By FRNash/PHX, AZ on Wednesday, November 21, 2007 - 03:51 pm: Edit Post
Somehow that statement triggered a rusty old corner of my grey matter something I have been intermittently ruminating on for several years, it just never gelled 'til now.
Maybe the ultimate explanation for that is that weather (and climate) are perhaps the most commonly encountered examples of chaotic systems, (see: Chaos theory) in that they are:
This realization in fact clearly demonstrates the problem of developing long-term weather forecasts, even with the most sophisticated computer models, running on the latest and greatest platforms.
Sensitivity to initial conditions, you say?
What and when were the "initial conditions" for the great "mother of all weather forecasting models"? Somewhere along the continuum from the initial "big bang" to relatively recent human history?
Not possible! So by definition the initial conditions used represent an approximation of some intermediate conditions rather far along along that continuum.
So there we are, right smack dab in the middle of the "butterfly effect" as in the work of Edward Lorenz, who in 1961 was using a numerical computer model to rerun a weather prediction, when, as a shortcut on a number in the sequence, he entered the decimal .506 instead of entering the full .506127 the computer would hold. The result was a completely different weather scenario!
Good luck nailing those "initial conditions", even for the "Maka daidai shōgi" (摩訶大大将棋) equivalent "big mother" computer model!
Not to mention that what we are attempting to model here is largely a stochastic process the behavior of pressure in a gas.
By John Dee on Wednesday, November 21, 2007 - 06:57 pm: Edit Post
I guess my main comment to this would be if Lorenz's point was to show the vast complexities involved with trying to model the atmosphere looks like he did it.