Friday, June 8, 2007

Pournelle on what to do about warming

Dr. Pournelle has a letter from someone who thinks that a couple of minutes with Google has led him to The Truth. Oh dear. From his reply:

My point, which I repeatedly make, is that until we know what the hell is going on, it is absurd to spend the money we could be spending to find out what is happening on "remedies" when we don't understand the problem. We spend what money we do get for research on computer models instead of better data. We ought to be taking more deep sea probes, launching satellites whose purpose is to get accurate temperature data at all levels of the atmosphere (it varies a lot) as well as ground and sea surface; get temperatures both in and outside of cities, upwind and down; measuring total glacier gains and losses (some glaciers are gaining); and so forth. We ought to know more about clouds. I don't claim to be the world's expert on what we ought to be doing to find out what is happening. I do claim to be enough of an old Operations Research guy to know that if you don't know what's happening, you're better off spending money to find out what's going on than you are in trying to fix something you don't understand.

Simple Bayesian analysis: if there are two possible trends, and dealing with each is expensive, and what you must do is pretty well determined by which trend is going to happen, you are better off spending money to reduce the uncertainty and pay the penalty for starting later on the right track than you are to start investing in remedies that may be for the wrong coming disaster. I could show you mathematically, but I am sure you can see the point.
Much more at the link.

No comments: