Which is, frankly, a good catch. Props to the collective journalistic effort of the blogosphere. We need more of that (including when temperatures are reported as very low; but this is probably unlikely, just due to basic human nature).
Of course, correcting data is as old as science itself. It is a fundamental part of science, and won't ever, ever go away. Science doesn't proclaim the truth -- it zeros in on it.
Still, though, suspicious things are going on. Anthony Watts at Watt's Up With That says that the NASA GISS data set is no longer reliable:
But, of course (and you know what's coming next) it was reliable enough a mere nine months ago when he used it to show "global cooling":
The GISS ΔT was -.75°C, which is larger than the satellite data from UAH ∆T of -.588°C and the RSS RSS ∆T of -.629°CThis is just one more example of what can only be considered intellectual dishonesty by some of the skeptics, which we have seen before.
GISS January Land-Sea Anomaly
click for larger image
The ΔT of -.75°C from January 2007 to January 2008 appears to be the largest single year to year January drop for the entire GISS data set.
This is yet one more indication of the intensity of planet-wide cooler temperatures seen in January 2008, particularly in the Northern Hemisphere, which has seen record amounts of snow coverage extent as well as new record low surface temperatures in many places.