The problem of climate models....
Dr. William Briggs, who specializes in the statistics of forecast evaluation...has some choice words about climate models...
I am not skeptical that man causes changes in his environment; in fact, I argue man must cause changes (see this post). I am only skeptical about the extent of these changes and about our ability to understand them. I am skeptical of the results from climate models that are used to posit large and harmful shifts in the earth’s temperature.
The vast majority of pronouncements about climate change are based on forecasts, guesses made about the future which are conditional on the multitude of assumptions underlying the models being true and on the forecasts having only small error. My specialty is in forecast evaluation (not just climate models, but any kind), and I do not feel that climate models have shown their ability to make accurate predictions thus far. This is why I said that the “error associated with climate predictions is also much larger than that usually ascribed to them; meaning, of course, that people are far too sure of themselves and their models.”
Overconfidence is a common human trait, and it holds in scientists just as much as it does with civilians. Typically, however, the excessive surety of scientists is tempered by the peer-criticism process, which has the effect of reducing, but never eliminating, prediction error. But this service won’t work well if experts are made to feel squeamish about making their critiques because of a public browbeating by autocratic scientists, politicians, and “activists.”
There is also a shade of “groupthink”—bandwagon research—not so much with climatologists, but with the mass of secondary and tertiary investigators who use climate model output as input to their own models of economics, public health, sociology, and so on. These models invariably show what they were programmed to show: that climate change of any kind is bad. This is, of course, physically impossible; but these are not physicists who are making these remarks—which of course quickly find their way into the press—and thus they are not held accountable in that sense.
Of course, if global climate models eventually show skill, then I will believe what they have to say.