Re: [asa] Creation Care

On 1/24/07, Bill Hamilton <williamehamiltonjr@yahoo.com> wrote:
>
>
> I agree that the hardware -- and the software -- for modeling have
> improved
> significantly. And that allows us to model more effects that we used
> heuristics
> for before. But climate models are still complex enough that if we miss
> anything, all the increased capability will do will be to enable us to
> produce
> garbage much faster. And I suspect that it's easy to miss effects in
> climatological models. Since the models are validated by comparison with
> historical data, an effect that shows up outside the time baseline may be
> missed. And given the chaotic nature of weather, _any_ omission --
> dynamical or
> boundary condition -- is likely to cause significant departure from the
> true
> behavior.

It is not the case that _any_ omission will cause a problem. A helpful
analog comes from control theory because you can view the climate a series
of positive and negative feedbacks. The question that is most at issue --
and cannot be computed because the system is too complicated -- what is the
Nyquist stability of the "system"? A classical example of an unstable
system is microphone feedback. You cannot eliminate positive feedback so
electrical engineers add negative feedback into the system to stabilize it.
An unstable system is sometimes referred to in the popular press as a
"tipping point".

Getting back to climate modelling. The post ice age climate has been stable
and the CO2 has been well bounded -- at least in the pre-industrial age.
Positive feedback from melting glaciers and methane from the permafrost has
been minimal. Thus, we have been able to effectively model the climate. If
we go back 3 Mya our models are not as good. The models seriously
underestimate the temperatures. You have to go back that far just to get
analogous CO2 levels with today, but when we plug in our current estimated
climate sensitivity the predicted temperatures are much lower than
the measured ones. We don't know why this is but maybe the models do not do
well modelling the rapid changes caused by the positive feedback. Another
modern example is the positive feedback of the polar ice melting and methane
release from the permafrost.

Climate scientists are well aware of this problem. Strategic Research
Question 4.3 of the Strategic Plan for the U.S. Climate Science Program is:

> What is the likelihood of abrupt chages [RDB note: read positive feedback
> mechanisms] in the climate system such as the collapse of the ocean
> thermohaline circulation, inception of a decades-long-mega-drought, or rapid
> melting of the major ice sheets?
>

So, yes, there is a danger that the models will miss predictions due to not
modelling all positive feedback effects. But, this just means that the
danger is mostly in underestimating global warming and not overestimating
it. This is not a theoretical risk either as recent experience has shown
that current sea level increases are on the high side of the models. It will
be interesting to see how this topic is dealt with in the upcoming AR4.

To unsubscribe, send a message to majordomo@calvin.edu with
"unsubscribe asa" (no quotes) as the body of the message.
Received on Thu Jan 25 14:08:12 2007