Should we make trillion-dollar social and economic policy based on playing a video game? You need to understand climate science for what it is.
We used to do science by assembling hypothesis into a falsifiable claim which could undergo a physical test, known as a reproducible experiment. Our cherished scientific method has served us well for hundreds of years. Why would we think we don’t need it anymore?
Climate science is essentially a simulation model in a computer. Climate scientists conduct experiments on models, not reality. They use models to create data and parameters for other models. Yes, you will find some physical science in these models. But mostly you will realize the models are highly dependent on parameters (approximations of physical processes which may be poorly understood) created to make the models sort of work by tuning and forcing them to fit with the late 20th century.
How well do these models represent reality? Short answer is that no one knows.
Here’s a thought experiment. On a scale of 0 to 9, a good flight simulator (simulation model) can score a 9. Flight simulators go through a rigorous certification process so we can trust them enough to train and test real pilots flying real airplanes in the real world. Where would a climate model fall on this scale? Close to zero could be a reasonable estimate.
You should not be surprised at this. We have a long way to go. Climate is a massively complex undertaking. Modeling climate is fraught with chaotic processes and non-linear relationships. Differentiating between causality, correlation, coincidence or coercion is problematic. Yes, I know there is some fundamental small and medium scale physical science around the effects of greenhouse gases and individual climate sub-processes. But with complexity, the sum of some parts does not make a coherent whole. Climate is an emergent property that, unlike short-term localized weather, does not lend itself to numerical, deterministic modeling.
Understand Climate Science – Nullius in Verba
I am a reasonably informed skeptic, which is the essence of scientific method. Back in 1660, Britain’s Royal Society came up with the motto Nullius in Verba, which essentially means “take no one’s word for it.” So, as I attempted to understand climate science, I was astonished to learn that it is impossible to verify or validate these models.
If you take the time to understand climate science, you must realize that if you can’t trust the models, you can’t trust the science, because it’s just a model.
If you refuse to take just my word for it (and rightly so) read Paul Edwards wonderful book A Vast Machine: Computer Models, Climate Data and the Politics of Global Warming, MIT Press, 2010. Edwards lays out the hundred and fifty year history and processes of climate science succinctly, and he is far from a denier. So I was astonished when he states:
“Ultimately, neither comparing climate models with observations nor comparing them with each other can prove that they are correct. Nor would any climate modeler ever make such a claim.”
I think this explains why we hear so much about “the consensus” and “the science is settled.” Consensus is a political, not scientific process. Real science is never settled. So, if the entire climate science project stands on a foundation of models that cannot be validated or verified – let alone certified – perhaps all you have left to fall back on is groupthink and exclusion of informed skepticism.
Climate Science – What I Know
Is climate changing? Yes, of course. It always has.
Do CO2 emissions affect climate? Probably, but how much is largely conjecture laden with uncertainty. Conversely, the effect of reducing emissions is unknown in both magnitude and timing.
Should we not be applying the precautionary principle, just in case? Okay, but the correct application of precaution in the first instance should be adapting to change rather than the hubris of thinking we know how to control it.
In its present form, climate science is largely a video game created to provide cover for a massive social engineering project. At very best, understand today’s climate science as a semi-empirical theory that cannot be tested. Let’s continue to evolve and improve the models using science that’s separated from ideology, imbued with integrity and objectivity i.e. the long-cherished scientific method.
In the meantime, let’s stop playing Sim Earth.
Value informed skepticism. Without it, we would still follow the consensus views that our planet is flat and orbited by the sun.
John, an excellent piece that rightfully puts climate modeling in a more balanced perspective than is commonly portrayed in the media and by advocacy groups. While some of these models can, to a certain extent, be validated by empirical observations, there is still a lot of uncertainty that the models cannot explain. The climate is changing as you say but formulating expansive and expensive policy prescriptions based on these models is fraught with problems and far too little attention is given to adaptation, as you point out. Complex systems are inherently unpredictable as you suggest and so a healthy degree of skepticism about these models is warranted. Hope to see more interesting posts from you in the future!