Feed on

I know you used to watch the Price is Right!



So, is our environment more like a train track or a plinko chip? If you were able to “set up” our climate, our oceans and our ecosystems with the same initial conditions in this same universe, say, 1 million times, and you let the movie run forward for 1,000 years each of those times and then you took a census of what the world looked like at the end of each and every one of those runs, what do you expect you would see? Would you be more shocked by the outcome of 1 million identical planet earths after your experiment or would you be more shocked by 1 million different ones? I think the answer is obvious and ecologists and real scientists know the answer is obvious. The answer is that given a set of initial conditions, there is absolutely no reason to expect the earth to look identical after multiple “runs of the game” so to speak.

Why does this matter? I’ve been studying up climate modeling on the side and it turns out that climate modelers think (well, they probably don’t think it, but they DO model it) the ecosystem is like a train track and not a plinko chip. Given a set of initial conditions, if they run their general circulation models a million times, it will show a million identical climates. That is alarming, no? So, for example, when climate modelers try to tell us that a warmer planet will cause problems because disease vectors will spread, most notably, malaria (see here by the way) what they do is set the initial conditions of their climate model. Then they simulate what will happen as CO2 concentrations are increased (along with a number of assumptions, debatable indeed, about feedback) and show how temperatures and precipitation will change.

Then what they do is take the existing area where bugs are located and where disease is located and extend it onto the “new” ecosystems that they project. That’s it. That’s all. So the climate modelers are assuming that bugs are in some sort of special steady-state and that the future range of disease and bugs will adjust smoothly as the temperature and precipitation patterns change. And they do the same thing for larger species, including us. But this is wholly absurd. All species adjust to changing conditions all the time, and have done so for millions of years. Wouldn’t they (we?) all be extinct otherwise?

Thus our design of legislation (and our beliefs) like the Endangered Species Act and thinking about wilderness set asides, or optimal fishery sizes and so on is entirely flawed. It is based upon a static concept of an ecosystem that has never and can never exist. Yet we use this idea not only in our everyday language (e.g. go to an “E”nvironmental website or just read any nature writing and count the times “natural” or “balanced” or “harmony” is mentioned) but we embed it in our policy. Even if 100% of the people on the planet were on board with the goals of these pieces of legislation, we are dooming ourselves to failure if we continue to believe the planet is like a train on the tracks and not like a Plinko Chip.

And it is in this vein that “E”nvironmentalists would do themselves well to read a little Hayek. No one can design a free or peaceful of fair society. The best we can do is to encourage institutions (such as the Rule of Law, an analog of “initial conditions”) and hope for the best. There is more than hope of course, but we don’t want to take us too far afield today. And does it mean that we can screw up the initial conditions so badly, say by spitting lots of extra CO2 out into the climate, that we increase our chances of a poor outcome? Sure, but the entire paradigm for how we think about the likelihood of bad outcomes or what a “bad” outcome is does not rest on bedrock.

6 Responses to “Is Nature Like a Train on the Tracks or a Plinko Chip?”

  1. ZT says:

    What’s your source for this? I think a generalizing claim about an entire scientific field needs to be pretty well supported. The fact that climate scientists use deterministic models certainly doesn’t mean they can’t account for error and variability in the initial conditions– that is, after all, how Lorenz “discovered” the “butterfly effect” (even though the general mathematical idea predated his discovery of it in a climate model.) Why would scientists just start being stupid and forget how to do things that were done properly 50 years ago?

    Since I hate it when “hard-science” people make generalizing and inaccurate claims about economics, (most recently some biology-related major decided to lecture me on where liability rules for pesticide over application should lie, because somehow his knowledge of glyphosate chemistry gave him an unwavering insight that Monsanto alone is responsible for whatever people do with its products), and I assume you have the same problem as well, we shouldn’t do the same to other fields. Climate scientists put up with a lot of sh**, and I think we’re the last ones who should chastise researchers for having imperfect answers to complicated problems.

    • wintercow20 says:

      (1) Daniel Botkin, opening chapter of the Moon and the Nautilus Shell

      (2) A climate change textbook: http://forecast.uchicago.edu/

      (3) The 4th IPCC Assessment Report (not for policymakers)

      This is an entirely fair claim to make. And the concern is not variability in initial conditions – it is variability in outcomes for the SAME initial conditions. And this has nothing at all with them being stupid and forgetting how to “properly” do things, quite the opposite. They are doing the science correctly, this in fact is a limitation of the GCM, not a criticism of the scientific approach to doing them. Hence the plinko analogy. This is not a general claim about an entire field, I don’t know where you come up with that. It is an observation about a particular model used, and technique applied, to estimate the future climate and changes in ecosystems.

      As far as climate scientists putting up with a lot of “sh*t”, if some persist in wishing to end industrial civilization, as some openly claim they need to (see Hansen’s death trains), then they ought to have a high bar. But this is far from being a chastising for their imperfect answers – in fact, it is a chastising for their open disregard for making it clear that their “answers” are anything but imperfect. That’s lesson #1 from economics, and in no way is unfair to expect climate scientists to respect.

      • ZT says:

        Can you point me to where those sources say the overwhelming majority of climate models are deterministic? A simple google search for “stochastic climate models” brings up a lot of results, such as http://math.uh.edu/~ilya/papers/cpam00.pdf or http://www.aos.princeton.edu/WWWPUBLIC/gkv/history/Hasselmann76.pdf . If memory serves, Michael Mann mentions stochastic modeling (specifically the Monte Carlo method) in his most recent book.

        Second, I don’t think I was clear about the butterfly effect. If the universe is so sensitive that it really is like a “plinko chip”– in other words, the same universe with the same initial conditions could generate wildly different plants, to use your description– and this carries over into the time-scale of climate models, then that same sensitivity could be expressed even in deterministic models by the fact that they’d give wildly varied results from slightly different starting points. Climate scientists have noted this in the past.

        Lastly, I’m a bit confused about your comments on Hansen– the “death trains” comment was hyperbolic and stupid (even though coal really does kill people http://solar.gwu.edu/index_files/Resources_files/epstein_full%20cost%20of%20coal.pdf) but how does it indicate a desire to end modern civilization? Has Hansen ever denied that climate models have some degree of uncertainty? Certainly the majority of climate scientists don’t make this case.

  2. Speedmaster says:

    Amazing hubris to believe that we know all of the variables in play. Or understand them properly.

  3. Steve says:

    Climate modelers have made the assumption that the climate is a servo system complete with fixed feedbacks. You can see this in the vast majority of the language they use to describe their models and the climate “system”. On top of that basic assumption, they have assumed that the feedbacks are mostly positive, and that the system is teetering on a knife edge and will fall if perturbed in the slightest way. Sorry for them that the climate hasn’t obliged with data to fit their model’s projections. The reason we are even alive now is due to the climate being stable around generally benign values, given extreme perturbations that we yet to understand, it will stabilize around less benign values, and in our recent geologic past, that is being much, much colder.


    • Harry says:

      Great points, Steve, and earlier, WC.

      Ever since the TI calculator replaced the slide rule and a few years later we got 64K Radio Shack computers that could do Multiplan, the Hubris grew out of control.

      Regardless of how capable one’s computer model is, we are stuck with data that are grossly imprecise and incomplete. For example, we have tree rings and ice cores as proxies for temperature where there is no control for rainfall and snowfall and other variables. We may know to a hundredth of a degree what the temperature was at noon on the roof of some lab at Oxford on July 1, 1798, but we do not know what the temperature was in what was to be the Wichita, Kansas, airport that day.

      As you say, the climate alarmists argue we are on a knife edge where 50 parts per million of CO2, however and wherever that is measured, will cause the poles to melt and submerge St. Mark’s Square, and therefore we should revert to a pre-Industrial Revolution way of life, at least most of us. This proposed solution has all the same appeal as Wahabbiism.

Leave a Reply