[Phys-L] Re: global warming

[dtmarx at ILSTU.EDU (David T. Marx)]

In response to John Denker's comments...

> David T. Marx wrote:
> > Do you find data like that shown at
> > http://www.koshland-science-museum.org/exhibitgcc/historical05.jsp
> > convincing?
>
> Well, let's think about that.  Suppose I roll a die, once,
> and it comes up "4".  Do I take that as evidence that the
> die is loaded?  No.  Do I take that as evidence that the
> die is not loaded?  No.

        I do not think temperature measurements are the same as a random process such as the rolling of a
die or the number of counts I get per observation interval from a beta decaying sample.  Either
there is a global warming trend or there is not.  It depends on the energy balance of the Earth.
If less EM radiation is reflected, due to decreasing albedo and more EM radiation is incident due
to increased solar output, then we might expect to observe some warming.  If increased carbon
dioxide in the atmosphere results in more infrared radiation being trapped, then we might expect to
see some warming.

> OTOH suppose I perform thousands of rolls, and they are,
> as a whole, overwhelmingly biassed toward higher than
> average numbers.  That starts to look like evidence.

That's my point.  There is not overwhelming evidence that global temperatures are increasing.  Some
methods show increases, others do not show significant increases.  And my point is that people will
look at one set of data and draw a conclusion just because a bias toward believing in the global
warming hypothesis.  People also do not bother to consider the methods used in measurement to make
an objective determination.  Many seem quite willing to accept the data on its face value.

> So ... It comes down to what you mean by "data like that".
> If you take _all_ the available data like that, it adds up
> to something pretty scary.

Not really.  I have looked at a lot of the peer reviewed literature and came to the conclusion that reliable, long term physical data is very hard to come by.

> The evidence for what's causing the warming is not as
> overwhelming than the evidence for the warming itself.
> It is also not entirely relevant.

It is entirely relevant as policy makers are considering whether or not we need to strongly
regulate "greenhouse gas" emmissions.  Please do not get me wrong on this - I do want to see
worldwide, reduced pollution emissions of all kinds - including carbon dioxide - but I'd like it to
be because we think poluution controls are a good thing, not because of some unproven, scary
warming scenario.

> > I might point out that meteroligists attempt to predict the weather
> > for the next five to seven days using such models.  They are right
> > about 20-25% of the time.
>
> That seems short on common sense and long on axe-grinding.
>
> If the task is to predict whether next week will be
> warmer or cooler than this week, I can get it right
> 50% of the time by tossing a coin.  If Prof. Marx's
> assertion is true, then I can get it right 75-80%
> of the time by taking the "meteoroligists" prediction
> and predicting the opposite.

I didn't make this up.  I read these numbers in a recent report from the American Meteorological
Society.  I was not just referring the temperature predictions, but also predictions of
precipitation, cloudy/sunny, etc. in the 5 to 7 day range.  In the past decade, the ability to
predict temperatures 3 to 5 days out has improved substantially.  For a better source on this, see

http://www.ametsoc.org/policy/weaforc.html

My point on this is that how can we rely on climate models to determine global temperatures 50 to
100 years in the future, when we can't even predict temperatures next week or next month?