Models have improved greatly in the past 30 years but still can't anticipate all the ways the atmosphere will respond as greenhouse gases climb. The dozen models in use today predict average temperature increases of 3 to 11 degrees by the end of the century.
Scientists and engineers love their models, but one should run a model against actual data and show that the model matches the data. This is not yet possible for geo warming models, hence the models can't be run backwards and predict things either. Expecially since we don't know what the sun was doing in the Eocene period.
With new evidence that the solar heat has been increasing by about the amount of the warming we should all realize that while climate change may be occurring, it may not be time to dump the family car. But it may be time to build higher levees.
Actually, this is just the tip of the iceberg (so to speak).
A while back I posted an article on simulation and its drawbacks. I am an electrical engineer, and I have see the pitfalls. I have also seen people tweak simulations to get what they want.
Now, if only that worked in the lab when you actually build the circuit.
http://pulse.typepad.com/countercolumn/2005/01/index.html Scroll down a little as the screen is not displaying properly.
20 years ago the same 'models' were warning us of an impending ICE AGE. These guys are always warning...just to keep the Federal Grants coming in. They could not find real productive work anywhere.