Evaluating the Models

There was a provocative piece in New Scientist in the spring (sorry, can't give you a citation, it doesn't seem to be on line) discussing the way in which scientists tend to evaluate new evidence based on their preconceived notions, giving stronger weight to evidence that is supportive and discounting evidence that is in conflict. This should in some sense be obvious, though it tends to rub scientists, with their belief in the purity of their own effort, the wrong way.

There's a marvelous demonstration of this principle in action today from our friends at Tech Central Station on potential meteorological effects of wind farms. The authors are Willie Soon and Sallie Baliunas, whose names will be well known to those who follow the climate wars:

A new simulation finds serious and previously unrecognized environmental threats from massive wind farms in the American Great Plains.

A recent study published in the Journal of Geophysical Research by scientists from Princeton and Duke Universities indicates massive wind farms would significantly increase local surface drying and soil heating, which in turn would impact agricultural or range use on or near the wind farm.

Soon and Baliunas are well known skeptics of results of computer climate simulations suggesting that fossil fuel burning could negatively affect climate, frequently saying such things as:

The answers must come from the application of the scientific method, which requires testing the models against good measurements from the real world. A model can make correct predictions if accurate observations validate it. (emphasis added)

The JGR wind farm study (which, as an aside, looks like a good and interesting study - I'm not trying to criticize it here) is entirely based on model runs, with no attempt made to empirically verify it against real-world data. It's not really possible, given that the size of wind farm they're trying to model does not exist. While that kind of science really seems to cheese off Soon and Baliunas when model results are used to criticize fossil fuels, that apparently is not the case in the wind farm study, where they are quick to conclude:

Wind farms may not be as benign to the environment and weather as its (sic) promoters say.

Posted by John Fleck at November 11, 2004 08:09 PM