WELL WRITTEN AND THOROUGH: When is a climate model useful?

When I was teaching -- and as a system engineering lead in industry -- I would always emphasize that a model was at best an abstraction of the most well-known, well-understood, and prominent characteristics of the system being modeled.

That being said, the point I always tried to make (and that most of my students and coworkers missed) was that the less well-known, less well-understood, and less prominent characteristics not modeled were exactly the ones most likely to make the model crash and burn in the real world.

Remember that as you read Charlie Martin's post.