Thursday, October 04, 2007

Climate Change (2) - Computer Models

The global weather system is incredibly complex. Scientists build computer models that attempt to represent the weather system, but these models by definition must be simplifications of reality.

I do not understand these models of the global weather system, but I am familiar with the computer models developed by economists to represent the economy of a nation. These economic modellers face the same problems as meteorologists. The economic behaviour of people and businesses is so complex that they cannot be fully represented in a computer model. Numerous simplifying assumptions have to be applied to get a model simple enough for a computer can handle.

The consequence of this simplification is that the results produced by an economic model are partly determined by the set of assumptions chosen. A different set of assumptions will produce a different set of results. This leaves the modeller in a position where they determine the results that there computer model produces. This is not very satisfactory for reaching the truth about the functioning of the economic systems.

Differing assumptions are one reason why economists frequently disagree with each other. They reach different conclusions by starting from a different place. This is also the reason why the predictions of economists are so often wrong. A factor that that is left out, or incorrectly specified in the model can have a greater effect than expected, causing some of the predictions of the model to be incorrect.

I am sure that the scientists attempting to model the global weather system face exactly the same problems. The results predicted by their models will vary according to the assumptions built into their models. The global weather system is far more complex that even the most complex economy, so their task is even hard. Most people understand that the predictions of economists are usually wrong. We should apply the same level of caution to the global predictions of the meteorologists.

Measuring the change in global temperatures is difficult enough. Determining the cause of changes is virtually impossible. Deciding what actions would reverse the change in temperature is well behond limited human capability.

No comments: