What’s wrong with microphysics in climate models, and how can we fix it?

I will show examples where microphysics tuning can have dramatic effects on climate projections (ERFaci and cloud feedback) in global models, and where the model default appears to conflict with what we know about microphysical processes. But in each of those cases, on closer inspection it turns out I don't have much confidence that the models are wrong and I am right. In general, the answer to “what's wrong with microphysics?” is “so many things”, but that doesn't necessarily give us a clear path to solving the climate problem.

In addition to “what's wrong”, I'll remind ourselves of two additional questions that, perhaps, will get us closer to solving the problem. First, how can we use global models with parameterized microphysics to cross-check other lines of evidence on climate projections and produce more reliable syntheses of multiple lines of evidence? And second, how “good” is “good enough” for parameterized microphysics so that we can fit it together with other good-enough pieces of the climate projections puzzle (observations, theory, and the other parameterized physics in models)? My gut tells me that we already have a collection of good-enough puzzle pieces for a much less uncertain picture of future climate to emerge than we think.