[cogitator]

Brief reply: the question regarding the current state of climate models is one of spatial and temporal resolution. The current state of the GCMs is that they have sufficient spatial and temporal resolution to ascertain the primary factors driving global climate with reasonable accuracy. They don't have sufficient spatial and temporal resolution to accurately reproduce annual meteorological variability. Does that make sense?

[palmer]

The GCM's have enough accuracy to model large scale phenomena that result from large scale continuous inputs. For example the ENSO phenomenon can be modeled from coarse simulations of the atmosphere, land and oceans. That coarseness is acceptable in both the space and time domains since ENSO takes years and happens over a huge area. The question is whether global climate can be modeled the same way. The key issue is whether forcing can be modeled in a coarse time and space model, or does forcing depend on accurate modeling of local meteorological phenomena. The answer is pretty simple when you look at what forcing is: it is the postulated increase in water vapor due to the increase in CO2. How that increase affects warming is the key question that that the models need to answer.

The leftover lenticular clouds from my MCS yesterday either had no effect on climate or they had an effect that a model can average over space and time by adding a parameter. Will there be fewer or more MCS's with increased water vapor? How will their cooling affect larger scale weather and climate? We don't get the answer from the model since they aren't modeled. We don't know how THE primary factor driving global climate leads to a mixture of cooling and warming over what areas. That does not lead to a global climate prediction of "reasonable" accuracy at all.