Replies

All the computer models agree. Has anyone answered the question why we need to have more than one computer model to get a consensus?

The truth is that computer modeling is only as good as the programmer who programmed it and the data that went into it.

If, for example, a call went out for people to create a computer model that will show how fast a cube of ice will melt in a glass, one group might program the model to calculate that .001 grams of the ice cube will melt every minute for every .5 degree increase in temperature, while someone else will program theirs so that .001 grams of ice will melt every minute for every .6 degree increase. The two models will then, obviously, diverge significantly at some point in time as the models are run.

On top of that, the people tasked with measuring the normal temperature change that occurs in the room where the ice cube is kept might record incorrect data. They might record that from 8:00am to 2:00pm, the temperature rose by .9 degrees every hour, when the real number is .87 degrees every hour. And if there are two separate groups recording data, the first might arrive at .9 degree per hour, and the second at .89 degrees per hour. Now, even if they went back and plugged that data into identical models, they would arrive at different conclusions, but they also have two slightly differing models, as above, and so the divergence is increased even more.

For something like weather or "climate change", the interpretation of how the multitudes of variables will interact with each other can have nearly infinite variation, and then the data gathering can be almost equally problematic.