Monday, August 20, 2007

Bookmark This Page

No, not this one. This one. It shows how NASA GISS had to revise down its figures for the hottest ten years since 1880. It turns out that there was a Y2K error in the data, of all things.

Ha! Our bad! Right? Well, the error would have been discovered sooner, except that the man who discovered the error had to reverse-calculate it out of GISS's data. Steve McIntyre, a Canadian mathematician, has been hammering on the statistical methods of Michael Mann's and James Hansen's Super-Secret Algorithms that they use to adjust raw data but that they won't release for scrutiny.

New top ten: 1934, 1998, 1921, 2006, 1931, 1999, 1953, 1990, 1938, 1939

And here is the famous McIntyre and McKitrick critique of Mann's "Hockey Stick."

Amazing stuff. I have no grasp of statistics, so I can't critique M&M's work, but criminey, Mr. Hansen, you won't release your algorithm?

UPDATE: Hansen has grudgingly released his algorithm and methods, or so he says. I guess if it's fake, McIntyre et al. will figure it out.

No comments: