OuterBanker wrote:The data is based on a model, the GFDL no less. We all know how inaccurate it is a few days out. Imagine projecting it out a hundred years.
No, this isn't the same GFDL model as is used to predict weather. Climate modelling is a whole different animal than weather modelling, and one can't draw any conclusions about the former from any perceived innaccuracy of the latter.
An analogy I've offered before:
Suppose one is modelling the flow of water through a segment of a river. All the computing power in the world wouldn't suffice to predict every little eddy for even a few minutes (even given perfect information regarding the initial state). That's the analog to modelling weather. OTOH, one can predict the total flow of water through this segment of river to quite high accuracy for months or even years, given reasonable information about the upstream forcings (rainfall and snowmelt). That's the analog to climate modelling.
Granted this analogy isn't perfect (we probably are better at modelling total flow of rivers than we are at climate, at least so far) but it illustrates the point that even though two problems appear to deal with the same issue on two different scales the intractability of the finer scale problem does not imply greater (or even similar) intractability of the large scale problem.