Thursday, September 02, 2004

US Climate Change Report 2004

The latest report is available here, and ought to make for interesting reading for anyone who'd like to know what the state of the evidence for global warming is. Needless to say, the evidence still seems to support the notion that the warming that the Earth has witnessed in the last 50 years is mostly man-made, but as this story points out, there are methodological difficulties with computer modelling that can't easily be resolved one way or another.

Those working in the field agree that the models need refinement. Most important, they say, is improving understanding of the aerosols spewed by smokestacks, unfiltered tailpipes and volcanoes. They were once presumed only to have a cooling influence. Now, however, aerosols are known to cause both cooling and warming, depending on their color and composition and how they affect clouds. Those properties are slowly being incorporated in the simulations.

But several experts on aerosols published a paper in the journal Science last year warning that such particles were so poorly understood that there was no way to incorporate them into models without adding greatly to uncertainty in the results.

Given the uncertainties, the authors, led by Dr. Theodore L. Anderson of the University of Washington, said the one-degree warming in the last century could just as easily have been caused by inherent variability in the climate system as by greenhouse gases.

Dr. Richard S. Lindzen, an M.I.T. meteorologist and longtime critic of climate models, said the designers essentially chose values that produced the desired result: significant future warming trends.

Model designers defend their work, saying there is no way to choose values for the forces affecting climate that produce a desired curve.

Ronald J. Stouffer, a modeler at the Geophysical Fluid Dynamics Laboratory in Princeton, N.J., noted that many competing simulations had been created by independent teams using different methods - and they all showed warming.

"It is the sum of all the papers and analysis that convinces me" that humans are altering the climate, Mr. Stouffer said.
As I've said before, I think the weight of the evidence is clearly on the side of those who propose that human activities are now the main driver of the rise in global temperatures, but as this excerpt shows, not all critics of the models used can be easily dismissed as ignorant know-nothings with axes to grind, as so many scientifically-illiterate activists are wont to do. The problem with modelling anything as complex as long-term global weather is that no matter how well one's model may fit with the data one already has, there's no easy way of knowing ahead of time whether the correspondence with reality will continue going forward, or whether all one has is just a fancier form of Ptolemy's epicycles, equants and deferents, lacking in any underlying physical justification and unlikely to hold up in future without ever more fudges to keep it in sync with the facts.