The Global Warming Test

One episode of A History of Ancient Britain, the recent BBC series, is about the Ice Age. If you know there was an Ice Age, you should grasp that the Earth varies in temperature a lot for reasons that have nothing to do with human activity. To measure the effect of recent human activity on global temperatures, you need to know what the Earth’s temperature would have been in the absence of human activity. Then you find the effect of humans by subtraction (actual temperature – predicted temperature assuming no human activity).

That’s hard to do. Because the non-human effects are so large, you need a really accurate model to “control” for them.  No such model is available. No current climate model has been shown to accurately predict global temperatures — the IPCC chapter called “Climate Models and Their Evaluation” (informal title: “Why You Should Believe Them”) is the most humorous evidence of that. Lack of accurate predictions means there is no good reason to trust them. (That the models can fit past data means little because they have many adjustable parameters. “With four parameters I can fit an elephant,” said John von Neumann.) The case against the view that humans have dangerously warmed the climate (sometimes called AGW, anthropogenic global warming) is that simple.

Because it is so simple, “the other side” consists of saying why 2+2 really does equal 20 or whatever. Sure, many people say it, so what? When I was an undergrad, I gave a talk called “The Scientific ____ “. I said usage of the term scientific without explaining what it meant was a sign of incompetence and a reader could safely stop reading right there. That isn’t terribly helpful, because few people use scientific that way. My grown-up version of this test is that when someone claims AGW is true, I stop taking them seriously as a thinker. I don’t mean they can’t do good work — Bill McKibben is an excellent journalist, for example. Just not original thought.




20 Replies to “The Global Warming Test”

  1. I’ve followed your global warming comments for some time now and, while I’m relatively ignorant, I’m intrigued. I assume that since we have no adequate models to “control” for non-human causes of global warming we don’t know what part of global warming can or cannot be attributed to human activity. So one can neither claim nor deny that AGW is true; all we can say with confidence is that it’s getting warmer for reasons we cannot confidently assert.

    Would you also dismiss as a thinker anyone who claims with authority that AGW is false? Or am I misreading your logic? Or are you claiming that the burden of proof in these matters falls completely on the shoulders of those who support AGW, since it is well established that non-human causes frequently alter climate? Could you tease this out more precisely for me?

    1. We don’t know if AGW is true. That’s what I believe. To claim AGW is false is less egregious, because there is no clear AGW “signal” (e.g., unprecedented rates of warming). But, yeah, to be sure it is false is also going beyond the evidence, in my opinion. As a practical matter, I think we should spend our attention on things that are undeniable problems. There are plenty of those. Perhaps AGW will someday be an undeniable problem, or maybe not. Quite apart from AGW, it’s important to develop new sources of energy and (what amounts to the same thing) conserve energy.

  2. I’ve asked this before: What evidence could be presented that would change your mind? Famously, J.B.S. Haldane (might have) said a pre-Cambrian rabbit fossil would refute evolution.

    Slightly less famously, Peter Duesburg says AIDS is not caused by HIV, and that anti-retroviral drugs are of no value in treating it. He persuaded the S. African govt to withhold them from AIDS patients, with imaginable consequences. I wonder how many people he would need to see pulled back from death’s door by such treatments before he would change his mind.

    What is your rabbit? And what should we conclude if there is none?

  3. There are two lines of argument advanced in favour of AGW.

    (i) The present warming is unprecedented. I have three counterarguments to that. (a) Your present day figures are faked: whatever the increase may be, you have exaggerated it with bogus “adjustments” to the data, and by failure to adjust properly for such effects as UHI, airport sitings of instruments, decline in the number of instruments, relocation of instruments, changes in instruments and inproperly operated instruments. (b) You have resorted to lies about earlier warm spells to allow you to claim that presently reported warming is unprecedented. (c) Your low levels of competence – for example your woeful grasp of statistics – lets me dismiss that part of your work that isn’t dishonest as largely dud.

    (ii) Our models require CO2 induced warming to allow them to model observed warming. I have three counterarguments (a) you are comparing your models with rubbish (see (a) above). (b) I have written, or supervised the writing of, mathematical models of physico-chemical systems since 1967. Consequently I am aware of the difficulty of modelling such systems, even when results from controlled experiments are available to guide the modeller in his task. Your task is orders of magnitude more demanding than any I have undertaken, and your guidance is from haphazard and incomplete sets of observations of dubious validity (see (a)). I find your hubris about your ability usefully to model the climate just risible. Or, as we say in Britain, “Oh Bollocks!”. (c) Not only do I find you hubristic, I am also struck by your generally low intellectual calibre (by the standards of the physical sciences), and inability to think critically. You are, in general, not fit for purpose.

    Another question, potentially important, is “Suppose that AGW is real – small, perhaps, but real. Then what?” To which I’d reply “Then let’s do intelligent economic analyses of the likely consequences rather than just assume that such a change must be a wicked and disastrous thing, to be avoided whatever the cost.”

  4. You’re not quite as clever as you think you are. AGW and “dangerously warmed the climate” are two separate propositions. Throwing in the “dangerous” is a common trick of climate skeptics. Even John Christie now accepts the AGW part but argues against the dangerous part. Have you stopped taking John Christie seriously as a thinker?

    1. “Trick”? I use “dangerous” because that’s what I mean. Whether humans have had tiny effects on the climate is a question that doesn’t interest me. I don’t know who John Christie is.

  5. “I’ve asked this before: What evidence could be presented that would change your mind? ”

    I realize this question is not addressed to me, but I will answer it anyway.

    (1) If somebody presented a climate simulation which consistently made interesting, accurate, and bona fide predictions, and that same model, when run with pre-1950 CO2 levels predicted a much cooler world than what we got over the last 50 years. In that case, I would accept that recent warming is largely due to increased levels of CO2.

    (2) If somebody demonstrated that recent temperatures (or rates of increase) are well outside the range of natural variation, I would accept that mankind’s activities are likely the cause.

    (3) If somebody developed a theory of climate which explained past temperatures and was sufficiently simple and elegant that it was highly unlikely to be just a coincidence, and that theory predicted significant (amplified) warming due to CO2.

    1. Thanks, sabril. I come close to agreeing with that. Basically my answer is your #1: (a) someone creates a climate model that makes accurate predictions and (b) that model — run with and without the effects of humans — shows that humans have dangerously warmed the planet. As for #2, I think it is clear that current temperatures and rates of temperature change are not unprecedented. As for #3, I think too many parameters need to be estimated from the data relative to the amount of data. Fitting past data will not convince me of anything.

  6. “As for #2, I think it is clear that current temperatures and rates of temperature change are not unprecedented”

    I agree 100%. You don’t even have to look back far or do anything sophisticated to find precedent. For example, melting glaciers in the Alps have revealed mountain passes used by humans during the Medieval Warm Period. The reasonable inference is that temperatures now are roughly the same as they were then.

    “As for #3, I think too many parameters need to be estimated from the data relative to the amount of data. Fitting past data will not convince me of anything.”

    Well I’m talking about a model without this problem of parameters. For example, suppose somebody came up with the hypothesis that global surface temperature change is the result of cosmic radiation. Suppose we let T represent global surface temperature anomaly (in degrees Celcius) and C represent cosmic ray intensity anomaly in watts per square meter. And suppose we knew T and C pretty accurately for the last 2000 years.

    What if it turns out that T = aC holds true from Year 1 to 1950, where a is a constant, the only parameter in use in the model. In that case, I would be reasonably convinced that global surface temperature anomaly really is caused by cosmic rays. Further, if temperatures after 1950 were significantly and increasingly higher than aC, I would agree that they are probably the result of man’s activities.

    Of course, the difference between this model and the simulations being pushed on us today is that my hypothetical model has one and only one parameter. Further, there are a priori reasons to believe that cosmic ray intensity might affect global surface temperatures. So the chances that the relationship is a coincidence are pretty low.

    By contrast, and as you point out, the situation where a model or simulation has dozens of parameters is very different. There’s actually a very good chance that the fit with history is just a coincidence.

  7. “We don’t know if AGW is true. That’s what I believe. To claim AGW is false is less egregious, because there is no clear AGW ‘signal”‘”

    I would actually go further than this and say it’s possible to be reasonably confident that AGW is false. The key point is that AGW is actually two completely independent hypotheses, which are as follows:

    (1) Increased levels of CO2 in the atmosphere will result in mild increases in global surface temperatures; and

    (2) Any such mild increase will be amplified to dangerous levels through the mechanism of water vapor feedback. i.e. temperature increases will cause more water to evaporate and since water vapor is itself a greenhouse gas, surface temperatures will increase further, and so on.

    If (2) is not true, then mankind does not need to worry about AGW since any temperature increases will be to small to make a big difference.

    Anyway, there is good reason to believe that (2) is false. For one thing, it posits positive feedback which would be unusual in a system which has been around for a long time. If you push on a system which has been around for a while, you would normally expect it to push back in the opposite direction. (It’s the same thing with your theory about diet and set points. If you start eating less in an effort to lose weight, your body system will push back by reducing its metabolism and increasing your appetite. Negative feedback. Which makes sense — systems which have been around for a while tend not to go into death spirals when they are stressed. Because if they did, they would have disappeared long ago.)

    So one would naturally expect that warming due to CO2 will be dampened by the climate system and not amplified.

    The second piece of evidence is that there is no “hot spot.” i.e. The theory of CO2 induced global warming enhanced by water vapor feedback predicts enhanced warming in the air 10km above the surface in the tropics. The actual measurements don’t seem to bear this prediction out.

    See here:

    Based on these two points, I’m pretty confident that AGW (as the term is popularly understood) is false, although I do agree that mankind’s CO2 emissions are likely to cause some increase in global surface temperatures.

    1. I agree, it is comforting that the system has been around a long time. On the other hand, the human race has been around a long time and just recently (around 1980) the obesity epidemic — a massive failure of a feedback system — began. Feedback systems can work fine for an extremely long time and then fail. In the case of climate, it is obvious, as you say, that the negative feedback has overwhelmed the positive feedback for a very long time (the entire history of the Earth). But that could change — e.g., due to changes in land cover. But I have had the same experience as you. The more I have looked at this, the more I encounter evidence suggesting that AGW is wrong.

  8. “On the other hand, the human race has been around a long time and just recently (around 1980) the obesity epidemic — a massive failure of a feedback system — began. ”

    That’s true, and it’s logically possible that CO2 is a stimulus which can overwhelm the system like large amounts of unhealthy foods can foul up a person’s health. However, there have been lengthy periods in the past when CO2 levels were far higher than current levels now and temperatures were not very different.

    1. I believe changes in ground cover (e.g., more asphalt, fewer trees, etc.) may affect the feedback system, causing it to work worse (= reduce negative feedback, increase positive feedback) than in the past. I’m not saying this is true, just that it is a possibility. It is a source of uncertainty that causes me to back away from strong statements either way about AGW.

  9. It strikes me that you may be applying a higher standard to the AGW question than to other issues.

    For example, what if your first sentence above was rewritten this way:
    To measure the effect of [diet change] on [weight], you need to know what the [person’s weight] would have been in the absence of [diet change]. Then you find the effect of [diet change] by subtraction (actual [weight] – predicted [weight] assuming no [diet change]).

    My understanding is that you would say this is wrong and that we don’t need a good model of weight in order to make at least preliminary conclusions. So help me out – why the different approach to climate questions? Why your emphasis on models? I agree that there may have been some (or even a lot) of bad climate modeling done. But there are also other approaches to studying climate questions.

    1. In the absence of exercise or diet changes, most people’s body weight is close to constant or increases linearly. So it is easy to see what the weight would have been in the absence of the diet. The diet has a sharp onset and large effect relative to noise (unexplained variation). With human activity and climate there is no sharp onset and no large effect.

  10. You responded to an apparent one-time 1% weight gain by dropping all experimentation with cold showers. Yet, increasing CO2 content of the entire planet’s atmosphere by 40%, with half of it occurring during your own lifetime, doesn’t count as sharp onset?

    Remind me not to ask you to call the fire department when I spot a house on fire.

  11. You seem to be saying that both presence of, and absence of, a sharp onset would weaken the case for AGW. What course of events would you not take as weakening the case for AGW?

    Do you disbelieve in global warming, or do you disbelieve in an industrial cause for observed global warming? Statements of one undermine those of the other, but I see both here. Is this an even-day/odd-day thing?

Comments are closed.