49 thoughts on “Climate Modeling”

  1. You can do anything with a model which makes them unreliable (as the people that abuse them.) Models are useful for explaining in the abstract but even then only if you follow dirty Harry’s advise and know their limitations.

    1. A model should only be trusted as far as it has been validated. It appears the only people trying to properly validate the climate models are finding them wanting.

  2. If you cannot make valid predictions the model is basically pointless. The IMO is a good reason to confine this particular model to the dustbin of history.

    1. You’re right, and I think we should just give up on computational fluid dynamics as well.

        1. I’ve looked at CFD runs, some are garbage, some are useful, you have to really
          have some people who really know the physics check them for singularities or
          sensitivities.

          1. some people who really know the physics

            I could see why you would need help in that area.

    2. Fairly simple models made in the 1970’s predicted that global temperatures would rise and that
      ice volume in the Arctic, Antarctic, Greenland and the various glaciers would shrink.

      Those predictions in general seem pretty good.

      While these models don’t give good quarterly predictions or even good annual predictions, they’ve been
      pretty good on a decadal basis.

      2. During the 1970s models of the approaching ice age outnumbered the forecasts of warming.

        1980 was Hansen’s big paper. And at the time that was brushed off as not “fitting the consensus”. There was a lot of new work in the 80s. And if you wanted to say the consensus among climate scientists in 1990 was for warming, I wouldn’t dispute that.

        But it wasn’t until Mann’s ridiculously erroneous hockeystick in 1998 that the historical record was flattened to the point of allowing the hyperbolic terms like “unprecedented”.

      3. Surely, you jest. The models have completely diverged from reality over the past decade+.

      5. Milanovic’s post addresses something one might almost call the metaphysics of climate modeling; what, if any, basis exists for supposing a useful model is even possible. He sets aside the whole issue of actually constructing a climate model once a putatively suitable theoretical foundation is in-hand. He doesn’t get into grubby implementation details and it’s here that the ClimateGate e-mails were revelatory. So much of “climate science” isn’t really climate science, but climate-related applications of math, stat and computer science. Unfortunately for the Warmist case, the record of climate modelers in applying these other free-standing areas of knowledge and practicum to the particularities of their field is not good. The statistical malpractice underlying Mann’s Hockey Stick graph is far from unique in climate science annals, unfortunately. And at the level of executable computer code, the models have been found to be a dog’s breakfast of no-no’s including – incredibly – unchecked overflows, underflows and divides-by-zero. It would be nice if I had some profound base of knowledge from which to arrive at the conclusions I have arrived at about so-called “climate change” research – namely that it’s a load of road apples baked into a pie – but I don’t need one. All I need is my personal experience as a software developer who did a bit of numerical methods implementation back in the day and knew enough to comb possible excursions to positive and negative infinity out of my code. Any “climate modeler” who didn’t/couldn’t do likewise has no claim to credibility on any deeper issue.

  3. Wow.

    Thomas Milanovic, short version: “Hey hockey team, I’ll see your poorly-applied rudimentary statistical models, and raise you roughly eight years of general engineering, chemical engineering, control theory, chaotic system modeling, and wild-hair math.”

  4. Well, I have been asking this for a while on climate forums. How do you know that you can make good predictions about any aspect of climate? I can certainly hear the possibility that increased greenhouse gases will increase surface temperature, conservation of energy and all that. But now that everyone is talking about heat flowing into and out of the deep ocean, that doesn’t make much sense to me: the surface climate is connected to this incredibly massive reservoir of far far more energy, which apparently can absorb or emit that energy. I wouldn’t be surprised if that fact just overwhelms everything else, and our climate really just looks like a random walk of heat entering and leaving the ocean.
    Anyhow, the way to disprove that suggestion (which is nothing but a guessed possibility) is to show me that you can pick an aspect of climate (e.g., global averaged surface temperature) and predict it accurately for a while. No back-predictions allowed, of course – you may use that to tune your model, but not to validate it.
    And no one has been able to do that yet. Their predictions have failed terribly. Which is fine – fix your model and try again! But it means I really don’t know how they’re sure it’s even possible.

  5. How do you know that you can make good predictions about any aspect of climate?

    Because some practitioners of science know enough math, physics, chemistry, etc., to be able to read and understand a lot of published scientific literature, and then follow the internet discussion of them carefully.

    It’s really not that hard, but is does require some effort and the willingness to at least try to learn. Not you, though, hand waving and histrionics seems to be sufficient for your understanding of the nature of reality.

    1. And yet, their predictions have failed.

      It really isn’t that hard. All you have to do is close your eyes, plug your ears, and cover your mouth, and you can be a wise monkey, too.

    2. “Because some practitioners of science know enough math, physics, chemistry, etc., to be able to read and understand a lot of published scientific literature, and then follow the internet discussion of them carefully.”

      If you dance harder, maybe the world wont end in a global warming apocalypse next year. We are all going to die in fire hurricanes if people don’t dance with more enthusiasm.

    3. “You people are too stupid, ignorant, and lazy” isn’t an argument you’re going to win on a website where there are a -lot- of engineering degrees floating around. But it is typical.

      There does seem to be a sense that climate scientists could use a lot more of the higher level engineering courses.

      1. “You people are too stupid, ignorant, and lazy” isn’t an argument you’re going to win on a website where there are a -lot- of engineering degrees floating around.

        I think that’s called an appeal to authority and credentialism, which of course I reject, based upon the evidence revealed by the comments here alone. Amazing how quickly so called freedom loving libertarians revert to their underlying belief in authority when pressed by the simplest observations or asked the easiest questions, and when caught with their pants down start waving their degrees.

        1. I think that’s called an appeal to authority and credentialism

          Amazing how you can type without removing your pacifier. As to reverting to an underlying belief in authority, that’s the exact opposite of what transpires here. And I’d rather be accused of [whatever] and getting caught with my pants down than being exposed for having never grown out of diapers. Your’s probably need changing.

    4. If the climate models don’t reflect reality – and the last 15+ years show they don’t – it isn’t reality that’s wrong. Computational fluid dynamics models produce useable results, unlike the climate models. That’s because fluid dynamics is much more understood than climate dynamics and there is better data to draw upon.

    7. follow the internet discussion of them carefully.

      What does being able to read have to do with using math, physics, and chemistry to actually develop a model that is accurate in its prediction? The models are known to be flawed because both the small temperature increase and the cataclysmic havoc anticipated have not occurred in the decade for which the various models claimed they would.

      You may know math, physics, and chemistry with a doctorates degree, but if you model cannot predict the empirical data, its not validated no matter how good you read. It’s not about following discussions, it is about following data.

      1. What does being able to read have to do with using math, physics, and chemistry to actually develop a model that is accurate in its prediction?

        A lot. For starters, you would understand what a model is and isn’t, what it is used for and what it is not, how to build one and test it, how to compare it to empirical evidence, and what needs to be done to tweak it, and when to know when to throw it out and start over, or engage in a complete rewrite or rebuilding of it to improve it. You have demonstrated none of those quantities even at the high school electronics enthusiast or computer programming level, yet you persist to feign knowledge and insight. Go figure. You haven’t even critically looked at other people’s models.

        1. And that will totally happen by “them” following the internet discussion reading the comments on blogs.

          It is a good way to get your marching orders from the activist class though.

        2. Heck, I started by reading the GCM source code. Of course I also read graduate books on ocean/atmospheric dynamics.

          1. I look forward to your definitive refutation of both the evidence and the theories that support the various models then. Go ahead, publish it right here.

            Be specific. Point by point. I need a good laugh.

          2. George has refuted talking points many times. You can search the website via your favorite engine and read for yourself without any more effort on his part. You claim to be a reader and researcher, the power is within your reach.

        3. You haven’t even critically looked at other people’s models.

          You don’t know what I have or have not done. That’s an assumption based on nothing at all that could possibly be real. Which is funny, because I suspect that is how you build your climate models as well.

  6. The sad state of affairs is that even if we had perfect models it wouldn’t matter because the data is junk. Temperature data is trash. Sea level data is bullshit, and covers too short a period of time. Humidity data is basically non-existent. Cloud cover data is so limited it’s practically useless. The amount of good data we have is so limited in geographical and temporal extent there’s no way it’s enough to either prime a model or realistically validate any models. And yet even with the garbage data the models are worse, amounting mostly to multi-variate regressions with too many free parameters (modern day epicycles), but even they don’t match the data anymore!

    Modeling of systems as complex as the climate is an extremely complex task, but it’s impossible with the massive amount of politicization that’s going on today. There’s too much push to assert more confidence of climate models than is possible and to produce the “correct” result.

      1. You prefer their expansion back to the Mississippi River Valley? They’ve been melting for a few thousands years. Do you want to blame that on the Industrial Revolution too? Or do you believe humans should live such a delicate life that the Earth must remain completely static for our survival? If so, do you not believe in evolution either?

        1. Evolution doesn’t mean survival of humans, it merely means survival of the fittest species for the conditions. The long term answer for survival is bacteria. The next long term answer
          is Dinosaurs.

      2. Some glaciers have been melting, sure. Is that enough to validate climate models? Seems a thin string to hold such complex models. Which models precisely do they validate and which do they falsify? How does soot from hydrocarbon burning (which is a big factor in ice sheet melting) factor into that? Is it accounted for by any models? Does the lack of that accounting affect whether the glacier melting validates the models? How could it not?

        Moreover, it’s incorrect to say that glacier melting would even be a prediction of current climate models, it would be more accurate to say that the models have been fit to that assumption.

        Ultimately it’s possible, but unlikely given the evidence, that the worst case global warming outcomes are an accurate forecast of future climate, but we cannot be sure of that to even the slightest amount because the models are so bad and the data is so sparse and poorly calibrated.

        1. Yes the data is sparse, but, i’ve looked at sparse data sets.

          The data for the Oort cloud and the Kuiper belt were very sparse yet
          the confidence in the science is very high.

          However there is no anti-Oort movement in the GOP so the science
          is accepted.

          Geology is mostly the science of sparse data yet scientists make the best of it.

          Some of the Models are crap too. The Monterey Oil Shale just got reduced some 96%
          in reserve capacity, once it got looked at hard.

  8. I received my Ph.D. studying numerical solutions to partial differential equations, spectral methods, fluid dynamics, heat transfer and chaos or in short, exactly what this article is about. Let me summarize: Because we don’t know everything (or don’t have a big enough computer to simulate every single molecule involved in global fluid dynamics) we better be careful in predicting things. Thank you Mr. obvious!

    He complains about turbulence and the closure problem, and lack of dissipation, etc. So what, anyone who knows what the Knudsen number is knows that the N-S equations are an approximation.

    I agree we have to be careful about accepting numerical results, but it doesn’t mean we shouldn’t attempt ever more sophisticated problems. I’d call global N-S solutions interesting but certainly not conclusive. In the case of global warming the simple physics are the most persuasive. The CO2 mass fraction in the atmosphere has gone up due to human activity, CO2 has a strong absorption band at 10.6 microns and the peak of the Earth black body radiation curve is 10.6 microns.

    I don’t trust these global models, I agree that humans are causing global warming, I don’t know if that’s 1 degree or 1 nanodegree, and I’m not at all sure we won’t be better off for any warming we cause.

    Recently people claim that since the global average temperature is stable or down it’s proof that AGW is a false theory. Wrong! Suppose we are entering a new ice age and thanks to our CO2 emissions we’re fighting that very bad possibility but despite our best efforts we’re losing. I don’t want Chicago under a mile of ice, burn mo

    1. “The CO2 mass fraction in the atmosphere has gone up due to human activity…”

      Hasn’t, actually. As the plot shows, the rate of change of CO2 in recent history is affinely related to temperature anomaly.The slope in that affine relationship is essentially precisely the factor needed to fit all the variability. That establishes that the slope is essentially a function of temperature. Human inputs also have a slope in rate. There is little to no room for it. It is already accounted for by the temperature relationship.

      It is apparent that the Earth regulates its CO2 in an extenisve, multi-faceted feedback loop, and that human inputs are effectively quashed by the feedback dynamics.

      In fact, since the temperature hiatus took hold, human emissions have continued accelerating, while the rate of change of atmospheric CO2 has plateaued along with the temperatures, and atmospheric CO2 concentration is only increasing approximately linearly. The two quantities are diverging from affine similarity. Assuming temperatures continue to stagnate, or even decline, that divergence should become stark in the not too distant future.

    2. “…CO2 has a strong absorption band at 10.6 microns and the peak of the Earth black body radiation curve is 10.6 microns.”

      Which means that the Earth will retain more heat than it otherwise would. But, this is a (mathematically) global, not a local, surety. It does not say that an incremental increase in atmospheric CO2 from a particular baseline will necessarily result in an incremental increase in heat retention. It is the difference between a secant line and a tangent line.

      In fact, given the CO2 temperature relationship above, it is not possible that the sensitivity function will be monotonic beyond a particular concentration, because that would lead to an unstabilizable, runaway positive feedback, which would have rendered the Earth uninhabitable for us eons ago. I.e., an increase in temperature leads to an increase in CO2, leads to an increase in temperature, and on and on. Such a dynamic is not stabilizable via SB radiation because of the integral relationship between temperature and CO2 above.

      It is also possible to see that this is necessary by considering the reductio at both ends. Consider the Earth as a black body. Zero GHG concentration means that the surface will stabilize at the temperature dictated by the Stefan-Boltzmann law. 100% GHG concentration means that the atmosphere can be considered an extension of the body itself, and the surface will stabilize at that the same temperature. In between, there can be a positive greenhouse effect, but it must inflect at some point so that the function can return to zero at the end point.

      Basically, it is a mistake to assume thermal transport between the surface and atmosphere is dominated by radiation. Air currents convect a great deal of surface heat directly to the atmospheric IR radiators. It is like looking at an automobile cooling system, and concluding that, because it obstructs the free path outward from the engine, the radiator must make the engine run hotter. That, obviously, entirely misses the transport of heat directly to the radiator via the coolant flow from the engine.

      1. Nice argument, unfortunately if CO2 levels rise high enough the oceans may slip towards a degree of acidity sufficient to severely damage coral reefs. That loss of reef productivity will lead to hundreds of millions of people starving.

        1. If, may, then will? There is no evidence of it. We should not be children letting the fear of monsters under the bed paralyze us.

    3. It appears we are at, or above, the inflection point in the present climate state, and that is why there is zero apparent evidence for any surface temperature sensitivity to rising CO2. If you look at surface temperatures over the last century, the rise from roughly 1970-2000 is almost precisely the rise from 1910-1940, well before rising CO2 could have been the culprit.

      We are witnessing a long term trend coming out of the LIA, plus a superimposed repeating pattern with a period of about 60-65 years. CO2 is just a bystander. (To preempt a frequently encountered objection, yes, I cut off the data before 1900 because the error bars for prior data explode, and there is no point in attempting to draw conclusions from them).

      “Recently people claim that since the global average temperature is stable or down it’s proof that AGW is a false theory. Wrong!”

      It proves that the climate establishment doesn’t understand the problem well enough to say with any assurance what is going to happen. We can’t panic over every imaginary bogeyman in the closet. Primum non nocere. If you don’t know what you are doing, you are as likely to make things worse as better. I am convinced, of course, by the arguments above that we would definitely be making things worse by taking action against this imaginary problem.

  9. It is amusing the people who denounce climate change models as worthless
    love econometric models which show Trillions in losses if we move away
    from coal.

Comments are closed.