RealClimate logo

The Physics of Climate Modelling

Filed under: — gavin @ 3 January 2007 - (Français) (Português)

This is just a pointer to a ‘Quick Study’ guide on The physics of climate modelling that appears in Physics Today this month, and to welcome anyone following through from that magazine. Feel free to post comments or questions about the article here and I’ll try and answer as many as I can.

132 Responses to “The Physics of Climate Modelling”

  1. 101
    John Dodds says:

    Re BPL’s 85.
    Yes , if you follow my idea out logically nothing can ever change its temperature by ITSELF. This is called ENTROPY. The 2nd law of thermodynamics, says (in one of its forms) that you can’t get an entity to raise its own temperature without adding work or energy from outside. Another version is: You can’t get something for nothing. Also there is the conservation of energy.
    To raise or lower an objects temperature you have to add or subtract outside energy – ie increase the solar energy-in, and we then come to a new equilibrium at a higher temperature. It happens every day when the sun rises- but it is a moving new equilibrium temp because the solar energy in is increasing or decreasing, except at night when it is zero.
    In our case the GCMs DO comply in the GHE. they take energy from the TOA & move it to the ground level but do not return to equilibrium. GHGs do NOT create energy to raise their temperature

    Re: Hank #87 comments.
    HAnk please read Eli’s comment in at 13 Jul 2005 @ 11:32 am which I tried to reference in #54. Same thought is in #50 but more info is in the earlier section.

    Yes I agree Hank we need a mean free path estimate or definition.
    Remember the minimum mean free path is the time to travel at the speed of light to the escape distance/TOA, since there are some small number of photons that bypass all absorptions & go directly to space. I think this transit time to space can be thought of as virtually zero.
    Eli says that for a CO2 absorption the photon is resident in the CO2 for several to 10 microseconds at sea level, before being returned to the global energy bath & the residence time varies with air density/height. So IF there are say 100,000 absorptions in a CO2 to air to CO2 to air to… to CO2 to space pathway then we have a 1 second transit time. If you use a convection pathway it is longer. Since most energy transport is radiation, the transit time is on the order of zero to a second. Eli – do you have a better estimate? about how many absorbtions are there really?

    Now following through, when a single added CO2 absorbs a photon and creates the GHE warming effect, which is an unstable dis-equilibrium where the ground is warmer, TOA is cooler and there is an energy disequilibrium, where more solar energy would be absorbed than emitted (per Hansen 2005) then it will take about 1 second or so for the warmer photon & GHE warming at ground level, to transit back to the TOA to return to equilibrium. The earth is now back at equilibrium, the CO2 has been added and energy absorbed AND returned to equilibrium. The mechanism for transporting energy out (all radiation, conduction and convection as driven by the Stefan Boltzmann law, -hotter objects radiate more- based on the temperature difference from equilibrium) must be equal to the radiation part that is reflected down from the added GHG absorption plus the extra warming from the GHE delays. Even when the added anthropogenic CO2 increases continuously, the atmosphere will return to equilibrium with only a slight delay. Yes, adding CO2 results in more energy moving back & forth (stronger hurricanes), but the end result equilibrium conditions required for balancing the sun (your ref) or for balancing the earth, still result in the stable equilibrium ground temperature, not net warming.
    This is similar to the water pipe analogy above. The system moves from its unstable disequilibrium when the recycle pipe is opened (ie CO2 absorbs) and returns to its 100gpm in and out equilibrium condition as soon as the recycle pipe fills up or the recycled energy rejoins the main pipe and moves to the exit point/TOA).
    It seems to me that the Earth will return to equilibrium at all points in the air, (no ground warming, no TOA cooling ) from a GHG addition almost immediately. If so then the accumulation of GHG absorbed energy and the ever increasing GHG forcing, to warm-up by a few degrees over a century is not possible. The GHG Forcing line is actually flat. Mother Nature’s forcing of equilibrium/balance to energy-in eliminates GHE warming.

  2. 102
    William Astley says:

    In comment to “[Response: The reason why we know CO2 is not following temperature this time is because we know that we’re putting more into the atmosphere .. Statistics from the ice ages (which show a strong correlation, not none!)…More to the point, where are these sophisticated Russian statistical models of which you speak? Find a paper and I’ll analyze it. Until then this is all smoke and mirrors.”

    Is there data, papers, and logic to support the Russian assertion? Seems to be.

    1) First what goes up will come down.

    The current solar activity is at its highest level in 8000 years. (see paper below). If the proxy climatic record can be used as an analogue: This warm period will end. The sun will enter a Maunder like stage and the earth will cool. (i.e. The proxy climate record shows periodic warming, followed by cooling.)

    The following is an excerpt from the above paper.

    “According to our reconstruction, the level of solar activity during the last 70 years is exceptional, and the previous period of equally high activity occurred more than 8000 years ago. We find during the past 11,400 years the Sun spent only of the order of 10% of the time at a similar high level of magnetic activity and almost all of the earlier high-activity periods were shorter than the present episode.”

    2) Could a change in the sun affect the earth’s temperature?

    Also in support of the Russian statement is the fact that the increase in solar magnetic activity correlates with the observed increase in temperature in 20th century.

    See figure 6 in the above linked paper that shows there is a close correlation between observed global temperature anomalies and the solar index “ak”.

    From that paper: “It has been noted that in the last century the correlation between sunspot number and geomagnetic activity has been steadily decreasing from – 0.76 in the period 1868-1890 to 0.35 in the period 1960-1982, … According to Echer et al (2004), the probable cause seems to be related to the double peak structure of geomagnetic activity. The second peak, related to high speed solar wind from coronal holes seems to have increased relative to the first one, related to sunspots (CMEs) but, as already mentioned, this type of solar activity is not accounted for by sunspot number. In figure 6 long term varations in global temperature are compared to the long-term variations in geomagnetic activity as expressed by the ak-index (Nevanlinna and Kataga 2003). The correlation between the two quantities is 0.85 with p< 0.01." 3) What is the possible mechanism that would enable the sun to affect the earth's temperature? And for those who are interested in how solar changes could affect planetary temperature, the following is a link to Tinsley and Yu's paper which provides an explanation of a mechanism that links solar wind changes (also changes to the earth's magnetic field and GCR changes) to cloud formation. (An increase in clouds cools the planet. A decrease clouds warms the planet.)

  3. 103
    Hank Roberts says:

    Dave, if you make up your own definitions, instead of using the ones the physicists use, you will create inordinate confusion among your readers.

    Look up “minimum free path” please.
    And “random direction” and “drunkard’s walk”

    You’re still imagining somehow that photons can all go in the same direction, and want to be free of the planet. For the same reason the Sun doesn’t collapse like a popped balloon, the Earth’s heat doesn’t all rush off into space. Please, read the standard definitions of the terms and use them in your logic.

  4. 104
    Hank Roberts says:

    Er, that should be addressed to “John” not “Dave” on the definition of minimum free path.
    And it’s not easy to find. Til a physicist comes along, I’ll try this once. After that, I really think you ought to start a website and put what you believe all in one place, so maybe someone can understand what would be helpful to you.

    I understand this process, as an amateur, to mean the distance on average that a photon travels between being emitted (in any random direction) and interacting again.

    A ‘drunkard’s walk’ is a random progression; on average it eventually returns to where it started, wandering around. If the drunkard happens to reach a curb, he falls off it. But he doesn’t make a straight line run for the edge. Each step is in a random direction. Nor do infrared photons in an atmosphere containing greenhouse gases make a run for the top of the atmosphere. Recall in the upper atmosphere the absorbtion is _more_ efficient. And re-emission is in random directions. The photons are repeatedly caught, and re-emitted in random directions. They don’t rush off the planet from Earth any more than they rush out of the Sun.

    The only time energy rushes out of a sun the way you are imagining photons rushing out of Earth’s atmosphere is during a supernova at the point in the process when a vast quantity of matter becomes neutrinos, which don’t interact significantly. Everything else we know of interacts and takes a very long time to get anywhere.

  5. 105
    Ike Solem says:

    You are ignoring, for one thing, the role of convection in the atmosphere entirely when it comes to calculating temperatures; you are also ignoring the role of water vapor and the sensible and latent heat transport (latent being the heat absorbed and released during phase changes; when water vapor condenses it releases heat); and you are also ignoring the fact that the main effect of increased CO2 is in the upper atmosphere; where adding CO2 leads to more absorption then if CO2 is added at sea level due to temperature and pressure effects, which means that the longwave radiation that reaches the Earth’s surface would come from lower in the atmosphere as CO2 levels increase.

    Add more CO2 to the atmosphere, in other words, and the ‘mean free path’ that an infrared photon from the surface will follow before being absorbed is shortened. Thus, it’s where the radiation is absorbed in the atmosphere that matters to the surface radiation as much as the total amount that’s absorbed – all the ice sheets are on the surface, after all. Imagine the atmosphere as a stack of translucent horizontal slices, as more CO2 is added each layer becomes more opaque to infrared radiation (not at all like a greenhouse!) and then you get more radiation from the lower layers.

    This is obviously complicated, but perhaps you should look at You can also think of it this way; if you wrap a blanket around yourself on a cold night, you will warm up and a faraway observer will initially see a cooler image through an infrared viewer – but over time, the net energy released will come into equilibrium (or you’d continue to heat up indefinitely) – but you’d still be warmer inside the blanket then if you didn’t have it. Now add more blankets, and you’ll get warmer – but which blanket is delivering the most warmth to your body, the inner layer or the outer layer? As you added more blankets, the far-off observer would again see an apparent cooling at the outer layer. So while there is no net warming (your body doesn’t start generating more heat), you stay warmer.

    As far as entropy goes, you misunderstand; you can get something to raise its own temperature without energy from outside – through conversion of internal stored potential energy – like when you strike a match; the friction is miniscule compared to the release of stored internal energy. This is why there is some concern about methane clathrates – warming ocean water might result in floods of methane to the atmosphere, though I have little idea of what the chances of that are.

  6. 106
    William Astley says:

    In reply to “Since the measured solar forcing (solar max to min) is so small (0.1% ~ 0.24 W/m2), any change of that magnitude is not going to make any difference in the next few decades.”

    The change in solar forcing (which is due to changes in the solar wind, that it is hypothesized in turn affects the amount of planetary cloud cover), if the solar cycle moves to a Maunder minimum, will be more than an order of magnitude greater than what you note.

    For example, the decrease in the earth’s albedo (due to a reduction in cloud cover) 1994/1995 to 1999/2001, is (based on the data from the Earthshine Project. A link to Palle’s paper is attached below) is minus 4.97% +/- 1.66% which converts to an increase in solar forcing of 7.5 +/- 2.4 W/m2, which is three times the estimate for the CO2 forcing, in the 20th century.

    The hypothesis that warming is followed by significantly cooling, is not a surprise, based on analogues with the past. The paleoclimatic proxy data provides a record of other similar warmings in the past that were followed by significant coolings.

    As to what is the possible mechanism by which changes in the solar wind, or changes in the geomagnetic field, or changes to Galactic Cosmic Rays (GCR) can change planetary cloud cover, the following is a link to Tinsley and Yu’s paper. Increases in planetary cloud cover result in planetary cooling. Decreases in planetary cloud cover result in planetary warming.

    What are your thoughts on the hypothesis, data, and analysis?

    (See my comment 102, for papers that provide additional support for this hypothesis.)

  7. 107
    Jeffrey Davis says:

    Any change in solar forcing is going to be in addition to CO2 forcing, right? So, if the solar forcing swamps the measured CO2 forcing, we’re going to fry. And soon.

    It’s a good way to test the accuracy of that, I guess.

  8. 108
    William Astley says:

    In reply to comment 107: “Any change in solar forcing is going to be in addition to CO2 forcing, right? So, if the solar forcing swamps the measured CO2 forcing, we’re going to fry. And soon.”

    If solar activity forced the climate 7.5 W/m2 (see comment 106), then the estimated warming due to increased CO2 has been over estimated.

    The papers that provide data and analysis to support the hypothesis that past climate changes are caused by (initiated by) changes in the geomagnetic field, or the solar wind, or Galactic Cosmic Rays (GCR) all of which it is hypothesized modulate the level of cloud cover, have a reduced estimate for the magnitude of planetary warming or cooling due to changes in atmospheric CO2.

    Interestingly and controversially it appears there have been periods when the planet was warm when CO2 levels where low and cold when CO2 levels were high.

    If the modulated cloud level hypothesis is correct and the sun follows a 8000 year cycle, the current warming event will be followed by a cooling event, based on what has occurred in the past. See the attached link to show what happened 8200 years ago.

    [Response: The 8.2kyr event has nothing to do with solar, and everything to do with huge lake discharges. The chances of it happening today are zero. Please keep it real. – gavin]

  9. 109
    John Dodds says:

    Hank you are misunderstanding my position.
    In summary I contend that the earth (and the sun) are forced by the Stefan-Boltzmann equation applied at all points from the center to space, to be in balance with energy-in equal to energy out. The SB equation identifies total energy flow which includes convection conduction and radiation. In the atmosphere a majority is by radiation.

    The calculations for the GCMs do not require a balance per their own statements, claiming that the addition of CO2 results in a net imbalance that continues and accumulates as long as the CO2 is continuously added. Hence warming at gound, cooling at TOA.
    It is my contention that earth’s natural feedback system (ie the Stefan Boltzmann law,) requires that when the CO2 absorbs a photon, delays its transport resulting in longer residence time in the air (and hence warming), then this warming causes the extra energy to be transported out by increased convecton conduction and radiation per the SB Law, such that the atmosphere will return to equilibrium almost as soon as the extra CO2 absorbtion warms it. The greenhouse warming effect will not accumulate, or you do not get an ever increasing GHG Forcing curve. The earth is always at or near equilibrium, which can only be changed by adding more energy. and adding CO2 doesn’t add energy.

    Re Entropy- adding energy by changing the nature of a component (eg nuclear power, the sun, methane clathrates etc) is defined as adding external energy. Entropy says that CO2 in the air can NOT warm itself up by staying as CO2 in AIR unless it gets energy from somewhere. The GCMs agree- the ground based air is warmed by taking energy from near the TOA (see Hansen et al 2005 figure 2e.

  10. 110
    Margo says:

    Gavin, I have a whole bunch of question, but I’d like to ask them one at a time–until comments are closed.

    In your article you say:

    “The physics in climate models can be divided into three categories. The first includes fundamental principles such as the conservation of energy, momentum, and mass, and processes, such as those of orbital mechanics, that can be calculated from fundamental principles. The second includes physics that is well known in theory, but that in practice must be approximated due to discretization of continuous equations. Examples include the transfer of radiation through the atmosphere and the Navier-Stokes equations of fluid motion. The third category contains empirically known physics such as formulas for evaporation as a function of wind speed and humidity.”

    I assume GCM’s don’t solve conservation of momentum from “fundamental principles”, since, taken literally, that requires solving Navier-Stokes equations from “fundamental principles” ). I also know if you actually do model the NS equations, the reason you approximate these is not remotely “due to discretization of the continuous equations”.

    My questions right now is:
    How is conservation of momentum modeled in a GCM?

    I don’t mean how do you discretize the continuous equations, I mean what PDEs do you start from (assuming you do.) Do you parameterize things like boundary layer? (If so how?) (A reference would be fine. But I’d like something more specific than what I found this NASA page —– that said nothing more than “the solution of the momentum equations is done within the DYNAM” which is rather vague. The turbulence model tells me you use a Turbulent Kinetic Energy equations– TKE.)


    [Response: What goes into the model at an algorithmic level is described in the published literature (i.e. Schmidt et al, 2006 and references therein). You main question is not very well posed though. Conservation of momentum is ensured by simply making sure that any process that affects the velocities does so in a way that momentum doesn’t change – this is fundamental. A perfect solution of the NS equations would do that of course – but so can all imperfect solutions – it’s just something you can take care of in the formulation. Conservation of energy is the same – any energy change in one reservoir/grid box must be balanced by an equivalent energy change in another reservoir/grid box. This too is fundamental (and much more so than the NS equations). I also don’t understand your point about how we solve the NS equations. At the large (synoptic) scale, the equations can be discretised and stepped forward in time with only minor adjustments to deal with unresolved variability. Boundary layer processess are more parameterised, but in our model expand out the Reynolds stresses out to thrid-order terms, but again you need to read the published literature (cited above) for the gory details. – gavin]

  11. 111
    Margo says:

    I can see we have different ideas about the exact meaning of “solving conservation of momentum from fundamental principles”. If someone replaced the momentum equations in a GCM with Stokes Equations — which do conserve momentum– would you say they modeled conservation of momentum from fundamental principles?

    I’d say “No”. But, if you’d say yes, then I’ll at least agree that by your definition of “fundamental principles” the GCM would be solving conservation of momentum from “fundamental principles”. Of course, the results would be pathologically wrong. . .

    Anyway, thanks for the paper. So far, I’m reading under d. Dynamics “The runs here use a second-order scheme for the momentum equations”. (This is immediately followed by information describing what you do to track the motion of passive scalars like heat and humidity. )

    I’m not actually seeing much specificity about how momentum transport is modeled over all. Presumably I can look up the papers in the references and eventually find more, since the text tells me much of the dynamics is described in an earlier paper. (Which, as we all know, not repeating what was already said in an earlier paper is common practice. )

    So, thanks.

    [Response: I’m not sure I follow you at all. Conservation of momentum is a fundamental principle, as is conservation of energy. Exact solutions of NS satisfy those principles but so must everything else including the discretised versions and the boundary layer scheme. -gavin]

  12. 112

    [[It is my contention that earth’s natural feedback system (ie the Stefan Boltzmann law,) requires that when the CO2 absorbs a photon, delays its transport resulting in longer residence time in the air (and hence warming), then this warming causes the extra energy to be transported out by increased convecton conduction and radiation per the SB Law,]]

    The SB law says nothing whatsoever about convection or conduction. It deals solely with radiation.

    [[ such that the atmosphere will return to equilibrium almost as soon as the extra CO2 absorbtion warms it. The greenhouse warming effect will not accumulate, or you do not get an ever increasing GHG Forcing curve. The earth is always at or near equilibrium, which can only be changed by adding more energy. and adding CO2 doesn’t add energy.]]

    If you’re saying the atmosphere doesn’t permanently store a given bit of energy, you’re right, but that has no effect on the temperature of the atmosphere or the ground, which do rise when you have more CO2 in the air. Equilibrium can be at any temperature. And the specific equilibrium arrived it will be a function, at least partially, of the amount of greenhouse gases in the atmosphere.

  13. 113
    Hank Roberts says:

    There’s an extra dot at the end of the Radmath link above; this works:

    When I read the AIP page linked above, it makes sense to me, I can read the papers linked, at an amateur level, and find that the basic radiation physics seems quite widely agreed on. It’s basic not just for Earth but for the Sun as well. You should put your argument on your web page where people can look at your math, maybe. It’s just scattered here, can’t follow it.

    From the Radmath page linked above:

    “The Earth must radiate back into space as much total energy as it receives, to stay in equilibrium. Adding gas to the atmosphere moves the site of this emission to higher levels, which are colder. Cold things radiate less than warm ones, so the system must warm up until it can radiate enough. For more, follow the link … to the “Simple Models” essay.

    “… Callendar … assembled measurements, made in the 1930s, which showed that at the low pressures that prevailed in the upper atmosphere, the amount of absorption varied in complex patterns through the infrared spectrum. …

    “Solid methods for dealing with radiative transfer through a gas were not worked out until the 1940s. The great astrophysicist Subrahmanyan Chandrasekhar and others, concerned with the way energy moved through the interiors and atmospheres of stars, forged a panoply of exquisitely sophisticated equations and techniques. The problem was so subtle that Chandrasekhar regarded his monumental work as a mere starting-point. It was too subtle and complex for meteorologists.”
    ————- end quote———-

    And that seems to be the problem.

  14. 114
    Margo says:

    Gavin: I agree conservation of momentum and energy are fundamental principles. I think we are differing on a matter of semantics.

    I guess this is going to be long, but let me explain how I parse the term “solving conservation of momentum form fundamental principles”.

    Let me start with an example: Suppose we examine steady fully developed flow or a viscous flow in a horizontal pipe. We can write down the NS equations. We can then simplify knocking out convective terms because the flow is steady and fully developed. Then we can solve the resulting equation in closed form. (We get a poiseuille flow solution. )

    In this case– by my definition– one has solved “conservation of momentum” from “fundamental principles”. This is because both of the following apply:
    a) momentum is conserved in the solution and
    b) we used fundamental principles to describe conservation of momentum.

    Now, supposed flow is turbulent and though fully developed and steady on average. We still know the Navier Stokes equations — but because the flow is inherently unsteady, we can no longer knock out the convective terms to obtain a solution.

    Of course, we can try schemes to develop models to estimate the effect of the convective terms. But we know any model we develop does not describe transport of momentum “from fundamentals”. The turbulence model is an approximation.

    Depending on the model we concocted, we may be able to solve the new set of approximate equations in closed form. Or we many need to use computational methods.

    But either way, once we introduce the turbulence model, while our system of equations do “conserving momentum” (which is a fundamental principle), our system of equations is no longer based on “fundamental principles govering transport of momentum”.

    Consequently, the way I see it, if you stuff these into a code the code– and the resulting solutions are not “fundamental principles governing transport of momentum.” (Note: the “non-fundamental” issue has nothing to do with discretization error, differencing schemes or any details involved in stuffing this into a code. I’m actually fine with that! )

    So to say it a different way: to my way of thinking, conserving momentum alone is a necessary, but not sufficient, condition to permit us to claim a code “solves conservation of momentum from fundamental principles”.

    Now, I know you may believe my definition is too strict. But if we permit the laxer definition for the meaning, then “momentum from first principles” type models could give widely inaccurate predictions for pressure drop vs. bulk velocity in pipe flow– and that’s a bulk feature. Needless to say if I extended this example to flow with heat transfer, and used the lax definition for “energy transport based on fundamental principles”, I make these sorts of “fundamental” models give atrociously bad predictions for the temperature gradient as a function of heat addition at the pipe walls. (And of course, pipe flow is much easier than climate modeling.)

    Of course, none of this implies that all models are bad or that climate models are bad. I’m fine with models– if they are used appropriately. It just means that I found your paragraph confusing.

    Mostly, reading the paragrpah I wanted to know what really happens in the code — so thanks for the reference.

    [Response: This all appears to be due to a misreading of what I wrote – go back to the original paragraph: “The first includes fundamental principles such as the conservation of energy, momentum, and mass…” – so far so good right? All of these are fundemantal principles – “…and processes, such as those of orbital mechanics, that can be calculated from fundamental principles.” – it is the processes that can be calculated from fundemental principles. The statement that you appear confused about appears nowhere in anything I’ve written. – gavin]

  15. 115
    Margo says:

    Gavin: ” it is the processes that can be calculated from fundemental principles.”

    The process of conservation of momentum is not calculated from fundamental principles describing transport of momentum in a GCM. It’s just not.

    [Response: Conservation of momentum is not a process. Convection is a process, and must conserve momentum. Drag from unresolved gravity waves in the stratosphere is a process, and must conserve momentum. I just don’t get what point you are trying to make. -gavin]

  16. 116
    Hank Roberts says:

    > Hansen et al 2005 figure 2e

  17. 117
    William Astley says:

    In reply to “[Response: The 8.2kyr event has nothing to do with solar, and everything to do with huge lake discharges. The chances of it happening today are zero. Please keep it real.”

    The above comment is meant to be taken literally. There is no possibility that a pulse of water from the Glacial Lake Agassiz, could today stop the THC. (There is no Lake Agassiz today.)

    The above comment does not question or address the data or analysis that supports the hypothesis that there will be a sudden increase in cloud cover, when the current high solar activity ends. If there is an increase in cloud cover, all agree the planet will cool.

    The assertion that there will be an increase in planetary cloud cover and cooling, would be correct, even if the 8200BP cooling event was not caused by solar or geomagnetic field changes (there was a drop in the magnitude of the earth’s magnetic field just before the 8200BP cooling event.)

    Two comments concerning the hypothesis that a pulse of water from Lake Agassiz stopped the THC and that the THC stoppage caused the 8200 BP cooling event.

    1) Concerning mechanism. I thought the THC currently was reduced by 30%. No cooling todate. Is the THC cooling or warming mechanism non-linear?

    2) Timing of the melt water pulse in relationship to the 8200BP cooling event. See figure 4 in the attached paper. The largest early Holocene melt water pulse 1B occurred 2 thousand years before the 8,200 cooling event. Is there cooling after each melt water pulse? Why does the cooling occur after the smallest pulse?

    There is an anomalous drop in the earth’s magnetic field recorded in volcanic flows, immediately prior to the 8200BP temperature drop.

  18. 118
    Margo says:

    Orbital mechanics is also not a process. Neither are fluid mechanics, solid mechanics, continuum mechanics nor just plain “mechanics”. Yet, you wrote:

    “The physics in climate models can be divided into three categories. The first includes fundamental principles such as the conservation of energy, momentum, and mass, and processes, such as those of orbital mechanics, that can be calculated from fundamental principles.”

    If you are saying that GCMs do not model conservation of energy, momentum or mass from fundamental principles, I agree. If you say conservation of mass, momentum and energy are not processes, I agree with that too. (I would go further and note that GCM’s don’t model “many of the dominant processes involved in transport of mass, momentum or energy”.)

    If you are saying your original two sentences convey the impression that GCMs do not calculate conservation of mass, momentum or energy from mathematical models derived from fundamental principles, because those three are “principles” and not “processes”, and that the only things you are claiming are calculated from fundamental principles are “the physical processes of orbital mechanics” …. well, I guess I’m not going to worry about that claim.

    After all, I think we’ve resolve this:
    GCM’s do not describe the all the physical processes involved in conservation of momentum in any way that could be characterized as ” solving conservation of momentum using mathematical representations that are derived from fundamental principals”. (Or are you still disagreeing on that?)

    If you need to understand the process or processes that are not modeled using mathematical models based on “fundamental principles ” they are: Momentum diffusion by small scale turbulent motions, and momentum diffusion by any sub-grid scale structures.

    Diffusion by these two processes leading order in at least some portions of the flow field — and they are modeled in
    GCMs. The reasons why GCMs don’t capture these have nothing to do with “discretization”.

    If you don’t understand this, I’m not going to worry about it further.

    [Response: If your point is that not everything is included in GCMs, then just say so (and you will have no argument from me). And if you want to point out that sub-grid-scale flows are not well dealt with, then again, no argument. But semantic parsings of distortions of statements I didn’t make is pointless. Processes in the GCM conserve momentum, energy and mass. Are all processes included? No. -gavin]

  19. 119
    Jim Cross says:

    Re #117

    William, what’s your source for a 8000y solar cycle? Your link only shows that abrupt cooling took place 8200 years ago. Even if solar cycles have more influence than is usually appreciated in this forum, it doesn’t mean every cooling or warming is the result of a solar cycle influence whether it be from GCR or some other mechanism.

    I seen a lot on 100Ky year cycle from Muller and more recently from this:

    I’ve also seen a lot of discussion of 1500y cycles in this forum and other places.

    But nothing much on 8000y.

  20. 120
    William Astley says:

    In reply to Jim Cross’ comment #119 to my comment 106.

    1) In my comment 106, I included a link to a paper that notes, cloud cover has decreased by 5% from the time of 1994/1995 to 2001/2002. The paper I linked to notes a reduction in cloud cover of 5% translates into an increase in solar forcing of 7.5 W/m2 or three times the estimated forcing, 2.5 W/m2, for the CO2 increase in 20th century.

    2) In my comment 106, I included a link to a paper that notes, the current solar activity is very unusual. The sun is at its highest activity level in 8000 years.

    3) I included in my comment 106 a link to a paper that describes the mechanism by which the current high solar winds that are associated with the current state of the sun, can reduce cloud cover.

    Jim Cross’ comment:
    “Even if solar cycles have more influence than is usually appreciated in this forum, it doesn’t mean every cooling or warming is the result of a solar cycle influence whether it be from GCR or some other mechanism.”

    Before discussing specifically the current state of the sun, the following is a link to a paper that provides data that supports the assertion that cyclic changes in the sun, are responsible for cyclic warming/followed by cooling in the Holocene.

    The following is a link to Bond’s paper “Persistent Solar influence on the North Atlantic Climate during the Holocene”

    Excerpt from the above linked paper:

    “A solar influence on climate of the magnitude and consistency implied by our evidence could not have been confined to the North Atlantic. Indeed, pervious studies have tied increases in the C14 in tree rings, and hence reduced solar irradiance, to Holocene glacial advances in Scandinavia, expansions of the Holocene Polar Atmosphere circulation in Greenland; and abrupt cooling in the Netherlands about 2700 years ago … Well dated, high resolution measurements of O18 in stalagmite from Oman document five periods of reduced rainfall centered at times of strong solar minima at 6300, 7400, 8300, 9000, and 9500 years ago.”

  21. 121
    Ike Solem says:

    Margo, you say: “If you are saying that GCMs do not model conservation of energy, momentum or mass from fundamental principles, I agree…”

    Well.. that’s a nonsensical string of words. Conservation of energy is an experimentally observed phenomenon, not something ‘calculated from first principles’; in fact if you look at the history of physics, conservation of energy was temporarily called into question in the 1930’s over the issue of neutrinos in particle physics; now that neutrinos have been detected, the case is that the principle of conservartion on energy, experimentally demonstrated by Joule in the 19th century, has never been challenged.

    A common theme on this thread seems to be attempts to use poorly considered semi-scientific but wordy arguments to confuse the issue.

    For a much longer introduction to General Circulation Models, see Read the whole article; it’s far longer and more detailed then the shorter piece this thread is based on, and includes the history of the development of climate and weather models.

    The bottom line is that the GCM outputs have to be compared to detailed data from the real world, and the scarcity of oceanic temperature and current profiles is a real problem, thus all the calls for a global network of ocean sensors. The oceans play a dominant role in the long-term climate, and the lack of extensive data on heat transport and other measures (alkalinity, say) in the oceans is as much a problem for climate models as the lack of weather stations and satellites would be for short-term weather forecasting models.

    While the three classes of models (short-term weather, mid-scale regional (i.e. El Nino, NAO), and planetary GCMs) all differ from one another, they all benefit from having very accurate data on current conditions – both to compare past predictions to, and to extrapolate forward into the future.

  22. 122
    English says:

    re #92

    “Statistics from the ice ages (which show a strong correlation, not none!)”

    I’ve looked for these in table form and found nothing, as I’ve often wondered how good the correlation is. Has anyone got a link to a table of temperature, CO2 (and other gases, if possible) from ice cores?

    Thanks in advance.

  23. 123
    Jim Cross says:

    Re #120

    Okay, the article deals with variations in solar irradiance but nothing about a 8000 year cycle and even the quote you cite has reduced irradiance occurring in various periods between 1100 and 500 years. True, one of them comes close to the 8200 years ago event that you cited originally.

    Also, nothing in this article ties to the GCR theories you are promoting.

    I’m fairly sympathetic to the idea that solar variation may be a much bigger part of longer term climatic changes than is generally accepted in this forum. And, it is an odd coincidence that the Sun has been unusually active in the last 100 years or so while we have been warming. Nevertheless, that doesn’t mean that the solar cycle accounts for much of our current warming (it might account for some of it) nor does it mean that drastic cooling is right around the corner.

  24. 124
    John Dodds says:

    Re 116, Hank, for someone who relies on google so much the failure to find this surprises me. It is widely quoted in RC (Gavin is a co-author) but alas one RC link in the archives to it no longer works. Figure 2e is a variation of ones that have been quoted in previous Hansen papers.
    Hansen et al 2005 is at:
    You can download the full pdf. I must give credit to NASA GISS. (Hansen & Gavin et al) They do a great job of documenting all they do. Even give you the program they use if you want it.

    However it really bothers me that they pay lip service to the concept that the earth is at equilibrium, like the sun, then have their GISS GCM tell us that anytime you add CO2 it pushes the earth out of equilibrium, warmer ground, cooler TOA, energy flow imbalance at the TOA (less energyout at lower temp) for decades, when quite clearly the energy flow imbalance is reversed at night on a daily basis, AND there is a driving force that would return the earth to equilibrium simply by the SB law transferring more energy from the warmer ground to the cooler TOA.
    Note to BPL 112 – The Stefan Boltzmann Law units are Joules/sec-m^2 – It says nothing about what the transport mechanisms are. If the mechanism moves energy then that is part of the total calculated by the SBL. In fact it is my contention that when the GHE moves the energy from the TOA to the ground (per hansen) that the return mechanism per the SBL actually increases the flow of all three, conduction convection and radiation, because hotter air rises- conduction, hotter objects radiate more – radiation. Hence I do NOT understand why the hotter air from the GHE at ground level does NOT naturally return to equilibrium by returning to the TOA, and thus just cancel out the Greenhouse warming effect. Even with a continually increasing GHG/CO2 GHE force trying to push us warmer! The driving force as calculated by the SBL says that the energy flow WILL return as long as the ground temrperature is warmer than equilibrium. (to solar energy-in- unless we identify some mag-field source of energy-in) If the air is returned to equilibrium every day, then the conclusion about a steadily increasing GHG Forcing curve can NOT be correct.

  25. 125
    Margo says:

    Eli re 121 you said:
    “Well.. that’s a nonsensical string of words. Conservation of energy is an experimentally observed phenomenon, not something ‘calculated from first principles’;”

    For what it’s worth, I didn’t say that conservation of energy is calculated from first principles. Also, I am familiar with Joule’s contributions thermodynamics and agree the first law of is not disputed. (And in any case, the 1930s neutrino issue would not be particularly germain to issues involved in climate modeling.)

    For what it’s worth, I have both substantive and semantic issues with Gavin’s paragraph, but I don’t think it’s worth delving into that further in comments. (Also, having issues with a paragraph is not quite the same as having major issue with the model; I may or may not have issues with the model: right now I simply don’t know.)

    Thanks for the link.

  26. 126
    Hank Roberts says:

    > why the hotter air from the GHE at ground level does
    > NOT naturally return to equilibrium by returning to the TOA …
    > If the air is returned to equilibrium every day, …

    The change you are imagining is actually much slower, measured as a change in the height of the tropopause, not TOA, and it’s documented:

  27. 127
    Ike Solem says:

    You say, Hence I do NOT understand why the hotter air from the GHE at ground level does NOT naturally return to equilibrium by returning to the TOA, and thus just cancel out the Greenhouse warming effect.

    Well, that’s a replay of Richard Lindzen’s theory of the dynamic atmosphere flipping itself over and dumping radiation to space. Lindzen has been thoroughly refuted, and is now claiming that Global-warming alarmists intimidate dissenting scientists into silence. Anyhow, Lindzen’s argument was that there is “excessive vertical diffusion of heat and moisture in GCMs”; i.e. it should just all go up to the top of the atmosphere…

    Of course, if that was the case what kind of temperature change would one expect to see in the upper atmosphere? A warming or a cooling? How about if solar forcing was responsible? Would the stratosphere cool or warm, in that case?

    Answers at

    (see for warming troposphere)

    (see for cooling stratosphere)

  28. 128

    The contention that because things “return to equilibrium,” greenhouse gases can’t elevate the surface temperature or the atmospheric temperature, is pure crackpot pseudoscience. Here’s how the process actually works:

    Sunlight hits the ground.
    The ground warms up.
    It radiates photons (I = ε σ T4) upwards.
    The greenhouse gases absorb some of the photons.
    The greenhouse gases heat up, because that’s what physical objects do when they absorb photons.
    The greenhouse gases radiate photons (Stefan-Boltzmann law again).
    Some of those photons go back to the ground.

    The ground therefore has both sunshine and “atmosphere shine” heating it up, and is warmer than it would be in the absence of an atmosphere.

    This is as certain as modern science gets about a subject. The contention that air molecules collide with other molecules and therefore spread around some of the energy rather than reradiating it is quite true, but completely irrelevant. As a result of the collision, the molecules are going a little faster. Their kinetic energy has been increased.

    Molecular motion is heat. The faster the molecular motion, the higher the temperature. Thus, whether by immediate absorption or by collisional excitation, the atmosphere heats up. It heats up exactly the same amount whatever the breakdown of the two processes. Any other result would violate conservation of energy.

    It is the contention that greenhouse gases can absorb photons and not heat the atmosphere that violates conservation of energy. Arrhenius was right. The “greenhouse gases can’t heat the atmosphere” people have the energy absorbed by the greenhouse gases disappearing. That doesn’t happen, folks. In every physical process we’ve ever observed, energy is conserved.

  29. 129
    Jim says:

    RE 57

    Actually air is a quite good heat conductor. I used to work with cryogenic experiements. The only real insulator is as intense of a vaccum as possible.

  30. 130
    Jeff says:


    Please excuse my naivety to science and climatology. I have, what seems to me, a very simple question. How current are climate models? I’ve heard it can take months, if not years to create climate models, so I wonder if they incorporate the dramatic changes we’ve observed in the last 3 to 5 years. If they haven’t yet incorporated the most recent data, to what degree is information obtained from them useful?

    I hope you can appreciate my need for a layman’s explanation, however I will read (and attempt to understand) anything already written on this if you prefer to just provide a link.


  31. 131
    Jeff Brown says:

    Re 78 Ike

    Thanks. That’s exactly what I’m getting at. Don’t worry about systematic errors right now. You can’t make large numbers of simulations to (presumably) remove random errors. I suppose you could finesse this by combining the simulations of other models (something like meta studies in medical research)but there aren’t that many models are there? So how are the predictions for say, a temperature rise of between (I’m making these up) 1.5 and 5 deg C w/ a most likely value of 3.2 derived, and is there a formal derivation for the connection between this and “standard” statistics? Any references you can suggest would also be greatly appreciated.

  32. 132

    Just a repetition that the conceptualization of increasing cloud cover, may actually have an opposite effect other than that desired, although natural processes might equalize said concept leading to net zero effect.Common knowledge that any area having cloud cover has a heat blanket over night periods that actually holds in heat, so must side with those above that doubt the potential of this concept. Of vital notation is the potential of these clouds movement over power generating plants and thereby holding in excess heat generated by man.

    Man made thermal energy is very often over looked in most global models, but from space satellites the hot spots At Night are obvious over generation stations and cities.This generation of thermal radiation needs more consideration in the big Thermodynamic balance picture.