Climate Science: Observations versus Models

by Richard K. Moore
Rkm documents & articles
January 8, 2010

Science and models

True science begins with observations and measurements. These lead to theories and models, which lead to predictions. The predictions can then be tested by further measurements and observations, which can validate or invalidate the theories and models, or be used to refine them.

This is the paradigm accepted by all scientists. But scientists being people, typically in an academic research community, within a political society, there can be many a slip between cup and lip in the practice of science. There are the problems of getting funding, of peer pressure and career considerations, of dominant political dogmas, etc.

In the case of models there is a special problem that typically arises. That is, researchers tend to become attached to their models, both psychologically and professionally. When new observations contradict the model, there is a tendency for the researchers to distort their model to fit the new data, rather than abandoning their model and looking for a better one. Or they may even ignore the new observations, and simply declare that their model is right, and the observations must be in error.

A classic example of this problem can be found in models of the universe. The Ptolemaic model assumed that the Earth is the center of the universe, and that the universe revolves around that center. Intuitively, this model makes a lot of sense. On the Earth, it feels like we are stationary. And we see the Sun and stars moving across the sky. “Obviouslyâ€