Climate Scientists Encounter Computer Models' Limits.

 Front page headline in Monday's Wall St Journal.  I could have told them that 20 years ago.  I used to write computer models.  A computer model is just a computer program written to predict what will happen in the future.  If the model fails to give the answer the model writers want, they change the model code until they get the answer they like.  

This is not really science, not as Roger Bacon described science back in the high middle ages.  Science makes observations and preforms experiments.  The results are used to suggest or back up general theories.  When enough good scientists deem the theories true and the evidence sufficient, the theory becomes part of generally accepted science.   The models don't do experiments, and they tend to "correct" the observations they get fed.  I cannot call a model's results scientific.  They are more like science fiction.