The "predictive sciences" are fascinating. I agree with Chuck about
how modelling programs can become so complex that they fail under
their own weight. The problem is that you start out with a limited set
of rules for a limited set of algorithms. As time goes on, you add new
rules and more algorithms. Almost without exception, instead of going
back and fixing the rules or algorithms, we add more to it.
The ancients had their astronomical clocks. The Eudoxan system added
much complexity to an otherwise simple system to try to accomodate for
the seemingly random motion of the planets as they scooted back and
forth across the sky. What happened is that additional algorithms were
added without the refining of the underlying rules. Although more
accurate, the problem was that the basic assumption of earth-centric
astronomical motion was wrong.
So, here we are with our climate change prediction models. They are
highly complex, but just like it was a couple thousand years ago, the
basic assumptions of the mechanics of the universe are likely wrong.
We're looking at the symptoms and are trying to apply causes to those
symptoms. That all makes sense, and everything, but just as with the
astronomical clocks of the ancients, at some point you've added so
much complexity to it that you lose sight of the basic motions
involved and you have to go all the way back to zero and start over.
It takes a guy looking at Jupiter through a new invention called the
Telescope to have the "DUH!" moment which completely changes the
science.
One assumption, which is likely very wrong is the long-term stability
of the ocean currents and exactly what the base-line is that we should
be setting our algorithms to.
--
Ken Norton
ken@xxxxxxxxxxx
http://www.zone-10.com
--
_________________________________________________________________
Options: http://lists.thomasclausen.net/mailman/listinfo/olympus
Archives: http://lists.thomasclausen.net/mailman/private/olympus/
Themed Olympus Photo Exhibition: http://www.tope.nl/
|