The scientists at the Met Office's Hadley Centre, have discovered an error in the way they calculate the global temperature. Up until now, the smoothed data for a given year has been based upon those months' data which is available - so the 2007 trend, as reported in February 2007, was based on data for January and February of that year. At the time, this gave an sharply rising temperature trend, which was rather convenient for the IPCC Working Group I which was meeting at the time, enabling all and sundry to conclude that we were all doomed.
Roll forward to 2008 and the data for January and February 2008 show temperatures falling at an equally startling rate.
And guess what? Suddenly the disinterested scientists at Hadley have discovered that their methodology is all wrong! They're going to change things to eliminate the bias. Well, bully for them!
Isn't it amazing that this indavertent error has gone unnoticed for all these years? Isn't it more amazing still how it only got noticed when temperatures were falling? Someone with a more suspicious mind than me might think these guys had an agenda!
It's amusing also to note that Professor Phil Jones, the head of the Hadley Centre (update: CRU) is a man with an extraordinary reluctance to release his data and code to outsiders, even refusing to acquiesce to requests under the Freedom of Information Act. We can reach our own conclusions as to why he might do this, but for the moment it's salutory to note that, had he adopted a position of openness, this embarrassing error, which surely calls into question his competence to occupy the position of trust he does, might never have happened.
(Updated to fix Jones' affiliation)