Climate scientists are the opposites of weathermen — the further out they’re asked to forecast, the more confidence they tend to have in their predictions.I think a lot of people don't understand this, including meteorologists used to wresting with the next day or two's forecast. In physics there are a great many things that you can calculate over large spatial scales and/or over long time periods, that you cannot calculate for small distances or short time frames -- like essentially all of thermodynamics and statistical mechanics. Or fluid flow. Or cosmology. Physics is very good at dealing with very simple systems, like a hydrogen atom or Mercury's orbit around the Sun, and large things, like the average temperature of a body of water or the properties of a universe. It's all the stuff in between that's difficult, and that's where most of climate science falls, once you get past the basics of energy balance. We probably knew enough to be wary of the dangers of fossil fuels shortly after Arrhenius did his calculation in 1896 -- and if he'd lived in India instead of Sweden, he might not have thought the warming would be so beneficial.
Physics leads to an argument like this, a comment on John Baez's blog Azimuth: