As usual Judy Curry has a very informative weblog post
In it, she writes the following
“What exactly does falsification of a prediction mean? For an ensemble prediction, the prediction is said to have no skill if the actual realization falls outside of the bounding box of the ensembles (or whatever skill score for whatever variable has been decided in advance). A prediction with no skill does not imply falsification or rejection of a model. Falsification of a climate model is precluded by the complexity of a climate model”
I agree with her that one cannot falsify models. However, one can falsify model predictions. With respect to climate, multi-decadal model predictions can be falsified as models themselves are hypotheses and can be tested. They are, after all, engineering code, as large parts of the physics, chemistry and biology are parametrized using tunable parameters. Only the dynamic core of these models (i.e. advection, the pressure gradient force, gravity) are expressed in terms of fundamental physics.
Thus, while models cannot be verified, they can be rejected (i.e. those that fall outside of a selected envelope around the observations). Climate models are just hypotheses like any other hypothesis and need to be tested against real world data.