Climate Change and the Scientific Method

P for Paul

For once and for all, let’s put to rest the relationship between climate change studies and the Scientific Method (hint: none.)

So frequently we hear from the so-called “Scientific Community” that perceived climate change is the result of human activity since the industrial revolution. This conclusion cannot be the result of application of the Scientific Method which requires that experiments be repeatable. The issue concerns the obvious, that there is only one record of climate change for planet earth.

Repeating the experiment of observing climate change requires, therefore, that the evolution of the earth be repeated, many times in fact, to accumulate reliable statistics. One could argue where these reruns should start, maybe 500 million years ago at the onset of life, or better, back to the origin of the solar system. Why stop there, however? Let’s say to rerun history from the Big Bang.

First, a digression: the word “change” so often part of the discussion, implies there is some underlying function which is seen to change. This can be temperature as a function of time, or ice cap thickness, again as a function of time, for examples. It is this underlying function which climate change students (as I refrain from saying “scientists”) wish to extrapolate. “Curve fitting,” the process by which such functions are created from data, is a highly suspect practice, especially when employed to make forecasts. Remember the “hockey stick”? a worthless piece of research if ever there were one.

Now, “change” as applied to a function normally amounts to taking the derivative, a function itself one step removed from the original, or level, function. Derivatives of fitted functions are deservedly infamous for attaining extreme values. In extrapolation, derivatives can instill fear easily in the minds of those who are not schooled in the theory and practice of “robustness,” as the subject is called. Suffice it to say that small changes in the fitted function as a result of selecting an alternative data set, or even changing a single point, can result in large changes in the derivative of the function, ringing the five-alarm bell in the minds of the innocent.

The late eminent Harvard paleontologist Stephen Jay Gould recognized that the record of life on earth is but a single instance of a stochastic process, and that therefore the Scientific Method cannot be employed to analyze it. Needless to say, he advocated other methods of inquiry to infer useful results, or else he would not have had a career.

The same is true for Albert Einstein who mainly employed inductive thought in his research, including his constructions for Brownian motion, which won him the Nobel Prize, and special and general relativity.

For an excellent short article on the use (and abuse) of the Scientific Method in modern research see this offering by the Tapestry Institute, and references therein: The Scientific Method Ain’t

Leave a Reply

Your email address will not be published. Required fields are marked *