Open Mind

Monckton part 1

November 16, 2006 · 32 Comments

An editorial was recently published in the English newspaper The Sunday Telegraph by Viscount Monckton of Brenchley, claiming that global warming is false. Monckton is not a scientist, but he did produc a lengthy document supporting his claims. The editorial and its accompanying document have stirred controversy in Great Britain.

I intend to examine Lord Monckton’s claims in detail, for two reasons: first, to illustrate the incorrectness of his claims, and second, to illustrate the level of reliability of denialist propoganda. I’ll start with his “Summary of the argument”:

ALL TEN of the propositions listed below must be proven true if the climate-change “consensus” is to be proven true. The first article considers the first six of the listed propositions and draws the
conclusions shown. The second article will consider the remaining four propositions.

1. That the debate is over and all credible climate scientists are agreed.
2. That temperature has risen above millennial variability and is exceptional.
3. That changes in solar irradiance are an insignificant forcing mechanism.
4. That the last century’s increases in temperature are correctly measured.
5. That greenhouse-gas increase is the main forcing agent of temperature.
6. That temperature will rise far enough to do more harm than good.
7. That continuing greenhouse-gas emissions will be very harmful to life.
8. That proposed carbon-emission limits would make a definite difference.
9. That the environmental benefits of remediation will be cost-effective.
10. That taking precautions, just in case, would be the responsible course.

Lord Monckton has begun his piece by claiming that all ten of his “propositions” must be proven true in order to consider global warming proven. This is a classic (and well-crafted) example of the logical fallacy known as the straw man. The essence of straw man arguments is to paint a false impression (the “straw man”) of your adversary’s position, then demolish that false impression. Some of his propositions are indeed central to the anthropogenic global warming (AGW) hypothesis, but some of them are irrelevant. What, for example, do #8, #9, and #10 have to do with the truth or falsehood of AGW?

Why #1 — does he really insist on unanimous agreement? You can’t get unanimous agreement on anything in science (or any human endeavor). Believe it or not, there are scientists who actually believe that relativity is balderdash. There are scientists who believe that evolution is both wrong and evil. It seems to me that Monckton has included #1 for two reasons: 1. to counter the inconvenient truth that there is a concensus among climate scientists, and 2. because it’s a straw man — easy to shoot down.

At least one of the “propositions” is decidedly contradicted by modern climate scientists. Monckton actually claims in #3 that we must prove “That changes in solar irradiance are an insignificant forcing mechanism” in order to believe AGW. In fact climate science acknowledges changes in solar irradiance as an important forcing mechanism. But again, Monckton has stated as a necessary proposition something which is not part of AGW, not claimed by the climate scientists who support AGW — but it’s easy to shoot down.

Think about this carefully. Monckton has painted a very false picture of the claims of climate scientists who believe AGW. Perhaps the most troubling thing about his mischaracterization is that any degree of critical thinking will uncover this. You don’t need a PhD in physics to get this — if you think about it carefully. Yet large numbers of people have been taken in by Monckton’s piece. Most of Monckton’s 40-page document is about technical issues related to climate science, and the general public can’t be expected to be knowledgeable about it. But his “summary of the argument” is so transparently false that everyone would see through it, but for the natural human habit to accept what is printed without critical thinking.

As for Monckton’s 40 pages of technical talk … stay tuned, there’s more to come.

Categories: Global Warming

32 responses so far ↓

  • Cobblyworlds // November 16, 2006 at 8:17 am

    1. That the debate is over and all credible climate scientists are agreed.

    As you say, why ‘all’?

    Anyway one issue Peiser’s retraction is addressed in links running from my post here: http://www.bbc.co.uk/dna/mbsn/F2801717?thread=3642766

    ABC’s piece is here: http://www.abc.net.au/mediawatch/transcripts/s1777013.htm

    Furthermore Dano’s “A Few Things Ill Considered.”
    http://illconsidered.blogspot.com/2006/02/there-is-no-consensus.html Lists the worldwide scientific bodies who endorse the concensus.

    2. That temperature has risen above millennial variability and is exceptional.

    No this definitely does not have to be proven. The theory shows that further warming is to be expected with an increase in GHGs.

    Jim Hansen’s 1988 model presented to the US Congress has actually turned out to track the observed global temperature increases since then. http://scienceblogs.com/deltoid/2006/06/pat_michaels_fraud_pure_and_si.php and http://rabett.blogspot.com/search?q=hansen , note in particular the post “Well lookee that….” Tuesday, September 26, 2006 and the graph which neatly shows the correspondence between observation and prediction.

    Furthermore there is no trend in the ‘Total Solar Irradiance’ as measured by satellites since 1978. e.g. Frohlich PMOD.
    http://www.pmodwrc.ch/pmod.php?topic=tsi/composite/SolarConstant

    Proxy studies suggest no change prior to that. e.g. Lean 2000 (sunspot numbers), http://www.junkscience.com/Greenhouse/irradiance.gif

    There has been no trend in neutron counts as measured by counters. The neutron counters measure neutrons that result from cosmic rays hitting the upper atmosphere. cosmic rays can feasibly affect cloud cover. But there has been no trend in neutron counts therefore no trend in cosmic rays.

    And the recent 3 decades of consistent increases in global average temperature have occurred against this background of a lack of matching trend in the Sun.

    3. That changes in solar irradiance are an insignificant forcing mechanism.

    See above, they are not an insignificant mechanism, but they do not explain the last 30 years, at least. I could mention attribution studies and that the ability of models to hindcast http://www.grida.no/climate/ipcc_tar/wg1/450.htm implies that the relative elements are corectly modelled, but there is that knee jerk ‘don’t trust models’ lobby out there who refuse to believe despite evidence (accepting the model’s limitations does not undermine the things they do well).

    4. That the last century’s increases in temperature are correctly measured.

    All datasets, ground instrumental, ocean surface, satellite and rocket (?)/ baloon sondes show warming trends at the surface and in the troposphere.

    The exact amounts of warming are difficult to reconcile between the various methods as they use different physical principles and deployment methods to measure temperature. The complexity of this can be appreciated to anyone who reads the 150 pages of the US Climate Science Programme’s report: “Temperature Trends in the Lower Atmosphere. Steps for Understanding and Reconciling Differences”. Again, to concentrate on the differences is to miss the important qualitative issue, the world is warming according to all measurements.

    5. That greenhouse-gas increase is the main forcing agent of temperature.

    Again, the major agent of past changes is arguably the Sun. But how does one explain the last 30 years of warming without CO2?

    6. That temperature will rise far enough to do more harm than good.
    7. That continuing greenhouse-gas emissions will be very harmful to life.
    8. That proposed carbon-emission limits would make a definite difference.
    9. That the environmental benefits of remediation will be cost-effective.
    10. That taking precautions, just in case, would be the responsible course.

    None of this addresses the science, which is what Mr Monckton is mangling. I don’t think we’ll do anything about AGW anyway, so I don’t argue these issues.

    But the science matters.

  • Neal J. King // November 17, 2006 at 2:14 am

    So far, what you have is structured already as a critique. In order to take the most advantage from collaborators’ inputs, I suggest that you re-arrange the material as follows, to keep the material segregated by target:
    - Parse Monckton’s articles into separate points. The newspaper article has lots of little points, and the back-up paper has several big points; but just create a separate webpage on your site for each point.
    - Invite comment on each point separately. That way, you can “grow” the responses to each point separately. If you want, you could have the input comments posted; or they could just go to the editor.
    - For each point, periodically (once every couple of days), post a carefully composed response, based on the inputs received to date. It need not contain every item anyone has mentioned, as long as it is complete enough to make it perfectly clear that there is no escape. When no one is suggesting more material, or proposing any further editorial changes, it’s cooked! (more or less)
    - This is probably the point at which Edward should be invited to look at the responses to the individual points.
    - After about a week or 10 days, it should be possible to take all these response and order them in the same ordering as Monckton’s articles. It makes sense to provide two separate documents, each responding to one of his. (OK, by now three documents.) The point of that is that anyone can tell, by putting the papers side by side, that every point raised by Monckton in one of his papers has been addressed by a discussion in the corresponding response document.
    - At this point, when three separate documents have been created, they can be posted for examination and final editing.

    I guess this sounds like a lot of overhead, with all the webpages; but it should save a lot of headaches when the editor is trying to separate out which material is responsive to which of Monckton’s points. It will also mean that your individual contributors can feel more free to respond to points in which they have a special interest or knowledge, without worrying about the other points that they don’t want to talk about; at least not at the same time.

    I hope this plan makes sense to you.

    [Response: I actually started this (and made this post) before the issue of a coordinated effort came up. Your plan makes perfect sense, and I think an organized, thorough refutation of Monckton, published in a high-profile location, is a far better approach than posting in my own blog.]

  • Geoff Wexler // November 17, 2006 at 10:55 am

    I intend to spend a little time checking the numerical values in Monckton’s so called paper (about sensitivity). A brief look is throwing up a variety of questions. The first one is the lack of error bars. I notice that he does show the IPPC chart for forcings together with some huge error bars but these get removed in his calculations. Thats only the start.

    Do others agree that this is worth while?

    Instead of a CV I shall give you a link to some essays I wrote a year or so ago:

    http://www. trumpington.co.uk/green

  • Neal J. King // November 17, 2006 at 6:31 pm

    Grant,

    It seems that Edward has also started up a Wiki on this project. It might be a really good idea to connect with him and find out how to merge/coordiate efforts, rather than compete! You can find it here: http://en.wikipedia.org/wiki/User:Dbuckner/climate

    The main problem I have with Edward’s approach is that, to pull it into one article, somebody is going to have to go through each one and pull out the relevant comments - and then try to synthesize this later. That will slow things down.

  • edward // November 18, 2006 at 10:36 am

    Edward here. The reason for setting this up on a Wiki was the editorial speed - footnotes, formatting, that kind of thing. No intention to compete. I am simply pulling together all the views in one place. If you look at the article now, you see I have organised it roughly as Monckton takes it. Anyone can edit this. And please do. Please put blog links (I’ve attached a few, including this one) at the end.

    Blog links should actually go in the discussion section, but I am busy tidying up at the moment.

  • Neal J. King // November 18, 2006 at 1:52 pm

    I have drafted two sections on Edward’s Wiki, one for the Monckton anti-article and the other for the Monckton anti-backgrounder. As I have explained above, I organized these to be parallel (anti-parallel?) to Monckton’s papers, to the extent possible; although sometimes he makes little points “off to the side” that need to be countered as well.

    I don’t know how long it will take me to get through a first-round of filling in the subsections. Large sections I may just “steal” from the material previously offered (at the Wiki and here) for this goal. If I do so, I may tidy up the language, so be sure to check in and make sure that it still says what was intended!

  • Neal J. King // November 20, 2006 at 2:25 am

    Cobblyworlds,

    You posted some shots at Monckton’s discussion on the role of the Sun. However, in preparing a response, I could really use a walk-through of that entire section: start to finish. That would be pages 20-22 of the backgrounder.

  • Cobblyworlds // November 22, 2006 at 6:12 pm

    Sorry Neal,

    I am trying to kick the habit of dicussing this with contrarists (apart from when I’m correcting them and taking the mick). So I hadn’t seen your requests.

    I’m very busy at work right now, but I’ll look at it and should make an intial post here, also with reference to the wiki page.

    Please note, I am an ex sceptic who decided to look into the science. I may have trouble getting references!

  • Cobblyworlds // November 22, 2006 at 8:20 pm

    I’ve looked at the Wiki page - but can’t work it out, and this has already taken me ages to sort out. I need to get a meal before bedtime!

    I will try to go over the Wiki page as there are things I’ve noted.

    I’ve popped in some explanatory comments in [brackets].

    [start]
    It is first worth noting that there is debate about the magnitude of the Sun’s role in changes in climate in quantitative terms. However there really can be no doubt that changes in both the activity of the Sun, and the Earth’s relationship to it, do have a significant effect of the climate of the Earth. However this in no way dismisses the theory that changes in CO2 and for that matter other factors, such as tectonic changes in distribution of landmasses on the Earth, also have a role in changes in climate.

    [I think chasing the straw man of little ice age/ medieval warm period etc is a waste of time. Accordingly I have not addressed Solanki and Fligge, given the discussion in the physics community about long term changes - e.g. RC 'Trouble with Sunspots http://www.realclimate.org/index.php/archives/2006/09/the-trouble-with-sunspots/ I really think it's pointless trying to adjudicate from the sidelines.]

    Mr Monckton questions the IPCC’s 2001 statement of 0.3W, which is based on the 11-year solar cycle minimum values in 1744 and 1996. Yet the IPCC themselves draw attention to this in section 6.11.1.2,” Clearly the starting date of 1750.. ..is crucial here: a choice of 1700 would give values about twice as large; a choice of 1776 would give smaller values..” The IPCC state a range of about 0.1 to 0.5 Watts per square metre. So his otherwise unsubstantiated claim that the actual value is likely to be higher turns out to be groundless. True there has been more variance within the last century, but 1900 onwards is self-evidently not the same as 1744 to 1996.

    Mr Monckton uses the ACRIM composite (Willson & Mordvinov 2003) in his first graph on page 21 of his references. However Claus Frohlich of the Davos Solar Observatory has produced a different dataset, known as PMOD. This is signifcantly different in that the PMOD composite shows no variation of the sort that the ACRIM composite shows. Indeed Frohlich found that a “detailed error analysis shows that the PMOD composite has a long-term uncertainty of less than about 90 ppm per decade (Frohlich, 2004), which makes the observed difference between the minima not significantly different from zero.”

    But why should we trust PMOD over ACRIM? Well on page 10 of his paper Frohlich compares 3 data sets of TSI, (specifically PMOD, ACRIM and IRMB) with the Kitt Peak Magnetogram data derived by Wenzler 2005. It is found that the PMOD composite accounts for 83% of the variance, whereas ACRIM only accounts for 62%. So Frohlich’s work agrees with measurements of the Sun’s magnetic activity to a far greater degree than Willson and Mordinov’s ACRIM dataset. This is important because the TSI and magnetic data are obtained by independent methods. On this basis Frohlich finds: “The close agreement with the reconstruction from Kitt-Peak magnetograms by Wenzler (2005), and with the 3-component proxy model supports the PMOD composite as the most reliable representation of the solar irradiance variability for the last 25 years.”

    Frohlich attributes this difference between ACRIM and PMOD to corrections applied during a gap in the ACRIM data saying on page 3 of his paper “The ACRIM composite neglects the corrections of the HF during the gap and this is the main reason for the claimed upward trend of TSI over the last 25 years.”

    This therefore raises the question, why does Mr Monckton prefer to use ACRIM rather than PMOD? Could this be because PMOD does not show that trend, whereas ACRIM does? If Mr Monckton used the PMOD dataset, which shows no trend, then he would not be able to rely upon solar changes to explain the recent warming, “0.6°C in the past three decades” according to NASA’s Goddard Institute

    So from 1978 to 2005 we can reasonably conclude that there has been no change in the amount of TSI received from the Sun. Yet from 1975 to 2005 there has been a steady warming trend of almost 0.2 degC per decade leading to an increase of about 0.6degC over that period.

    Mr Monckton’s discursive attempt to attribute changes over the 20th century is of course flawed. With the significant changes of contribution by factors such as the Sun, volcanoes, particulate polution and of course CO2 over the 20th century. It is simply too complex a picture for such simple reasoning to throw light on the interactions involved. The only attainable option remains the use of climate models in attribution studies, as published in scientific literature and incorporated into the IPCC’s Assesment Reports.
    [As I have already said I don't get involved in this fatuous number crunching - GCMs are the only way to do it. So I am ignoring his numbers. I'm not sure that TOA forcing figures are interchangeable with figures of TSI.]

    [end]

    My refs:

    Frohlich, Solar Irradiance Variability Since 1978, Revision of the PMOD Composite during Solar Cycle 21. C. Frohlich, Space Science Reviews 00: 1–13, 2006. Available from here http://www.pmodwrc.ch/pmod.php?topic=tsi/composite/SolarConstant

    NASA, GISS Surface Temperature Analysis, Global Temperature Trends: 2005 Summation.
    http://data.giss.nasa.gov/gistemp/2005/

    Willson, R. C., and A.V. Mordinov, Secular total solar irradiance trend during solar cycles 21 and 22, Geophys. Res. Let.,
    30, 1199-1202, 2003.

    Wenzler, T.: 2005, ‘Reconstruction of Solar Irradiance Variations in Cycles 21-23 based on Surface Magnetic Fields’. Ph.D. thesis, ETH Nr 16199, Eidgenössische Technische Hochschule, Zürich.

  • Cobblyworlds // November 22, 2006 at 8:47 pm

    I promise I won’t spam you with any more points tonight. I’ve figured out the Wiki page and have posted the above there. (without bracketed comments).

  • Neal J. King // November 23, 2006 at 12:09 am

    Cobblyworlds,

    Thanks, I will look at your comments in review of Monckton’s article.

    I’ll let you know as more help is needed.

  • Geoff Wexler // November 25, 2006 at 12:06 pm

    Comments on Monckton’s revised estimate of global warming
    (via his Stephan Boltzmann approximation).

    Summary. Gavin Schmidt in his article Cuckoo Science has provided at least two criticisms. I pick out some of the lines of Monckton’s paper to which these apply. I think that the logical status of these criticisms is different. One of them is defensive (it shows that Monckton’s paper has not disproved the standard theory) and the other shows that there is a flaw in Monckton’s new theory. I also try to interpret Monckton’s discussion to make it easier to read.

    In the lead article Monckton in the Sunday Telegraph shows his familiarity with Boltzmann by writing :

    ”You don’t need computer models to “find” lambda. Its value is given by a century-old law, derived experimentally by a Slovenian professor and proved by his Austrian student (who later committed suicide when his scientific compatriots refused to believe in atoms). The Stefan-Boltzmann law, not mentioned once in the UN’s 2001 report, is as central to the thermodynamics of climate …..”

    Then in the pdf draft paper he takes up this point again :

    ”THE THIRD ASSESSMENT REPORT of the Intergovernmental Panel on Climate Change (UN 2001). does not refer to the Stefan-Boltzmann equation. Yet this equation is central to answering the key question in the discussion of climate change: how great will be the temperature response to radiant energy forcings such as elevated greenhouse-gas concentrations? ”

    Comment.
    To put this into context, the approximation based on the Stefan Boltzmann equation for grey body radiation is not disregarded by the standard theory but has been presented in elementary introductions to climate theory (see e.g. Hartmann p.231 for the black body version and Leeds notes for the grey body version). The normal approach moves on from this type of simple model, by incorporating much more climate physics. Monckton’s strong claim which is made in the first sentence of the first quotation amounts to the assertion that it is possible to deduce the so-called climate sensitivity (lambda) by attending the beginning of a course in climate physics and dropping the rest.

    ”total energy flux E integrated over all frequencies or wavelengths as a function of emissivity and temperature:
    E=(epsilon) (sigma) T^4
    where E is radiant energy in watts per square metre per second ;
    epsilon is emissivity, equivalent by Kirchhoff’s Law to absorptivity;”

    Comment: This (SB) equation is a law if applied to a grey body such as ice or snow. But when applied to the whole Earth it is a simple model because it deliberately excludes much of the physics. In the first line of p.25 of the pdf file Monckton refers to the emissivity of the Earth/troposphere system. This is consistent with his choice 0.6135 (discussed at the end of his Appendix A) which is considerably lower than the typical values for the ground (at infra-red wavelengths). As in other models there is a deliberate simplification; the physics tells us that most of the infra-red radiation radiated by the ground is intercepted by greenhouse gases and radiated from higher altitudes where the temperature is considerably lower than the ground. In this model however it is required that the temperature to be inserted is the ground temperature. It is the job of this reduced ‘Earth troposphere emissivity’ to compensate for this inaccuracy. There are of course other simplifying features i.e that this version of the equation involves averaging over different latitudes and types of ground. The most important issue, however, is feedback, which I shall consider later.

    ”‘The UN (2001) briefly discusses the value of = T / E, the equilibrium response of global mean surface air temperature to a change in net tropopausal irradiance.”

    Comment.
    Here Monckton introduces the most important quantity normally called the climate sensitivity. If for example the Earth’s loss of heat is reduced by greenhouse gases , it will be equivalent to the Earth receiving a few watts per square meter of additional power ; this effect is called a forcing. This raises a crucial question ‘by how much does 1 watt/m^2 of forcing raise the surface temperature?’ The value of this quantity (called lambda) will affect the future warming of the planet. It is not of course true that the TAR refers to this briefly. What Monckton means is that it does not use the simplified SB model.

    Just what is Monckton’s model for lambda? So far the SB equation has been extended (by introducing a fictitious epsilon) to provide an estimate for the total power of the infra-red radiation emitted to outer space. The next stage is to assume that the Earth and its atmosphere start in energy balance i.e the absorbed solar radiation is exactly balanced by the infra-red energy estimated by the SB equation. A forcing of 1 W/m^2 means that the outgoing infra-red is reduced by 1/Wm^2 . The incoming energy will no longer balance the outgoing and the Earth will warm up. As a result the outgoing infra-red will increase (as predicted by the SB model) until a new balance is achieved. If we assume that the absorbed solar radiation is unchanged during the warming then lambda can be determined from the SB equation alone.

    ”The quoted passage is followed by a lengthy consideration of the “remarkable” invariance of which the general-circulation climate models have treated as though it were an output.”

    Comment.
    Lambda can be estimated in two ways, by careful appeal to observations OR from theory. Monckton uses both approaches just like the climatologists ; it is just that he uses a different highly simplified theory and a different way of approaching the observations. We shall have to look separately at both of these.

    ”if the emissivity of the Earth/troposphere system is held constant, as it must be to permit like-for-like comparison of radiative forcings over time.”

    This is a most important constraint on Monckton’s model (i.e theory) Since one of the roles of the parameter called the ‘emissivity of the Earth/troposphere’ has been to allow for the effects of the troposphere , a constant emissivity builds in an assumption of a constant troposphere. But a small amount of warming increases the amount of water vapour and alters the cloud formation thus changing the troposphere. Any effects caused by a temperature dependent atmosphere are called feedbacks and will affect the actual warming. Whatever else it does , the SB model with constant emissivity is not designed to allow for this kind of feedback. There is also the question of feedback caused by reduced albedo; more on this later.

    before the table on p.25 of the pdf document he states:

    ”The reason for the near-invariance of is that over the narrow range of astronomically-low temperatures relevant to climate, the small Stefan-Boltzmann constant offsets the biquadratic T4, rendering the temperature
    response to radiative perturbation over that range near-linear.”

    The final phrase referring to linearity is correct, although the explanation based on the value of the Stefan Boltzmann constant is uneccessary or faulty. The expansion parameter is (delta T)/T which is approximately the same as (delta E)/4E which does not contain the Boltzmann constant. Anyway the value of this parameter is about (0.6/287) over the last century which is much less than 1. One effect of this linearity is that it hardly matters whether the author uses the differential value of the sensitivity defined in e.g. Hartmann or the more accurate incremental version based on finite temperature and energy differences.

    The first table on page 25 of the pdf document asserts that the “actual” value of lambda is 0.303 and that the corresponding value for a totally absorbant (black) Earth would be 0.223 (“base value”). There is an error or a misunderstanding here because the estimates are not inversely proportional to the emissivity as they should be.
    0.303/0.223 =1.359
    whereas
    1/0.6135 = 1.63
    As far as I can see this discrepancy is harmless because it is the so-called base value which looks incorrect and it is never used. Why is the base value entered into the table? Why does Monckton call his estimate of lambda the actual value? Perhaps it is a claim that it has been obtained from observations? Even if we assume that lambda is just the quantity defined by the SB equation it is still not unique, because that equation depends on the
    value assumed for the ‘emissivity of the Earth/troposphere’

    The third and fourth columns of the table are labeled so as to suggest to the reader that everyone else’s estimates of the twentieth century warming disagree with the observations whereas his own in the second row is spot on ! The third column specifies the discrepancy in warming and the last column expresses takes the ratio of calculated to observed warmings.
    These results come from the equation:

    (delta T) = lambda (forcing)

    for a single value of the forcing = 1.98 watts/metre ^2, and on his estimate 0.6C for the warming between 1900 and 1998. Clearly his estimate of forcing is crucial and I shall return to this.

    Just below the table Monckton continues as follows:

    ”The UN’s temperature projections to 2100 assume that a Clausius-Clapeyron exponential rise in water vapour pressure with temperature causes a climate feedback described as a “near-doubling” of base forcings.”

    Comment. This needs clarification. It is not the forcings which have to be doubled but the response to them. This is due to the fact that extra water vapour is a strong greenhouse gas. This mechanism has been central to global warming theory for about 110 years.

    He continues:
    ”In addition, reductions in albedo are thought to amplify base forcings by 20% (Houghton, 2002). Houghton (2006) explains that values > 0.303 allow for such climate feedbacks.”

    Comment:
    This is another effect which has been included in global warming theory from the start. An initial warming converts ice into water which reflects only about one fifth of the incoming solar energy compared to ice; this creates more warming.

    He continues:
    ” the warming of the oceans – in effect a large heat-sink – may explain why observed air temperature rose very much less in the past century than the elevated values for would mandate.”

    Comment. This time effect is normally supposed to be responsible for the so called committed warming which will occur whatever we do.

    The paragraph ends with a strong claim i.e

    ”calculation using a simple model shows that neither the proposed elevations of nor the explanation in Hansen (2006) is needed, for there is no discrepancy between observation and the calculated value ~ 0.303.”

    Comment.
    He is asserting that his model does not need to be corrected by feedbacks or time delays because it agrees with the observations. Either the effects mentioned are negligible or they have already been included in his simple model. The second option is ruled out because of the assumptions made by the model. It is obvious that the SB equation allows for no time delays; the only way it could do this would be to have a time dependent emissivity instead of the value .6135. In addition as I mentioned earlier the fact that this is held fixed, independent of temperature, means that the model omits water vapour feedback. Finally the idea behind the BS model is to estimate the rise in temperature required to increase the thermal radiation by just enough to compensate for the additional power (forcing) in the system. As discussed above, the model for lambda is based on the assumption that the absorbed component of the sunshine is held fixed while the Earth warms up to create a new balance. There has been no allowance for the fact that melting ice alters the fraction of incoming sunshine which is absorbed. This leaves us with the other possibility i.e that the feedbacks and time delay do not need to be included because they are too small to worry about.

    Forcing estimates. He now leaves the SB model for a while and includes some review material on forcing.

    The 1st. sentence of the next para. asserts:

    ”The UN (2001) attributes its estimated 1900-1998 increase of 0.6C in temperature chiefly to the forcing effect of rising concentrations of greenhouse gases in the atmosphere: all its other forcings are far smaller, less well understood, and broadly self-cancelling.”

    The part of this conclusion which is true is the description “less well understood” . But the forcings arising from the aerosols are not necessarily “far smaller” than those from the greenhouse gases and the effects of the large uncertainty ranges do not necessarily cancel. (see coloured diagram which follows figure).

    This part of Monckton’s account is written as if he is using the IPCC TAR data. But he has omitted to mention that he is ignoring their estimate for direct aerosol cooling i.e. between 0 and –2 watts/m^2 and a possible further term of similar magnitude accounting for indirect effects of aerosols on rain and clouds. By disregarding aerosols Monckton may be boosting the positive forcing effect that existed in the twentieth century and thus reducing his observational estimate for lambda = (warming / forcing). What Monckton is doing here is starting to select the part of the uncertainty range which favours his desired outcome i.e that his SB model should agree with the warming of the twentieth century. I shall come back to this point later.

    Just below the coloured diagram on p.26 Monckton asserts:

    ”Calculation shows that the UN’s table of historic forcings must include all climate feedbacks.”

    This remark must be set against the following formal defnition

    Item 6.1.1. of the TAR provides the following definition:
    “The radiative forcing of the surface-troposphere system due to the perturbation in or the introduction of an agent (say, a change in greenhouse gas concentrations) is the change in net (down minus up) irradiance (solar plus long-wave; in Wm-2) at the tropopause AFTER allowing for stratospheric temperatures to readjust to radiative equilibrium, but with surface and tropo-spheric temperatures and state held fixed at the unperturbed values”. In the context of climate change, the term forcing is restricted to changes in the radiation balance of the surface-troposphere system imposed by external factors, with no changes in stratospheric dynamics, without any surface and tropospheric feedbacks in operation (i.e., no secondary effects induced because of changes in tropospheric motions or its thermodynamic state), and with no dynamically-induced changes in the amount and distribution of atmospheric water (vapour, liquid, and solid forms). ”

    Notice how the TAR definition specifically excludes surface and tropospheric feedbacks which accounts for most (perhaps all) of the feedbacks which I have mentioned above. I don’t understand Monckton’s assertion in the previous quote. It appears that he is claiming that many independent theorists have violated their own definition of forcing.

    The table of forcings on p.26 refers to the period 1750 –1998. Monckton wants to consider the shorter period 1900 to 1998 so he has to reduce the estimates in the table. He has used data for the concentrations of CO2 in the year 1900 as well as 1750 and 1998. No mention is made of the graphs in the TAR which show how some of the other greenhouse gases have varied over the last 1000 years. (I have not followed this up). Instead he attempts to obtain an adjustment from the one gas CO2. His first step is to convert concentrations into forcings and this involves the logarithmic equations below the coloured chart on p.26 of the pdf file. Having worked out a reduction factor for the CO2 forcing he then applies the same factor to the rest of the table . Instead of scaling all the forcings by a single factor he could have scaled the concentrations. This would have given a different answer because of the logarithmic dependence of forcing on concentration. I have not tried this step. Both methods are arbitrary and involve guesswork. I have not performed a complete numerical check of this paragraph. His final estimate for the forcing is 1.99 W/m^2.

    The next paragraph asserts:

    ”This centennial forcing E from all sources can also be directly computed from the Stefan-Boltzmann equation.with emissivity 0.6134 at 14.3C, tropospheric radiant-energy flux is 237.50wm-2. Deducting 235.52wm-2 for the energy flux in 1900 gives the centennial forcing of 1.98wm-2, a near-identical result.”

    Comment:
    This comparison is a basically a repeat of that shown in line 2 of the previous table on p.25 of the pdf file. Both use his SB model and both claim remarkable agreement. It just depends on whether you start with the forcing as given and work out the warming or the other way round.

    So this is the summary so far, Monckton appears to have accounted for the warming between 1900 and 1998 by means of a well known simple model which excludes feedback and time delays. Applying the same method to the future yields much lower projected warmings. Is the standard theory wrong?

    So what is Gavin Schmidt’s answer? First he argues that Monckton has omitted the time delay in the climates’ response to the forcing. Because of this effect Monckton’s so-called observational estimate of the warming is wrong because it neglects the committed warming which will be coming later. But this won’t satisfy Monckton who might argue logically that he has managed without a time delay and is therefore skeptical about it. Monckton is not interested in the rest of the physics, just in the observations over the period 1900-1998 and he has managed to account for them quite well without a time delay without feedbacks and without taking account of aerosols.

    So what is the significance of Gavin’s first criticism? From the narrow standpoint of an observational test its effect is to defend the standard theory from the charge that it does not agree with the observations. If you consider the standard theory you have to accept that there is a committed warming and hence the observed warming so far is an under-estimate.
    We seem to have two alternative ways of accounting for the observations over the period 1900 to 1998. You can either accept all the physics on the one hand or Monkton’s simplifed story with its lower warming. The observations between 1900 and 1998 have not been able to decide between the two. But we have not finished.

    Gavin Schmidt’s second criticism of Monckton is that it neglects aerosols. In my view this is the killer blow as we shall now see. It is a question of cherry picking.

    A very hot cherry.
    Monckton takes the forcing to be 1.98wm-2. This depends on his choice of zero for the cooling effect of the aerosols. Lets take a different value for the effect of the aerosols , instead of Monkton’s choice zero I shall take -1.75 wm-2 . This appears arbitrary but it is within the range of uncertainty for the direct cooling and takes no account of the indirect cooling which is also negative. The new value for the forcing is 0.23 wm-2 which is now in poor agreement with Monckton’s theoretical SB value of 1.99. You may object that my choice of aerosol cooling is unjustified . That would be right, but so is Monckton’s; both are examples of cherry picking. The new value of lambda would be 0.6/0.23 = 2.61 degs m^2 w^-1
    instead of Monckton’s “actual value” of 0.303 degs m^2 w^-1. But it does not stop there. If I had chosen a slightly higher estimate for the aerosols I could have obtained infinite or even negative values for lambda. There is something wrong. That point has been discussed by Gavin Schmidt in Realclimate in an article called ‘Climate sensitivity and aerosol forcings’ of 6th.July 2005 where he argued that the twentieth century is not a good century for estimating the climate sensitivity, partly because you might obtain excessively HIGH values for the climate sensitivity. Monckton has done the opposite obtaining a very LOW value. The logical conclusion of this criticism is different from the effect of the time delay. Cherry picking invalidates the comparison between observation and theory. Monckton’s theory is no longer a valid alternative because he has omitted to estimate the error bars. As a result of this criticism Gavin Schmidt’s first criticism is strengthened. The one justification for ignoring the rest of the physics was the apparent agreement between the SB model and the observations. If this agreement is based on a flawed argument then this justification collapses and we have to go back to the physics.

  • Neal J. King // November 26, 2006 at 12:29 am

    Geoff Wexler,

    I believe there are three main problems with Monckton’s M-model:

    a) The equation,
    energy-flux = emissivity * sigma * T**4,
    simply has no meaning when the emissivity is a function of frequency (as is certainly true for the atmospheric system). This becomes extremely clear when you follow the derivation, starting from Planck’s blackbody radiation law. If emissivity depends on frequency, you cannot pull out the dependence on T to get T**4 (rather, you can pull it out, but what is left over is not independent of temperature, so you don’t get a constant).

    b) Therefore, the “emissivity” Monckton uses is just that value needed to make the LHS of the equation equal the RHS of the equation, without a real derivation. So the only real meaning of the equation is the assumption (unjustified) that flux is proportional to T**4. What this boils down to is the relationship of any change in energy flux to the resulting change in temperature. According to this postulated relation, one must expect:
    delta-energy flux =
    4* emissivity * sigma *delta-T* T**3
    = 4*(delta-T/T)*(energy flux)

    or:

    (delta-energy-flux/energy-flux) = 4*(delta-T)/T

    c) I haven’t sat down to do the arithmetic, but I assume that this is essentially his “prediction” of lambda. To check how well this M-model works, we would have to compare this equation to the results. It is not yet clear to me what he settles on as the value of the delta-energy-flux: Is it just the CO-2 forcing, or does he include other effects? (I have only looked at the paper superficially, but I wanted to encourage your thoughts even though I haven’t finished my own considerations!)

    At any rate, if it turns out that a reasonable value for delta-energy-flux and the measured delta-T fit into the equation above, I think that would be a surprising coincidence (surprising because I see no theoretical justification for it). What I suspect instead is that Monckton finds a way of selecting his value of delta-energy-flux in order to make the equation fit.

    What do you think?

  • Geoff Wexler // November 26, 2006 at 5:12 pm

    Neal J. King
    (I just accidentally erased my “quick response” so shall have to retype)

    Initial reply to para(a)
    Yes I agree with your para(a) except that it does not include my point about a variety of sources at different temperatures. This means that I was too hasty in referring to the Stefan Boltzmann LAW for a grey body. I have a book which asserts that the radiant power emitted by a gas is the product of a well defined factor T^4 and an ill defined pair of factors both of which depend on temperature (not worth providing the reference). This is consistent with what you write at the end of your para(a) which I prefer to the phrase “simply has no meaning” at the beginning of the para.

    Incidentally we agree that he is not actually relying on the SB equation. It is the number 4 which crops up in both your and my comments which is crucial. A better computer calculation (based on similar assumptions of neglecting feedback and time delays) would probably replace this 4 with something else.

    Now it is not always simple to evaluate a highly simple model because the various errors sometimes compensate, particularly if you fit it partially against observations which is what is done here and elsewhere by choosing an emissivity. As I wrote earlier , the comparison of the model with the observed warming is seriously flawed because it is based on assuming zero aerosol effects i.e on cherry picking.

  • Neal J. King // November 30, 2006 at 2:43 am

    Geoff Wexler,

    In mulling things over, I’ve come to another realization: If one assumes literally Monckton’s equation (the one that isn’t the real Stefan-Boltzmann equation, because it uses the fake epsilon) to be true, it gives the wrong sign for the lambda factor!

    Why? If you believe what is says literally, it says that higher temperature is related to higher emitted radiation - period.

    Now, look at the forcing due to more CO2: it cuts down on the outgoing IR flux, so it reduces the cooling. The fact that it reduces the cooling makes it a “heating” force; but the fact that it does this by reducing the IR flux means that the delta-energy-flux is negative, which means that (according to the M-model) delta-T is negative.

    So that would mean that the forcing due to CO2 was negative forcing, which is definitely not true!

    How did we end up with this contradiction? It’s because the S-B formula applies only in cases of thermal equilibrium: in that case, the radiation intensity is going to be controlled by the temperature. But in the case of interest (earth + atmosphere), we don’t have thermal equilibrium, and so the radiation distribution (which is non-Planckian) plays all sorts of energy-redistribution games. This is exactly why the S-B formula does not apply: the earth+atmosphere system is just not acting alike a rock. Specifically, a reduction in energy-flux, that would have to be due to a decrease in temperature for something at thermal equilibrium, means no such thing in this case: it is because the point of radiation has been moved up to a higher altitude and thus to a lower temperature.

    Another way of thinking about it: in thermal equilibrium, everything is balanced and perfect, so you can think of the temperature of the body as controlling the radiation fluxes, or the of the radiation fluxes as controlling the temperature. It doesn’t make any difference. But in the earth+atmosphere system, what is clearly going on is that the relative amounts of the radiation fluxes is driving heat exchange and thus temperature. The system does not make any sense if you try to understand the radiation fluxes as driven by the temperatures: there’s not enough information.

    This is definitely going into the M-model section of the anti-article. The only thing I want to understand better is the way in which he manufactures “agreement” between his model and the “measured” lambda: it could be a coincidence (because the model is clearly bonkers), but I’m suspicious that it’s really due to a method of justifying the selection of a value of the flux that just happens to match the desired lambda and measured temperature increase.

    I’ll write this up in the wiki.

  • paulh // November 30, 2006 at 3:31 pm

    Hi Grant,

    I followed this over from Real Climate and I’ve just posted a small contribution onto the Wiki. I hope no one minds me joining in like this.

    Regards
    Paul

  • Geoff Wexler // November 30, 2006 at 6:10 pm

    Hi Neil,

    I came on to put on a revised version and have encountered your most intereresting argument. I shall do my best to save the SB-lambda approach because it is not due to Monckton anyway and I have other comments. But here is my initial reaction to your last post.

    It seems to me that you have made a non-trivial criticism of the SB-lambda (my terminology) model but that you have taken it too far.

    Lets start in energy balance (not equilibrium) and add a bit of greenhouse gas. Because of time delays we no longer have an energy balance and we shall find that the outgoing infra-red is reduced by delta(E) because the source has moved to higher altitude which is cooler. So seen from the outside there is no doubt the Earth/atmosphere has cooled. The fact that the input power is now less than the output takes time to have an effect.

    According to SB the temperature will also have cooled immediately which has the correct sign.
    First change in temp. = delta1(T)

    The effect of delta(E) is that everything will go on warming until a new power balance is achieved. SB will yield its usual estimate for the warming delta(T). |So it looks as if the final temp. is now

    - delta1 (T) + delta(T)

    which is zero to first order. I might try 2nd. order later. But this would be a completely useless model ! We have two different IR outputs for the same temperature i.e

    (a) The initial temperature with no greenhouse perturbation
    (b) The final temperature (almost the same) with greenhouse perturbation.

    What’s wrong? Clearly the perturbation changes the atmosphere and must be treated initially as a change in emissivity. So here is my patched up model for now:

    E(1) = k T(1)^4 = S (absorbed short wave)
    E(2) = (k-delta(k)) T(1)^4

  • Geoff Wexler // November 30, 2006 at 6:12 pm

    Hi Neil,

    I came on to put on a revised version and have encountered your most intereresting argument. I shall do my best to save the SB-lambda approach because it is not due to Moncton anyway and I have other comments. But here is my initial reaction to your last post.

    It seems to me that you have made a non-trivial criticism of the SB-lambda (my terminology) model but that you have taken it too far. (By the way this is not Monckton’s model).

    Lets start in energy balance (not equilibrium) and add a bit of greenhouse gas. Because of time delays we no longer have an energy balance and we shall find that the outgoing infra-red is reduced by delta(E) because the source has moved to higher altitude which is cooler. So seen from the outside there is no doubt the Earth/atmosphere has cooled. The fact that the input power is now less than the output takes time to have an effect.

    According to SB the temperature will also have cooled immediately which has the correct sign.
    First change in temp. = delta1(T)

    The effect of delta(E) is that everything will go on warming until a new power balance is achieved. SB will yield its usual estimate for the warming delta(T). |So it looks as if the final temp. is now

    - delta1 (T) + delta(T)

    which is zero to first order. I might try 2nd. order later. But this would be a completely useless model ! We have two different IR outputs for the same temperature i.e

    (a) The initial temperature with no greenhouse perturbation
    (b) The final temperature (almost the same) with greenhouse perturbation.

    What’s wrong? Clearly the perturbation changes the atmosphere and must be treated initially as a change in emissivity. So here is my patched up model for now:

    E(1) = k T(1)^4 = S (absorbed short wave)
    E(2) = (k-delta(k)) T(1)^4

  • Geoff Wexler // November 30, 2006 at 6:15 pm

    (this bit was chopped by software)
    where delta(k) is caused by greenhouse gas perturbation
    Response is to warm up to T(final) such that
    E(3) = (k- delta(k)) T(final)^4 = S

    Now delta(k) is determined by the forcing - delta(E) i.e by
    - delta(E) = - delta(k)T(1)^4
    Solving for T(final) gives the same result as usual for this model.

    I have not had time to check.

  • Geoff Wexler // November 30, 2006 at 9:36 pm

    Footnote to tatty messages.
    At first sight the two parts of my previous reply to Neil appear to contradict.
    I argued that the effective temperature of the IR radiation will really fall (initially) if a small addition to the greenhouse gases is made. Then I proposed that the SB-lambda model should be slightly modified by modeling the extra greenhouse gases by an equivalent reduction in emissivity. This would have the effect of reducing the IR as required but it would not reproduce the initial cooling which I have just described i.e. it appears not to model the physics correctly in this respect.

    Answer: The SB model uses the surface temperature not the effective temperature of the IR as seen from outer space. The emissivity is defined as the ratio of the IR emerging into outer space to the IR which would be emitted by a black body at ground temperature. According to this definition the emissivity ought to be reduced when modeling the greenhouse perturbation as I suggested last time.

    (I hope that I can avoid the truncation of the message which occurred last time)

  • Neal J. King // November 30, 2006 at 9:37 pm

    Hi Geoff,

    I find two points in your analysis where I see a problem:

    1) We need to be clear on which temperature we’re talking about: the temperature at the earth’s surface or the temperature at the top of the troposphere. It seems clear to me that Monckton is talking about the temperature at the surface: these are the temperatures he cites in his tables, over the years. And these temperatures increase over time. Therefore, he cannot be talking about the temperature at the top of the troposphere, which has indeed decreased due to the rising up of the (Optical Depth = 1) level of the spherical shell of CO-2. So I agree that the temperature at (OD = 1) has decreased, and it is even related to an effective temperature of the CO-2 sphere; but the temperature at the earth’s surface has increased, otherwise we would not be talking about global warming!

    2) In your analysis, you are changing the effective emissivity when you go from (k) to (k - delta(k)). This is bad in two ways:
    a) It kills the model. The model assumes a constant emissivity => constant k.
    b) It contradicts what Monckton says a couple of times. For example, at the bottom of Appendix A of his backgrounder, in the description of “Earth-troposphe emissivity”: “the Stefan-Boltzmann equation gives epsilon ~ 0.6135 in 1900, pegged to allow like-for-like comparison of forcings over time.”

    So, my conclusion is that you’re doing more to save Monckton’s model than he did. If you’re going to change constants, they’re no longer constant; and in that case, you no longer have any version of the S-B law, you have a (very simplistic) global climate model.

    I invite you to go through your analysis again, keeping a clear distinction between the two temperatures, and then try to make sense of Monckton’s presentation. I don’t think you will be able to make it make sense!

    (On a separate point: I’m looking at the 25-line calculation of the different cases in Appendix B. I don’t understand them as well as I would like yet, but I suspect some inconsistencies there. Why am I looking at this?
    a) I’ve proven to my own satisfaction that there is absolutely no reason to sully the names of Stefan and Boltzman with this model. But it is always possible to find one value of epsilon that makes the equation true at one temperature.
    b) I’ve proven to my own satisfaction that even the relationship between forcing and temperature change is wrong (because of the simplistic nature of the model), even with the wrong sign. Nonetheless, passing over that point, he claims to get the “correct” value of the lambda, in the sense of the measured delta-T divided by the correct delta-flux. The delta-T is a matter of historical fact; while I don’t believe his model, I’m curious to see if his value for delta-flux is an honest measure, or if it is pre-cooked: set up to give the right answer. I’m still not sure yet.)

  • Geoff Wexler // November 30, 2006 at 10:17 pm

    Geoff Wexler

    Comments on Monckton’s revised estimate of global warming
    (via his Stephan Boltzmann approximation).

    [Thanks to Neil for item (b) below; but I have omitted all reference to Neil's latest discussion]
    Revised Nov. 30th.

    Summary. Monckton uses a well known but very simple model of the climate in order to claim that the projected warmings in the IPCC TAR are far too high. Gavin Schmidt in his article Cuckoo Science has provided at least three criticisms namely that Monckton’s model ignored feedbacks , time delays and incorrectly allows for changes in solar energy.
    The approach I adopt here is to highlight those places in Monckton’s document where such errors arise and to explain other parts of his reasoning.

    Introduction. Highly simplified models are very common in science. They can be used to produce rough answers without having to bother with all the details. They must however be used correctly. Of course we might not expect too much accuracy. But that is not always the case. There are well known occasions when a simple model can produce better answers than a refined one.

    A more serious point is that the model should not be used to make claims about effects which have been deliberately disregarded for the sake of simplicity. This remains true even if the results appear to agree with the observations. This is especially true if those observations require theoretical interpretation before they can be used. To put it another way simplifications can sometimes lead to theories which are hard to disprove because they predict too little. If this is the case they should be treated with skepticism. They may have loose joints which can be manipulated. Thus if the model omits all reference to time, there is nothing to stop you from guessing that the climate responds instantly (or very slowly which ever is more likely to produce agreement with the observations). A better model produces detailed time dependence and is thus easier to falsify (check).

    In the lead article Monckton in the Sunday Telegraph shows his familiarity with Boltzmann by writing :

    ”You don’t need computer models to “find” lambda. Its value is given by a century-old law, derived experimentally by a Slovenian professor and proved by his Austrian student (who later committed suicide when his scientific compatriots refused to believe in atoms). The Stefan-Boltzmann law, not mentioned once in the UN’s 2001 report, is as central to the thermodynamics of climate …..”

    Then in the pdf draft paper he takes up this point again :

    ”THE THIRD ASSESSMENT REPORT of the Intergovernmental Panel on Climate Change (UN 2001). does not refer to the Stefan-Boltzmann equation. Yet this equation is central to answering the key question in the discussion of climate change: how great will be the temperature response to radiant energy forcings such as elevated greenhouse-gas concentrations? ”

    Comment.
    To put this into context, the approximation based on the Stefan Boltzmann equation for grey body radiation is not disregarded by the standard theory but has been presented in elementary introductions to climate theory (see e.g. Hartmann p.231 for the black body version and lecture notes by David Archer for the grey body version). The normal approach moves on from this type of simple model, by incorporating much more climate physics. Monckton’s strong claim which is made in the first sentence of the first quotation amounts to the assertion that it is possible to deduce the so-called climate sensitivity (lambda) by attending the beginning of a course in climate physics and dropping the rest. In principle this is plausible, but Monckton’s claim is much stronger i.e he argues that using a simple model like this one is capable of disproving the results obtained by the whole climatological community.

    ”total energy flux E integrated over all frequencies or wavelengths as a function of emissivity and temperature:
    E=(epsilon) (sigma) T^4
    where E is radiant energy in watts per square metre per second ;
    epsilon is emissivity, equivalent by Kirchhoff’s Law to absorptivity;”

    Comment: This (SB) equation is a law if applied to a grey body such as ice or snow. But when applied to the whole Earth it is a simple model because it deliberately excludes much of the physics.

    The definition of emissivity of a grey body or gas (Hartmann p. 24) is the ratio of the actual emission of the body or gas to the blackbody emission which it would emit at the same temperature. For grey bodies the result is independent of temperature (over a reasonable range) but in the case of a gas the emissivity defined in this way varies with temperature which indicates that the SB equation (i.e energy proportional to fourth power of absolute temperature) is only approximate. The simplifications in the SB model include:
    (a) The variations of temperature e.g. with height are ignored.
    (b) Even if the variation of temperature were to be ignored, the atmosphere does not radiate like a grey body. It emits strongly at special wavelengths characteristic of the greenhouse gases. This is analagous to the radiation emitted by a coloured body rather than a grey one except that the radiation is in the invisible (infra-red).
    (c) There is no account of feedback.

    In the first line of p.25 of the pdf file Monckton refers to the emissivity of the Earth/troposphere system. Although the ground is nearly black at infra-red wavelengths the atmosphere is partially transparent and it is the greenhouse gases which radiate much of the infra-red radiation which escapes to outer space. This discrepancy is consistent with his choice 0.6135 (discussed at the end of his Appendix A) which is considerably lower than the typical values for the ground (at infra-red wavelengths). Monckton explains his choice at the end of Appendix A where he writes

    ‘Earth-troposphere emissivity: With E = 235.52wm-2 (cf. 236wm-2 in Houghton, 2002) at 13.7C, the Stefan-Boltzmann equation gives epsilon ~0.6135 in 1900, pegged to allow like-for-like comparison of forcings over time.’

    In 1900 the SB equation is fitted to the observation and thus provides no information. The crucial assumption is contained in Monckton’s last phrase in which he pegs epsilon over the whole of the twentieth century.
    This is what makes the equation useful i.e. predictive (for subsequent years) but also what makes it only a rough model.

    The physics tells us that a large component of the radiation is from greenhouse gases which are considerably cooler than the ground. In this model however it is required that the temperature to be inserted is the ground temperature. It is not easy to correct for this and other simplifications by an adjustment to the emissivity which does not depend on temperature. There are of course other simplifying features i.e that this version of the equation involves averaging over different latitudes and types of ground. The most serious consequence of assuming a constant emissivity is that it neglects feedback effects which I shall consider later.

    ”‘The UN (2001) briefly discusses the value of lambda= deltaT / delta E, the equilibrium response of global mean surface air temperature to a change in net tropopausal irradiance.”

    Comment.
    Here Monckton introduces an important quantity normally called the climate sensitivity. If for example the Earth’s loss of heat is reduced by greenhouse gases , it will be equivalent to the Earth receiving a few watts per square meter of additional power ; this effect is called a forcing. This raises the problem of estimating the sensitivity of the climate to this additional power. To be precise ‘by how much does 1 watt/m^2 of forcing raise the surface temperature?’ The value of this quantity (called lambda) will affect the future warming of the planet. It is not of course true that the TAR is brief about this topic. What Monckton means is that it does not use the simplified SB model.

    Just what is Monckton’s model for lambda? So far the SB law has been extended (by introducing a fictitious and approximate epsilon) to provide an estimate for the total power of the infra-red radiation emitted to outer space. The next stage is to assume that the Earth and its atmosphere start in energy balance i.e the absorbed solar radiation is exactly balanced by the infra-red energy estimated by the SB equation. A forcing of 1 W/m^2 means that the outgoing infra-red is reduced by 1/Wm^2 . The incoming energy will no longer balance the outgoing and the Earth will warm up. As a result the outgoing infra-red will increase until a new balance is achieved. If we assume that the absorbed solar radiation is unchanged during the warming then lambda can be determined from the SB equation model on its own.
    I shall call this determination the ‘SB-lambda model’ because it contains the extra assumption just mentioned. You can also read about it in introductory accounts to climate physics (see e.g. David Archer’s notes). There is a serious limitation with the SB- lambda model, which is that it provides no guidance about the time it takes to reach a new energy balance.
    I shall return to this point later.

    ”The quoted passage is followed by a lengthy consideration of the “remarkable” invariance of lambda which the general-circulation climate models have treated as though it were an output.”

    Comment.
    Lambda can be estimated in two ways, by careful appeal to observations OR from theory. Monckton uses both approaches just like the climatologists ; it is just that he uses a different highly simplified theory and a different way of approaching the observations. We shall have to look separately at both of these.

    ”if the emissivity of the Earth/troposphere system is held constant, as it must be to permit like-for-like comparison of radiative forcings over time.”

    This is one of the simplifications which define the SB-lambda model. I have already mentioned some of the effects which it ignores. I shall now consider feedback. Since one of the roles of the parameter called the ‘emissivity of the Earth/troposphere’ has been to allow for the effects of the troposphere , a constant emissivity builds in an assumption of a constant troposphere. But a small amount of warming increases the amount of water vapour and alters the cloud formation thus changing the troposphere. Any effects caused by a temperature dependent atmosphere are called feedbacks and will affect the actual warming. Whatever else it does , the SB-lambda model with constant emissivity is not intended to allow for this kind of feedback. There is also the question of feedback caused by reduced albedo; more on this later.

    before the table on p.25 of the pdf document he states:

    ”The reason for the near-invariance of lambda is that over the narrow range of astronomically-low temperatures relevant to climate, the small Stefan-Boltzmann constant offsets the biquadratic T4, rendering the temperature response to radiative perturbation over that range near-linear.”

    The final phrase referring to linearity is correct, although the explanation based on the value of the Stefan Boltzmann constant is unnecessary or faulty. The expansion parameter is (delta T)/T which is approximately the same as (delta E)/4E which does not contain the Boltzmann constant. Anyway the value of this parameter is about (0.6/287) over the last century which is much less than 1. One effect of this linearity is that it hardly matters whether the author uses the differential value of the sensitivity defined in e.g. Hartmann or the more accurate incremental version based on finite temperature and energy differences. Apart from the factor of 4 in the term
    (delta E/4E) mentioned above, the linear approximation means that the Stefan Boltzmann fourth power law is not really being used, but this point
    has no effect on the subsequent discussion. Because of the argument above, I don’t think that you have to assume the BS approximation is valid in order to conclude that lambda is constant. Conversely a constant lambda cannot be used to justify the BS approximation.

    The first table on page 25 of the pdf document asserts that the “actual” value of lambda is 0.303 and that the corresponding value for a totally absorbant (black) Earth would be 0.223 (“base value”). There is an error or a misunderstanding here because the estimates are not inversely proportional to the emissivity as they should be.
    0.303/0.223 =1.359
    whereas
    1/0.6135 = 1.63
    As far as I can see this discrepancy is harmless because it is the so-called base value which looks incorrect and it is never used. Why is the base value entered into the table?

    So far there is nothing new in Monckton’s work. It is when he comes to present the results obtained from more elaborate models that he starts to spin. Why does Monckton call his estimate of lambda the actual value? It appears to pre-suppose that the SB-lambda model gives an exact account of the warming. The third and fourth columns of the table are labeled so as to suggest to the reader that everyone else’s estimates of the twentieth century warming disagree with the observations whereas his own in the second row is spot on ! The third column specifies the discrepancy in warming and the last column displays the ratio of calculated to the so-called observed warming.
    These results come from the equation:

    (delta T) = lambda (forcing)

    for a single value of the forcing = 1.98 watts/metre ^2, and on his estimate 0.6C for the warming between 1900 and 1998. As we shall see there are problems with both of these values.

    Just below the table Monckton continues as follows:

    ”The UN’s temperature projections to 2100 assume that a Clausius-Clapeyron exponential rise in water vapour pressure with temperature causes a climate feedback described as a “near-doubling” of base forcings.”

    Comment.
    It is normally assumed that the warming is proportional to the forcing. It therefore makes no difference if you assume that the forcing is doubled or the warming is doubled. According to the standard definition the forcing must not include feedback so it is unconventional to suggest that it is the forcing which is doubled. This point sounds trivial here, but will lead to confusion later on.

    Incidentally the effect itself is due to the fact that the extra water vapour is a strong greenhouse gas. This mechanism has been central to global warming theory for about 110 years but as we shall see it will be ignored by Monckton.

    He continues:
    ”In addition, reductions in albedo are thought to amplify base forcings by 20% (Houghton, 2002). Houghton (2006) explains that values
    lambda> 0.303 allow for such climate feedbacks.”

    Comment:
    This is another effect which has been included in global warming theory from the start and will be ignored by Monckton. An initial warming converts ice into water which reflects only about one fifth of the incoming solar energy compared to ice; this creates more warming.

    He continues:
    ” the warming of the oceans – in effect a large heat-sink – may explain why observed air temperature rose very much less in the past century than the elevated values for lambda would mandate.”

    Comment. This delay is normally supposed to be responsible for the so called committed warming which will occur whatever we do. It will also be ignored by Monckton.

    The paragraph ends with a strong claim i.e

    ”calculation using a simple model shows that neither the proposed elevations of lambda nor the explanation in Hansen (2006) is needed, for there is no discrepancy between observation and the calculated value lambda ~ 0.303.”

    Comment.
    This is where Monckton announces his policy of ignoring feedbacks and time delays. It marks the real beginning of his work. He has put numbers into the SB-lambda model and claims that they have been confirmed by the observed temperatures in 1900 and 1998. He argues that this justifies a low estimate for lambda and that no further adjustments for time delays (Hansen) or feedbacks should be applied. We need to examine these conclusions quite closely.

    Let us start by accepting the assertion that the SB-lambda model has been checked by the observed warming. He concludes from this that the model need not be adjusted to include time delays (Hansen) or feedbacks. This sounds like the dodgy logic mentioned in the introduction. To reach his conclusion reliably he would need to put these effects into a theory and estimate if they really are small. But I find his text is ambiguous at this stage; the so-called agreement between model and observations could occur either if the effects mentioned are negligible or if they have been included in the model. One way of deciding is to try to determine from the text what the author believes, I’ll come back to that later; the other way is to analyse the physics of the model and the observations.

    A brief glance at the SB-lambda model reveals that the time variable is absent. What the model tells us is whether there is an energy balance. If the absorbed solar power is greater than the emitted infra-red, the model tells us that the Earth will warm but it does not tell us fast. It might reach a new balance instantaneously or after a thousand years. The model contains no dynamical information. At the opposite extreme, consider a recent type of model called a coupled atmosphere-ocean general circulation model (AOGCM) used by the Hadley Centre, this explicitly considers the oceans and sets out to account for the observed warming over time. Monckton however sets his goals much lower, he only aims to account for the total warming after an interval of 98 years without bothering with intermediate times or with the duration of the interval. He must ignore the duration 98 years because the model is silent about this matter. But he needs to check his model with observations and so he has to have some idea of the delay.
    His choice is effectively zero i.e the response is assumed to be instantaneous. This is what he means by his rejection of the need to bother with Hansen (2006) who, with other scientists, refer to the effect of the oceans in slowing down the response of the climate to an external forcing. But what is the justification for Monckton’s rejection? His argument is that the SB-lambda model appears to agree with the observed warming. But the observed temperatures do not provide an estimate of the warming produced between 1900 and 1998, not unless you neglect the committed (delayed) warming by assuming a very fast climate response. The argument is circular. The best that we can say is that the SB-lambda model has not been disproved by the observations. But that is not a virtue. As we are seeing this model is hard to check.

    In addition as I mentioned earlier the fact that the emissivity is held fixed, independent of temperature, means that the model omits water vapour feedback. Finally the idea behind the BS-lambda model is to estimate the rise in temperature required to increase the thermal radiation by just enough to compensate for the additional power (forcing) in the system while the absorbed solar radiation is assumed to remain unchanged. There has been no allowance for the fact that melting ice alters the fraction of incoming sunshine which is absorbed.

    To summarise so far. Whatever Monckton claims, the physics of the SB-lambda model does not include time delays or feedbacks. The existence of time delays would mean that the observed warming in 1998 cannot be used to test the model. It is therefore impossible to use such a model to make claims about the magnitude of such effects. As we shall see later, Monckton appears to believe that there are no significant time delays, but that there are substantial feedbacks and that these have already been included in the SB-lambda model. There is no evidence for the first claim and the second one is wrong.

    Forcing estimates. He now leaves the SB model for a while and includes some review material on forcing.

    The 1st. sentence of the next para. asserts:

    ”The UN (2001) attributes its estimated 1900-1998 increase of 0.6C in temperature chiefly to the forcing effect of rising concentrations of greenhouse gases in the atmosphere: all its other forcings are far smaller, less well understood, and broadly self-cancelling.”

    The part of this conclusion which is true is the description “less well understood” . But the forcings arising from the aerosols are not necessarily “far smaller” than those from the greenhouse gases and the effects of the large uncertainty ranges do not necessarily cancel. (see coloured diagram which follows figure).

    This part of Monckton’s account is written as if he is using the IPCC TAR data. But he has omitted to mention that he is ignoring their estimate for direct aerosol cooling i.e. between 0 and –2 watts/m^2 and a possible further term of similar magnitude accounting for indirect effects of aerosols on rain and clouds. By disregarding aerosols Monckton may be boosting the positive forcing effect that existed in the twentieth century and thus reducing his observational estimate for lambda = (warming / forcing). What Monckton is doing here is starting to select the part of the uncertainty range which favours his desired outcome i.e that his SB model should agree with the warming of the twentieth century. I shall come back to this point later.

    Just below the coloured diagram on p.26 Monckton asserts:

    ”Calculation shows that the UN’s table of historic forcings must include all climate feedbacks.”

    This remark must be set against the following formal defnition

    Item 6.1.1. of the TAR provides the following definition:
    “The radiative forcing of the surface-troposphere system due to the perturbation in or the introduction of an agent (say, a change in greenhouse gas concentrations) is the change in net (down minus up) irradiance (solar plus long-wave; in Wm-2) at the tropopause AFTER allowing for stratospheric temperatures to readjust to radiative equilibrium, but with surface and tropo-spheric temperatures and state held fixed at the unperturbed values”. In the context of climate change, the term forcing is restricted to changes in the radiation balance of the surface-troposphere system imposed by external factors, with no changes in stratospheric dynamics, without any surface and tropospheric feedbacks in operation (i.e., no secondary effects induced because of changes in tropospheric motions or its thermodynamic state), and with no dynamically-induced changes in the amount and distribution of atmospheric water (vapour, liquid, and solid forms). ”

    For the present purpose we can ignore the remarks about the stratosphere.
    Notice how the TAR definition specifically excludes surface and tropospheric feedbacks which accounts for most (perhaps all) of the feedbacks which I have mentioned above. Monckton’s assertion in the previous quote from him makes it clear that he thinks many
    many independent theorists have violated their own definition of forcing.
    I have never found any evidence for this claim. It will be central to what he has to say about solar forcing.

    The table of forcings on p.26 refers to the period 1750 –1998. Monckton wants to consider the shorter period 1900 to 1998 so he has to reduce the estimates in the table. He has used data for the concentrations of CO2 in the year 1900 as well as 1750 and 1998. No mention is made of the graphs in the TAR which show how some of the other greenhouse gases have varied over the last 1000 years. (I have not followed this up). Instead he attempts to obtain an adjustment from the one gas CO2. His first step is to convert concentrations into forcings and this involves the logarithmic equations below the coloured chart on p.26 of the pdf file. Having worked out a reduction factor for the CO2 forcing he then applies the same factor to the rest of the table . Instead of scaling all the forcings by a single factor he could have scaled the concentrations. This would have given a different answer because of the logarithmic dependence of forcing on concentration. I have not tried this step. Both methods are arbitrary and involve guesswork. I have not performed a complete numerical check of this paragraph. His final estimate for the forcing is 1.99 W/m^2.

    The next paragraph asserts:

    ”This centennial forcing delta E from all sources can also be directly computed from the Stefan-Boltzmann equation.with emissivity 0.6134 at 14.3C, tropospheric radiant-energy flux is 237.50wm-2. Deducting 235.52wm-2 for the energy flux in 1900 gives the centennial forcing of 1.98wm-2, a near-identical result.”

    Comment:
    This comparison is a basically a repeat of that shown in line 2 of the previous table on p.25 of the pdf file. Both use his SB-lambda model and both claim remarkable agreement. It just depends on whether you start with the forcing as given and work out the warming or the other way round.

    So this is the summary so far, Monckton appears to have accounted for the warming between 1900 and 1998 by means of a well known simple model which excludes feedback and time delays. Applying the same method to the future yields much lower projected warmings.

    A cool cherry or a hot one?
    I mentioned above that Monckton has made an arbitrary choice of forcing when he took the forcing to be 1.98wm-2. This depends on his choice of zero for the cooling effect of the aerosols. Lets take a different value for the effect of the aerosols , instead of Monkton’s choice zero I shall take -1.75 wm-2 . This appears arbitrary but it is within the range of uncertainty for the direct cooling and takes no account of the indirect cooling which is also negative. The new value for the forcing is 0.23 wm-2 which is now in poor agreement with Monckton’s theoretical SB value of 1.99. You may object that my choice of aerosol cooling is unjustified . That would be right, but so is Monckton’s; both are examples of cherry picking. The new value of lambda would be 0.6/0.23 = 2.61 degs m^2 w^-1 instead of Monckton’s “actual value” of 0.303 degs m^2 w^-1. But it does not stop there. If I had chosen a slightly higher estimate for the aerosols I could have obtained infinite or even negative values for lambda. There is something wrong. That point has been discussed by Gavin Schmidt in Realclimate in an article called ‘Climate sensitivity and aerosol forcings’ of 6th.July 2005 where he argued that the twentieth century is not a good century for estimating the climate sensitivity, partly because you might obtain excessively HIGH values for the climate sensitivity. Monckton has done the opposite obtaining a very LOW value. This is a second reason why Monckton’s claim to have checked the SB-lambda model is flawed.

    pa.3 p.27 of pdf.
    To update the calculation to 2006, temperature has risen by ~ 0.14C since 1998, equivalent to 0.46wm-2, and CO2 concentration rose from 365 to 380ppmv. Forcings rose by 5.35g ln(380 / 365) = 0.35wm-2, or 0.11wm-2 below observation. However, between the 1998 solar minimum and the 2004 maximum, TSI rose by ~1.1wm-2

    Comment.
    He now extends the observational interval by another 8 years up to 2006
    and has started to discuss the variation of solar power.
    It looks as if his first estimate for the additional forcing 0.46 wm-2 is based on his BS-lambda model. This does not justify the description ‘observation’.
    Solar forcing appears to be taken to be as the increase of absorbed solar energy compared to 1998. For some reason this does not include solar forcing before 1998. I have not followed this up.

    In next para. (p.25 pa.5)he states:
    projection. If the lower centennial temperature increase estimated by NCDC
    were applied, it would be legitimate to infer that the all-greenhouse-gases forcing equation, E = gz ln(C/C0), itself produces overstated forcings, and that z should be further cut from the UN’s current value of 5.35 to 4.71.

    and then
    Centennial
    increases in solar surface-energy flux above UN 2001’s central estimate of 0.30wm-2 since 1750 would demand yet further reductions in z.

    Comment: This is a continuation of Monckton’s unconventional way of scaling forcing rather than say the climate sensitivity.

    In line 3 up on p. 27 of the pdf file Monckton writes
    Solar forcing from 1750 to the present may have been only 0.3C,

    Comment. Perhaps this should have been 0.3 W/m^2.

    On p. 28 Monckton changes course without warning the reader (as Gavin has pointed out in his comments). At the end of the first para. we read

    When temperature-induced climate feedbacks arise, they do so as much from solar as from greenhouse-gas forcings. Therefore, multiplying 1.17 by 2.67 (Houghton, 2006) for climate feedbacks would yield 3.04wm-2 after climate feedbacks – more than one and a half times the observed warming of 1.98wm-2.

    Comment. This is where Monckton’s unconventional treatment of forcing begins to look very confusing. Why are the solar and non solar contributions treated differently? Only the first value 1.17 is boosted by the feedback factor 2.67. To make sense of this I have to assume that Monckton is asserting that both the IPCC tables of forcing and the SB-lambda estimates include a boost of a factor of 2.67 to allow for feedbacks. I cannot accept either of these possibilites for reasons already given. It looks suspiciously like a blunder.

    ‘Deducting the UN’s 0.30 wm-2 allowance for solar forcing (already
    cancelled by minor, negative forcings in the UN’s table) gives 1.2 wm-2 additional solar forcing, leaving only 0.79 wm-2 for centennial greenhouse-gas forcing and requiring the coefficient z to be cut by more than half.’

    Comment.This continues in the same vein. I did not manage to find the Solanki article in Nature 2005 when I last visited the library. Gavin was much more concise. Are these comments any use to anyone? I could abbreviate them but my motive was to help someone actually reading Monckton’s argument.

  • Neal J. King // November 30, 2006 at 10:59 pm

    Geoff,

    In no particular order, some comments:

    - You’re suggesting that Monckton gets the value of delta-flux of 1.98 by dividing the known temperature change by his “correct” value of lambda. He compares that to the IPCC’s value of 1.99. In my book, 1.98 is certainly very close to 1.99 - and this seems like the only achievement of the theory. Are you suggesting that the 1.99 is cherry-picked? If so, then my perspective is: 3 strikes and he’s out!

    - I think I see that you are also noticing the fact that his model has room only for one temperature, whereas the earth + atmosphere system needs at least 2 separate ones.

    - To be fair, Monckton claims (anyway) that the adding together of the other GHG effects with CO-2 was taken from the IPCC report. Whether or not this is true, it seems a perfectly useless thing to do: there’s no reason to believe that these different gases GHG-effects would be similar, so they won’t scale alike (and that’s why the factor slides around). Better just to add them up. (Reason: I think there’s enough CO-2 that the IR in those bands is near-saturated, whereas that is not true of methane. So they won’t behave the same way.)

    - The difference between forcings and heating is significant. Conceptually, I like the idea of adding the forcings, although I guess it’s true that if the heating for each forcing is not proportional, it still gets messy. What the BS/SB model does is to make the earth+atmosphere system try to act like a rock, and then you really do get only one kind of forcing; and a reduction in outwards flux really must be related to cooling, it cannot be related to warming. It don’t got no moving parts to accommodate both flux-reduction and warming.

  • Geoff Wexler // December 1, 2006 at 12:14 am

    Neil :
    Many thanks for your reply. I have not finished reading or thinking about it but have to go to a one day meeting on cutting down on carbon emissions. Meanwhile here’s my reaction:

    1. I have kept the two discussions apart for the time being.
    2. As I have said Monckton is not responsible for this model.
    3. In my view you could regard it as a formula for estimating lambda (hence my name SB-lambda). This is basically the derivative of the SB equation. So you can START the whole argument with the equation for lambda where the temperature is the surface temperature. That’s all that is required from a crude model. You can’t attack it for being approximate; that is not what you do with simple models. Yes you must define a unique emissivity to put into it but remember that its sole purpose is to estimate the warming between 1900 and 1998. You do not have to follow out my suggestion of altering the emissivity but it helps in providing a rough and ready justification for the starting equation. I don’t think this is inconsistent because there is only one equation to use.

    4. You can attack it for the way Monckton tries to use it which is to push it to absurd lengths as suggested by my long winded discussion. In addition his claim to have checked it is invalid.

    Analogy: Einstein’s theory of specific heat. Very approximate; it works but omits lots of things. Einstein would not have dreamt of pushing it very hard.

  • Neal J. King // December 1, 2006 at 8:34 am

    Geoff,

    If I were just reading Monckton’s paper to look for insight, I would have tossed it a long time ago: he’s made too many fundamental errors for me to expect much. However, since we want to rebut his arguments, it helps to know what they mean in detail.

    Because of the looseness involved, it’s my impression that he hasn’t demonstrated anything interesting by defining an effective value of emissivity to make the SB equation true at one temperature. The only point that I’m curious about is the legitimacy of his claim to have arrived at the “correct” value of lambda without having done all of the physics behind it. I am suspicious that this is just a matter of cherry-picking the cases to get a number that matches his model.

    (I call it “his” model because he seems to refer to it in his article and in this backgrounder as the “M-model”. Any guess as to what the “M” stands for? )

    In any event, ultimately we have to prepare a description of the situation. I think it fair to describe the conceptual failing behind the model, and to briefly describe the numerical failure as well. We can’t go into a blow-by-blow of the whole paper (although it’s useful to me), because it represents about 15% of the entire backgrounder, and we’ll lose our readers.

    Finally, wrt comparison with Einstein’s theory of specific heat: an important difference between the two is that Einstein’s theory had a well-defined physical meaning, even though it left a lot of important things out. You could still learn something from the theory. My view is that the M-model (or whatever you want to call it) does not have a well-defined physical meaning: when he applies the CO-2 forcing onto his formula, he is ignoring the fact that this forcing is actually a reduction of outgoing radiation flux, which can only be produced within the physical conception of his model by a temperature reduction. If you ignore that, then you are saying that there is no physical picture behind his model at all, it is just games with an equation in which a “constant” is allowed to change.

    I would be most interested in what you could figure out about the 1.98 vs. 1.99 question: Regardless of the basic foundations, is this a legitimate agreement, or is it a manufactured coincidence?

  • Geoff Wexler // December 2, 2006 at 5:27 pm

    Neil,
    (Thanks for your replies and the link you put on the Realclimate web site I’m sorry I won’t have time to reply to all your points until later but meanwhile here are some points which require no further work on my part. I’m sorry that I don’t agree with all of your consclusions.

    Many of your comments are correct and interesting but I don’t think that they are all focussed on Monckton’s work. I may be repeating myself, but Monckton starts with a well known model and you tend criticise that rather than his use of the model. The contrarian attacks on climate physics are usually based on the same kind of reasoning i.e that the models make approximations. I have spelt out what I think is wrong with Monckton and it is based on logic rather than approximations. Your ‘negative lambda discussion’ appears to be logical but even this may not be focussed on Monckton but on the point before he starts work. In any case I still have doubts about the consistency of that part of your argument. (Please see later).

    Now for a few more detailed replies. (To your Nov 30th.9.37PM comment)

    ‘I find two points in your analysis where I see a problem:’

    I suppose you mean my reply to your previous comment about –ve lambda. I have written another version of that which I shall include later.

    ‘1) We need to be clear on which temperature we’re talking about: the temperature at the earth’s surface or the temperature at the top of the troposphere. It seems clear to me that Monckton is talking about the temperature at the surface:’

    Good. I assumed that in the first version of my comments. I also pointed out that one of the main assumptions of the model was to take a single temperature.

    2) In your analysis, you are changing the effective emissivity when you go from (k) to (k - delta(k)). This is bad in two ways:
    a) It kills the model. The model assumes a constant emissivity => constant k.

    The standard way of dealing with this problem does not use this asumption. It was only a way of removing the inconsistency you encountered when you developed your –ve lambda argument. (If you don’t follow yet , I hope it will be a bit clear later)

    ‘So, my conclusion is that you’re doing more to save Monckton’s model than he did’

    I am not referring to Monckton’s model but just to the simple energy balance model introduced by climatologists.

    ‘If you’re going to change constants, they’re no longer constant; and in that case, you no longer have any version of the S-B law, you have a (very simplistic) global climate model.’

    Comment.
    I thought I had been careful to use the term SB-model rather than SB law.
    Yes very simplistic. It is probably possible to avoid changing constants
    (please see later)
    The change was to try to fill a void in the reaaoning in your argument about negative lambda. Even that change does make sense however because it describes the perturbation as internal rather than external and it specifies it.

    ‘I invite you to go through your analysis again, keeping a clear distinction between the two temperatures, and then try to make sense of Monckton’s presentation. I don’t think you will be able to make it make sense!’

    My long account only refers to the ground temperature at two times 1900 ,1998. Have I missed something? There are parts of it which don’t make sense such as his interpretations and also his adjustments to deal with solar forcing. I think Gavin got it right. I have just spelt it out more.

    a) ‘I’ve proven to my own satisfaction that there is absolutely no reason to sully the names of Stefan and Boltzman with this model.’

    Comment. I would put it another way. I wouldn’t give Monckton the credit.
    It DOES start with SB. He has applied the model in a particular way i.e by choosing a value for epsilon (even that is not unusual). He has compared the results with a cherry picked forcing (THAT is unusual) and has made some quite false inferences.

    ‘b) I’ve proven to my own satisfaction that even the relationship between forcing and temperature change is wrong (because of the simplistic nature of the model), even with the wrong sign.’

    Comment.
    This is the one remark of yours which I think is probably wrong. Lambda involves rigorous adherence to energy balance and I don’t think that you have done that. Please see later argument which still needs to be checked.

    ‘I’m curious to see if his value for delta-flux is an honest measure, or if it is pre-cooked: set up to give the right answer. I’m still not sure yet.’

    Gavin has already answered this point and I thought that I had clarified it. I think the answer is (a) Neglect of aerosols (b) Neglect of time delays. These are sufficient. But after having read his treatment of solar forcing I would not be surprised at anything. So good luck with your quest to find still more !

    Reply to your Dec 1st comment.
    ‘Because of the looseness involved, it’s my impression that he hasn’t demonstrated anything interesting by defining an effective value of emissivity to make the SB equation true at one temperature.’

    Comment.
    Of course but this is not a criticism of him or of the model. This type of fit is quite normal.

    ‘The only point that I’m curious about is the legitimacy of his claim to have arrived at the “correct” value of lambda without having done all of the physics behind it. I am suspicious that this is just a matter of cherry-picking the cases to get a number that matches his model.’

    Please see my long comment. Yes it does rely on cherry picking. I obtained another value(s).

    ‘I think it fair to describe the conceptual failing behind the model,’

    I think Gavin has done that if a bit briefly. So what is needed is perhaps a slightly longer explanation of Gavins’s remarks.

    General comments about your remarks.

    ANY radiation vs temperature equation can be used to provide a model with conceptual significance because it illustrates what some people describe as a negative radiative feedback. Starting with a system in energy balance, it will respond to a forcing by modifying its radiation in the direction towards restoring energy balance. It follows that such an equation may be a used to estimate this feedback by ignoring all other feedbacks and time delays. This is what happens if you use the radiation equation on its own but constrain it so as to maintain energy balance. Ideally you should use an EXACT radiation formula for this step in which case you would get a zero order approximation for lambda which would still be very inaccurate because it would exclude every other kind of feedback effect. It would also be difficult to compare with observations because of the time delay. If you only use an approximate radiation equation you won’t even obtain an exact estimate for the negative radiation feedback, but at least you will have made a start.

    The next point concerns your point about the so called cooling which you attribute to the model. First this argument departs from the constraints of the energy balance model. In spite of that, there is an apparent contradiction which needs to be understood. I think it arises because you are comparing two different prodecures (a) apply an external forcing to an unperturbed system. (b) Do not have an external forcing. You must then be careful. You can (b1) either have an internal forcing to make up for the lack of the external forcing (b2) have no internal forcing. (b2) is the original system which will not warm. (b1) which I described in earlier comment. But these procedures don’t aim at what we need which is described by the previous paragraph.

    (This is what I wrote earlier; it might still be worth reading?)

    You have to be very clear about the perturbation and the question you want to ask and why.

    Unperturbed problem.
    A system in energy balance driven by fixed input solar power.
    There is more than one way to describe the perturbation. Standard way is to suppose the system is forced by an increase of input power F. The system itself must not be adjusted for additional greenhouse gases. That would double count the forcing. (Keep epsilon fixed).
    Constraint. The system must remain in energy balance.
    Question. Calculate new temperature.

    Your (Neil’s) version.
    Unperturbed problem. As before.
    Question. By how much would the temperature have to be reduced in order to reduce the OLR (outward long wave radiation) by F.
    Comment. (a). You have relaxed the constraint of energy balance.
    (b). One motive for asking this question is to see if the model can explain the forcing. But it does not go back far enough to explain the forcing and the interpretation of the answer is unclear . Certainly not a value of lambda.
    (c) There is a possibility of an inconsistency if you fail to include the greenhouse gases properly as suggested later. The answer is either to avoid enquiring into the cause of the forcing or to model the greenhouse gas effect.

    Tracing the source of the forcing. This is a different goal from calculating the radiative feedback.
    Unperturbed problem.
    In energy balance with unperturbed value of greenhouse gases.
    Perturbation.
    Add greenhouse gas.
    Try to model this with a one temperature model. The only way I think I can do this is by increasing thermal resistance by reducing emissivity.
    Now ask for new temperature under constraint of energy balance

    I think the first version is simple and adequate. The main problem in applying this model is to choose a numerical value for the emissivity. Monckton relies on existing estimates (e.g from Houghton) . I am not sure where Houghton’s solar power comes from.

    Thermal equilibrium.
    You are more or less right.
    But you need that assumption all over the place e.g in deriving Kirchoff’s law for equating the spectral emissive power to the spectral absorptivity of a gas. But I don’t see the point of raising it. In fact the standard radiation physics sometimes makes a similar approximation. The use of temperature itself involves the assumption of thermal equilibrium which is usually but not always justified. It depends on the pressure. Is the temperature of excited CO2 molecules the same as the surrounding air? Why bother with all of that discussion? That would just get into a battle of the approximations.

    Incidentally you could develop an analogy with an electrical circuit. Instead of the Sun I would have a constant current generator. Current is analagous to power flow. The analogue of temperature is voltage. The analogue of Ohm’s law is the SB law. Notice that the voltage (temperature) is now determined and energy balance is automatic. Now force it by increasing the current . It is harder (as stated above) if you insist on trying to explain where the the forcing comes from. You have to introduce at least one more current generator. Perhaps it would be easier to suppose that an extra resistor is placed in series to model the rise in greenhouse gases.

    I’m sorry for some repetition in the above but have no time to rewrite it,

  • Neal J. King // December 3, 2006 at 9:45 am

    Geoff,

    Thanks for your comments.

    During a recent flight, I have had time to study Monckton’s calculations in more detail, and I find even more questions. In addition to the problems I had already mentioned above, I find the cherry-picking in his estimate of the forcing.

    Rather than respond immediately to all of your comments above, I will update the write-up in the anti-article in the wiki, and you can properly consider all factors brought into the discussion.

  • Geoff Wexler // December 3, 2006 at 11:20 am

    To Anyone:
    Where and waht is the anti-article in the wiki? please?

    (I appear to have ruled out Wikipedia)

  • Neal J. King // December 4, 2006 at 2:03 am

    Geoff,

    http://en.wikipedia.org/wiki/User:Dbuckner/climate

    Be sure to register with Wikipedia and sign-in: There has been an unidentified editor trying to sneak a few curve-balls in. If you read the introduction, I think you can figure it out.

  • Geoff Wexler // December 5, 2006 at 1:46 pm

    Neil.
    Many thanks.

  • Neal J. King // December 16, 2006 at 5:46 pm

    The Wiki site for the creation of the article in response to Monckton’s has been deleted. Two issues:

    - The good Viscount apparently protested to the management of Wikipedia, threatening legal action; and
    - There was a query raised as to the appropriateness of using Wikipedia to generate an article intended for a purpose other than use within Wikipedia.

    So the site is no more, but the material generated has been saved. My opinion is that we need to set up a better process, not involved with Wikipedia.

    In any case, I have not felt entirely comfortable with a totally open-ended editing process. We have had a few ringers trying to tilt the intent of the article away from a response to Monckton’s article.

    More later.

  • Ian Walsh // January 10, 2007 at 7:40 pm

    “An editorial was recently published in the English newspaper The Sunday Telegraph by Viscount Monckton of Brenchley, claiming that global warming is false. ”

    NO the above is not true and is a deliberate attempt to deceve…….Monckton makes is VERY clear warming is happening but says it may well be NATURAL WARMING….

    DONT YOU GET IT…

    [Response: Here's a direct quote from pg. 16 of Monckton's "accompanying document":

    "I conclude that today's temperatures are not exceptional, and that the mediaeval warm period was at least as warm as the present and probably up to 3C warmer."

    Despite your protestations to the contrary, Monckton does deny global warming in every way he can. Why else would he go to such lengths to cast doubt on the thermometer record, but a vain attempt to "prove" in every way he can think of that the modern warming is only a small fraction of what it really is. He is a denialist of the first magnitude.

    It's Lord Monckton who is guilty of "a deliberate attempt to deceive." Multiple counts.]

Leave a Comment