Y2K Bugs Still Kicking, and Messing with Climate Data

Interesting news from the climate front: One global warming skeptic appears to have found a real error in the algorithm used to measure the temperature of the Detroit area lakes. That error, it's said, was due to our old friend, the Y2K bug.

Revised figures released by NASA in response to the correction show slightly lower temperatures than previously reported (less than 2%), but does knock 1998 out of its old "hottest year ever" spot, which now goes to 1934 (but just barely). (Disclaimer for purists: The bug may actually not be a classical 1900 vs. 2000 error, but it's still a bug nonetheless.)

Now don't jump all over me: I'm an avowed greenie with deep concerns over climate change and greenhouse gas emissions, but I'm also concerned with accurate data being given to the public to work with. If we can't have accurate data, how can we know whether changes we might make are doing any good? How many other bugs are lying undiscovered in unpublished code? And if we can't fix our algorithms, how can we have any hope of fixing the planet?

The changes here are ultimately minor and fairly irrelevant, but they do bring up an important point: NASA refused to publish the code to calculate these figures, for reasons I don't fully understand. NASA isn't Microsoft: This is a public agency accountable to the people, and if we want to see some source code, it really ought to oblige us.

Seriously NASA, don't you have enough problems already? One would think you could use all the help you could get.

LINK: Blogger Finds Y2K Bug in NASA Climate Data
LINK: Watts Up With That? 1998 no longer the hottest year on record in the USA