In message
"Alastair McDonald" k wrote:


My point is that chance and chaos are not the same, but I must admit I
have not explained it well, if at all :-( Briefly what I am suggesting is
that chaotic is more random than random!

This is an idea I have been developing over the last year, but not having
fully explained it to anyone it is not surprising you do not get it. In fact
many may not acept it even after I have explained it!

One analogue for chaos is turbulent flow as opposed to laminar flow. I
like to think that mountain streams (Scottisn burns) are a good example
of chaos whereas the slow moving English River Stour is a good example
of laminar flow.

When I was in Scotland at Christmas I examined a burn formed by the
heavy rain. As I had expected, while the general flow was down hill, but
when it hit a rock the water rose up in a stationery wave. Eddies resulted
in some water travelling in the opposite direction to the general flow. Just
as it is with water in a burn, so it is with temperature during the the global
warming of the planet. Most areas slowly warm, but some nearby cool.
Others leap about, being warmer than average one year and cooler the
next.

It is this leaping about which distinguishes chaos from random. If the
climate was to warm randomly, one might expect a 0.1C increase in
one year and 0.2C the following year, and 0.15 in the next year. What
we are getting is the 0.2C in the first year and 0.15C in the following
year, but -0.2 C in the year that followed that one. (1997 - 1999). It is
the chaotic nature of weather and climate, which seems random, that
makes it so much more difficult to spot the trends.

Does this make any sense?


I'm no statistician, but I believe that mean temperatures will conform to a
"normal" distribution.

You are simply saying (I think, and without quantifying it)) that the
standard deviation of a set of "chaotic" samples is greater than that of
"random" samples. But "standard deviation" should be meaningless if the
distribution is random.

The fact is that the distribution of mean temperature values for a given
month is NOT random, since the greatest number of samples are close to the
mean, so that a frequency curve has a maximum at that value. If the
distribution was random, the frequency curve would be close to a straight
line, and any value would be just as likely as any other. That this is not
the case is clear from looking at any set of monthly mean temperatures.

Martin


--
Created on the Iyonix PC - the new RISC OS computer.