Comments should be relevant to the subject matter of the paper, and normal courtesy all round would be much appreciated. To conserve space, quoted text should be kept to a minimum and kept relevant to the topic under discussion. - JD

Name

Richard Courtney
Hans Erren
Jack D. Black
Dave Dardinger
Richard Courtney
Onar Åm
Dave Dardinger
Peter Dietze
Jack Barrett  
Craig Wheelock   
Heinz Hug & Jack
    Barrett   
Andy Harris    
Richard Courtney   
Jim Clarke     

Country

Britain
Netherlands
USA
USA
Britain
Norway
USA
Germany
Britain

Germany &
   Britain
USA
Britain
USA

Date

17 Dec 2001
18 Dec 2001
18 Dec 2001
19 Dec 2001
21 Dec 2001
22 Dec 2001
23 Dec 2001
23 Dec 2001
17 Jan 2002
27 Jan 2002
11 Feb 2002

8 Mar 2002
9 Mar 2002
14 Mar 2002

Remarks

A question re `relevant processes'
Suggested amendment to sensitivity equation
Relates Hug-Barrett paper to MSU-surface conflict
Questions relating to atmospheric heat distribution
Atmospheric thermal transfer mechanisms
Suggestion for a validation test 
Modes of atmospheric heat transfer
Discussion of atmospheric radiative properties
Discussion of sources of atmospheric heat transfer
Challenge for any evidence of re-radiation by CO2
Their reply to Peter Dietze's criticisms.

Convective v. radiative transport of heat - definitions
Response to Andy Harris re' modelling procedures
Questions the error ranges of the models

Date: Tue, 18 Dec 2001 08:04:44 GMT
From: "h erren" < erren21@zonnet.nl >

Nice paper 
I think I discovered a typo:

Insertion of two brackets might clarify the sensitivity equation
dT/dE = 1/(4 [sigma] T^3)

instead of
dT/dE = 1/4 [sigma] T^3

Hans Erren

Date:  Tue, 18 Dec 2001 08:33:39 -0600
From: "BLACK, JACK D. (JSC-NE) (GHG)" < jack.d.black1@jsc.nasa.gov >

I don't lay claim to any particular expertise in the field of photometric energy absorption but the Hug-Barrett postulation seems to me to be rationally consistent with the general concepts of atomic energy absorption and dissipation. It also seems to offer a rationally consistent approach to explaining the apparent discrepancy between the near-surface temperature anomalies and the Microwave Sounding Unit/Radiosonde Balloon measurements in the lower troposphere.

The US National Academy of Science review of these discrepancies in temperature logs essentially found all three to be accurate but offered no guidance whatsoever for resolution of the discrepancies. It seems reasonable to at least consider the prospect of energy loss, in the lowest strata of infrared absorbing atmospheric gases, via collisional energy transfer which would preclude direct reradiation of absorbed photon energy as is apparently assumed in the current GCM models. The result of these energy transfers would, it seems to me, be to produce a near surface layer of higher kinetic energy or temperature while the higher layers responded much more slowly and function essentially to smooth local variations.

J. Black       banshee13@earthlink.net 

Date: Wed, 19 Dec 2001 20:33:43 -0700
From:  "DAVE DARDINGER" < dedmod@email.msn.com >

I'm having a bit of trouble with the 19%, 61%, 20% division of atmospheric heating from the surface proposed. I see where it comes from, but it looks to me like not all the requisite figures are being considered. Thus there are 67 W/m2 absorbed directly from the atmosphere, which presumably is converted into thermal motion and thence into IR radiation, some of which contributes to the 324 W/m2 back radiation. A more difficult situation is the radiation emitted by clouds amounting to 30 W/m2. Part of this would come from the latent heat of condensation from rising water vapor, and part might come from equilibration with atmospheric gases in general. I'd assume that most all latent heat released by low level clouds would be reabsorbed by the higher levels of the atmosphere, while much latent heat released at high level clouds would be released to space.

The point is that given these other inputs / outputs I'm not sure such precise divisions of contribution to atmospheric warming by various surface processes can be made.

Dave Dardinger

Fri, 21 Dec 2001 10:36:31 EST
From: Richard Courtney  < RichardSCourtney@aol.com >

In 'Open Review' of the Hug-Barrett paper, Dave Dardinger says, "The point is that given these other inputs / outputs I'm not sure such precise divisions of contribution to atmospheric warming by
various surface processes can be made." 

I agree, and I suspect Hug and Barrett agree too. The significant point is that Hug and Barrett make this distinction whereas IPCC assumes all thermal exchange in the atmosphere is radiative. This was first explained in IPCC 1990 page 49 where it says; 

"The surface, planetary boundary layer and the free troposphere are tightly coupled via air motions on a wide range of scales, so that in a global mean sense they must be considered as a single thermodynamic system. As a result it is the radiative flux at the tropopause, and not the surface, that expresses the radiative forcing of the climate system." 

Only radiative thermal transfer occurs at the tropopause. And, as the above quotation states, the IPCC makes no distinction between various thermal effects lower in the atmosphere but assumes they summate to the radiative effect at the tropopause. Hence, the IPCC asserts that all atmospheric thermal transfer can be emulated as being radiative. Indeed, the IPCC considers effects of water vapour to be a "feedback" despite water vapour being the major greenhouse gas. 

The IPCC assertion is clearly nonsense because if it were true then rain would never happen. But the IPCC has not retracted the mistake it first made in 1990. 

Hug and Barrett may not have partitioned the thermal transfer effects accurately. But they have partitioned them, and this
partitioning is why their analysis provides very different conclusions from those of the IPCC. Large changes to the values Hug and Barrett assign to the partitions make relatively small change to their results. 

All the best      Richard 

Date: Sat, 22 Dec 2001 13:26:55 +0100
From:  Onar Åm < onar@netpower.no >

I have no particular expertise in radiative physics, but the Hug-Barrett hypothesis may be relevant to a little mentioned problem with current climate models, namely the fact that they appear to COLD. That is, too cold in absolute temperatures. There is no clear understanding of this phenomenon and one assumes that it does not significantly affect predictions of warming, i.e. relative temperature changes.

However, the Hug-Barrett hypothesis could potentially explain why models are too cold. The GCMs assume that the thermal transfer in the troposphere is radiative only.  But if some of this radiation translates into molecular heat, then that would not only change the radiative behavior of the atmosphere but also the absolute temperature. Presumably it would be WARMER, consistent with observation.

If my hunch is correct, then it should be possible to partially test the Hug-Barrett model by implementing it into a GCM and check if the absolute tropospheric temperature increases to match observation.

Onar.

Date: Sun, 23 Dec 2001 08:54:00 -0700
From: "DAVE DARDINGER" < dedmod@email.msn.com >

errata from my last note. In " 67 W/m2 absorbed directly from the atmosphere" should read " 67 W/m2 absorbed directly from the sun by the atmosphere."

Thanks for the clarification by Courtney.

Since my other point might imply that the % of heat transfer within the atmosphere by way of radiation is higher than 20% I'll mention a mechanism this time in the opposite direction. Since the 61% transfer from evaporation/transpiration only is concerned with the net transfer of heat from the surface, it presumably doesn't include additional vertical transport of heat within the air column. But in addition to the immediate transport of latent heat via water vapor there's also the indirect transport by way of re-evaporation of precipitation before it reaches the ground.  This is known as virga when none of the rain or ice reaches the ground, but even when it's actually raining, there will be some evaporation on the way down as long as the underlying air is warmer (which is generally the case). This results in a net transfer of heat upward as the evaporation cools the lower layers of air at the expense of the upper layers. I have no idea exactly how large this effect is, nor whether or not it's offset by transfer of heat from water vapor to other components of the atmosphere. (This last seems unlikely, however, since water vapor would presumably be equilibrated with the atmosphere as it evaporated.)

Dave Dardinger

Date: Sun, 23 Dec 2001 22:09:45 +0100
From: < p_dietze@t-online.de > (P. Dietze)
Reply-To: < 091335371@t-online.de >

Jack Barrett and Heinz Hug are to be commended for this paper. But I would like to make a few objections and additions:

> The IPCC explains the phenomenon in terms of radiation and
> imply that molecules do not obtain kinetic energy (heat) by
> absorption of radiation by greenhouse gases.

The Schwarzschild radiative transport equation (which IPCC uses for model calculation) does not cope at all with what happens to the energy.  Indeed, after absorbing an IR photon, a molecule's excitation is vibrational or rotational. But after collision with other molecules this energy is converted into kinetic and thus thermal energy. The absorbed energy is calculated using HITRAN spectra. The emitted energy is calculated from a given temp profile, using Planck's formula for black body radiation and Kirchhoff's law (emissivity = absorptivity) for a number of atmospheric layers.

As the radiative transport equation has no term related to spontaneous re-emission (contrary to the authors conviction), the emitted energy must stem from kinetic (thermal!) energy by which the layer temperature is defined. Btw, the radiative transport equation has no mechanism to force the emitted energy to be equal to the absorbed energy (i.e. to fulfill the local thermal equilibrium LTE). IPCC thus may violate the LTE condition if the temp profile is not preset adequately.

> The IPCC state that the air is warmed by contact with
> the warmed soil/ocean surfaces.

I never heard this. I suppose this may be either an erroneous statement from some textbook or a conclusion of the authors, based on their impression that all IPCC's absorbed energy is spontaneously re-radiated and thus the atmosphere does not warm up. Actually the re-radiation (and thus back-radiation to ground) from CO2 and the other GHGs is calculated based on assumed atmospheric warming and the CO2 concentration increment. But in fact more CO2 leads to an increasing lapse rate (see Ch.3 of my official IPCC TAR Review at http://www.john-daly.com/forcing/moderr.htm). This results in warming near ground but radiative cooling starting within the lower troposphere and progressively increasing up to the tropopause. This increased lapse rate (which could so far not be simulated with atmosphere models) should be the reason why satellite MSUs measure no significant warming trend in the range of 1-5 km height.

> The IPCC assumes that photons are recycled by CO2 causing 90%
> of the absorbed radiation to return radiatively to the surface

This is actually not IPCC's figure but stems from Graedel/Crutzen p.47 (Chemie der Atmosphäre, 1994) and deals with near ground radiative fluxes for the basic natural Greenhouse effect including other GHGs and clouds. My HITRAN calculation for the anthropogenic impact yields a total absorption of some 7.4 W/m² for CO2 doubling. IPCC says, the forcing (at tropopause level) is 3.7 W/m². This is 50% and not 90%.

> whereas under the non-equilibrium conditions occurring
> daily in any part of the globe at any time, the [Kirchhoff]
> law is inapplicable

I am quite sure that Kirchhoff's law (i.e. emissivity is equal to absorptivity) which is used in the radiative transport equation, holds under any circumstances and does not require thermal equilibrium. As long as cool air is warming up during a sunny day, the emitted energy is less than the absorbed energy, but still emissivity = absorptivity (denoting a material property). The law does not say that absorbed energy = emitted energy. The authors may have misinterpreted this.

Heinz Hug's lab results in Fig.1 are very interesting, revealing that the absorption of CO2 is considerably enhanced by IR-neutral gases like nitrogen - which so far has not been coped for by HITRAN and which can by no means be explained by pressure broadening or equipment fallacies. But the problem whether this may lead to more or less residual absorption (for CO2 doubling) at the fringes, could not be solved. Heinz uses this finding rather to prove that thermalization by molecular collision really takes place. But I think, this debate versus assumed spontaneous re-radiation of IPCC is not very fruitful as it makes (macro-energetically) not much difference whether the absorbed energy is directly re-radiated by CO2 or thermalized and then re-radiated. Of course, thermalization would yield a different heat and spectral distribution within the troposphere, but any difference in ground warming is difficult to quantify.

Jack Barret's chapter 4 about atmospheric sensitivity really reveals IPCC's exaggeration factor 3.7 - this has been excellently presented at the DECHEMA colloquium. In chapter 5 Table 2 Jack shows that the delta absorption for a 100 m layer and 2*CO2 without vapor is 1.7 %, but with vapor it is only 0.7 %. But the wavenumber interval for which this was integrated, is missing.

> Hug emphasized that at a height above the ground of
> 1000 m 97% of the carbon dioxide bands are saturated

This only holds for the linestrengths (molecular extinctions) which are not representative for energy absorption. There are always fringes that are not saturated and this is why we have the same forcing in W/m² each time CO2 is doubling.

> that the spectral overlap, ignored by the IPCC, is the
> reason for the sensitivity being exaggerated by the IPCC

It is not only the vapor overlap, but another fact is that IPCC relates the 3.7 W/m² to 255 K (yields 0.98 K) whereas if it is related to 288 K near ground, we get 0.68 K only. Apart from double the forcing (by not coping with vapor overlap) IPCC uses about a 2.5 times vapor feedback.  But acc. to Bengtsson et al. (JGR Feb 1999) most of the troposphere is cooling with more CO2 (as observed). The lower regions where the clouds are, may be hardly affected. As the whole troposphere may even contain less vapor, the strong vapor feedback should rather be a fairy tale. Only near ground we will have a vapor increment. Alltogether IPCC's CO2 sensitivity is probably by a factor four too high and thus the vapor increment near ground is considerably reduced. The great differences in cloud forcing shown in Fig.3 are really eyestriking. 

Thanks to Jack.

Peter

Date:  Thu, 17 Jan 2002 12:05:59 EST
From: Jack Barrett   Jack1Barrett@cs.com

Richard Courtney's contributions are welcome. Hans Erren's clarification of the derivative of the Stefan-Boltzmann equation is good; you could get a very funny result if the T^3 term was understood to be above the line. Dave 
Dardinger's comments are valid. The 67 W m^-2 absorbed by the atmosphere are discounted in our calculation of the contribution to warming by radiation. The 67 W m^-2 are absorbed in the stratosphere (10) and the upper troposphere. Some (but how much) will contribute to downward radiation. We could do a 'Solomon' on it and say half comes down and the other half goes up, but the main point is that not 100% of the warming is due to radiation. The 30 W m^-2 cooling from the cloud-tops is part of the diagram, but I don't know its origin. If Stefan-Boltzmann is applied, the cloud tops are very cool indeed.

Jack

Date: Sun, 27 Jan 2002 15:30:31 -0800
From: Craig Wheelock < craige@QNET.COM >

Readers and Staff,

I hereby challenge all readers of this note to provide me with quantitative and reproducable evidence of the re-radiative properties of CO2. I have been searching for some time, and have come to the conclusion it is non-existant.

Craig Wheelock    < craige@qnet.com >

1. From H & B point 2.

(a) We have no quarrel with the Schwarzchild equation or its applicability. Its two terms are connected with the heating (absorption of photons) and cooling (emission of photons) of the atmosphere as radiation passes from the surface to space. The IPCC object to our interpretation of the heating rate in terms of thermalization of the absorbed energy, but it seems to us that there is no difference in understanding of the equation. The emitted energy is approximately calculated by using the Planck emission moderated by the absorption coefficients for the molecules participating and at the various temperatures along the optical path. There are uncertainties in such measurements.

(b) The Planck emission equation contains both Einstein A and B factors and if either was to have a zero value the Schwarzchild equation would not work. In a very simplified version the Planck emission can be formulated as: L = [A21(n)/B21(n)] * C(n, T). The term C(n, T) describes especially the radiative thermodynamic equilibrium and includes the Boltzmann factor. If the spontaneous emission A21(n) were zero there would be no emission – also no "thermal emission". If the stimulated emission in the radiation field were zero (B21(n) = 0) then the emission would be infinite. Instead of L (radiation to space) the term B (brightness) of a layer is also in use by some authors. In both cases Planck emission is assumed.

(c) Local thermodynamic equilibrium means that the collision rate is sufficient to maintain the Maxwell-Boltzmann distribution of velocities, and the values of the Boltzmann factors for the molecules of the atmosphere. It does not mean that there is anything like a true equilibrium in existence. LTE breaks down only at high altitudes (>85 km) when the fraction of vibrationally excited CO₂ molecules is less than that given by the Boltzmann factor and where radiative emission predominates over collisional deactivation.

(d) We can only repeat that the IPCC people at the Frankfurt conference maintained that the atmosphere in contact with the surface was mainly heated by the physical contact [whether they meant thermal conduction or that plus convection and evaporation we are not sure] rather than by any radiative processes. We strongly disagree with their point of view.

A similar testimony can be found in the book by Gary E. Thomas and Knut Stamnes "Radiative Transfer in the Atmosphere and Ocean". [This book is highly recommended by us.] Prof. Raschke and other modellers refer to it, and it describes the fundamentals of radiative transfer as assumed by the IPCC. The authors call the thermalization or quenching an "inelastic collisional" process and say: By inelastic collisions "energy is transferred from kinetic to internal excitation energy, or vice versa. An elastic collision is one in which there is no net transfer of energy from kinetic to excitation energy (p. 99)." At the same page, figures are given which assume that thermalization (quenching by inelastic collisions) is of less interest: "Elastic cross sections are of order 10^-19 m² Inelastic cross sections are much smaller; of order 10^-23 to 10^-25 m² (p. 99). There is a remark (Note 15 at p. 128) about the significance of these numbers: "If inelastic cross sections were not much less than the elastic, there would be drastic consequences. Consider air at sea level, for which ~10⁹ collisions occur [per molecule, our addition] per second. If a significant fraction of these collisions were inelastic – leading to an eventual radiative loss of heat – the air would cool down almost instantaneously!" Thus, the only conclusion remains: The basic theory of modelling assumes that the radiation that had been emitted by the soil escapes through the "open window" or is absorbed by greenhouse gases. In the latter case the absorbed energy is transported in the air only by radiation. Thermalization is not assumed in any case. The only way in which kinetic energy of air molecules can be attained is by the direct contact between air and soil (or water). Thermalization is excluded by the theory of radiative transfer. This had been confirmed at the DECHEMA by the modellers. Reading only the IPCC papers for "policy makers" or other papers in Science, Nature, etc., should not be the base on which the main error can be discussed.

(e) You say that the energy budget reference should be to Graedel and Crutzen. They may have the same diagram, but our reference is correct. The diagram (and that of Graedel/Crutzen presumably) is based on the original data given by Kiehl and Trenberth, Bull. Am. Met. Soc., 78, 197, (1997). The diagram indicates values for all fluxes occurring in the atmosphere, not just those near the surface. Your model does not reproduce the experimental observations, and the fraction of 50% radiation returning to the surface value is your choice.

There is more than one explanation of the greenhouse effect. One is based on calculating the increase of long-wave radiation remaining in either cloudless or cloudy atmosphere (G. Myrhe, E. J. Higwood, J. P. Shine, F. Stordal (1998) Geophys Res Lett 25, 2715. Another concept compares the thermal radiation from the ground with emission to space over the same area. The values range between 100 and 250 W m^-2 depending on the geographical region and season (E. Raschke, Fresenius J Anal Chem (2001) 371, 791. The last theory copes with that what H & B are saying. They mean that the GE is caused more by thermalization and delayed emission (delayed by previous collisional and bulk transfer processes) to space by the atmosphere over the whole day (day & night) and the Earth’s surface. The solar energy absorbed by the Earth’s atmosphere has a residence time of around 124 days, not a figure that would be consistent with processes all occurring at the speed of light.

(f) We agree with Richard’s comments on the Kirchhoff equation.

2. From H & B point 3.

The linestrengths reported are from real spectra and are a measure of absorption; they are integrals of the absorption coefficients over the spectrum. This is the conventional way of representing the intensity of a transition rather than just the absorption coefficient at the frequency of maximum absorption, the latter measurement being subject to the pressure of the system.

3. From H & B point 5.

Jim Clarke