Originally posted by dang65
View Post
As far as CO2 goes the simple argument is that as a radiatively active gas, if you stick some in a box and shine a light on it, it will warm up. Stick some more in, and it will warm up proportionately more.
Ice cores that go back 700,000 years show CO2 and temperature waxing and waning as the ice ages come and go, but at no point did the concentration change as quickly as it has since the industrial revolution.
Most climate models predict the predominant affect to be a shift in circulation patterns, rather than a direct warming as such. Changing the poleward temperature gradient has non-linear affects in this respect that aren't entirely predictable. GCMs are mostly dodgy, but they're the best and only tool to try and use the knowledge of physics to model a highly complex system. Most 'averaged' predictions you see in documents like the assessments by the IPCC use an average of 20 or so of the worlds best models, but look at them individually and where one shows a positive value another shows a negative, and the average results in an intuitive residual that's just about believable.
Saying that though, observation length required for statistical significance is just about getting there with some remote and in-situ methods, so I have much more faith in measured data now than say, 10 years ago.
Comment