On July 15, 2008, the editors of an obscure newsletter that I've become rather fond of over the years, "Physics and Society", posted the first two entries in a "debate concerning [...] conclusions of the International Panel on Climate Change (IPCC)".
This wasn't the first time the editors had courted somewhat "contrarian" views and encouraged strong debate, though the primary topics of discussion in past years have been more concerned with nuclear weapons proliferation than with the environment. They even hosted a brief debate back in 2004 between myself and Steven Fetter on the practicality of Space Solar Power.
But this time the editors seem to have gone well out of their depth. The second article was by the notorious climate change denier "Christopher Monckton of Brenchley", a British Viscount who has little training in science. After the brouhaha that followed, Catherine Brahic at "New Scientist" interviewed one of the editors to find out "what they were thinking":
and this was naturally picked up by half the right-wing blogs, distortions and all.
Naturally, the American Physical Society objected to this mischaracterization of the newsletter, Monckton's article, and the size of the society. (APS happens to be my employer). Last fall APS adopted a strongly worded statement in support of the IPCC conclusions on anthropogenic global warming. APS publishes some of the most respected and highly cited peer reviewed journals in physics, but "Physics and Society" is not one of them; rather it is an independent publication of one of the APS subunits, the Forum on Physics and Society, and the editors do not claim to hold the articles they publish to the "peer review" standard of the formal journals. This was clarified on the website of the newsletter in the days following, which led to more press releases and attacks from Monckton and the SPPI and blogs of various stripes chimed in...
My involvement started earlier, however. As a long-standing member of the forum I'd been considering accepting a nomination for its executive committee, encouraged by the outgoing president, Larry Krauss, at the time not worrying about my other existing connection to APS. However, on July 15th, as a forum member, I received notice of the posted articles and, taking a look, quickly realized there was something very disturbing about Monckton's article. The level of outright deception that was evident in the lengthy text was enough to almost make me physically sick. I started taking notes for a rebuttal, and when I got wind of the fuss that had erupted a few days later, I redoubled my pace, sending in a draft response on July 23, and, out of habitual politeness, cc'ing Monckton.
One of the problems was that everybody's critique in that first week after the article was posted was limited to a handful of highlights of issues in Monckton's article, when in reality there was barely a paragraph in the piece that didn't contain some glaring bit of nonsense, and some had error piled upon error. In hopes of settling things a little more firmly, I decided to undertake a comprehensive critique of this supposedly "major peer-reviewed" "proof", and try to settle whatever nonsense people were still getting out of it once and for all.
It was a much bigger task than I'd bargained on when I started, with my original list of notes. Actually looking up many of the papers and other sources Monckton referenced (some of which he provided highly incomplete references for), I found misquote after misquote, and misinterpretations galore. Trying to make sense of his supposed "proof" I found logical inconsistencies, red herrings, and some nearly unbelievable leaps of incomprehension. In all it took me some six weeks of my spare time, from July 24 to September 6, to get through the entire article in detail, with a final list of 125 erroneous or nonsensical statements. Read the full list here.
It took me over 40 days to comprehensively critique his article. Yet in less than a day Monckton had produced a "rebuttal" to my first response which, aside from the stupid ad hominems, sounded reasonable, but he mostly just made up. Is this 40:1 effort ratio typical for the fight between science/truth and the forces of denial? And, Monckton had SPPI and those other sites to publicize his attack - his original press release spread and was distorted around the denialosphere to the point that the Drudge report claimed the American Physical Society had changed its position on global warming! No wonder so many scientists just give up in despair. I have to hand it to the folks at RealClimate and the other science-based blogs for putting in the effort, it must be incredibly frustrating.
Not out of my frustration, but to avoid further mis-perception on the association between APS and the Forum, in the middle of all this I withdrew my name from consideration for the Forum executive committee and suggested some other people in my place. Maybe I'll run next year. In the meantime, here's an excerpt from my full critique, errors #49-58, on the question of the value of radiative forcing and Monckton's discussion of the "tropical troposphere hotspot":
E49
The models draw upon results of laboratory experiments in passing sunlight through chambers in which atmospheric constituents are artificially varied; such experiments are, however, of limited value when translated into the real atmosphere, where radiative transfers and non-radiative transports (convection and evaporation up, advection along, subsidence and precipitation down), as well as altitudinal and latitudinal asymmetries, greatly complicate the picture.
Wrong, Red Herring, Confused: The start of this sentence appears to be a reference to spectroscopic measurements, though the phrasing significantly downplays the unparalleled experimental precision that has been routine in spectroscopy for many decades. The "models" in fact draw upon large databases of the spectroscopic parameters of atmospheric constituents to understand their effects on thermal radiation. One of the most frequently used is the HITRAN database from Harvard-Smithsonian, which contains details on close to 2 million spectral lines for 37 different molecules. A summarized version of this detailed spectroscopic data is used in the "moderate resolution" ModTran program, which calculates radiative forcings based on specific atmospheric profiles and can be run online here. Modern climate models may use either the HITRAN data (line by line calculation) or a simpler version as ModTran does.
In particular, there is no question that these spectroscopic details and calculations used by the models very accurately portray the behavior of thermal radiation as it passes through the atmosphere, and satellite and air-borne measurements confirm that they indeed apply to the "real atmosphere". Monckton's list of non-radiative terms - "convection", "evaporation", "advection", "subsidence", "precipitation" - is completely irrelevant to radiative transport and forcing: those factors certainly play a role in determining the temperature profiles of the atmosphere at different "latitudes" and "altitudes", but those temperature profiles are
input to the radiative calculation. Remember, the definition of forcing is the effect on the balance of energy flux at the tropopause when the term at issue (here CO2 concentration) changes, all else being held constant.
In short, the forcing from CO2 or other greenhouse gases is determined by taking the current profile of temperatures, pressures and constituent concentrations throughout the atmosphere, averaged over suitable climatic time periods, along with averages of cloud cover and similar factors, and applying these well-known detailed spectroscopic databases. The uncertainties (at about the 10% level) come from uncertainty in measuring those atmospheric temperature profiles and similar factors, not from the inapplicability of the spectroscopic data, nor from any influence of "convection" or other non-radiative factors.
E50
Using these laboratory values, the models attempt to produce latitude-versus-altitude plots to display the characteristic signature of each type of forcing.
Confused, Cherry Picking: there is a vast quantity of information that can be gleaned from the models, both statistical (means and distributions), and more detailed (regional changes in temperature, precipitation, etc.). Much of the IPCC AR4 WG1 content (chapters 8 through 11) is devoted to what models can tell us about the climate response to changes in various forcings. The earlier chapters of the WG1 report discuss what observations tell us about changes thus far, and some of the later sections compare models with observations in some detail. This "latitude vs altitude" plot is but one of hundreds of indicators of change that can be compared between models and observations. This particular comparison is discussed in section 9.2.2.1 (p. 674-676), "Spatial and Temporal Patterns of Response" in the WG1 report.
E51
The signature or fingerprint of anthropogenic greenhouse-gas forcing, as predicted by the models on which the IPCC relies, is distinct from that of any other forcing, in that the models project that the rate of change in temperature in the tropical mid-troposphere - the region some 6-10 km above the surface - will be twice or thrice the rate of change at the surface
Wrong and Red Herring: Monckton's Figure 4 is a relabeled copy of the IPCC's Figure 9.1 (p. 675) and as he himself notes in his caption to the figure, it is the modeled response to the actual forcings believed to be relevant for the period 1890 to 1999. Since the forcing from well-mixed greenhouse gases during this period (the third image) is many times larger than any other forcing, it will of course show the strongest response. Nevertheless, the fact is, all the images show an enhancement in warming or cooling in the mid-troposphere region in response. This is hard to see in Monckton's reproduction of the images where some of the color contours are washed out, but in the original IPCC figure 9.1, the small solar forcing shows a stronger response in the tropical mid troposphere; the ozone forcing also shows patches of additional warming in the mid troposphere, and the direct aerosol forcing shows enhanced cooling in the same place, the tropical mid troposphere. In particular, the contours in the original figure 9.1 for the solar forcing show a warming of 0 to 0.2 K at the surface and 0.2 to 0.4 K in the tropical mid-troposphere, an amplification perfectly compatible with the 2 to 3 factor Monckton claims is unique to greenhouse gas forcing. When the solar forcing is scaled to the same value as the greenhouse gas forcing, the tropical mid-troposphere region looks almost identical, as can be seen for example in the images shown in this RealClimate discussion of the mid-troposphere predictions.
So, far from being a unique signature of greenhouse gas forcing, an enhanced response in the tropical mid troposphere is a characteristic of all of the forcings. It is in fact a consequence of the change in the adiabatic lapse rate (the decline in temperatures with altitude) expected for an atmosphere with higher water vapor levels. The more water vapor, on average, the slower the rate of temperature decrease as you rise through the atmosphere, on very general basic physical principles. Therefore, if the surface temperature rises and the lapse rate declines, the temperature for some distance above the surface will rise faster than at the surface. The exact same mechanism arises whatever the cause of the surface warming. The comparison of this basic theoretical prediction with observations is made in section 9.4.4.4 "Differential Temperature Trends" in the IPCC report; more on this below (E53).
However, there is a unique signature for one of the forcings - not greenhouse gases, but solar forcing. As the text of section 9.2.2.1 states: "Solar forcing results in a general warming of the atmosphere with a pattern of surface warming that is similar to that expected from greenhouse gas warming, but in contrast to the response to greenhouse warming, the simulated solar-forced warming extends throughout the atmosphere." You can see this signature in the images of figure 9.1 and perhaps more clearly in the RealClimate images linked just above. For solar forcing, the warming is everywhere from surface to stratosphere. For all the others, whatever the surface does, the stratosphere (above about 20 km) does the opposite. In particular, the greater emissivity associated with increased greenhouse gas concentrations directly causes a cooling of the stratosphere, even before any warming seen at lower altitudes. The observation of stratospheric cooling is strong proof that the sun cannot be responsible for recent changes in global surface temperatures.
E52
The fingerprint of anthropogenic greenhouse-gas forcing is a distinctive "hot-spot" in the tropical mid-troposphere. Figure 4 shows altitude-vs.-latitude plots from four of the IPCC's models: [...] - Figure 5 - Fingerprints of anthropogenic warming projected by four models All show the projected fingerprint of anthropogenic greenhouse-gas warming: the tropical mid-troposphere "hot-spot" is projected to warm at twice or even thrice the surface rate.
Wrong, Confused and Red Herring: Aside from getting his own figure number wrong, see the commentary above (E51) on why Monckton has completely confused the nature of the "fingerprint" shown in these images. All four images do clearly show the true fingerprint: cooling of the stratosphere. Warming of the tropical mid-troposphere is, as discussed above, due to changes in the lapse rate predicted from an increase in water vapor levels in a warmer world. It is unrelated to the cause of the warming and is in no way a "fingerprint" of any one of them, so this is a complete red herring.
E53
the projected fingerprint of anthropogenic greenhouse-gas warming in the tropical mid-troposphere is not observed in reality. [...] In the tropical mid-troposphere, at approximately 300 hPa pressure, the model-projected fingerprint of anthropogenic greenhouse warming is absent from this and all other observed records of temperature changes in the satellite and radiosonde eras: - Figure 6 - The absent fingerprint of anthropogenic greenhouse warming
-
[...] The greater rate of warming in the tropical mid-troposphere that is projected by general-circulation models is absent in this and all other observational datasets, whether satellite or radiosonde.
Wrong, Nonsense and Inconsistent: Monckton repeats his unfounded assertion that enhanced warming of the tropical mid-troposphere is a "fingerprint" of greenhouse warming half a dozen times here and in the previous few paragraphs, indicating the centrality of this false notion to his whole argument in this section.
On the actual observations, first note that they do demonstrate the actual fingerprint of greenhouse (non-solar) warming: cooling of the stratosphere while the surface and lower atmospheric layers warm. So we can be quite certain just from these observations that the sun is not responsible for recent warming of the Earth.
Second, note that the time period for these observations is much shorter than in the graphs of projections (figure 4 and 5): Figure 4 referred to the forcings over about 100 years, from 1890 to 1999. Figure 5 looked at the modeled responses to a doubling of CO2, roughly 300 years from from 1780. Figure 6, in contrast, refers to observational data from "the satellite era", the late 1970s to the present, or at most 30 years. As discussed in E2, 30 years of averaging is barely long enough to define climate, and for detailed features such as these altitude vs latitude averages, one would still expect significant noise. Error bars and uncertainties for these radiosonde and satellite observations are discussed extensively in the IPCC reports - see section 9.4.4 and section 3.4.1 ("Temperature of the Upper Air: Troposphere and Stratosphere") of IPCC AR4 WG1:
Within the community that constructs and actively analyses satellite- and radiosonde-based temperature records there is agreement that the uncertainties about long-term change are substantial. Changes in instrumentation and protocols pervade both sonde and satellite records, obfuscating the modest long-term trends. Historically there is no reference network to anchor the record and establish the uncertainties arising from these changes - many of which are both barely documented and poorly understood. Therefore, investigators have to make seemingly reasonable choices of how to handle these sometimes known but often unknown influences. It is difficult to make quantitatively defensible judgments as to which, if any, of the multiple, independently derived estimates is closer to the true climate evolution. This reflects almost entirely upon the inadequacies of the historical observing network and points to the need for future network design that provides the reference sonde-based ground truth. Karl et al. (2006) provide a comprehensive review of this issue.
The radiosonde record extends back much further than the satellite one (starting in the 1940s) but there is much uncertainty about long-term calibration, for instance (sec 3.4.1.1, p. 266):
Sherwood et al. (2005) found substantial changes in the diurnal cycle in unadjusted radiosonde data. These changes are probably a consequence of improved sensors and radiation error adjustments. Relative to nighttime values, they found a daytime warming of sonde temperatures prior to 1971 that is likely to be spurious and then a spurious daytime cooling from 1979 to 1997. They estimated that there was probably a spurious overall downward trend in sonde temperature records during the satellite era (since 1978) throughout the atmosphere of order 0.1 C per decade globally. The assessed spurious cooling is greatest in the tropics (0.16 C per decade for the 850 to 300 hPa layer) and least in the NH extratropics (0.04 C per decade). Randel and Wu (2006) used collocated MSU data to show that cooling biases remain in some of the LKS and RATPAC radiosonde data for the tropical stratosphere and upper troposphere due to changes in instruments and radiation correction adjustments. They also identified problems in night data as well as day, indicating that negative biases are not limited to daytime observations. However, a few stations may have positive biases (Christy and Spencer, 2005).
The radiosonde data set is limited to land areas, and coverage is poor over the tropics and SH. Accordingly, when global estimates based solely on radiosondes are presented, there are considerable uncertainties (Hurrell et al., 2000; Agudelo and Curry, 2004) and denser networks - which perforce still omit oceanic areas - may not yield more reliable 'global' trends (Free and Seidel, 2005).
Similar calibration issues, though not as severe, are a concern for deriving long-term changes from the shorter satellite record. The most complete summary analysis of all these records presented in the IPCC report is found in figure 3.17 (p. 268), broken down by the major altitude regions, but not by latitude. The variability of the average upper atmospheric temperatures is clearly much higher than that at the surface, but that makes the magnitude of trends harder to determine. Figure 3.18 (p. 269) shows global and tropical trends from 1979 to 2004, with their error bars, clearly large relative to the signal for most of the trends.
Section 9.4.4.4 (p. 701 of IPCC AR4 WG1) discusses the central issue of the apparently missing warming in the tropical mid-troposphere in some more detail:
Since 1979, globally averaged modelled trends in tropospheric lapse rates are consistent with those observed. However, this is not the case in the tropics, where most models have more warming aloft than at the surface while most observational estimates show more warming at the surface than in the troposphere (Karl et al., 2006). Karl et al. (2006) carried out a systematic review of this issue. There is greater consistency between simulated and observed differential warming in the tropics in some satellite measurements of tropospheric temperature change, particularly when the effect of the cooling stratosphere on tropospheric retrievals is taken into account (Karl et al., 2006). External forcing other than greenhouse gas changes can also help to reconcile some of the differential warming, since both volcanic eruptions and stratospheric ozone depletion are expected to have cooled the troposphere more than the surface over the last several decades (Santer et al., 2000, 2001; IPCC, 2001; Free and Angell, 2002; Karl et al., 2006). There are, however, uncertainties in quantifying the differential cooling caused by these forcings, both in models and observations, arising from uncertainties in the forcings and model response to the forcings. [...]
A systematic intercomparison between radiosonde-based (Radiosonde Atmospheric Temperature Products for Assessing Climate (RATPAC); Free et al., 2005, and Hadley Centre Atmospheric Temperature (HadAT), Thorne et al., 2005) and satellite-based (RSS, UAH) observational estimates of tropical lapse rate trends with those simulated by 19 MMD models shows that on monthly and annual time scales, variations in temperature at the surface are amplified aloft in both models and observations by consistent amounts (Santer et al., 2005; Karl et al., 2006). It is only on longer time scales that disagreement between modelled and observed lapse rates arises (Hegerl and Wallace, 2002), that is, on the time scales over which discrepancies would arise from inhomogeneities in the observational record. Only one observational data set (RSS) was found to be consistent with the models on both short and long time scales. While Vinnikov et al. (2006) have not produced a lower-tropospheric retrieval, their estimate of the T2 temperature trend (Figure 3.18) is consistent with model simulations (Karl et al., 2006). One possibility is that amplification effects are controlled by different physical mechanisms on short and long time scales, although a more probable explanation is that some observational records are contaminated by errors that affect their long-term trends (Section 3.4.1; Karl et al., 2006).
This is clearly a scientific question with considerable uncertainty in its resolution at the time of the writing of the latest IPCC report in early 2007 (and for the most part papers had to be published well before the report came out to be fully considered). There have been several more recent papers on the topic, both re-evaluating the calibrations of the different data sets and looking additionally at wind shear numbers from the radiosonde measurements to get better long-term temperature reliability. This is discussed in considerable detail in this RealClimate article on the subject, and there is good reason to believe that the observations, not the models, need to be corrected to reach agreement on this.
In fact, Monckton refers directly to one of these new papers, "Allen et al (2008)" in subsequent paragraphs which illustrate why the observations are certainly not clearly definitive on this (red herring) issue at this point.
One final note here - Monckton is relying on satellite measurements of the temperatures of layers of the atmosphere. This can only be done through measurements of thermal emissions observed at satellite altitudes, and comparing that in detail to the "results of laboratory experiments in passing sunlight through chambers" - the spectroscopic characteristics of the emitting species in the atmosphere (molecular oxygen lines near 60 GHz are typically used). Translating the observed spectra into an averaged atmospheric temperature profile is a tricky business, but Monckton expresses no doubt in its validity. So how can one possibly explain, except through willful ignorance, his doubts about applying essentially the same theoretical analysis to calculation of the forcing from greenhouse gases in the first place?
E54
There are two principal reasons why the models appear to be misrepresenting the tropical atmosphere so starkly.
Nonsense: the models aren't, see above (E53).
E55
the concentration of water vapor in the tropical lower troposphere is already so great that there is little scope for additional greenhouse-gas forcing
Wrong, Confused, and Inconsistent: the limit on water vapor concentration is determined by the Clausius-Clapeyron relation under which the saturation vapor pressure increases roughly exponentially with temperature. Monckton himself points this out later in this very article, in the discussion on non-linearity: "The increase in water-vapor concentration as the space occupied by the atmosphere warms is near-exponential". So there is plenty of scope for more water vapor as the surface (and near-surface atmosphere) warms.
As to the effect of this increase in water vapor on greenhouse-gas forcing, changes in the water vapor component of the greenhouse effect are never considered as a part of the forcing, rather they are part of the feedback response. Additional water vapor in the atmosphere most certainly does have an effect on the radiative properties of the atmosphere; the water vapor component of the atmosphere is an important part of the total greenhouse effect, and changes in water vapor are the most important part of the feedback process.
But more than this, the "hot spot" from the models is, as discussed above (E53), caused by changes in the "lapse rate", the temperature decrease with altitude that balances the pressure reduction with height due to gravity. The higher the water vapor content of the atmosphere, the lower the lapse rate. This has absolutely nothing to do with radiative forcing or the greenhouse properties of water vapor. Monckton continues to completely misrepresent the basic questions here.
E56
Secondly, though the models assume that the concentration of water vapor will increase in the tropical mid-troposphere as the space occupied by the atmosphere warms, advection transports much of the additional water vapor poleward from the tropics at that altitude.
Invalid: There is no reason changes in advection in a warmer world would somehow transport the "additional water vapor" only (or mostly). In any case, this would be an issue for estimation of feedbacks, not radiative forcing, the supposed topic of this entire section.
E57
the great majority of the incoming solar radiation incident upon the Earth strikes the tropics
Wrong and Red Herring: The tropics are usually defined in climate discussions as from 20 degrees S to 20 degrees N. The fraction of incoming sunlight striking that part of the globe at the top of the atmosphere is about 43%, less than half. Even using the full tropical range of 23.4 degrees N and S gives only about 49%. Hardly the "great majority".
Moreover, the relevant factor for radiative forcing (supposedly the topic of this section of the article) is the outgoing thermal flux, not incoming solar, and that comes simultaneously from all parts of the globe, day and night. You need to expand the definition of tropics to 30 degrees S to 30 degrees N (as is sometimes done) to include half of Earth's surface area; 20 S to 20 N is just 34%.
E58
any reduction in tropical radiative forcing has a disproportionate effect on mean global forcings. On the basis of Lindzen (2007), the anthropogenic-ear radiative forcing as established in Eqn. (3) are divided by 3 to take account of the observed failure of the tropical mid-troposphere to warm as projected by the models
Invalid and Nonsense:
Lindzen's claim (hardly supported by much evidence, given the above discussion (E53) on the tropical troposphere observations) is concerned with the feedback effect of water vapor levels in the tropical atmosphere, and has nothing to do with the radiative forcing calculated for greenhouse gases. So the factor of three applied here might be more correctly applied to the feedback factor discussed later - it clearly has absolutely nothing to do with the value of the greenhouse gas forcing number ΔF<sub>2x</sub> that Monckton attributes it to here.
Even if this factor of 3 is real it cannot be applied to the entire globe - it applies only to the tropics. As noted above (E53), the models and observations agree well outside the tropics. Monckton is claiming the unlikely factor of 3 reduction in tropical water vapor feedback can be applied globally based on the preceding erroneous claim about the "great majority" of incoming solar radiation. One would have to slice this reduction at least in half since the "great majority" is itself less than half. In the end, if the models are wrong about the tropical troposphere by overestimating water vapor levels as Monckton is arguing here, the final effect on the water vapor feedback term would be on the order of a 30% reduction, rather than the claimed factor of 3. Such a 30% reduction is well within the existing range of estimates of the water vapor feedback from different climate models anyway - see figure 8.14 in IPCC AR4 WG1 (p. 631).
In any case, Monckton's equation 17 is wrong and of no value; the arguments here have absolutely no bearing on the forcing from greenhouse gases.
In retrospect, I might have saved myself a lot of time just listing the statements in Monckton's article that were not wrong in some way. Of course that would have avoided actually explaining why all the things he got wrong were wrong. At times in the article he appears to be understanding some of the central issues with respect to modeling and feedbacks, but at other times it's just amazingly far off reality. Can you be this wrong without it being deliberate? It reminds me of the title (and related discussion) in Robert Park's book "Voodoo Science: The Road from Foolishness to Fraud" - at some point these guys have to realize that what they're doing is simply wrong, but they keep it up, and compound error upon misquote, red herring on cherry pick, lie upon lie.
Ok, maybe this is about McCain and Palin after all...