There is a saying in science and engineering, "All models are wrong but some are useful." For example, if one creates a mathematical model to represent some phenomenon based on data, one usually finds the "best fit" equation. Rarely will any data point actually fall on the line or curve of the equation, so the model is wrong. However, using the model can be useful to help predict behavior within a degree of error as long as one stays within its useful range. For example, I could write an equation that matches the data for housing prices for say the time period from 1990 to 2000 and that equation would be useful to predict housing prices for 2004. But try to use it for 2005 and it would not be useful at all - remember the pop of the housing bubble? One could also use the model to estimate the price of a house (+ or - some degree of error; perhaps $3000 for example) in the past; again as long as one stays within the applicable range.
The problem with climate science is that it is based on models that haven't demonstrated any predictive value. Their predictions for the future - even near-term - have not proven to be accurate. Further, they should be able to estimate such things as past warming and cooling periods but they don't.
"All models are wrong but some are useful." Unfortunately, climate models aren't useful. Basing "settled science" on them is anything but scientific.
No comments:
Post a Comment