Global warming and structure versus behaviour
Posted: Fri Aug 01, 2003 2:25 pm
Thanks to Coyle (SD4377) for an opportunity to make, what I think is a
very important point about global warming in light of the philosophy of
system dynamics.
Coyle utters scepticism about global warming ("if that exists except in
the minds of the global warming industry"). He refers to a 20000 year
long time-series to make the point that the recent temperature increase
cannot be seen as anything but a natural variation in world
temperatures. By itself that comment is correct and it is a point often
missed by those who use the recent temperature increases as a proof of
manmade global climate change.
However, as system dynamicists, our first priority should be to look at
the structure of the system. For more than hundred years it has been
known that carbon dioxide works as the glass of a greenhouse, trapping
heat. This is a well established fact. The burden of proof lies on all
those that claim that this is wrong or that other natural mechanisms
(feedback loops) will take over the control of the climate and keep the
temperature at the current level. As far as I know, there is far more
uncertainty about the working of such mechanisms than about the heat
effect of greenhouse gases.
In statistical language, think of the problem in light of Bayesian
decision making. Our prior knowledge suggest global climate change.
Uncertain measurements of temperatures, over long intervals, imply a
likelihood function with a wide spread. Thus, we need more data about
temperature changes before the data start to dominate the prior give by
the physics of greenhouse gases.
In more practical language, think about the following example. You enter
a cold cottage and light a fire in the woodstove. After a few minutes
you measure the change in room temperature, and you notice no
difference. Should you use your time-series data to conclude that the
woodstove does not work? No! Both your experiential knowledge and if you
want, established laws of physics, suggest that if you wait, you will
see that the temperature increases. While waiting for a delayed effect
(the stock to fill up), the time-series data are of little help.
Still we see that both those that believe and those that do not believe
in global warming use time-series to make their points. The discussions
are at times very heated. To rephrase Shakespeares "Much ado about
nothing", one could say: "Much ado about the wrong thing".
The above is a point we stress in our system dynamics education in
Bergen, because it is so natural, even for those with a system dynamics
background, to accept limiting boundaries for discussions about dynamic
issues. Not only must we train ourselves to understand that structure
leads to behaviour, we must also repeatedly remind ourselves that one
cannot discuss one in isolation from the other.
My best regards,
Erling Moxnes
From: Erling Moxnes <erling.moxnes@ifi.uib.no>
very important point about global warming in light of the philosophy of
system dynamics.
Coyle utters scepticism about global warming ("if that exists except in
the minds of the global warming industry"). He refers to a 20000 year
long time-series to make the point that the recent temperature increase
cannot be seen as anything but a natural variation in world
temperatures. By itself that comment is correct and it is a point often
missed by those who use the recent temperature increases as a proof of
manmade global climate change.
However, as system dynamicists, our first priority should be to look at
the structure of the system. For more than hundred years it has been
known that carbon dioxide works as the glass of a greenhouse, trapping
heat. This is a well established fact. The burden of proof lies on all
those that claim that this is wrong or that other natural mechanisms
(feedback loops) will take over the control of the climate and keep the
temperature at the current level. As far as I know, there is far more
uncertainty about the working of such mechanisms than about the heat
effect of greenhouse gases.
In statistical language, think of the problem in light of Bayesian
decision making. Our prior knowledge suggest global climate change.
Uncertain measurements of temperatures, over long intervals, imply a
likelihood function with a wide spread. Thus, we need more data about
temperature changes before the data start to dominate the prior give by
the physics of greenhouse gases.
In more practical language, think about the following example. You enter
a cold cottage and light a fire in the woodstove. After a few minutes
you measure the change in room temperature, and you notice no
difference. Should you use your time-series data to conclude that the
woodstove does not work? No! Both your experiential knowledge and if you
want, established laws of physics, suggest that if you wait, you will
see that the temperature increases. While waiting for a delayed effect
(the stock to fill up), the time-series data are of little help.
Still we see that both those that believe and those that do not believe
in global warming use time-series to make their points. The discussions
are at times very heated. To rephrase Shakespeares "Much ado about
nothing", one could say: "Much ado about the wrong thing".
The above is a point we stress in our system dynamics education in
Bergen, because it is so natural, even for those with a system dynamics
background, to accept limiting boundaries for discussions about dynamic
issues. Not only must we train ourselves to understand that structure
leads to behaviour, we must also repeatedly remind ourselves that one
cannot discuss one in isolation from the other.
My best regards,
Erling Moxnes
From: Erling Moxnes <erling.moxnes@ifi.uib.no>