One problem with Millers (and his ilks) climate modelling skepticism is that, at its base, it confuses two quite distinct issues. The question “Are climate models reliable?” is misleading if we read it as meaning there is only one purpose for a climate model. Even a very simple model, one we could write down, solve on a home PC and constrained by little more than known physical laws and the variation in global climate measured and recorded over the last 50 years (or so) may well be reliable enough for making mitigation decisions (e.g., regarding when and to what extent we should decrease emissions if we want to avoid more than 3 degrees of warming). However, even the most complex and sophisticated models, like those used in IPCC AR5 ensembles, we have available today may still be inadequate for adaption decisions (e.g., how high to build sea walls or how to adjust agricultural crop management policies, etc.). This is important because we clearly need to keep developing climate models to make reliable predictions for adaptation, in particular for regional adaption decisions, but it is unlikely that the next generation of models (or the few generations thereafter) is going to tell us anything radically different from todays models about the need to cut emissions in order to avoid dangerous climate change. Scientists are never done improving scientific understanding, and climate science is no different. Climate scientists are always trying to improve the predictive power, sensitivity and resolution (both temporal and geographical) of climate models, but that does not entail that we are not obliged to base mitigation decisions on todays best models (which, as far as global temperature projections are concerned, give pretty much the same answer as models constructed in the 1980s).
Posted on: Fri, 02 Jan 2015 01:16:42 +0000
Trending Topics
Recently Viewed Topics
© 2015