RealClimate logo

Simple Question, Simple Answer… Not

Filed under: — group @ 8 September 2008 - (Français) (Japanese) (Finnish) (Czech)

Guest commentary by Spencer R. Weart, American Institute of Physics

I often get emails from scientifically trained people who are looking for a straightforward calculation of the global warming that greenhouse gas emissions will bring. What are the physics equations and data on gases that predict just how far the temperature will rise? A natural question, when public expositions of the greenhouse effect usually present it as a matter of elementary physics. These people, typically senior engineers, get suspicious when experts seem to evade their question. Some try to work out the answer themselves (Lord Monckton for example) and complain that the experts dismiss their beautiful logic.

The engineers’ demand that the case for dangerous global warming be proved with a page or so of equations does sound reasonable, and it has a long history. The history reveals how the nature of the climate system inevitably betrays a lover of simple answers.

The simplest approach to calculating the Earth’s surface temperature would be to treat the atmosphere as a single uniform slab, like a pane of glass suspended above the surface (much as we see in elementary explanations of the “greenhouse” effect). But the equations do not yield a number for global warming that is even remotely plausible. You can’t work with an average, squashing together the way heat radiation goes through the dense, warm, humid lower atmosphere with the way it goes through the thin, cold, dry upper atmosphere. Already in the 19th century, physicists moved on to a “one-dimensional” model. That is, they pretended that the atmosphere was the same everywhere around the planet, and studied how radiation was transmitted or absorbed as it went up or down through a column of air stretching from ground level to the top of the atmosphere. This is the study of “radiative transfer,” an elegant and difficult branch of theory. You would figure how sunlight passed through each layer of the atmosphere to the surface, and how the heat energy that was radiated back up from the surface heated up each layer, and was shuttled back and forth among the layers, or escaped into space.

When students learn physics, they are taught about many simple systems that bow to the power of a few laws, yielding wonderfully precise answers: a page or so of equations and you’re done. Teachers rarely point out that these systems are plucked from a far larger set of systems that are mostly nowhere near so tractable. The one-dimensional atmospheric model can’t be solved with a page of mathematics. You have to divide the column of air into a set of levels, get out your pencil or computer, and calculate what happens at each level. Worse, carbon dioxide and water vapor (the two main greenhouse gases) absorb and scatter differently at different wavelengths. So you have to make the same long set of calculations repeatedly, once for each section of the radiation spectrum.

It was not until the 1950s that scientists had both good data on the absorption of infrared radiation, and digital computers that could speed through the multitudinous calculations. Gilbert N. Plass used the data and computers to demonstrate that adding carbon dioxide to a column of air would raise the surface temperature. But nobody believed the precise number he calculated (2.5ºC of warming if the level of CO2 doubled). Critics pointed out that he had ignored a number of crucial effects. First of all, if global temperature started to rise, the atmosphere would contain more water vapor. Its own greenhouse effect would make for more warming. On the other hand, with more water vapor wouldn’t there be more clouds? And wouldn’t those shade the planet and make for less warming? Neither Plass nor anyone before him had tried to calculate changes in cloudiness. (For details and references see this history site.)

Fritz Möller followed up with a pioneering computation that took into account the increase of absolute humidity with temperature. Oops… his results showed a monstrous feedback. As the humidity rose, the water vapor would add its greenhouse effect, and the temperature might soar. The model could give an almost arbitrarily high temperature! This weird result stimulated Syukuro Manabe to develop a more realistic one-dimensional model. He included in his column of air the way convective updrafts carry heat up from the surface, a basic process that nearly every earlier calculation had failed to take into account. It was no wonder Möller’s surface had heated up without limit: his model had not used the fact that hot air would rise. Manabe also worked up a rough calculation for the effects of clouds. By 1967, in collaboration with Richard Wetherald, he was ready to see what might result from raising the level of CO2. Their model predicted that if the amount of CO2 doubled, global temperature would rise roughly two degrees C. This was probably the first paper to convince many scientists that they needed to think seriously about greenhouse warming. The computation was, so to speak, a “proof of principle.”

But it would do little good to present a copy of the Manabe-Wetherald paper to a senior engineer who demands a proof that global warming is a problem. The paper gives only a sketch of complex and lengthy computations that take place, so to speak, offstage. And nobody at the time or since would trust the paper’s numbers as a precise prediction. There were still too many important factors that the model did not include. For example, it was only in the 1970s that scientists realized they had to take into account how smoke, dust and other aerosols from human activity interact with radiation, and how the aerosols affect cloudiness as well. And so on and so forth.

The greenhouse problem was not the first time climatologists hit this wall. Consider, for example, attempts to calculate the trade winds, a simple and important feature of the atmosphere. For generations, theorists wrote down the basic equations for fluid flow and heat transfer on the surface of a rotating sphere, aiming to produce a precise description of our planet’s structure of convective cells and winds in a few lines of equations… or a few pages… or a few dozen pages. They always failed. It was only with the advent of powerful digital computers in the 1960s that people were able to solve the problem through millions of numerical computations. If someone asks for an “explanation” of the trade winds, we can wave our hands and talk about tropical heating, the rotation of the earth and baroclinic instability. But if we are pressed for details with actual numbers, we can do no more than dump a truckload of printouts showing all the arithmetic computations.

I’m not saying we don’t understand the greenhouse effect. We understand the basic physics just fine, and can explain it in a minute to a curious non-scientist. (Like this: greenhouse gases let sunlight through to the Earth’s surface, which gets warm; the surface sends infrared radiation back up, which is absorbed by the gases at various levels and warms up the air; the air radiates some of this energy back to the surface, keeping it warmer than it would be without the gases.) For a scientist, you can give a technical explanation in a few paragraphs. But if you want to get reliable numbers – if you want to know whether raising the level of greenhouse gases will bring a trivial warming or a catastrophe – you have to figure in humidity, convection, aerosol pollution, and a pile of other features of the climate system, all fitted together in lengthy computer runs.

Physics is rich in phenomena that are simple in appearance but cannot be calculated in simple terms. Global warming is like that. People may yearn for a short, clear way to predict how much warming we are likely to face. Alas, no such simple calculation exists. The actual temperature rise is an emergent property resulting from interactions among hundreds of factors. People who refuse to acknowledge that complexity should not be surprised when their demands for an easy calculation go unanswered.

623 Responses to “Simple Question, Simple Answer… Not”

  1. 1
    Mark says:

    Gilbert’s figure of 2.5 thought, I think, a very good starting point.

    The question being asked is “how does greenhouse gasses affect the earths temperature?”. And this one answers “If you consider only CO2 (which we are responsible for), 2.5 degrees per doubling.”

    Now if someone wants to gainsay that, they are no longer asking the question, are they. They are asking for an answer that they like.

    Now you can just tell them “well, you do the maths then and we’ll look over the result, but by default, this is what would happen and you must take in any changes that occur whether they increase sensitivity or reduce it”. Or just ignore them and let them know that if they wanted a different answer, they should have asked a different question.

  2. 2
    Rick Kennerly says:

    I’m trying to do some laymen-level research on various change models predicting climate winners and losers in North America, particularly related to areas to retire to in, say, 20 years. Where do you think I could find some understandable middle of the road graphics, etc?

  3. 3
    Andrew says:

    Thanks! This is a great (and eloquent) description of the incrementing complexity required of models to go from the interactions between atoms and photons to those between Earth system components.

    As someone who’s used to working with complex (= simple; compared to reality!) models of biogeochemistry, I’ve gotten so used to requiring simulations to explore behaviour that I no longer give it a second thought. This has given me an institutionalised mindset that, at times, blinkers my understanding of why others misunderstand. This article is a helpful, if obvious, reminder that not everyone else is in the same boat! ;-)

    Best regards,


  4. 4
    Spencer says:

    #3 Andrew: I’m trying to figure out whether your question is off-topic or not! To get reliable predictions for climate on regions as small as a US state will take more computing power and understanding of climate processes than we have at present. So far, experts are pretty sure that the Southwest from Calif. to Texas will have worse heat waves and drought (=forest fires), which may have started already. This is pretty robust; people have been predicting it for decades. Most other US regions aside from Alaska don’t seem too sensitive to near-term climate change… ironically, the country most responsible for the greenhouse gases in the atmosphere is going to feel the effects later than many innocent countries. We do expect everywhere the generalities of climate change: hotter summer nights and later frost, more variability such as more intense droughts and rainstorms. And don’t get beachfront property near high tide if you are looking more than a couple of decades ahead. All that was predictable from simple hand-waving considerations, and indeed has been predicted for half a century.

    If you have access to Science magazine, see the report Aug. 15 p. 909 of an article on this in press in Geophysical Research Letters, with a nice graphic.

  5. 5
    Jeffrey Davis says:

    The degree of complexity takes the issue into a region where people not intimate with the complexity are simply outsiders. Since there are consequences to the issue that are the opposite of “academic”, those who want to stir the world to action are on the horns of a dilemma. What they want in the way of response and what’s possible for them to communicate to the layman are at such a remove that it will be a miracle if anything actually happens to stem climate change. In an ideal world, “Trust us” coupled with a well-organized primer (like the IPCC reports) would be sufficient, but as we all know and as Fate would have it, there are Other Interests involved. And as Upton Sinclair heartrendingly pointed out, “It’s hard to get a man to understand something when his paycheck depends upon him not understanding it.” And there are lots and lots of paychecks involved.

  6. 6
    Gsaun039 says:

    As an engineer, and one that has worked his entire career on issues involving energy and the environment, and most specifically on air/atmospheric issues, I take some exception to the characterization of “engineers.”

    While I realized long ago that simplistic equations and approaches can get you nonsense as an answer, I now tell people who attempt this that only when they’ve gone through the process line by line, layer by layer will they have some sense of what is involved. As noted, many will simply give up realizing that they have not the scientific prowess or time to go through the exercise.

    That said, I do often point out the most straight forward of the simple laws that govern the processes involved and point out that the theory does not rely on correlations of measurements to create GHG theory. Rather, I point out that the correlations support the underlying physics and are not just some random correlations without some underpinning science to them. One need only go to such sites as and look at the relationship between global temperature and the number of pirates.

    As an engineer attempting to deal with the consequences, I like to know what future I am planning for.

  7. 7
    Ron Taylor says:

    Thanks, Spencer, for a very lucid explanation. Most senior engineers, however, should understand this level of complexity and the confidence one must have in solutions obtained by numerical integration using digital computers. Otherwise, we never could have launched a single missile or space vehicle.

  8. 8
    Michael says:

    “Physics is rich in phenomena that are simple in appearance but cannot be calculated in simple terms. Global warming is like that. People may yearn for a short, clear way to predict how much warming we are likely to face. Alas, no such simple calculation exists. The actual temperature rise is an emergent property resulting from interactions among hundreds of factors. People who refuse to acknowledge that complexity should not be surprised when their demands for an easy calculation go unanswered.”

    Ah! But then people who offer “simple” solutions to this complex problem shouldn’t be so surprised when they are asked to provide the simple proof, should they? It is a double edged sword, this complexity. If you are asking for change, shouldn’t you be required to accurately predict what your change will produce? But apparently that isn’t possible.

  9. 9
    Hank Roberts says:

    > If you are asking for change, shouldn’t you be required to accurately predict
    > what your change will produce?

    You’ve stated the precautionary principle. But as Dr. Weart points out, accurate prediction was available.

  10. 10

    Would it be possible to have an actual senior engineer present their (presumably, mainstream) views of anthropogenic climate change and of the use of models?

    As a (senior? and scientifically trained) engineer myself, I can guess what Mr Weart is aiming at, but he’s still using a language that brings down no barrier. For example, a statement such as

    “Gilbert N. Plass used the data and computers to demonstrate that adding carbon dioxide to a column of air would raise the surface temperature”

    will and does definitely make people suspicious.

    You see, I have seen dozens, and I am sure there are out there hundreds of thousands of designs that have been “demonstrated” in a computer only to fail miserably when put into practice.

    In fact, one point that I don’t think Mr Weart realizes (and likely, it’s all part of the miscommunication) is that it’s the engineers that have to deal with the actual world out there, and all its complexity, starting from but having to go beyond what calculations (formulae and/or models) suggest.

    It really is the job of engineers to understand the complexity of the real world, and to make things work within that complexity.

    There is little point in arguing to your manager that, say, in the computer your revolutionary design of a car needs only 2 gallons per 100 miles, when the actual thing is measured as drinking much more than that.

    The one rule common to all engineered system is, the more stuff you put in, the higher the chances that something will go wrong. In Mr Weart’s case: the more factors need to be made to interact using models and supercomputers to calculate “global warming”, the higher the chance that the computed answer won’t be the right one.

    Therefore, rather than accusing engineers of looking for simple answers (likely, misunderstanding them), Mr Weart should try to bridge the gap.

    An example of another scientific endeavor, apart from climate change, where extremely complex, just-made modelling has been successfully applied as-is into an engineering project, would definitely be a good starting point.

  11. 11
  12. 12
    Chris Dudley says:

    For those who want a “simple” place to start, the work of the 1920’s on understanding the structures of stellar interiors is worth review. A number of approximations, many developed by Eddington, made calculation tractable with the methods of those days, often a room full of woman calculators. Much more detailed opacities are used now and the atmospheres of stars above their photospheres also get detailed treatment. It turns out that the physics is not controversial, just complicated. It is a fun study, but give it a few months to start to get a grip on it.

  13. 13
    Bryson Brown says:

    Michael (#8): Change of some kind or other is going to happen– it always does; it’s not a choice between ‘keeping things the same’ and ‘changing them’. Why do those who advocate a reduction in our GHG emissions owe you a ‘simple’ proof of what the consequences of their proposal would be, while you don’t owe them a simple proof that continuing our GHG emissions won’t have disastrous consequences? Are you foolish enough to just say, ‘well, there’s no disaster yet’, so maybe it’s OK? If so, consider this old joke: a man falling from the top of the Empire State building falls past the 20th floor and someone yells at him, ‘how are you’– he answers, ‘alright so far’… The evidence that serious problems are already developing, and that much worse are yet to come, is very strong.

  14. 14
    crust says:

    Michael (#8), just because you can’t predict something in a page doesn’t mean you can’t “accurately predict” it.

    Consider an example from pure mathematics: Most results that are at all deep haven’t (and likely can’t) be proven in one page or ten or in many cases even a hundred. But that doesn’t undermine the validity of the proof, say, of the solvability of groups of odd order (a famous example from 1963 of a very simple, very important result that took about 250 pages of very dense reasoning and hasn’t been materially condensed in the four decades since then).

  15. 15
    Ray Ladbury says:

    Michael #8: It would appear that we have what is actually a pretty simple argument here, though based on complex diagnostics: The patient says, “Doctor, Doctor, it hurts when I raise CO2 levels.”
    To which the Doctor riplies: “Don’t do that.”

    Once we have established a causal link between CO2 and climate–and by any reasonable standard, we have–if we remove the cause, we avoid the effect. It would be much more difficult to justify any sort of geoengineering aproach without a huge level of effort by your standards, but reducing CO2 emissions is a no-miss solution.

  16. 16
    Gösta Oscarsson says:

    Reading the paper I came to think of an experience I had, a (large) number of years ago, with what then was called micro simulation. A number of social scientists attached probabilities to each individual in a region as to for instance their demographic and economic behaviour and made assumptions as to how they all would react to the behaviour of all the others. (Obviously very complicated calculations, far beyond the grasp of a simple central bureaucrat = me.) What made me suspicious (and here we have my link to the present paper) was that the scientists said that they after each computation checked the aggregate results against macro data on the region in question and when they got “monstrous” and “unrealistic” descriptions of the future region they changed their assumptions untill the result was “reasonable”. This “method” made me very doubtful. Do you see the parallel?

  17. 17
    Mark says:

    Mauritzio, #10.

    OK, so please furnish us with an Earth Mk II so we can avoid having to use a mathematical simulation.

    Or do you have any explanation as to WHY the simulations are incorrect?

    Or are you just looking for an answer you’re happy with?

  18. 18
    Doug H says:

    Great post! Does anyone know where can one obtain a fairly detailed laypersons description of modern bio-geo-chemo-climate models anywhere that lists all the positive and negative feedbacks?

  19. 19
    Jeffrey Davis says:

    If you are asking for change, shouldn’t you be required to accurately predict what your change will produce? But apparently that isn’t possible.

    In this country, we’re terribly familiar with acting upon insufficient information about a potential catastrophe. Our current bill for that “1% possibility” will reach more than $1 trillion dollars before the last veteran maimed by the war is laid to rest.

    Compare the transparency of the science with the political shenanigans which prefaced the war. Yes, there’s complexity with the science of AGW, but complexity is a “mote” compared with the “beam” of the 3 Card Monty game involved in manipulating intelligence.

    Much of the change encouraged by scientists ought to dovetail nicely with the necessity brought about by Peak Oil and the good involved in extricating our political choices from the swamp of Mid Eastern grievances. The solution to all these things looks remarkably similar: burn a lot less carbon.

  20. 20
    Ray Ladbury says:

    Maurizio Morabito, I would not attach great importance to Spencer’s use of “engineer” as questioner. He is merely taking an engineer as an example of an educated professional not acquainted with details of climate science. As indicated by debacles in the APS Forum on Physics and Society and elsewhere, one could equally take “physicist,” “statistician,” or even “meteorologist.”
    The point stands that without concerted effort to understand details of the science, even an educated person will come to incorrect conclusions. Nobody is saying that the models reproduce reality, but rather that they give us insights into how the real world is working as long as we have prepared ourselves by decades of study.

  21. 21

    “You don’t need a weatherman to know which way the wind blows”

    Is there a similar phrase for climatology?

    Maybe, ‘You don’t need a climatologist to tell which way the heating goes’

    Doesn’t quite have the rhythm and poetry of Bob Dylan. We all want elegant and simple statements that reveal truth about change.

  22. 22
    Andrew says:

    #4: Oops. I don’t think I asked a question. I was just shamelessly offering praise. ;-)

    I thought that your piece was a very clear articulation of why one can’t simply jump from a handful of first-order equations to a prediction of the consequences of a doubling of CO2. Aside from the plethora of processes that even complicated climate models exclude, there’s a big chunk of simulation necessary to tease out the all of the emergent properties necessary for answering even seemingly simple climate questions.

    Anyway, great article.


  23. 23
    austin says:

    I’d like to see the models take into account Hurricanes. These are vast dissapative systems. But they may be the least of them.

  24. 24
    Douglas says:

    Following up on #11. If modeling problems are so complex that they defy closed-form solution, a sound engineering response, in my experience, is to establish sufficient correlation of the model to reality before using it to project new scenarios. If it is not capable of predicting reality, the usual request is to go look for the missing variables or relationships until it does correlate sufficiently. 3 questions then:

    (1) How well have climate models of 10-25 years ago predicted climate-scale change over the last 10-25 years?

    (2) Do climate models contain the physical models/factors that recently resulted in the toggle of the PDO and resulting projections by some that we might see a delay in global warming for 10 years or so? And if so, why did they not predict it?

    (3) If these factors are not present, what efforts are ongoing to address this?

    Can you point to any papers or discussions on this topic if the answer is too involved to give here?

  25. 25
    George Musser says:

    A question about the basic physics explanation:

    We understand the basic physics just fine, and can explain it in a minute to a curious non-scientist. (Like this: greenhouse gases let sunlight through to the Earth’s surface, which gets warm; the surface sends infrared radiation back up, which is absorbed by the gases at various levels and warms up the air; the air radiates some of this energy back to the surface, keeping it warmer than it would be without the gases.)

    Earth is in radiative equilibrium, at least over the long term. In a basic-physics explanation, it is better to describe the present warming trend as a transient response (as the system seeks out a new equilibrium at higher temperature) or a steady-state response (as a new value of opacity enters into the radiative balance)?


    (forgive me if this is a duplicate posting – I’m having trouble with the captcha)

  26. 26
    John Farley says:

    Spencer Weart, it’s a good question: how can you calculate global warming mathematically?

    Here’s my contribution: a link to an essay that I published recently for nonscientists, “The Case for Modern Anthropogenic Global Warming”. It’s a reply to the contrarian arguments advanced by Alexander Cockburn in 2007 on the CounterPunch website and in the pages of The Nation.

    Because the essay was written for a nontechnical audience, all equations have been banished to endnotes.

    In this essay, endnotes 12 and 15 give some simple calculations.

    It IS possible to calculate, using the Stefan-Boltzmann equation, the temperature of a hypothetical Earth with no greenhouse gases in the atmosphere. It’s below the freezing point of water. This shows that the greenhouse effect is a BIG effect.

    People are often surprised that global warming, caused by the enhanced greenhouse effect, is large enough to cause a problem, even though carbon dioxide is only present in the atmosphere at a level of hundreds of parts per million. Instead, the real surprise is that global warming is a small as it is!

    -John Farley
    Professor of Physics, UNLV

  27. 27
    Jim Eager says:

    Re Michael @8: “If you are asking for change, shouldn’t you be required to accurately predict what your change will produce? But apparently that isn’t possible.”

    And that impossibility is what makes it so risky: we know in general what will happen, but we don’t know exactly what will happen, which is why we can’t rule out the worst case scenario. Sounds like precisely a time to invoke the precautionary principle, no?

  28. 28
    John Lang says:

    Since the GHG sensitivity estimates are completely based on the results of climate models, a method of testing of how accurate the computer models are must be undertaken.

    There is a clear (moral, social, economic and scientific) responsibility to do so given what is at stake.

    [Response: That isn’t the case (sensitivity estimates are also made via paleo-climate and modern changes), and secondly, that testing is already being undertaken (read the IPCC AR4 report for an assessment of that work. – gavin]

  29. 29
    anna says:

    as far as simple a proof goes, not being a scientist myself i’m pretty happy with ockham’s razor (though final year of high school physics is also pretty useful). i find it entirely plausible that if you pump 37% more of a potent and important GHG into the atmosphere, then you can expect it to have an impact.

    as for engineers, i work with them regularly (civil and mechanical mainly) and they reliably and usefully design structures with a safety margin built in. so, if a beam will be reasonably expected to carry X load, it will actually be designed to carry say a 10X load. this is basically the “just to be on the safe side” factor, because complex systems are necessarily difficult to predict.

    the important and entirely reasonable thing here though is that they design things on the safe side because the consequences of not doing so could be devastating.

    and this is only for one beam on one floor of one building, we’re talking about the very planet that sustains us.

    i’m not particularly worried about super accurate predictions, the stakes are just so high and human behaviour and the climate system so complex, i’d rather play it safe with my future (and my children’s, your’s, your children’s, etc, etc).

    the quest for super accurate predictions actually reminds me of a Borges story where a people are so obsessed with making the perfect map they end up with a map as big as the land it maps. we can wait for perfectly accurate predictions but they will be useless because they will only ever be perfectly accurate *after* the event has been observed.

  30. 30
    Ray Ladbury says:

    Jeffrey Davis says: “Compare the transparency of the science with the political shenanigans which prefaced the war.”

    Oh, let’s not, shall we? I’d really like to hold science to a higher standard than politics–particularly the politics of a kleptocratic regime.

    [Response: Agreed. The war is OT. – gavin]

  31. 31
    Ray Ladbury says:

    Gösta Oscarsson says: “What made me suspicious (and here we have my link to the present paper) was that the scientists said that they after each computation checked the aggregate results against macro data on the region in question and when they got “monstrous” and “unrealistic” descriptions of the future region they changed their assumptions untill the result was “reasonable”. This “method” made me very doubtful. Do you see the parallel?”

    Actually, not at all. Big difference between this and the dynamical approach used for climate modeling.

  32. 32
    Jim Eager says:

    Re John Farley @26, your Monthly Review critique of Cockburn’s arguments is a thorough and valuable resource, but I fear your rebuttal of Contrarian Claim 3, the lag of CO2 increase behind temperature rise at the end of an ice age, has a serious deficiency in that it does not describe the subsequent role of CO2 as a feedback in adding more warmth, and thus the ability of CO2 to act as both a feedback in the case of deglaciation, and as a forcing when added directly to the atmosphere as at present.

  33. 33
    Hank Roberts says:

    John Lang Says:
    8 September 2008 at 1:44 PM

    ” Since the GHG sensitivity estimates are completely based on …” and then states his misunderstanding.

    John, where did you get this mistaken idea? Did you read it somewhere? Could be you’re misreading a good source, or you’re being fooled by fake “facts” — ??

  34. 34
    John Mashey says:

    “These people, typically senior engineers, get suspicious”.

    Please, can we get deeper than “senior engineers” – that really isn’t improving insight. If we want to do that, we need to probe a lot deeper than just “senior engineers”.

    Let me offer a speculation, although not yet a serious hypothesis:

    Amongst technically-trained people, and ignoring any economic/ideological leanings:

    1) Some are used to having
    a) Proofs
    b) Simple formulae
    c) Simulations that provide exact, correct answers, and must do so to be useful
    d) And sometimes, exposure to simulations/models that they think should give good answers, but don’t.

    2) Whereas others:
    a) Are used to missing data, error bars,
    b) Complex combinations of formulae
    c) Models with varying resolutions, approximations, and that give probabilistic projections, often only via ensemble simulations.
    d) Models that are “wrong”, but very useful.

    My conjecture is that people in category 1) are much more likely to be disbelieving, whether in science, math, or engineering.

    1) In this thread, a well-educated scientist (Keith) was convinced that climate models couldn’t be useful, because he was used to models (protein-folding) where even a slight mismodel of the real world at one step caused final results to diverge wildly … just as a one-byte wrong change in source code can produce broken results.

    See #197 where I explained this to him, and #233 where light dawned, and if you’re a glutton for detail: #66, #75,l #89, #1230, #132, #145, $151, #166 for a sample.

    2) See Discussion here, especially between John O’Connor & I. See #64 and #78. John is an EE who does software configuration management. When someone runs a rebuild of a large software system, everything must be *perfect*. There’s no such thing as “close”.

    Also in that thread, Keith returned with some more comments (#137) and me with (#146), i.e., that protein-folding was about as far away from climate modeling as you could get.

    3) Walter Manny is a Yale EE who teaches calculus in high school. He’s posted here occasionally (Ray may recall him :-), and participated in a long discussion at Deltoid, and has strong (contrarian) views. In many areas of high school/college math, there are proofs, methods known for centuries, and answers that are clearly right or wrong.

    4) “moonsoonevans” at Deltoid, in #21 & #32 describes some reasons for his skepticism, #35 is where light dawns on me. He’s in financial services, had experienced many cases where computer simulations done by smart people didn’t yield the claimed benefits. In #35 I tried to explain the difference.

    All this says that if one is talking with an open-minded technical person, one must understand where *they* are coming from, and be able to give appropriate examples and comparisons, because many people’s day-to-day experience with models and simulations might lead them to think climate scientists are nuts.


    1) Electrical engineers (a *huge* group, of which only tiny fraction are here)

    Many EEs these days do logic design, which requires (essentially) perfection, not just in the design, but (especially) in simulation.

    Design + input =>(logic simulator) => results

    At any step, the design may or may not be bug-free, but the simulator *must* predict the results that the real design would do given the input, exactly, bit for bit. Many test-cases have builtin self-checks, but the bottom line is that every test-case yields PASS or FAIL, and the simulator must be right.

    Many people buy simulators (from folks like Cadence or Synopsys), and run thousands of computers day and night simulating millions of test-case inputs. But, with a million test-cases, they’re not looking for an ensemble that provides a distribution, they’re looking for the set of test-cases to cover all the important cases, and for EVERY one to pass, having been simulated correctly. This has some resemblance to the protein-folding problem mentioned above.

    Now, at lower levels of timing and circuit design, it isn’t just ones and zeroes (there’s lots of analog waveforms, probabilistic timing issues, where one must guarantee enough margin, etc). When I’d tease my circuit designer friends “Give me honest ones and zeroes”, they’d bring in really ugly, glitchy HSPICE waveforms and say “so much for your ones and zeroes”. (This is more like the molecular “docking” problems that Keith’s colleagues mentioned.)

    At these levels, people try to set up rules (“design rules”) so that logic designers can just act at the ones-and-zeroes level.

    If one looks at EEs who worry about semiconductor manufacturing, they think hard about yields, failure attribution, and live with time-series. (Standard answer to “We got better yield this month, how do you think it looks?” was “Two points don’t make a trend.”

    2) Software engineers

    Programs often have bugs, but even a bug-free program can fall apart if you change the wrong one byte of code, i.e., fragile. (I don’t recall the source, but the old saw goes something like: if skyscrapers were like software, the first woodpecker would knock them over.)

    Configuration management / software rebuilds are fairly automated these days, and they must be correct. One cannot include the wrong version of code, or compile with incompatible options.

    Performance engineering and benchmarking tend to be more probabilistic-oriented, and although a lot of people want to believe in one number (once the mythical “MIPS” rating), we’ve (mostly) fixed that over the last 20 years. Good performance engineers have always given relative performance ranges and said YMMV (Your Mileage May Vary).

    3) Mechanical engineers

    This, I expect, varies. In some cases, closed-form equations work pretty well. In other cases, one is using big structural dynamics and computational fluid dynamics codes to obtain “good-enough” approximations to reality before actually building something. For example, automobiles are extensively modeled via “crash codes”.

    4) Petroleum engineers

    It’s been a while, but certainly, people who do seismic analysis and reservoir modeling *start* with data from the real world, analyze it to make decisions, so ought to be a little more accustomed to probabilistic analyses.

    5) Financial engineers (Google: financial engineering)

    Not having physics to constrain simulations yields some wild results, although at least, some people are very comfortable with risk, uncertainty, and ensemble projections. I especially like Sam Savage’s Flaw of Averages”.
    On the other hand, when Nobel Economists lose $B (LTCM), I’m not surprised there is skepticism about climate models.


    That’s a speculative start. I do *not* think lumping a large group together as “senior engineers” helps progress, because I have at least anecdotal evidence that the sources of skepticism tend to be attached to the kinds of models and (maybe) simulations that someone does day-by-day. The problem is that many people tend to generalize from their discipline to others, and especially if they have trouble getting useful models, they tend to be suspicious of others’.

    At one extreme, people have long-established mathematical proofs, and answers that are clearly right or wrong.

    At the other extreme, people have to make decisions based on the best approximations they can get, and if their discipline has good-enough approximations, they tend to think one way, and if the approximations aren’t so good, they may think another about equations and climate models.

  35. 35
    Jeffrey Davis says:

    re: 28

    “given what is at stake.”

    If the likelihood of a ruinous car crash were greater than 2.5% per year, the insurance companies wouldn’t let me in the door. By the time we get to a rise of 2C, I think the issues of precaution and mitigation are going to be moot. One can’t divvy up probabilities exactly, but the IPCC pegs a rise in temps of between 1.5-4.5C at the 95% level of certainty. Do you think you could get car insurance at that level of risk?

  36. 36
    Fred Jorgensen says:

    Re 27: Jim Yeager:
    “Sounds like the time to invoke the precautionary principle.”

    It depends on the cost of the precaution, the magnitude of adverse consequences, and the degree of certainty.

    If, to bring down CO2 levels to 275 ppm or so, we need a global WWII type mobilization, rationing, and restriction of freedoms, then we better have one heck of a solid case!

  37. 37
    mugwump says:

    John Farley, #26. I followed your link, and read your footnote 15 which is of particular interest to those (like myself) who are looking for methods to estimate climate sensitivity without using GCMs:

    15. Someone who distrusts all models can examine the change in temperature and solar forcing between glaciation (ice age) and interglacial (no ice age) periods. The change in temperature, revealed in ice core samples, is 5˚C, while the change in solar forcing is 7.1 W/m2. The computed climate sensitivity is therefore 5/7.1 = 0.7 K(W/m2)-1. We can use this empirically derived climate sensitivity to predict the temperature rise from a forcing of 4 W/m2, arising from a doubling of the atmospheric CO2 from pre-industrial levels. The result is a predicted temperature increase of 3˚C, which is perfectly consistent with the IPCC predictions of 1.5 to 3.5˚C (Seinfeld and Pandis, Atmospheric Chemistry and Physics, 1102).

    The problem with this argument is the glacial – interglacial temperature increase is accompanied by a large decrease in the amount of ice and snow covering the earth, and hence in the amount of sunlight reflected back towards space. Today there is relatively little ice left, and what there is is near the poles which contributes little to reflection (the poles are dark for half the year and for the other half the sun is on the horizon so they present very little “cross-section” to the sun). Thus, we would expect a much stronger positive feedback from glacial to interglacial than we would today from a similar increase in forcing.

    [Response: That is already factored in since the ice sheets are imposed as a forcing in this kind of calculation. In general, you are probably correct – the existence of extensive snow/ice cover increases sensitivity, but for the ranges of climate change we expect in the next 100 or so years that does not seem to be a big effect. – gavin]

  38. 38
    Jim Eager says:

    Re Fred Jorgensen @36: “It depends on the cost of the precaution, the magnitude of adverse consequences, and the degree of certainty.”

    Fred, you left out the cost of not taking the precaution, and the magnitude of adverse consequences of not taking the precaution, yet you specifically set the bar at 275 ppm.

    Now why would you do that?

  39. 39
    Guenter Hess says:

    Dear Maurizio Morabito #10,
    I think there is actually a more encouraging explanation to Spencer Weart’s observations, which also supports your statement.
    Firstly, at our excellent universities we train scientist and engineers in the basics of their respective fields during their undergraduate studies.
    Secondly, during their graduate studies we encourage them to go beyond the basics, discover new things, and go beyond the current knowledge and acquire new knowledge on their own. One way to do this is challenging hypothesis, theories, assumptions of complex models and I guess even sometimes the consensus.
    That is how progress is made at the universities and in the six sigma breakthrough methodology in the industry.
    Thirdly, in industry we also train them to look for simple rules and models that can be implemented with ease in a production process. This is sometimes necessary to make money.
    Actually, I think the linear feedback model that was adapted by climate science is just a typical linearization you would apply in industry to model a more complex problem.
    I also agree with you, many of them therefore understand the complexities of real world problems.
    Therefore, I am very optimistic that with the help of all those excellently trained scientists and engineers mankind will survive the warming.

  40. 40
    David B. Benson says:

    Fred Jorgensen (36) — Don’t need any of that, just 1–2% of the world’s gross product for 70–100 years. Probably it can be done for quite a bit less eventually. That shouldn’t stop us from starting now.

  41. 41
    Chris Colose says:

    Spencer Weart, thanks for the great writeup.


    concerning your last response to mugwump, I guess I assumed incorrectly that Milankovitch forcing was amplified by reflective and GHG feedbacks… but instead ice sheet responses are considered a forcing? I may just be getting tripped on words rather than anything radiatively significant, (I’m assuming that the “forcing” part is a reduction in surface albedo, though the semantics make no sense to me). How exactly are ice sheets to be considered a forcing, unless we consider freshening of the THC or something along those lines? I do agree with the comment though, that under situations of a same radiative forcing, two Earth’s (one with ice, one with little ice) would expect differences in sensitivity. But you’d have better insight to the actual quantitative aspects of that than me.

    [Response: It’s the difference between the Charney sensitivity (fast feedbacks included, ice sheets, vegetation change not) and the Earth System Sensitivity (all feedbacks included). Most discussion has focused on the former, and that is the context for the LGM calculation. The ESS is different – and indeed if you try and calculate the ESS to orbital forcing you get extreme values (because the regionality/seasonality of the orbital forcing is so important for the ice sheets). – gavin]

  42. 42
    Hank Roberts says:

    Doug H. wrote:
    > Does anyone know where can one obtain a fairly
    > detailed laypersons description of modern
    > bio-geo-chemo-climate models anywhere

    Yes, the IPCC, or the Start Here links at top of pag> that lists all the positive and negative feedbacks?

    Nope, it’s not such a simple question that all the answers are completely known even for a single moment in time, and they will change over time.

    Example (you can follow articles like this forward in time to see how the questions get worked on):

    from twenty-one years ago (before ocean pH change was noticed as an issue):

  43. 43
    Fred Jorgensen says:

    Re 38: Jim Eager:
    The cost of not taking the precaution would be the ‘Adverse Consequences’ in the broadest term.
    NASA’s Jim Hansen recently set the bar at 350 ppm, but there have been calls for pre-industrial,
    historically stable levels of 280 or so.

    Re. 36: David B. Benson: But the world’s GDP isn’t the benchmark! China and India want US
    standard of living, – with the premium for low CO2 technology paid by the developed world.
    Much more than 1-2 % of our GDP I would think! And we ‘may’ only have 10 years or so before the
    speculative ‘tipping point’! Looks like an enormous project, so we’d better NOT start down that path
    without a very high degree of certainty, since the cost of ‘precaution’ (insurance), could
    be used to solve more critical problems elsewhere.

  44. 44
    John Lang says:

    Hank Roberts Says:
    8 September 2008 at 2:24 PM
    “John Lang Says:
    8 September 2008 at 1:44 PM

    ” Since the GHG sensitivity estimates are completely based on …” and then states his misunderstanding.

    John, where did you get this mistaken idea? Did you read it somewhere? Could be you’re misreading a good source, or you’re being fooled by fake “facts” — ??”

    That is what the commentary by Spencer Weart says.

  45. 45
    David B. Benson says:

    Fred Jorgensen (43) — Well, when I worked it out, 1–2% was enough to

    (1) Deeply sequester, as biochar or torrified wood, enough carbon to offset the annual addictions of excess carbon to the active carbon cycle;

    (2) Just enough funding left to do the same with about 1% of the existing excess carbon; hence the need for 70–100 years.

    There are many variations on the above theme; it could be combined with some CO2 CCS; some of the torrified wood could directly replace coal; etc.

    The idea is simply to show that it can be done; we ought to start doing some of it right away. Moving away from fossil fuels will not alone suffice; just concrete production will surely continue to contribute about 0.4 GtC per year.

  46. 46
    The Wonderer says:

    John Mashey: Let me offer another speculation. “Senior Engineers” and other Senior Technical Leaders who are worth their salt, think more broadly and have a knack for telling scat from shinola. The people to whom you refer are stuck in a very narrow line of thinking and are unable to see the broader picture.

    If anyone out there thinks we engineers are a tough crowd, watch out for the bakers and steelworkers ;)

  47. 47
    Richard Hill says:

    The linked presentation should appeal to engineers,
    it explains surface temperature in an understandable way
    using only gravity, radiation and convection.
    The GHG effect is proved unnecessary.
    It would be appreciated if someone could point out the error
    or provide a link to the rebuttal.

    The Thermodynamic Atmosphere Effect
    – explained stepwise
    by Heinz Thieme
    German Version:
    Using a set of technical models of planets with and without an atmosphere the reasons are

    explained for differences in surface temperature of the planet without an atmosphere

    compared with the temperature in the ground layer of atmosphere of the other planet. The

    differences are caused by thermodynamic characteristics of these gases, which cause the

    mass and the composition of the atmosphere, and the atmospheric pressure, which results

    from gravity.
    Trying to avoid the spam filter, link to english version is at…

    [Response: Nonsense, I’m afraid. He assumes that the planet has an adiabatic lapse rate (heating with increasing pressure), a very specially chosen amount of mass and radiation only at the edge of the planet. This might work, but does not resemble the real world in any respect. – gavin]

  48. 48
    captdallas2 says:

    Interesting, there is no simple answer. I am curious why you didn’t reference the brilliant math used (2 plus 4 divided by 2 equals 3 degrees) to determine climate sensitivity. Tsonis et al has a paper on dynamical models that isn’t all that simple to understand. I do recommend it as a good read though.

  49. 49

    Thanks for your clear explanation of some of the basic concepts Spencer. Your book is very clear and well written too.

    One aspect of the “basic” physics that puzzles me is the lag time of CO2 from burning fossil fuels in the atmosphere.

    Archer (2007: 123) suggests, “The bottom line is that about 15-30% of the CO2 released by burning fossil fuel will still be in the atmosphere in 1000 years, and 7% will remain after 100,000 years.”

    Why is there “only” a lag of 800 years in the icecore record between the earth coming out of a glacial period and the response of CO2?

    Reference: Archer D (2007) “Global Warming: Understanding the Forecast” (Blackwell Publishing).

  50. 50

    re #46: “That is what the commentary by Spencer Weart says.”

    Not anywhere that I could see; there is a lot about what is needed to model the greenhouse effect, to be sure, but nowhere a statement about what other evidence may or may not exist.

    For a bit more about the other evidence, see Gavin’s inline response to the original query at #28, with more at #37.