General Discussion

I am pretty sure that any model considering this would be awash in unconfirmed and probably anthropologically biased guesses

When building a model about how Moose move about their home range we made many wags, which for the uniformed means wild ass guesses

I am pretty sure that mathematical skill wouldn't and couldn't correct for any BAD assumption that went into the building of the computer code.

I suspect, but never looked at the model, that what it most likely does is produce projections built around common sensical guesses.

Analytical models have few variable, lack reality, but can be solved by applying calculus. I can't imagine this problem with a handful of variables which whould make the calculus tractable.

Numerical models are often plush in variables, for which there is little data to characterize mathematical relationships with other variables. I can imagine a group of modellers thinking they were including all the variables THEY felt were important.