Sunday, September 14, 2008

Global Warming and Complex Models

Actually, the people at realclimate.org do have a disdain againts engineers, or it is just this one physicist (though he is a guest commentator, so it figures).
Actually, he thinks that engineers are so stupid that they only can calculate a very straightforward one-iteration equation problem.

He dismisses a whole field of engineering called: Numeric simulations. He later adds that this is the current method used. He never refers to numerical simulation, but rather describes what it is. You simulate the earth as a given set of layers (Elements) which have different properties and governing laws and then the computer solves the differential equations resulting from them with a specific method.

And finally, he cannot (or will not) accept that any model needs an experimental proof. However, he says that it suffices to have a model that includes all the factors someone can think about, without experimentational results.
I hope this guy never has led an engineering problem or other real world constructions, because I would fear for all users.
You can't just take a mathematical-physical model stuff in your idea and take the results as a given fact.

For example.:
I was involved in 3D-broaching simulations and heat transfer modelation, which is fairly good understood in a relatively closed environment. A closed environment is here a production facility and a clean work environment and fairly good understood means +/- 5 Kelvin deviation.
I mean, we can get results that are in the range fo +/- 5°C of the real-world example and we can get our data directly from a thermo-cam in a closed environment with good boundary conditions and material data.
So, we have this deviation of 5 °C from the actual temperature rise and this is only when we assume that there is a stable equilibrium, because the process during the heating is virtually unkown (due to its complexity).

And ultimately, this is what this scientist acknowledges in the last paragraphs. Climate is too complex for a few set of straight equations and this is also not what engineers complain about. He obviously doesn't have a lot of conversations with engineers, or he wouldn't have phrased his response like that, especially engineers who worked in a simulation department.

No comments: