in a 247-page specialist programmer's log, unearthed among the thousands of files hacked from East Anglia University, which is at the centre of a climate change email scandal.
Labelled "HARRY-READ-ME", the log catalogues problems with the raw, historical climate data sent from hundreds of meteorological stations around the world.
The Australian data comes in for particular criticism as the programmer discovers World Meteorological Organisation codes are missing, station names overlap and many co-ordinates are incorrect.
At one point the programmer writes about his attempts to make sense of the data. "What a bloody mess," he concludes. In another case, 30 years of data is attributed to a site at Cobar Airport but the frustrated programmer writes: "Now looking at the dates. something bad has happened ... COBAR AIRPORT AWS [automatic weather station] cannot start in 1962, it didn't open until 1993!"
In another he says: "Getting seriously fed up with the state of the Australian data ... so many false references ... so many changes ... bewildering."
The log spans four years of work at the university's Climatic Research Unit, the British keeper of global temperature records. The programmer rails that the information has "no uniform integrity".
His criticisms relate solely to the construction of the database and do not question the validity of historical temperature records or analyses that suggest the impact of human activity on global warming trends.
"I am very sorry to report that the rest of the databases seem to be in nearly as poor a state as Australia was. There are hundreds if not thousands of pairs of dummy stations, one with no WMO and one with, usually overlapping and with the same
station name and very similar co-ordinates. I know it could be old and new stations, but why such large overlaps if that's the case? Aarrggghhh! There truly is no end in sight."
Michael Coughlan, the head of the National Climate Centre at the Australian Bureau of Meteorology, said it was difficult to comment without knowing the source of the raw data. It was unlikely to have come directly from the bureau's centre because unchecked, raw data was rarely requested for climate analysis. The bureau had a network of more than 100 specially selected weather stations to monitor climate change, and a century of records from them had been checked.
"We've put an enormous effort into developing a high-quality reliable climate record for Australia and all that data is freely available," Dr Coughlan said.
But he said that if the British programmer had been using raw weather data, which is sent around the world in real time for weather forecasting, it would not be surprising that it contained errors. This raw data could have come from countries other than Australia, and would have been difficult to correct without access to information in Australia, such as the original field books.
"A computer programmer sitting in England won't have the resources to make those corrections. I can understand their frustrations," Dr Coughlan said.
The programmer's log is one of the most read files worldwide since the email archives were leaked. The log has been treated particularly sympathetically as it reveals his blow-by-blow frustrations, which seemed to be unfolding as his scientist colleagues, including the head of the Climatic Research Unit, Phil Jones, appeared to discuss via email ways to avoid freedom-of-information requests for raw data and to denigrate their critics.
Professor Jones, who has denied a conspiracy to manipulate global warming statistics as "complete rubbish", has stood down from his post while the university investigates the leaks.
The Herald attempted to contact Professor Jones and spoke to the computer programmer we believe to be the author of the file. The programmer did not deny his name but referred queries to the university's media unit. Professor Jones has not responded.
RealClimate, a website run by climate scientists, confirms the log as the work of a specialist charged with upgrading data.
"Anyone who has ever worked on constructing a database from dozens of individual, sometimes contradictory and inconsistently formatted datasets, will share his evident frustration with how tedious that can be," it says.
