Once again, we skirt around the real issues by nitpicking a small change in slope of a very noisy dataset - one in which moving the starting point by a single year can have a drastic effect on the trend. We also muddy the issue by coming up with ‘sound-bite’ arguments or by latching onto non-scientific arguments if they happen to support ‘our’ side while heaping scorn on the opposition when they do the same.
For example, it’s common to hear the ‘consensus’ community trumpet stats like, “X of Y hottest years happened in the last decade! That proves it!”. But of course, we all agree that the earth is warming overall, and has been for a long time (long before greenhouse gases would have been a major factor). So even if the warming was just the natural warming you’d expect at this point in the intra-glacial phase, you would still expect each decade to set new records for temperature. It tell us nothing whatsoever as to whether humans are causing global warming.
And of course, making the leap from ‘global warming is happening’ to assuming that the only proper response is to tax and otherwise curtail energy is not warranted at all. There are plenty of scenarios under which it’s perfectly reasonably to argue that global warming is happening, and humans are helping to cause it, yet the proper response is to do nothing. Yet the global warming activists act as if it’s a binary choice - if global warming is real, you must do the things we advocate in order to save the planet. They simply do not want to defend those policy proposals - they would rather act as if they are the automatic choice once we assume that global warming is happening.
Finally, I still have a real problem with global warming models, for the following reasons:
-
They assume that we understand the major carbon sources and sinks and the feedbacks that predominate. And yet, almost daily we read about new research that has massive implications for our understanding of climate and the carbon cycle. Usually, as our understanding of a system matures, new discoveries happen on the margins. When new discoveries are being made that have major impacts on core understanding, you know there are still a lot more out there to be discovered. The ‘unknown unknowns’ become the dominant factor. This is especially true in complex adaptive systems (see below). Also, this type of uncertainty is not really captured by the IPCC. The uncertainty they focus on is in the known effects and their amplitude. Systemic uncertainty is another breed of cat.
-
The models are ‘tuned’ based on past data, and the predictive nature of ‘tuned’ models is highly suspect.
-
The past predictive track record of these models is not very good - especially outside the region containing the data used to tune the models in the first place.
The test for these models is often done by sequestering some of the climate data, using the rest of the data to build the model, then see if the model predicts the data it never ‘saw’. That’s a valid approach in some cases, but in a complex adaptive system it just means that you were able to predict the output of the system when it was in a past, known state. Or in the case of a highly tuned model, you may just be engaged in an exercise of curve fitting by adding terms and fudge factors.
Since the nature of a complex adaptive system changes in unpredictable and unknown ways, this isn’t much help for predicting future responses. Ask the Keynesian economists who predicted a massive recession in the U.S. at the end of WWII, or who said that having inflation and high unemployment at the same time was impossible. Or, ask the economists on the right who said that all the quantitative easing in the last decade would result in high inflation. Both used similar models (general equilibrium macro models) to make these predictions. Models that seemed to have predictive ability when applied to the past. Both were spectacularly wrong.
- They assume a certainty about climate responses to inputs that are unwarranted, given that climate is a complex adaptive system and our attempts to predict the behaviour of such systems has been pretty dismal. Consider the characteristics of complex adaptive systems: They are sensitive to tiny changes in initial conditions, they have stochastic elements which means their behaviour is not conventionally predictable, and their behavior is dominated by fat-tailed events and various feedbacks. Our attempts to model these systems in macroeconomics, ecology, the immune system, and climate have shown dismal results to date. Economists cannot predict GDP growth even a year in advance with any accuracy better than a simple model that assumes next year’s GDP will be the same as this year’s.
It’s funny that the same environmentalists who totally buy into complex systems theory when it comes to ecosystems completely reject the same theory when it comes to climate models. The Gaia hypothesis and the precautionary principle come straight out of early work in complex systems. Environmentalists glommed onto it because it seemed like a strong argument for not trying to control or interfere with nature. But the same logic applies to the economy and to climate, and there because the conclusions favor their ‘opponent’s’ arguments, the left simply ignores it.