An inconvenient update

Comment Last week, statistician and amateur meteorologist Steve McIntyre notified NASA of an error in its climate data. The results of the hasty correction mean that as far as the US is concerned, 1998 is no longer the hottest year on record. 1934 is.

Headline-grabbing statements that nine out of ten of the hottest years on record were in the last decade are no longer correct, for the US, at least (bad news for Mr Gore, certainly). And those who remain sceptical about the nature of the link between human activity and global warming were delighted, as the Goddard Institute for Space Studies had to quietly admit the mistake and publish corrected data.

But what does this mean for the rest of us? What was the glitch? Where was the miscalculation? And do we need to check our data? Can we all hop into our Humvees and barrel around town, untroubled by our carbon emissions?

Goddard itself says the change is not significant enough to change the overall trends associated with global warming. Is it right?

Richard Allen, environmental systems scientist at the Centre for Atmospheric Science, thinks the revision is not worth getting too agitated about.

"The US only provides two per cent of the data, so it is not important as far as global temperature change goes," he told us.

But this is only true if we assume the rest of the world is not suffering from a similar (or entirely different) glitch. So what was the problem?

"What happens is that station data [the raw temperature readings from US weather stations] are corrected for slight changes, such as urbanisation. NOAA [National Oceanic and Atmospheric Administration] in the US normally does this, it supplies it in near real time to GISS and NASA. But for some reason, they stopped doing this, and some stations didn't have their corrections. Now they do." Allen explains.

Allen argues that discontinuities like this do tend to get picked up pretty quickly, and says this whole episode is a good example of how this happens. "Nothing is perfect," he says. "But there are a lot of scientists out there who are working very carefully. So it is unlikely that it is a big problem."

Because the error in the US data is so specific to the way the US manages its figures, it seems unlikely that the data from the rest of the world will be afflicted by the same problem.

Meanwhile, McIntyre is unhappy with the way Goddard handled the situation. He says that the failure to put out an official announcement of the update left GISS open to accusations that it being less than frank.

The rather taciturn handling of the change has provoked some to wonder whether a revision in the opposite direction might have been given more prominence. And who can honestly say that it wouldn't?

That alone should be a sobering thought for those working in the field.

But botched PR doesn't prove that global warming isn't happening. That changes had to be made needs to be taken seriously. The scientific community should take note and make sure the rest of the data is in order.

For those who are interested, the old data is here, and the recalculated data is here. The old and new temperatures differ by a hundredth of a degree. ®