« Able Danger and Constitutional rights | Main | Web 2.0 visualized »

Let's just melt, then.

The Arctic is melting, and it's melting fast. According to MSNBC, "scientists stopped short of directly blaming the melting trend on global warming but said they have few other explanations at this point." It's important to note here that scientists hypothesize rather than proclaim, which is why they "stop short." That doesn't mean that they don't think that the global climate is changing - it's obvious. Laymen will try to tie specific events to that change, but most scientists won't leap without more data, whatever they might suspect. (The Union of Concerned Scientists is a little more forthcoming: "Global warming is one of the most serious challenges facing us today. To protect the health and economic well-being of current and future generations, we must reduce our emissions of heat-trapping gases by using the technology, know-how, and practical solutions already at our disposal."

While we're here, I should mention the post at Realclimate yesterday on pseudo-scientist Michael Crichton's visit to the U.S. Senate, joining global warming skeptic Senator James Inhofe and hurricane forecaster William Gray in a parade of "half-truths and red herrings."

First, let's be clear where there is agreement. Climate science doesn't deal in certainties - it deals in probablities and the balance of evidence. We agree with Crichton's statement that 'Prediction is not fact'. That certainly doesn't mean, however, that projections of possible future climate changes are not meaningful or useful, as Crichton claims.

Crichton seemed to imply that "prediction" (such as that provided by weather or climate models) is useless in the decision making process. (As an aside, we wonder how Gray, who is largely known for prediction of hurricane behavior based on (statistical) modeling, felt about this?). We fundamentally disagree. All science is about observation, understanding and prediction. When those predictions work, you make new predictions. When they don't, you revisit the observations, attempt to improve your understanding of the underlying processes, and make a new prediction. And so on. In the case of climate models, this is complicated by the fact that the time scales involved need to be long enough to average out the short-term noise, i.e. the chaotic sequences of 'weather' events. Luckily, we have past climate changes to test the models against. Even more to the point, successful climate predictions have actually been made in past Senate hearings. The figure at the end of this comment by Jim Hansen demonstrates that projections of global mean climate presented in a 1988 senate hearing (17 years ago) have actually been right on the money ...

jon posted this at 4:55 PM
Share on Facebook| email to a friend

 Subscribe to RSS feed for Weblogsky