Monday, February 21, 2005

Does a Flawed Hockey Stick Invalidate Kyoto?
Although I know I've loaded up the blog with a number of postings on climate change recently, I also don’t want to ignore the controversy stirred up by the Wall Street Journal's recent article on the famous (or infamous) climate change “hockey stick” graph. The Journal's editorial last Friday is spot on in its concerns about the politicization of science—though this should hardly be news—and the possibility that we may be embarking on expensive efforts to mitigate a phenomenon that is not entirely proven. But what should we do instead?

In a nutshell, the chart in question has provided a sort of “smoking gun” visual argument in support of the idea that climate change is both real and, coinciding as it does with the rise of industrialization, largely man-made. This chart has come under fire from a small number of scientists and non-scientists on the grounds that it cannot be exactly reproduced from the raw data in question. Reproducibility of results is at the very core of all science. So does this undermine the entire notion of global warming, making the Kyoto Treaty worse than irrelevant?

It's entirely possible that Dr. Mann's graph will turn out to be fatally flawed, though my statistics background isn't quite up to forming a solid opinion. But the chart in question is only one piece—albeit a visually compelling one—in a much larger mosaic of scientific observations adding up to a comprehensive picture of a global climate that appears to be shifting out of its normal range of variance. The evidence includes some compelling recent studies. Just as it is probably no coincidence that this shift seems to correlate with our rising consumption of fossil fuels, neither does it seem coincidental that the rise of our present civilization should occur during a relatively stable and benevolent climatic phase.

As I've indicated before, for me it boils down to risk management. If we accept the Journal’s estimate of the cost of the “insurance” at $150 billion/year, what price tag would we assign to the events against which we are insuring? What would be the cost of irreversible changes in the location of the Gulf Stream and temperate zones, or the permanent loss of even a quarter mile of coastline globally, due to rising sea levels? I don't know, but I can visualize a number with many more zeroes than in the figure cited above. None of this is certain, but neither is the “null hypothesis” that the climate isn't changing--or that it is, but without any human influence. What are we willing to bet?

No comments: