Climate Science Glossary

Term Lookup

Enter a term in the search box to find its definition.


Use the controls in the far right panel to increase or decrease the number of terms automatically displayed (or to completely turn that feature off).

Term Lookup


All IPCC definitions taken from Climate Change 2007: The Physical Science Basis. Working Group I Contribution to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change, Annex I, Glossary, pp. 941-954. Cambridge University Press.

Home Arguments Software Resources Comments The Consensus Project Translations About Support

Bluesky Facebook LinkedIn Mastodon MeWe

Twitter YouTube RSS Posts RSS Comments Email Subscribe

Climate's changed before
It's the sun
It's not bad
There is no consensus
It's cooling
Models are unreliable
Temp record is unreliable
Animals and plants can adapt
It hasn't warmed since 1998
Antarctica is gaining ice
View All Arguments...

New? Register here
Forgot your password?

Latest Posts


Is the CO2 effect saturated?

What the science says...

Select a level... Basic Intermediate Advanced

The notion that the CO2 effect is 'saturated' is based on a misunderstanding of how the greenhouse effect works.

Climate Myth...

CO2 effect is saturated

"Each unit of CO2 you put into the atmosphere has less and less of a warming impact. Once the atmosphere reaches a saturation point, additional input of CO2 will not really have any major impact. It's like putting insulation in your attic. They give a recommended amount and after that you can stack the insulation up to the roof and it's going to have no impact." (Marc Morano, as quoted by Steve Eliot)


This myth relies on the use (or in fact misuse) of a particular word – 'saturated'. When someone comes in from a prolonged downpour, they may well exclaim that they are saturated. They cannot imagine being any wetter. That's casual usage, though.

In science, 'saturated' is a strictly-defined term. For example, in a saturated salt solution, no more salt will dissolve, period. But what's that got to do with heat transfer in Earth's atmosphere? Let's take a look.

Heat-trapping by CO2 in the atmosphere happens because it has the ability to absorb and pass on infra-red radiation – it is a 'greenhouse gas'. Infra-red is just one part of the electromagnetic spectrum, divided by physicists into a series of bands. From the low-frequency end of the spectrum upwards, the bands are as follows: radio waves, microwaves, infrared, visible light, ultraviolet, X-rays, and gamma rays. Gamma rays thus have a very high-frequency. They are the highest-energy form of radiation.

As our understanding of the electromagnetic spectrum developed, it was realised that the radiation consists of particles called 'photons', travelling in waves. The term was coined in 1926 by the celebrated physicist Gilbert Lewis (1875-1946). A photon's energy is related to its wavelength. The shorter the wavelength, the higher the energy, so that the very high-energy gamma-rays have the shortest wavelength of the lot.

Sunshine consists mostly of ultraviolet, visible light and infra-red photons. Objects warmed by the sun then re-emit energy photons at infra-red wavelengths. Like other greenhouse gases, CO2 has the ability to absorb infra-red photons. But CO2 is unlike a mop, which has to be wrung out regularly in order for it to continue working. CO2 molecules do not get filled up with infra-red photons. Not only do they emit their own infra-red photons, but also they are constantly colliding with neighbouring molecules in the air. The constant collisions are important. Every time they happen, energy is shared out between the colliding molecules.

Through those emissions and collisions, CO2 molecules constantly warm their surroundings. This goes on all the time and at all levels in the atmosphere. You cannot say, “CO2 is saturated because the surface-emitted IR is rapidly absorbed”, because you need to take into account the whole atmosphere and its constant, ongoing energy-exchange processes. That means taking into account all absorption, all re-emission, all collisions, all heating and cooling and all eventual loss to space, at all levels.

If the amount of radiation lost to space is equal to the amount coming in from the Sun, Earth is said to be in energy balance. But if the strength of the greenhouse effect is increased, the amount of energy escaping falls behind the amount that is incoming. Earth is then said to be in an energy imbalance and the climate heats up. Double the CO2 concentration and you get a few degrees of warming: double it again and you get a few more and on and on it goes. There is no room for complacency here. By the time just one doubling has occurred, the planet would already be unrecognisable. The insulation analogy in the myth is misleading because it over-simplifies what happens in the atmosphere.

Please use this form to provide feedback about this new "At a glance" section. Read a more technical version below or dig deeper via the tabs above!

Further details

This myth relies on the use of a word – saturated. When we think of saturated in everyday use, the term 'soggy' comes to mind. This is a good example of a word that has one meaning in common parlance but another very specific one when thinking about atmospheric physics. Other such words come to mind too. Absorb and emit are two good examples relevant to this topic and we’ll discuss how they relate to atmospheric processes below.

First things first. The effect of CO2 in the atmosphere is due to its influence on the transport of 'electromagnetic radiation' (EMR). EMR is energy that is moving as x-rays, ultraviolet (UV) light, visible light, infrared (IR) radiation and so on (fig. 1). Radiation is unusual in the sense that it contains energy but it is also always moving, at the speed of light, so it is also a form of transport. Radiation is also unusual in that it has properties of particles but also travels with the properties of waves, so we talk about its wavelength.

The particles making up radiation are known as photons. Each photon contains a specific amount of energy, and that is related to its wavelength. High energy photons have short wavelengths, and low energy photons have longer wavelengths. In climate, we are interested in two main radiation categories - firstly the visible light plus UV and minor IR that together make up sunshine, and secondly the IR from the earth-atmosphere system.

The Electromagnetic Spectrum

Fig. 1: diagram showing the full electromagnetic spectrum and its properties of the different bands. Image: CC BY-SA 3.0 from Wikimedia.

CO2 has the ability to absorb IR photons – it is a 'greenhouse gas'.So what does “absorb” mean, when talking about radiation? We are all familiar with using a sponge to mop up a water spill. The sponge will only absorb so much and will not absorb any more unless it's wrung out. In everyday language it may be described, without measurements, as 'saturated'. In this household example, 'absorb' basically means 'soak up' and 'saturated' simply means 'full to capacity'. Scientific terms are, in contrast, strictly defined.

Now let's look at the atmosphere. The greenhouse effect works like this: energy arrives from the sun in the form of visible light and ultraviolet radiation. A proportion reaches and warms Earth's surface. Earth then emits the energy in the form of photons of IR radiation.

Greenhouse gases in the atmosphere, such as CO2 molecules, absorb some of this IR radiation, then re-emit it in all directions - including back to Earth's surface. The CO2 molecule does not fill up with IR photons, running out of space for any more. Instead, the CO2 molecule absorbs the energy from the IR photon and the photon ceases to be. The CO2 molecule now contains more energy, but that is transient since the molecule emits its own IR photons. Not only that: it's constantly colliding with other molecules such as N2 and O2 in the surrounding air. In those collisions, that excess energy is shared with them. This energy-sharing causes the nearby air to heat up (fig. 2).

CO2 heat transfer

Fig. 2: The greenhouse effect in action, showing the interactions between molecules. The interactions happen at all levels of the atmosphere and are constantly ongoing. Graphic: jg.

The capacity for CO2 to absorb photons is almost limitless. The CO2 molecule can also receive energy from collisions with other molecules, and it can lose energy by emitting IR radiation. When a photon is emitted, we’re not bringing a photon out of storage - we are bringing energy out of storage and turning it into a photon, travelling away at the speed of light. So CO2 is constantly absorbing IR radiation, constantly emitting IR radiation and constantly sharing energy with the surrounding air molecules. To understand the role of CO2, we need to consider all these forms of energy storage and transport.

So, where does 'saturation' get used in climate change contrarianism? The most common way they try to frame things is to claim that IR emitted from the surface, in the wavelengths where CO2 absorbs, is all absorbed fairly close to the surface. Therefore, the story continues, adding more CO2 can’t make any more difference. This is inaccurate through omission, because either innocently or deliberately, it ignores the rest of the picture, where energy is constantly being exchanged with other molecules by collisions and CO2 is constantly emitting IR radiation. This means that there is always IR radiation being emitted upwards by CO2 at all levels in the atmosphere. It might not have originated from the surface, but IR radiation is still present in the wavelengths that CO2 absorbs and emits. When emitted in the upper atmosphere, it can and will be lost to space.

When you include all the energy transfers related to the CO2 absorption of IR radiation – the transfer to other molecules, the emission, and both the upward and downward energy fluxes at all altitudes - then we find that adding CO2 to our current atmosphere acts to inhibit the transfer of radiative energy throughout that atmosphere and, ultimately, into space. This will lead to additional warming until the amount of energy being lost to space matches what is being received. This is precisely what is happening.

The myth reproduced at the top – incorrectly stating an analogy with roof insulation in that each unit has less of an effect - is misleading. Doubling CO2 from 280 ppm to 560 ppm will cause a few degrees of warming. Doubling again (560 to 1130 ppm) will cause a similar amount of additional warming, and so on. Many doublings later there may be a point where adding more CO2 has little effect, but recent work has cast serious doubt on that (He et al. 2023). But we are a long, long way from reaching that point and in any case we do not want to go anywhere near it! One doubling will be serious enough.

Finally, directly observing the specific, global radiative forcing caused by well-mixed greenhouse gases has - to date - proven elusive. This is because of irregular, uncalibrated or limited areal measurements. But very recently, results have been published regarding the deep reinterrogation of years of data (2003-2021) from the Atmospheric Infrared Sounder (AIRS) instrument on NASA's Aqua Satellite (Raghuraman et al. 2023). The work may well have finally cracked the long-standing issue of how to make finely detailed, consistent wavelength-specific measurements of outgoing long-wave radiation from Earth into space. As such, it has opened the way to direct monitoring of the radiative impact (i.e. forcing + feedback) of greenhouse gas concentration changes, thereby complimenting the Keeling Curve - the longstanding dataset of measured CO2 concentrations, down at the planet's surface.

Note: Several people in addition to John Mason were involved with updating this basic level rebuttal, namely Bob LoblawKen Rice and John Garrett (jg).

Last updated on 31 December 2023 by John Mason. View Archives

Printable Version  |  Offline PDF Version  |  Link to this page

Argument Feedback

Please use this form to let us know about suggested updates to this rebuttal.

Related Arguments

Further reading

V. Ramanthan has written a comprehensive article Trace-Gas Greenhouse Effect and Global Warming.

Further viewing

Video by Rosh Salgado on his "All about Climate" YouTube channel in which he debunks Will Happer's claim that the CO2 effect is saturated in the atmosphere:


Prev  1  2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  17  18  19  20  Next

Comments 226 to 250 out of 715:

  1. stealth @224, given that it is the Top Of Atmosphere radiative forcing that we are discussing, the proper comparison is not with the back radiation (which is of secondary importance) but with the Outgoing Longwave Radiation (OLR).  Granted that 1.3 W/m^2 is just 0.5% of the OLR, but then, just 0.5% of the Global Mean Surface Temperature (GMST) is 7 C.  Percentages without perspective are not very informative here.

    To calculate the temperature impact of a given radiative forcing prior to any feedbacks, you must recognize that a positive radiative forcing represents a reduction in the OLR.  In order to restore the TOA energy balance, and assuming no feedbacks, the OLR must be restored to its original value.  That requires an increase in the effective temperature of radiation to space.  Assume that 240 W/m^2 OLR is required for the energy balance, then the effective temperature or radiation to space must be (240/(5.67x10^-8))^0.125, or 255 K.  A radiative forcing of 1.3 W/m^2 then, reduces the effective temperature to 254.7.  Consequently a 0.3 K increase, ignoring feedbacks, is required to restore radiative balance.

    For the full 3.7 W/m^2 from a doubling of CO2, the reduction in effective temperature is 1 K, and hence a 1 K increase is required to restore radiative balance, ignoring feedbacks.  Finally, because atmospheric temperatures within the troposphere are locked together by convection so as to follow the lapse rate, any change in temperature at the top or middle of the troposphere the results for the need to restore radiative feedback will result in a change in surface temperature of the same size.  After that occurs, the increase in back radiation will be larger than the radiative forcing, but the energy balance at equilibrium will still be neutral because heat transfer from the surface by convection and latent heat will increase to make up the difference.

    Finally, the most recent surface and TOA energy balance diagram is from Stevens et al  2012:

  2. Also, just looking at the change in temperature from increased CO2 isnt that meaningful. You cannot change temperature without also invoking the water vapour feedback. Calculating the other feedbacks is complex (hence the range in estimates for climate sensitivity) but Planck feedback plus water vapour feedback should be the baseline for considering the effects on increased CO2.

  3. scaddenp @227, it is incorrect to think of the water vapour feedback as a singular factor.  To illustrate this, consider the procedure for estimating the planck response plus water vapour feedback using Modtran.  I will do so just using the 1976 US Standard Atmosphere with no clouds for illustrative purposes.  To do it properly, you should do it for a representative sample of environmental and cloud conditions, and take a weighted average, something it is not strictly possible to do with the University of Chicago Modtran model due to the limited number of environmental conditions specified.  Bearing that caveatte in mind, however, we proceed as follows:

    1. We determine the upward IR flux at 280 ppmv with all other values unadjusted (260.02 W/m^2).
    2. We increase the CO2 concentration to 560 ppmv, thereby reducing the upward upward IR flux.
    3. We increase the temperature offset until the upward IR flux again matches the initial value (Offset of 0.86 C required.)  That represents the Planck response.
    4. We increase the water vapor scale to equal ((288 plus offset)/288)^4 to allow for the increased water vapour pressure at the higher temperature (1.012 scale factor).
    5. We again increase the temperature offset to restore the upward IR flux to the original value (Offset of 0.96 C required).  This represents the increased water vapour pressure due to the initial water vapour response.
    6. You repeat step five until the value stabilizes.  You have now calculated the Planck response plus the water vapour feedback to the Planck response.

    Now, at this stage we may want to calculate the snow albedo feedback to the Planck response plus WV feedback to the Planck response.  That will again increase the offset temperature required, which will inturn result in another round of WV responses, and a further reduction in snow cover and so on.  

    It is because feedbacks iterate like this that it is not correct to talk about the WV feedback as a singular factor.  Supose, for example, that the total cloud feedback were slightly negative rather than (as is more likely) positive.  Then the total WV feedback will be less.  On the other hand, if the snow albedo feedback is stronger than expected (as is known from observation), that will result in a stronger WV feedback.

    Because feedbacks interact in this way, I think it is conceptually better to determine the Planck response, and then determine the feedback factors as a group to the extent that is possible.

  4. Tom, I do realise that H2O isnt that straightforward a feedback - especially if you take into account clouds, (I've worked through the excellent series at SoD ) but my understanding from th IPCC reports is that uncertainties with GHE of water vapour are in the "well understood" category with good agreement between theory and experimental/observational data. (unlike say clouds, ice sheet loss, clathrates etc) Ignoring clouds, I understand the effect to be effectively double planck response. For that reason, I think claims of "only" 1K for double CO2 are particularly spurious. You can argue about the feedbacks from clouds and melting ice, and especially ocean saturation and methane release, but you cant argue too much about the water vapour.

  5. scaddenp @229, it is not that the water vapour feedback is not well understood.  Rather, because feedbacks are responses to warming of cooling, other feedbacks which also warm (or cool) will also result in an additional WV feedback response.  Therefore you cannot quantify the WV feedback without quantifying all feedbacks.  The IPCC recognize this.  They quantify the WV feedback to the Planck response alone, but note that "... because of the inherently nonlinear nature of the response to feedbacks, the final impact on sensitivity is not simply the sum of these responses. The effect of multiple positive feedbacks is that they mutually amplify each other’s impact on climate sensitivity."  Consequently, while the WV+Lapse Rate feedback increases the temperature response by 50% of the Planck response ignoring other feedbacks, their total contribution to climate sensitivity will be greater than 0.5 C.

    Ignoring cloud feedbacks, the IPCC indicates the other feedbacks will result in an increase of temperature of 1.9 C for a doubling of CO2.  If the cloud feedback would increase the temperature response by 50% by itself, then the final climate sensitivity will be > 2.85 C, with a combined WV, Lapse rate and Surface Albedo feedback greater than 0.9 C.  If, however, it is -10%, the resulting climate sensitivity will be less than 1.71 C, and the contribution of the non-cloud feedbacks will be less than 0.9 C.

  6. Hello,

    I have a question that I was hoping might be answered here.  I've read through the comments and admit that most of what is being discussed are not things I understand well.  It seems that the results of difference spectra reported by Harries et al. are a smoking gun.  IR measurements from space over time provide concrete, easy to interpret proof that the composition of the atmosphere has changed with time in such a way that more IR is captured.

    In trying to understand the methodology better, I came across this more recent publication by the same author using the same approach.  It included data from another satellite in 2003.  Here is the result:

    The paper states that "The CO2 band at 720 cm-1 ... shows some interesting behavior, with strong negative brightness temperature difference features for 1997-1970 ... whereas, the 2003-1997 ... shows a zero signature." (Edited for clarity relative to my question--the essence of it is captured)

    The "expanation" offered in the paper is essentially that there most be some compensating effect since it is known that CO2 concentrations increased between 1997 and 2003.

    I'm willing, in my ignorance, to grant that that's true.  However, I wonder, if the presence of a difference between 1970 and 1970 is seen as proof that CO2 isn't saturated, why is zero difference between 1997 and 2003 not powerful evidence that it is?  It just seems to me that if the former evidence is enough to make one feel sure CO2 is absorbing more that the later evidence should convince the same person that CO2 is not absorbing more (between those dates).

    Any insight would be appreciated!


    [RH] Fixed image width.

  7. Sorry for my various typos.  I should have read it over before submitting!

  8. basnapple @232, I have difficulty reconciling your description of the "explanation" by Griggs and Harries with that which they offer in the preprint version of their paper.

    Specifically, in that version they show (Fig 3 a&b) that two popular reanalysis products do not predict the observed changes in OLR.  They also show, however, that there is a profile constrained by observations that does predict the observed changes in OLR (Fig 3 c).  That means the observed changes in OLR are consistent with the expectations of radiative physics plus observed changes in gas and temperature profiles, even though the observations of those profiles is of insufficient resolution to permit accurate prediction of these small changes in OLR.  As they put it,

    "Simulations created using profiles merged from a number of datasets show that we can explain the differences seen in the CO2 and ozone bands by the known changes in the those gases over the last 34 years." 

    This contrasts sharply with your claimed "explanation", which is of course no explanation at all.  Converting a claim that changes in OLR lie within those expected given known limits of observation, and hence that there is no discrepancy, to a claim that a discrepancy exists for which there is no explanation is very substantial.  I doubt that editorial review would have forced so large a change on the paper.  Nevertheless I ask that you quote the original sections of the paper as published to show that you have indeed fairly represented Griggs and Harries. 


  9. Hello Tom,

    Thanks for the reply.  I read through the preprint as best I could.  I am unfamiliar with the models used and some of the terminology.  Is it true that, in their analysis, the temperature profile of the atmosphere is the fitting parameter?

    As far as how well I represented their claims, I will quote with some (perhaps unnecessary) context:

    First quote, commenting on the difference spectra figure I included in my previous post:

    "An initial inspection indicates that the processing of the data has not caused any major artifacts. In all cases the difference spectra are seen to have consistent and reproducible
    features. The only sign of asymmetry (which could indicate a mismatch of wavenumber scales between the spectra) is in the CO2 (0110 → 1000) band at 720 cm-1, which may be due to its position on the very steep high frequency wing of the CO2 fundamental centered
    at 667 cm-1."

    Second quote, in reference to the same:

    "A negative brightness temperature difference is observed in the CO2 band at 720 cm-1 in the IMG–IRIS (1997–70) and the AIRS–IRIS (2003–1970) difference spectra, indicating increasing CO2 concentrations, consistent with the Mauna Loa record (Keeling et al. 1995). However, this channel in the difference is also sensitive to temperature, and we note that in the 2003–1997 difference, despite a growth in CO2 between these years, there is no signal at 720 cm-1."

    Third quote is where the portion in my previous post comes from, now with context:

    "The CO2 band at 720 cm-1, though asymmetric for the reasons stated earlier, nevertheless shows some interesting behavior, with strong negative brightness temperature difference features for 1997–1970 and 2003–1970: whereas, the 2003–1997 (a much shorter period, of course) shows a zero signature. Since we know independently that the CO2 concentration globally continued to rise between 1997 and 2003, we must conclude that the 2003–1997 result must be due to changes in temperature that compensate for the increase in CO2. This would mean a warming of the atmosphere at those heights that are the source of the emission in the center of this band. This is somewhat contrary to the general (small) cooling of the stratosphere at tropical latitudes." (emphasis mine, I'm only trying to show where my "explanation" came from)

    Fourth quote, regarding differences between model results and observations:

    "Finally, there exists a marked gradient in the simulated spectrum between 800 and 700 cm-1, which is absent in the observations. This coincides with the far wings of the strong CO2 band centered at 667 cm-1. In sensitivity tests, this gradient showed sensitivity to the amount of CO2, and is therefore related to the strong CO2 band, and may reflect reanalysis uncertainties in temperature."

    The paper includes appendices detailing the temperature profiles used to coerce the models to the data.

    I would appreciate some help digesting this.

  10. basnappl @234, first a correction.  I thought the copy of Griggs and Harries 2004 was a preprint of the paper you were looking at.  In fact, you were looking at Griggs and Harries 2007, which is a seperate (although related) paper.  Based on that, my comment @233 is correct so far as it goes.  That is, observations of changes in OLR match those predicted by models,  within observational limits of ghg, temperature and H2O profiles.  That is, the slight discrepancy you have pointed out results from our limited knowledge conditions within the atmosphere rather than any failing of Line By Line (LBL) or Band models of radiative transfer.

    One thing we do know is that those radiative transfer models are extraordinarilly accurate.  This is shown by the match between one particular model and observations shown in the graph below:


    (For more examples, see my discussion here and my article here.)

    Because of these tests of accuracy when the conditions are well known, and because radiative transfer models are based on very well known physics, the discrepancy you point to is almost certainly the result of atmospheric conditions rather than deficiencies in the radiative transfer models.  That being the case, there is no question of the greenhouse effect being saturated, for all radiative transfer models predict the logarithmic relationship between CO2 concentration and radiative forcing, ie, that you get approximately the same increase in forcing for each doubling of CO2.

    It is interesting, however, to look at the reason for the "zero signature" between 1997 and 2003.  Griggs and Harries are more explicit in 2007 than in 2004, attributing the lack of signature to temperature profile.  As it happens, using the lapse rate it is possible to use the brightness temperature as a rough indicator of altitude.  If we do so, we see that in order to have zero influence in the band in question, the increase in CO2 between 1997 and 2003 would need to be matched by an increase in temperature at 7-10 km altitude greater than that at the surface, and that expected by the models.  In other words, it appears that Griggs and Harries have found a tropospheric hotspot.

  11. KR @222. I went and read Myhre 1998 that produced the ΔF = 5.35*ln(C/C0) W/m2 estimate for CO2 forcing. The issue I have with this is that the alpha coefficient of 5.35 is derived from three different models, which assume that the models are accurate relative to the global atmosphere.

    Being a software modeler myself (for 30+ years) dealing with RF energy through atmosphere, I understand that what is measured in the lab, what is model in software, and what the real world does are almost always very different. The real world is so noisy and chaotic that I have found models are almost useless in predicting what will really happen in the real world. I would be stunned if this is not also true for this forcing equation, and GCMs in general.

    Trying to measure this value for the real world, on average, is probably impossible given that the atmosphere is so different moment to moment and place to place, and changes in long term trends may be hard to determine since we have so little empirical measurement data. So while this is the “best go-to reference” we have, I doubt it is realistic or correct relative to what is really happening in the real world. I admit it *might be* correct, but I cannot prove or disprove it, nor can anyone else. This isn’t meant as a criticism of experts in this field, only a realization of my experience that the atmosphere is impossible to model accurately.

    Tom Curtis @226: I think you made a minor math error. If the average global temperature is 15C, or 288K, then 0.5% of this is 1.44K. I love the energy balance diagram by Stevens et al 2012. I went and read the paper and I found it quite interesting. Since Kevin Trenberth has generated a topic on the energy budget (, I am going to take my questions and comments about the energy budget over there.

    Scaddenp @227. Just because warmer air can hold more water vapor, doesn’t mean it will. Charts I have seen ( of humidity for the atmosphere over time has been about the same at low levels, but at higher altitudes relative humidity has been decreasing. If these charts are correct, this seems to suggest that water vapor has not increased as the atmosphere has warmed over this time period (65 years).


    [TD] For your comment on water vapor, please read the counterargument to "Humidity is Falling," and if you disagree with the peer-reviewed evidence presented there, please comment there.  Regarding water vapor in the stratosphere, see "What is the role of stratospheric water vapor in global warming?"  Anyone who responds to Stealth's comment here about water vapor, please, please do so on those other threads, not here.  Everybody must help to keep the conversations on the appropriate threads.  Thank you.

    [TD] Thank you for recognizing that another post is the right place to talk about energy budget!

  12. Stealth - Regarding Myhre 1998:

    "I doubt it is realistic or correct relative to what is really happening in the real world."

    You would be wrong. Those model results have been proven out, empirically measured by the satellite observations, such as those discussed in the opening post (Harries et al 2001 in particular). Have you read the opening post of this thread?

    Yes, the Myhre results are based on numeric models of radiative absorption/emission - using column estimates from three multiple latitudes, three different models to minimize bias and atmospheric variation. And they have been confirmed - the satellite spectra show the same outgoing radiation as predicted by those models. And therefore the modelling of slightly different atmospheric compositions is trustworthy. There is really no doubt about them, no significant uncertainties in direct forcing calculations. 

    If your model reproduces observations from basic physics, it's a good model. Your issues about uncertainties are unwarranted. 

  13. KR @237 To be honest, I’m fairly stunned by your response and I am not sure how to address it. Just out of curiosity, what is your background?

    You seem to be asserting that the alpha coefficient of 5.35 of the CO2 forcing function is both accurate and precise because it has been empirically measured. That cannot be true, otherwise people would not be building models as a way to attempt to arrive at these values. Correct? Why build a model when you can just measure it. Your assertion that this is basic physics and models match the real world simply cannot be true. I understand physics (I have a physics degree) and I build software models for living (I also have a computer science degree) so I think I am qualified to speak to models and physics.

    The climate is not that simple – far from it – so my assertion of uncertainty is, I believe, completely accurate and true. As further evidence to “prove” that there is enormous uncertainty in the climate, just read the TOA energy balance paper Stevens et al that was referenced by Tom Curtis @ 226. This is great paper! It is peer reviewed. At the end it states: “The net energy balance is the sum of individual fluxes. The current uncertainty in this net surface energy balance is large, and amounts to approximately 17 Wm–2. This uncertainty is an order of magnitude larger than the changes to the net surface fluxes associated with increasing greenhouse gases in the atmosphere.”

    Think about that – the uncertainty in the energy budget is ten times larger than the fluxes associated with GHGs. This is clearly proof that my assertions of uncertainty are completely warranted.

  14. @StealthAircraftSoftwareModeler:

    Out of curiousity, which climate models have you analyzed in depth? 

  15. Stealth - Perhaps you should re-read just what you have quoted: energy balance... This uncertainty is an order of magnitude larger than the changes to the net surface fluxes associated with increasing greenhouse gases in the atmosphere.

    Since what we are discussing WRT Myhre 1998 are radiative transfer codes, and the change in forcings due to changes in atmospheric composition, we are indeed speaking of the 'net surface fluxes' which have far lower uncertainties. You seem to be conflating uncertainties in accounting for multiple energy flows into a total budget with uncertainties in computing atmospheric spectral response - applying an entire collection of uncertainties to a tiny portion of the puzzle.

    If you wish to discuss the total energy budget, the sum of individual components (and their uncertainties) of the energy budget, there is an appropriate thread. However, the radiative transfer codes are well proven, giving results within under 1% of observations (Chen et al 2007) including dealing with compositional changes - arguing any significant uncertainties in that regard (as you have) is quite frankly unsupportable. 

  16. Stealth:

    Appeals to personal qualifications and arguments from incredulity such as on display in #236 and #238 are not terribly convincing.

    All this:

    Being a software modeler myself (for 30+ years) dealing with RF energy through atmosphere, I understand that what is measured in the lab, what is model in software, and what the real world does are almost always very different. The real world is so noisy and chaotic that I have found models are almost useless in predicting what will really happen in the real world. I would be stunned if this is not also true for this forcing equation, and GCMs in general.

    Trying to measure this value for the real world, on average, is probably impossible given that the atmosphere is so different moment to moment and place to place, and changes in long term trends may be hard to determine since we have so little empirical measurement data.

    I admit it *might be* correct, but I cannot prove or disprove it, nor can anyone else. This isn’t meant as a criticism of experts in this field, only a realization of my experience that the atmosphere is impossible to model accurately. [Emphasis mine.]

    Your assertion that this is basic physics and models match the real world simply cannot be true. I understand physics (I have a physics degree) and I build software models for living (I also have a computer science degree) so I think I am qualified to speak to models and physics.

    The climate is not that simple – far from it – so my assertion of uncertainty is, I believe, completely accurate and true.

    strikes me as practically equivalent to:

    I know what I'm talking about, trust me & not the data.

    I don't believe this is true, therefore it is not true.

    although I am sure it was not your intent to communicate such sentiments.

    (I have highlighted in the quotes from your comments the three words that are often the butt of jokes on medical blogs: "in my experience" or variants there of are sometimes called "the most dangerous words in medicine". I see no reason why this maxim should not generally be applicable to other scientific domains, particularly when the person asserting it is arguing against the weight of evidence, as you are.)

  17. Stealth - you say. "That cannot be true, otherwise people would not be building models as a way to attempt to arrive at these values."

    I think you are confusing different models here. I dont think anyone is doing much on work on refining the 5.35 value from Myhre. As pointed out, the uncertainities are low and matches observation.

    By comparison global GCMs are trying to model what will be climate response to this deltaF. These models do have significant uncertainities resulting in  varying estimates for climate sensitivity from 2-4.5. Dont confuse the difficulties with modelling climate response with the modelling required for calculating the forcings. Different models.

  18. ps. After getting the full picture on humidity from the link indicated, I'd be interested in your assessment of tactics used by Climate4you to mislead.

  19. Stealth

    A very important distinction needs to be made and clarified here. The models that are being referred that calculate the 5.35 ln(C/C0) result are not climate models! They are Radiative Transfer Codes; solutions to the equation of Radiative Transfer. As such what they do is, given a known state for a column of gas - temperature, pressure, humidity and composition profiles - they calculate the instantaneous radiative state at any point in that column. As such, the underlying maths is actually relatively simple. And they work from databases of very well established spectroscopic data. There are no assumptions or time based modelling or projections, they calculate a single snapshot.

    A bit like engineering stress analysis programs, where they do the same simple calculations many times over for small cells to build up the composite picture. And their results are used in a wide range of applications, Climatology is only one of them. They are used in astronomy, military,, satellite communications modelling, a whole host of different domains. And their results have been extensively tested in the field and in the lab.

  20. John Hartz @239: I would love to examine the source code of some GCMs to see if Dr. Freeman Dyson’s claim that GCMs are full of fudge factors is true. I suspect that it is true because software modelers always have to make assumptions and design trades in order to get software to run in a reasonable amount of time. Can I get the source to any of the GCM models? I doubt that I can, but it is worth a try. I know there are a dozen so models, and if source code is available, which would you recommend I look at? I don’t have time to look at them all, and they give widely different projections based on the spaghetti graphs I’ve seen, so I’d only like to see the one that is considered the best.

  21. Source code is available for many of them. See here for GISS ModelE. Weather is chaotic so the same climate model will different wiggles for different initialisations. They dont pretend to be able to predict weather. 20-30 years are what climate is about. There is a very useful article on interpretation here.

    Understanding the real differences between different modelling approache is what CMIP5 (and its predecessors) is about.

  22. Stealth, the fancy computer models merely fine tune the basic projections that have turned out to be pretty accurate, starting in the 1800s, and the early ones certainly did not involve computer code because computers had not been invented yet. You can try some of those simple models yourself by getting an introductory textbook such as David Archer's Global Warming: Understanding the Forecast, or by taking notes while watching his free online lectures from his class at the University of Chicago.

    Tamino has illustrated a simple climate model you can run without a computer if you have a lot of time, or with a spreadsheet if you don't mind using a computer. He also has a followup that's only a bit more complicated.  

    There are a bunch of other climate models that are simple enough for learning and teaching. One list has been compiled by Steve Easterbrook.

    Code for slightly more complex or narrow models also is freely available.  RealClimate's "Data Sources" page has a handy but short list.  Even the full-blown General Circulation Models (GCMs) have freely available code. RealClimate's Data Sources page also has a handy list of links to those codes.  Steve Easterbrook has a three-year old list of GCMs with links to whatever info he could find about getting their codes.  See also Tamino's Climate Data Links.

  23. Stealth - RealClimate has links to both climate data and a number of model codes here, including GCMs and others. 

    "...GCMs are full of fudge factors..."

    Um, no. They are full of physics. There are parametric approximations of small-scale phenomena, and for limitations of sampling and scale - but those are anchored in physical measurements, they are not "fudge factors" or tuning knobs for giving a specific result. Temperature projections and climate sensitivity are outputs of the models, not inputs, and a great deal of the variation between individual model runs comes from differing initial conditions. That variation is in fact part of the results, indicating to some extent the range of potential weather we might see around climate trends. See scaddenp's link above for more discussion. 

    You do, I hope, realize that "fudge factor" claims are essentionally accusations of fraud aimed at the model authors? And unsupportable ones, to boot?

  24. Stealth,

    I don’t have time to look at them all, and they give widely different projections based on the spaghetti graphs I’ve seen, so I’d only like to see the one that is considered the best.

    This demonstrates a fundamental misunderstanding of what climate models are doing.

    Weather is chaotic. The timing of even signficant events like El Nino/La Nina cannot be predicted years into the future. In order to distingish between long term climate change, and the effects of internal variability, it is essential that repeated runs of the same climate model exhibit different realisations of that internal variability. This allows them to be averaged together so that the random variations cancel out leaving behind the systematic changes that will dominate in the longer term.

    Even then, there is a risk that the forecasts of individual climate models inadvertently contain systematic biases that won't be cancelled out by this procedure due to the choices that were made in designing them and potentially even software bugs. So different models from completely different groups are also combined, to see what is common in their forecasts and what varies between them. Given their varied nature, if they all predict the same thing to within some tolerance, then we can have a certain amount of confidence that the true answer lies within that range; if they disagree about something, then our confidence is reduced.

    Of course, since they are all embodying known physics to varying degrees, they will all contain systematic biases towards "reality, as we understand it" in that regard.

    Anyway, far from being an indication of failure, those spaghetti graphs are an essential element of determining the reliability of the forecasts and trying to predict the range of possible outcomes.

    Can I get the source to any of the GCM models? I doubt that I can, but it is worth a try.

    Perhaps you should check a little bit harder before beginning to "doubt"?

  25. Stealth - "Can I get the source to any of the GCM models? I doubt that I can..."

    From the first two pages of a Google search on "code for climate model":

    Not too difficult to find...

Prev  1  2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  17  18  19  20  Next

Post a Comment

Political, off-topic or ad hominem comments will be deleted. Comments Policy...

You need to be logged in to post a comment. Login via the left margin or if you're new, register here.

Link to this page

The Consensus Project Website


(free to republish)

© Copyright 2024 John Cook
Home | Translations | About Us | Privacy | Contact Us