Saturday, December 16, 2006

An Inconvenient Bayesian Truth?

James Annan feels his (and Hargreaves') paper, which says that the high end of predicted temperature changes from global warming is unlikely for a doubling of CO2 levels, has been unjustly rejected for publication, possibly because it is not an alarmist result. Stoat agrees.

I don't have anything like the math chops to make my own judgments, but I accept James' statement that multiple referees may have had criticisms but didn't consider his work to be fundamentally flawed. Given that, and because Annan gets a result that eliminates possibilities no one else has eliminated, it seems pretty obvious the paper should be published. Let the debate then go on as to whether the paper's right.

Several points worth mentioning though - Annan thinks the resulting temperature increase is still enough to warrant emission reductions. His result doesn't support the next line of defense by fossil fuel industry - the "forget reductions, let's just adapt" nonsense. It also only deals with doubling CO2 - if we don't control emissions, they'll more than double and cause much more trouble. Eli Rabett also points out that what Don Rumsfeld would call "known unknowns" are mostly dangerous - at best, they won't make things worse, and at worst, they will. Those unknowns weren't (and couldn't be) included in Annan's result. Not a flaw in the paper, but another reason to do something about greenhouse gases.

And one of Annan's commenters notes that it's hard to draw conclusions about peer review of a single paper in terms of demonstrating institutional bias. The only conclusion Annan may be able to draw from being screwed over is that he's been screwed over. Hopefully, the paper will get published somewhere appropriate.

No comments:

Post a Comment

Note: Only a member of this blog may post a comment.