--APlYHCtpeOhspHkB
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
Content-Transfer-Encoding: quoted-printable

On Wed, Apr 22, 2009 at 10:35:09PM +0900, Phlip wrote:
> Eleanor McHugh wrote:
> 
> >My main criticism of climate science's attempts to model global  
> >warming is that it's fundamentally flawed in its methodology. The  
> >climate is a chaotic non-linear system and hence the only way to make  
> >accurate predictions is to understand its exact starting conditions.
> 
> The _weather_ is chaotic. The _climate_ is the walking average, beneath the 
> turbulence. The point of the Butterfly Effect aphorism is you _can't_ 
> predict whether the butterfly's wings flapping in Africa will create a 
> hurricane in the Atlantic, even if you _did_ know how its wings flapped in 
> microscopic detail.

Climate is an emergent property of weather trends.  Climate is an average
of a chaotic nonlinear system.

. . . and while you may not be able to predict such a hurricane with 100%
accuracy (there's some question about the predictability of anything in
the universe, thanks to quantum probability effects), you can certainly
get close enough for government work if you know the starting state of
all contributing factors and have time and processing power to build the
model.  This, of course, ignores the possibility of free will having
unexpected influence, but I don't think butterfly wings conform to the
definition of "free will", so that's not really the point here.

The problem, as Eleanor pointed out, is that climatologists' models are
nowhere *near* that level of precision and comprehensiveness.  The state
of the art of climate modeling is something more like witchcraft, or
horse race betting at best, at this point.  Maybe in a hundred years
we'll have achieved all the accuracy of card counting at the blackjack
tables in Las Vegas.


> 
> However, you can predict trends over time, and you can predict the effect 
> of forcing (such as more sunspots, or more CO2) on those trends. For 
> example, Katrina occurred at the peak of a decadal cycle in hurricane 
> activity. That doesn't mean CO2 didn't have a role...

You can make predictions about trends over time -- but the less starting
state data you have at your fingertips, the less those predictions will
actually tend to accurately represent what will happen.


> 
> >Instead researchers take patchy historical data over a brief  
> >geological period and feed that into incredibly complex climate models  
> >to produce predictions.
> 
> That's the prediction phase of the experiment. Then they confirm their 
> predictions by correlating them to historical data, such as ice cores in 
> glaciers containing the predicted amounts of certain chemicals at certain 
> depths.

I wonder how many climatologists haven't figured out yet that correlation
doesn't strictly imply causation.


> 
> >Am I the only one reminded of Ptolemy's model of the heavens? A  
> >stunning intellectual achievement, but as Copernicus demonstrated so  
> >effectively, a complete fiction which prevented astronomy from  
> >progressing for a good fourteen hundred years.
> 
> You have both the history and details wrong. Copernicus did not 
> _demonstrate_ anything - he simply published the alternate view, which had 
> already existed. And (the ghost of Carl Sagan notwithstanding), Ptolemaic 
> models did not "prevent astronomy from progressing". You could still 
> predict, over time, where Mars and Jupiter would appear in the skies. 
> That's as old as Astrology.

The publication of an alternate view that effectively debunked the
preceding view was, in effect, a *demonstration* that Ptolemy's model was
a load of hooey.

It's true, though, that Ptolemy's model didn't "prevent" astronomy from
progressing.  It just significantly hindered such progress.

-- 
Chad Perrin [ original content licensed OWL: http://owl.apotheon.org ]
Quoth Anonymous: "It is easier to measure something than to understand
what you have measured."

--APlYHCtpeOhspHkB
Content-Type: application/pgp-signature
Content-Disposition: inline

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2.0.10 (FreeBSD)

iEYEARECAAYFAknwxwgACgkQ9mn/Pj01uKVknQCfYwp5xIzWF1tYXRRG6FuX0I4h
fmwAn1S6GbQOtsnBzJ1OWzItkCy6EHWu
74
-----END PGP SIGNATURE-----

--APlYHCtpeOhspHkB--