There is video available here of a lecture given by Professor Mike Hulme entitled "How do Climate Models Gain and Exercise Authority?". Hulme asks whether deference towards climate models is justified and whether we should have confidence in them. I think the answer is "We don't know".