[onewhippedpuppy]

Originally Posted by RWebb

wow - I thought you were an engineer??

They used advanced statistical analysis to partition the variance in an observed phenomenon and were able to explain 50% of the variance.

Even look at crack propagation? That is one similar thing off the top of my head.

or turbulence - eddy sizes, yada yada

Yup. Do you want to fly on an airplane in which I can accurately predict 50% of a cause and effect relationship?

The entire field of climate science is fuzzy because it's all theory backed by a very small sample size (relative to the entire data set aka the life of the earth). There are relatively few ways to actually test or prove a hypothesis, it is more an attempt to establish links between variables through statistics and extrapolation. We use analysis to do initial airplane sizing and design, then test the hell out of it. More often than not we discover that the analysis was not perfect and drive design changes due to test results. In this case, there is no way to definitively test the analysis. Hell, meteorology struggles to accurately predict yearly temperature and rainfall patterns.