Some thoughts on radiative transfer and GHG's

Absorptions bands in the Earth's atmosphere cr...

Absorptions bands in the Earth’s atmosphere created by greenhouse gases and the resulting effects on transmitted radiation. (Photo credit: Wikipedia)

Guest post by Reed Coray

The following example illustrates the issues I have with reasoning often used to argue that increasing the amount of CO2 in the Earth’s atmosphere will increase both the Earth’s surface temperature and the Earth’s atmosphere temperature. Immediately following is a direct quote from URL

The present situation is that there has been an increase in infrared-absorbing gases in the atmosphere, such as carbon dioxide (CO2) and methane (CH4). Energy that would normally escape into space is absorbed by these molecules, thus heating the atmosphere and spreading through convection currents. The average temperature of the atmosphere has increased 0.25 °C since 1980, mainly attributed to an increase in infrared-absorbing gases in the atmosphere.

Although the above statement makes no direct reference to Earth surface temperature, I believe it carries the implication that greenhouse gases in the Earth’s atmosphere increase the Earth’s surface temperature.

I make two comments: the first is relevant only if the above implication is valid, the second is relevant independent of the validity of the implication. First, placing matter adjacent to a warm surface such that the matter is capable of absorbing/blocking radiation to space from the warm surface can lead to a decrease in the warm surface’s temperature. Second, increasing the amount of the absorbing/blocking matter can lower the temperature of the absorbing/blocking material.

Take for example an internal combustion engine whose metal surface is exposed to a vacuum. In addition to doing useful work, the engine produces thermal energy (heat). That thermal energy will produce a rise in the temperature of the engine’s surface such that in energy-rate equilibrium the rate energy is radiated to space from the engine’s surface is equal to the rate thermal energy is generated within the engine. By attaching radiating plates to the engine’s surface, some of the energy radiated to space from the engine’s original surface will be absorbed/blocked by the plates; but because thermal energy can be transferred from the engine to the plates via both radiation and conduction, the temperature of the engine’s original surface will be lowered. This is the principle of an air-cooled engine[1]: provide a means other than radiation of transferring heat from an engine to a large surface area from which heat can be removed via a combination of conduction, convection and radiation, and the engine’s surface temperature will be lowered.

If plates at a temperature lower than the original engine surface temperature are attached to the engine, it’s true that the temperature of the plates will increase to establish energy-rate equilibrium. Once energy-rate equilibrium is established, however, increasing the plate radiating area (adding additional matter that blocks more of the energy radiated from the original engine surface) will likely lower the plate temperature.

Thus, blocking the amount of surface radiation escaping to space does not necessarily increase the surface temperature; and increasing the amount of radiation blocking material does not necessarily increase the temperature of that material. In both cases (the Earth/Earth-atmosphere and the internal combustion engine in a vacuum), the heat eventually escapes to space–otherwise the temperature of the Earth’s surface and the engine would continue to rise indefinitely. The difference isn’t that the energy doesn’t eventually escape to space (it does in both cases), the difference is in the path the energy takes to reach space. The amount of generated thermal energy in conjunction with the path the thermal energy takes to get to space determines temperatures along the path; and adding more material may increase or decrease those temperatures. To say that “Energy that would normally escape into space is absorbed by these molecules, thus heating the atmosphere…” by itself is unwarranted; because an equivalent statement for the case of adding extra plate material to the engine would be “Energy that would normally escape to space from an engine with small attached plates is absorbed by additional plate material, thus heating the plates…” For air-cooled engines, this statement is not true—otherwise the plate surface area of air-cooled engines would be as small as possible.

It’s fairly easy to visualize why (a) adding thermally radiating plates to an air-cooled engine might decrease the engine’s surface temperature, and (b) increasing the area of the radiating plates might decrease the plate temperature. It’s not so easy to visualize, and may not be true, why (a) adding greenhouse gases to the Earth’s atmosphere decreases the Earth’s surface temperature; and (b) increasing the amount of atmospheric greenhouse gases lowers the temperature of the Earth’s atmosphere. I now present one possible argument. I do not claim that the argument is valid for greenhouse gases in the Earth’s atmosphere, but I do claim that the argument might be valid, and can only be refuted by an analysis more detailed than simply claiming “Energy that would normally escape into space is absorbed by these molecules, thus heating the atmosphere.”

If we assume that (a) matter cannot leave the Earth/Earth-atmosphere system, and (b) non-greenhouse gases radiate negligible energy to space, then for a non-greenhouse gas atmosphere the only way thermal energy can leave the Earth/Earth-atmosphere system to space is via radiation from the surface of the Earth. The rate radiation leaves the surface is in part a function of both the area and temperature of the surface. For a greenhouse gas atmosphere, energy can leave the Earth/Earth-atmosphere system to space both via radiation from the Earth’s surface and radiation from greenhouse gases in the atmosphere. Suppose it is true that the density of greenhouse gases near the Earth’s surface is such that radiation emitted from low-altitude greenhouse gases does not directly escape to space, but is in part directed towards the Earth’s surface and in part absorbed by other atmospheric greenhouse gases. As the atmospheric greenhouse gas density decreases with increasing altitude, radiation emitted from high-altitude greenhouse gases can directly escape to space.

Now it’s not impossible that since (a) in addition to radiation, heat is transferred from the Earth’s surface to greenhouse gases via conduction, and (b) convection currents (i) circulate the heated greenhouse gases to higher altitudes where energy transfer to space can take place and (ii) return cooler greenhouse gases to the Earth’s surface, that the process of heat transfer away from the Earth’s surface via greenhouse gases is more efficient than simple radiation from the Earth’s surface. Many engines are cooled using this concept. Specifically, a coolant is brought into contact with a heated surface which raises the coolant’s temperature via conduction and radiation, and the coolant is moved to a location where thermal energy transfer away from the coolant to a heat sink is more efficient than direct thermal energy transfer from the heated surface to the heat sink.

One way to realize increased thermal transfer efficiency would be to use a coolant, such as greenhouse gases, that efficiently radiates energy in the IR band (i.e., radiates energy at temperatures around 500 K). Another way would be to spread the heated coolant over a large surface area. Since surface area increases with increasing altitude, thereby providing expanded “area” (in the case of a gas, expanded volume) from which radiation to space can occur, it’s not clear to me (one way or the other) that greenhouse gases won’t act as a “coolant” reducing both the temperatures of the Earth’s atmosphere and the Earth surface.


[1] It’s true that for most air-cooled engines the main transfer of heat from the engine plates is via a combination of (a) conduction of heat to the air near the plates, and (b) convection that replaces the warm air near the plates with cooler air. To aid this process, a fan is often employed, or the engine is located on a moving vehicle and the vehicle’s motion through an atmosphere provides the flow of air across the plates. Although conduction/convection may be the primary means of heat dissipation from the plates, radiative cooling also dissipates heat.


newest oldest most voted
Notify of
Peter Dunford

An interesting alternative take. And convection is certainly ignored in the “Standard Model”. However, I can’t see the likes of Trenberth even giving it a second glance, the science is “settled”.

Old England

Convection via thunderclouds – as Willis has noted – would fit alongside this. There is a lot of commonsense in this thinking and it deserves study. Intuitively it begins to show how part of earth’s ‘thermostat’ may function.


This ought to rattle some cages!

Getting close to a better theory.
Heat cannot be stored by these ‘so called’ GHG’s as 2nd law states that heat must be lost, an increase in entropy, but adsorbed SIR will increase the molecular kinetic energy, increasing the temperature, but this kinetic energy will be transferred to the other gasses not directly affected by the SIR. The GHG’s will radiate LIR but at a reduced energy level, the frequency change from short to long wave IR is the evidence that this happens. Also the 1st law dictates that it must. If it did not then we would be driving in cars using perpetual motion engines which violate both 2nd and 1st laws.
The reradiated heat is in fact a reduction of solar heat reaching the surface so it cannot raise the temperature more than that reached without the intervention of the GHG molecule. If the theory of reradiated heat were to be true then warm liquids placed into a vacuum flask, with its mirrored internal surfaces, would raise the liquid’s temperature by a considerable amount. We all know from experience that this does not happen but a vacuum flask only reduces heat loss slowing cooling. It insulates well but not perfectly.
We must also ask ourselves why the earth’s temperature is fairly even in that max. temperatures rarely exceed 50C nor do we get any lower than -85C. The moon, receiving the same solar insolation as earth but having zero atmosphere, achieves over 150C in sunlight and less than -150C in shadow. So much for an atmosphere increasing temperature by reradiation.

The average temperature of the atmosphere has increased 0.25 °C since 1980, mainly attributed to an increase in infrared-absorbing gases in the atmosphere.”

But has it really? If you look at the warming from 1900 to present, look particularly at the warming from about 1910 to about 1945. Temperatures then cooled until about 1975. Beginning in 1976, things began warming again but it took several years to warm back up to where temperatures were in the 1930’s and 1940’s. We can not really consider that period to really be overall “warming” in the context of recovery from the LIA, we can only consider that to be recovery from a spate of cooling after temperatures reached their peak in the late 30’s early 40’s. So we must wait until temperatures get back to that point before we consider that we would have any additional overall warming of the planet.
Whether or not we ever reached the levels of the 1940’s is a matter of some debate, but I would say that the period since 1980’s saw NO temperature rise attributed to CO2 and only a recovery of temperatures to NEAR what they were in the 1930’s and 1940’s. From my point of view, the second half of the 20th century actually saw no or very little “warming” at all in the context of additional recovery from the LIA. It only saw some natural variation where temperatures declined from the 1940’s through the 1970’s and then recovered.


Took a long way getting there, but I understand it in the end. Adding greenhouse gas may be like adding blades to an air-cooled engine.
If our concern is surface temperature then perhaps the “more blades” equivalent on Earth might be adding more mountains?
In a still environment you have conduction and radiation emission. The surface temperature at night will always be coolest in a vacuum won’t it? Any atmosphere, even non-greenhouse, will cause heat to linger around the surface – the buffer – which raises the average temperature at the surface. Greenhouse gases create a thicker buffer.
Just thinking of the mountain example above. Incoming radiation is constant and if we add more surface area the radiation is spread across a greater area. Each unit area absorbs less heat and hence doesn’t radiate as fast – thermodynamics says the higher the difference the faster the loss. With that in mind perhaps a flat Earth endures greater extremes than a bumpy Earth? Perhaps that applies even where more finer details are concerned e.g. trees add to the surface area.

Dr Burns

Most heat transfer from the Earth’s surface is not directly to space, as is so often assumed in such models. You should consider clouds, which cover 70% of the Earth’s surface. Their temperature is set by the lapse rate.

michael hammer

I am extremely sceptical of CAGW but I have to strongly disagree with the above analysis. Adding cooling fins to a motor decreases its surface temperature because it increases the surafce area availabel to radiate that heat away, In the case of the earth the surface area is not increased. The point that energy can be lost to space from the surface at all thermal IR wavelengths and from the atmosphere but only at the GHG wavelengths is true in principle. However, because the GHG effect is so strong over the atmospheric column in effect the surface can only lose enegy at the non GHG wavelengths while the atmopshere can only lose energy to space at the GHG wavelengths. What increasing the GHG concentration does is to slightly increases the range of GHG wavelengths so the surface can lose energy over a slightly smaller range of wavelengths and the atmosphere over a slightly larger range of wavelengths. Sine the atmosphere is cooler than the surface it loses less energy than would the surface at the same wavelength. Thus the actio of the GHG increase is to slightly reduce the energy loss to space and to restore balance the temperature of the system must slightly increase.
However – what is at issue is how much. By how much does the temperature have to increase to compensate? A simple calculation shows doubling CO2 would lead to an increase of about 1C in the absence of feedbacks. That’s not serious so are the feedbacks positive or negative? This is the crux of the debate but its worth noting that every naturally stable system shows strong negative feedback and that means the rise will be less than 1C not more than it.
Where does the negative feedbakc come from here? Higher temperatures means more evaporation which must mean more rain but rain comes from low clouds so it must mean more low cloud either in density or in coverage. Both increase Earth’s albedo and reduce temperatures.


This is a very baffling post Reed, it doesn’t make sense. An air cooled engine does not cool by blocking radiation. It cools by conducting heat away from the engine into the fins and, because they provide a larger surface area, more heat is subsequently radiated away (or convected away). You have to be very careful with analogies, and this one is too contrived to be useful. The atmosphere, clearly, doesn’t work like this.


As an ex-motorcyclist, first hand experience leads me to agree with much of what Reed Coray writes. As a layman in the sciences I must rely on educational web sites that explain heat transfere and IR radiation and such like. I am led to believe that gases in the atmosphere can absorb OR radiate specific radiation bands dependant on the local temperature. It cannot do both at the same time. Wein’s Law will give the peak temperature at any specific IR wavelength. Using Wein’s Law to look at CO2 I find that the 2.7 micron band peaks at ~800C, the 4.3 micron band at ~400C and the 25 micron band at about -80C!! I understand only limited areas of the Earth’s surface might radiate at up to 50C so the 2.7 and 4.3 micron bands will NEVER be exited enough to absorb any energy from the surface. They might absorb a very little from the sunlight but that is working as a coolant. The so called standard surface temperature of the Earth is said to be 15C, well above the the -80C temperature level of the 15 micron band for CO2. The problem now is most of the CO2 molecules in the atmosphere will be at a temperature comensurate with the adiabatic lapse rate starting at the surface. So assuming a drop of 10C per kilometre altitude air temperature should be down to -80C at about 9.5 kilometres altitude, almost the tropopause. Only then will the CO2 molecules be cool enough to absorb radiation at 15 microns.
BUT! There is indeed nothing to stop the CO2 radiating at 15 microns and some of that radiation reaching the surface. Now another BUT! The surface, except at possibly a small area at the south pole, is well above -80C!! Any element, black body or not, does not absorb radiative energy below its peak temperature.
A CO2 molecule IS a black body with rather specific characteristics. And so is any other gas molecule in the atmosphere.
Since I am completely unable to see any ‘greenhouse’ effect in the atmosphere I need more education. Please post links that will this layman.

ursus augustus

What about the additional LH of Vapourization drawn from the surface by both any direct additional evaporation due to CO2 GH effect but also the increased transpiration from plants as their metabolism increases due to more CO2 (double -ve feedback)? H2O vapour up into the atmoshpere causes some more GH effect , true (+ve feedback) but then recondenses as clouds ( albedo => -ve feedback) and releasing the LHV at altitude where convection/cnduction takes it upwards and into space ( -ve feedback). What is the net feedback effect? Who knows but there is no hot spot at altitude so it cannot be much.


Is Trenberth’s approval a hurdle this has to clear?


Tcha!! Educate, educate, educate!


Those spectral intensity curves are wrong. The formula I = kT^4 gives the area under the curves (in W/m^2) (k = Stefan-Boltzmann constant).
Temperature Area under curve
210K 110 W/m^2
260K 260 W/m^2
310K 523 W/m^2
Solar 1367 W/m^2 at top of atmosphere.


The simple fact is that convection trumps radiation every time. Just hold you hand in front of a working ‘radiator’ and then above it. (This is why radiators are very often placed below windows, in fact).
There is no proof that I am aware of that more CO2 does not in fact cool the atmosphere by convection.


I keep looking at this: air and ground temperature recorded in the N. African desert during a solar eclipse.


Could someone please show me how to use Plank’s Law of Radiation (I use Open Office) – as shown here:
As I keep getting the “ultraviolet catastrophe”.

Andre Bijkerk

This is inline with the real null hypothesis that I posted before.
We read all over internet that the black body temperature of the Earth would have been -18C, but the actual average temperature is +15C; consequently this 33 degrees difference is supposed to be the greenhouse effect. But is this true?
Is the blackbody situation the “null hypothesis”? I don’t think so. The black body calculation assumes a sphere with a constant flux of light energy, uniformely distributed over the surface, using the Stefan Boltzman equation to derive it’s temperature like this.
But the earth is nowhere near a blackbody and if we want to really look at the null hypothesis, we would have to look at an earth without greenhouse effect, but still with an (inert) atmosphere and still rotating in 24 hrs, with seasons and all.
Now instead of using an average steady state solar radiation, we need to realize that we have the diurnal cycle with max insolation radiation at noon and no radiation incoming when the sun is below the horizon. So during daytime the earth surface warms up and much more than the according the average radiation. Equilibrium temperature at the equator in a steady state with the sun in zenith, using the full incoming 1365 w/m2 (albedo 30%) would be 360K or 87C. This follows from applying the Stephan Boltzman equation for the spot directly under the sun, instead of a uniformely distributed radiation.
So this much higher temperature of the earth surface is transmitted via conduction to the lowermost boundary layer of the atmosphere. This heated air gets is less dense, and it becomes buoyant so it rises up; Convection, the very basics of meteorology. So at daytime the atmosphere receives thermal energy of the earth. How can it lose this energy again? Remember we are in the null hypothesis, no radiation, no greenhouse effect, so the inert atmosphere cannot lose the energy by radiation.
Now, at night time the Earth does not receive radiation energy from the sun but it radiates energy out and cools quickly, obviously much more quickly in the null hypothesis even than with the greenhouse effect, which would have directed (“reflects”) some radiation back to earth. Now the cooler earth also cools the boundary layer of the atmosphere by conduction again, however there is no negative convection as the cool air gets more dense and tends to stay put; the inversion; also very basic meteorology. So despite the cooling of the earth, the missing radiation from the atmosphere prevents it from cooling at night and the next day more conducted energy is convected into the atmosphere, that stays there again.
Obviously we have an unbalance. And equilibrium can only be reached, maybe after thousands of years, when the convection at daytime has reduced so much to balance heat loss at night time via conduction back to the surface. For that the lower level atmosphere needs to be at the same temperature / density than the boundary layer would reach due to the conduction of heat from the surface.
Conclusion, in the null hypothesis, without greenhouse effect, the average temperature of the lower atmosphere would be considerably higher than the black body temperature of the surface. How much I don’t know. But the main point is that a certain portian of the temperature difference between black body and actual atmospheric temperature is not due to greenhouse effect but to the inability of the inert atmosphere to cool down by radiation.

Stephen Wilde

Pretty much fits my previous contention that when molecules in the atmosphere absorb more energy then the circulation changes so as to accelerate energy to space faster.
The effect on the energy content of the system being at or near zero but the price to be paid is that circulatory change.
Then the only question is whether the circulation change from human emissions is measurable as compared to the natural changes caused by sun and oceans which gave us the MWP, LIA and current warm period.


I do claim that the argument might be valid, and can only be refuted by an analysis more detailed than simply claiming “Energy that would normally escape into space is absorbed by these molecules, thus heating the atmosphere.”

You do realise that this quote you dug up is from an educational website aimed at schoolchildren, right? And that more detailed analyses have been being published in the scientific literature since Victorian times? Your argument is “not even wrong”. Presumably you don’t have any idea why the Moon is colder than the Earth.

convection is certainly ignored in the “Standard Model”

If it was, then the predicted surface temperature would be about 45&degC. It’s one thing to be ignorant of the science, most people are ignorant of the science. Believing in your own ignorance is another thing.


I think Reed Coray is on the ball.
The IPCC assertion that nitrogen and oxygen are not GHGs appears to conflict with the findings of Tyndall’s physical experiments (1861) which showed that N2 and O2 neither absorb nor radiate heat in the longwave infrared radiation (LIR) spectrum, even though they are transparent to incoming solar shortwave radiation. Modern spectroscopy reveals a total absence of N2 in the LIR, and barely any O2 relative to the atmospheric H2O and C2O, which dominate the infrared spectrum. Yet H2O and CO2 comprise only about 1% of the atmosphere, as against the 99% consisting of N2 and O2. Thus if the former are blanketing the earth, that is an achievement when they comprise so little of the atmosphere – most of us prefer blankets that are close to 100% wool.
Tyndall’s physical laboratory experiments found no evidence for any significant absorption of heat by nitrogen and oxygen in the longwave spectrum, and that meant for him they could not radiate heat to space. His experiments showed that air comprising only [H2O] and [CO2] both absorbed and radiated 15 times as much as air consisting only of N2 and O2:
“Air without [water vapour and CO2] produced an absorption of about 1.
Air direct from the laboratory, containing therefore its carbonic acid [CO2] and aqueous
vapour, produced an absorption [and radiation] of 15”.
(Lecture 1861:28).


Comparison of the Brazilian rainforest and the N. African Desert.,_Amazonas…0.0.l9hxb0L6cLs&oq=Barcelos+amazonas,_Algeria…0.0.PUZtKMJOlKc
For May 2012, Barcelos, Brazil (Lat: 1 South)
Temp: monthly min 20C, monthly max 33C, monthly average 26C
Average humidity 90%
For May 2012, Adrar, Algeria (Lat: 27 North)
Temp: monthly min 9C monthly max 44C, monthly average 30C
Average humidity around 0%


GHG does not block radiation, it absorbs and then re-transmits … a better term than block would be slows …


If the energy input to the Earth system (from the Sun) remained unchanged, then the fingerprint of an enhanced GHE as the culprit of tropospheric warming would be a gradual reduction in OLR at TOA. This is not what we observe. We observe the opposite. This suggests rather that the Earth system is working towards balancing an INCREASED energy input. Which is also observed. But in this case, the increased energy IN (heat gain) is clearly what causes the warming. The increased energy OUT (heat loss) is Earth’s attempt to keep pace.
Where’s the evidence of a greenhouse gas-driven warming?

David Ross

Very interesting post Reed, but I have my doubts (hey, it’s what we do here).
The additional CO2 added to the atmosphere does not significantly increase its volume. A thicker atmosphere would certainly have a greater greenhouse effect (see Venus).
Keeping the volume of the atmosphere constant and increasing the surface area exposed to space (i.e. a bigger earth with thinner atmosphere) would be a more appropriate comparison to adding more metal plates to an engine.
The physical properties of the atmosphere are the sum of its constituent parts. Adding CO2 changes those physical properties. Consequently, I think a better analogy might be: adding CO2 to the atmosphere is like changing the composition of the metal (i.e. an alloy) used in the engine plates. Which is a very different situation from adding more plates.


GHG does not block radiation. They absorb and re-radiates … and they must be cooler than the hot surface to do that …

Sorry Reed, you miss the point, but you are not alone. Eli remembers eminent analytical chemists who missed the same point in print many years ago.
What happens is that effectively GHG block radiation from reaching space across most of the IR.
This includes IR emitted from the GHGs low in most of the troposphere.
Increasing concentrations of GHGs raises the altitude that GHGs can radiate to space in the blocked regions of the spectrum
Because of the lapse rate, the higher you go in the troposphere, the lower the temperature
This slows down the rate at which the Earth emits to space because it is now radiating at higher, colder altitudes
To maintain radiative balance (sun in, IR out) the entire Earth system warms until the temperature rises enough in the mid troposphere to restore the balance.
In somewhat more detail you can read about this at RR and there are links to other, more mathematical explanations there plus Science of Doom probably beat this to death somewhere

Baa Humbug

@ Richard111 says:
July 21, 2012 at 4:01 am
Try this bloke at the link Richard. Only a half dozen posts including a small pdf.
let me know what you think.


Very helpful and clear. I like the focus on radiation because that is the only mechanism of interest to the warmists: they can’t account for convection and it hurts their case badly, so they ignore it. This analysis of radiation suggests that convective processes are, ultimately, like the “blades” of material on a radiator: in both cases, there is a presentation of the warmer material to as much of a cooler material as geometry or fluid transport will allow. A thunderhead is like a giant temporary radiating surface. Of course I am not giving credit to heat transport from phase changes of water in the thunderhead, but ultimately they release thermal energy through radiation to space as well. The role of water vapor in transporting heat to the upper atmosphere where it can dissipate through IR radiation is simply gigantic (see the diagram of absorption/emission spectra: CO2 and other GHGs are dwarfed by H2O).
So much of the warmists’ thinking rests on –dare I say it?– thin air.


This is a most baffling, confused post which conflates many things and adds nothing.


Andre Bijkerk says: July 21, 2012 at 4:34 am
“We read all over internet that the black body temperature of the Earth would have been -18C, but the actual average temperature is +15C; consequently this 33 degrees difference is supposed to be the greenhouse effect. But is this true?”
No, the -18C must be ground temperature while the +15C is ait temperature: comparing apples and orangutans.
“The black body calculation assumes a sphere with a constant flux of light energy, uniformely distributed over the surface, using the Stefan Boltzman equation to derive it’s temperature like this.”
And that doesn’t work because the relationship between radiation intensity and temperature is not linear. Take a blackbody “Moon” with 1400 W/m^2
on one side and 0 W/m^2 on the other side: that give two temperatures – 400K on one side and 0K on the other: average 200K. Now cast 700 w/m^2
on both sides and you get 332K on both sides: average 166K.
So the temperature of 700 + 700 W/m^2 does not equal the temperature of 1400 + 0 W/m^2 .


JeffC says: July 21, 2012 at 5:15 am
“GHG does not block radiation, it absorbs and then re-transmits … a better term than block would be slows …”
I believe JeffC to be correct. CO2 almost instantly re-radiates the outgoing IR radiation it intercepts, with around 50% of this radiated back towards the Earth’s surface. The attenuation of IR passing through CO2 does not mean the CO2 molecules have increased in temperature, but rather they have scattered the IR.
Outgoing IR radiation radiated back to the Earth’s surface could slow the radiative cooling of surface materials. However by physical experiment I have found that the effect is negligible with regard to liquid water that is free to evaporatively cool. If the grey body figure of 1.2c for radiative forcing via a doubling of atmospheric CO2 used by the IPCC is adjusted for the partial exclusion of 71% of the Earth’s surface, climate sensitivity to a doubling of CO2 would appear to be around 0.3c to 0.4c. However even the grey body figure of 1.2c should itself be in doubt given the ability of radiative gasses to radiate as IR any energy they have acquired conductively.


Arrgh! Not this again … back to basics!


@ trcurtin Isn’t the aurora borealis oxgen and nitrogen emitting in the visible spectrum? If they can emit visible light, surely they can absorb visible light.


michael hammer says:
July 21, 2012 at 3:52 am
The “blocking” by GHGs would only exist as long as the stratification is not disturbed in any way. As soon as convection rears its, IMO, beautiful head, you get the wonderful heat pipe effect which transports heat from hotter to colder. Willis has in extensio presented, demonstrated via Argo temperature measurements, and discussed this.
Circulation, even when it’s horizontal, works much like convection, however we have the adiabatic lapse rate causing high pressure areas (descending air mass) with heating, and low pressure areas (rising air mass) with cooling. This might be what confuses some of the people saying we’ve got global warming. Low pressure areas are moving large quantities of heat upwards, particularly so if they’re humid air masses. On the other hand, the high pressure areas are low humidity and therefore reach higher air temperatures, but with lower heat content.
All of this regulates the heat content of the air. It is particularly noticeable here on the Gulf Coast. One is quickly disabused of the idea that air controls water temperature as soon as a south wind kicks in following a north wind. The temperatures rise quickly in the Winter, in the summer they drop and we get that wonderful “feels like” temperature measurement = hot anyway.

Bob Shapiro

A couple of points.
1. Non-GHGases DO emit radiation. Both Oxygen & Nitrogen emit in the UV range, but they do lose some energy to space through radiation.
2. We’re interested in the lower atmosperic temperature (supposed to be 6 feet off the ground for weather stations) rather than earth land/water surface temperatures. I expect this makes some difference to your analysis.

Rober Doyle

A novel, “out of the box” hypothesis. Good post.
Further, data may exist to test it.
My assumption is:
For AGW models and despite atmospheric mixing, there would exist
gravitational bands of concentration for CO2. Otherwise, there is no
greenhouse. This reflecting model should show, due to reflection
a temperature at the earth side of the gravitational band a temperature
labeled x.
If this convection hypothesis is correct, the atmosphere below the gravitational
band would produce a temperature labeled y.
y would be measured to be less than x due to the convection transfer
into the gravitational band. There would be less reflected energy

I may have more to say about this later, but a couple things trouble me:
Since surface area increases with increasing altitude
Area of sphere’s surface is proportional to r^2. Convection reaches up to the tropopause, for the general limit, so for an Earth diameter of 8000 miles and a tropopause of height of 10 miles, the ratio of the increase is 8010^2/8000^2, is 1.0025. What’s the average radiation, some 400 w/m^2? That means an extra watt. Worth including, but not enough.
[1] It’s true that for most air-cooled engines
This bouncing between IC engine in a vacuum and air cooled IC engine gets confusing. The plates, err fins, of an air cooled engine provide more surface area for conduction to air. Fins don’t work as radiators – you need an unobstructed view of the cold surface to work well. That why the radiators on the Space Shuttle were just flat plates on the payload doors. The air cooled engine isn’t helpful to the discussion, the radiatively cooled engine is to far outside most peoples’ experience to be a good analogy. Besides, a lot of the heat from either engine is carried away by exhaust gases. I’d be just as happy with talking about radiation from a hot plate.


There are serious flaws in your analogy. You have your fins made from the same sort of material as the engine. Try thinking of the fins as being made of wood so you wind up with insulation rather than merely additional radiator area.
A parcel of gas at a given pressure and temperature will absorb a fraction of the power being transferred through it, essentially depending upon the ghg concentrations and not so much on its temperature. It will also radiate outward (and radiate inward) a given amount of power that depends upon its temperature and the concentrations of ghgs. If the gas is the same temperature as a surface, then the emission rate of the gas outward will be equal to the absorption rate of the original surface radiation continuum going through it. If the gas is hotter, then there will be emission lines superimposed upon the continuum energy and if the gas is cooler, there will be absorption lines.
If you add more ghgs to the parcel, there will be a stronger absorption from the continuum but there will also be a greater emission occurring for a given temperature. The gas temperature will adjust to balance the emitted and absorbed energy and that includes radiation, conduction, and convection energies. The same thing goes for adjacent or near by parcels of gas except they are emitting energy in a spectrum. Ultimately, it takes less of a T increase to radiate the same amount of power than it would if one does not take the added emissivity into account but the added absorption does require some T increase.
The top of the troposphere is the tropopause and this is where convection stops being an extremely important part of the heat transfer. Below this, convection is very significant.
The whole problem with looking just at this is one misses the bigger picture. around 62% of the Earth’s surface is covered by clouds and a goodly portion of this is on the sunlit side as a lot of the cloud cover tends to form because of sunlight and tends to dissipate at night. Clouds and particulates provide a continuum and it radiates at a continuum rate for the temperature of the material. Also, clouds and scattering and particulates provide about 80% of the Earth’s albedo which is the dominant factor in determining Earth’s temperature. It also impacts the outgoing radiation significantly, blocking surface radiation and substituting it’s own characteristic radiation temperature.
Ultimately, the ghg contributions as you are looking at and as explained and discussed everywhere is only for clear skies and when clouds are involved, it changes quite a bit.
Based on people like hansen who are at the bottom of this CAGW pyramid, one can see serious flaws and nonphysical ideas and concepts being peddled which lead to erroneous results. An example is the characteristic radiating altitude that supposedly changes with ghg concentrations which has nothing to do with reality and hence cannot provide conclusions for changes.


Reed: Eli Rabett ( ) and Arthur ( ) hit the nail on the head. The explanation that you critique is not even the best of the non-mathematical simple explanations for the greenhouse effect. Furthermore, underlying this explanation are radiative transfer calculations that show that you are wrong and the accepted explanation is correct. (The actual surface temperature of the Earth compared to the highest surface temperature it could have in the absence of an IR-absorbing atmosphere and still obey conservation of energy is also evidence that the natural greenhouse effect warms the Earth.)
And, these radiative transfer calculations are well-verified by, among other things, the modern field of remote sensing. If you want to be consistent in your skepticism, you would also have to disbelieve all of Spencer and Christy’s work reconstructing tropospheric temperatures from satellite remote sensing along with much of the other work in remote sensing (perhaps even the weather satellites).
Jerome says:

The simple fact is that convection trumps radiation every time. Just hold you hand in front of a working ‘radiator’ and then above it. (This is why radiators are very often placed below windows, in fact).
There is no proof that I am aware of that more CO2 does not in fact cool the atmosphere by convection.

The simplest way to understand why saying the word “convection” doesn’t magically slay the greenhouse effect is to understand that convection only equalizes temperatures to a point. In particular, convection drives the lapse rate in the atmosphere down to the adiabatic lapse rate but not further. This is why convection, while reducing the natural greenhouse effect from what it would be in its absence (by close to a factor of two as I recall) , does not eliminate it. It also explains why Nikolov and Zeller in their very mistaken work ( ) were able to demonstrate that convection added to a simple model eliminated the greenhouse effect: They mistakenly added convection to the model in a way that drove the lapse rate to zero. It is easy to take the model, add convection in a more correct manner and show how it doesn’t eliminate the greenhouse effect.

Now it’s not impossible that since (a) in addition to radiation, heat is transferred from the Earth’s surface to greenhouse gases via conduction, and (b) convection currents (i) circulate the heated greenhouse gases to higher altitudes where energy transfer to space can take place and (ii) return cooler greenhouse gases to the Earth’s surface, that the process of heat transfer away from the Earth’s surface via greenhouse gases is more efficient than simple radiation from the Earth’s surface.
Not only not impossible, but true! And recognized even by the Evil Dr. Hansen in papers dating all the way back to the late 80s! It is one of the processes moderated by moisture, especially, since water vapor carries enthalpy aloft, such that the dry air adiabatic lapse rate is quite different from the moist air adiabatic lapse rate. Assuming that moist/humid air is dominant — even all the way back in Hansen’s earliest papers — leads to a much more moderate total greenhouse forcing on a doubling of CO_2.
The point is that it is really not smart to assume that climate scientists — even ones that have perhaps gotten carried away with “save the world” passion to the point where they are no longer sufficiently objective to avoid the demon of confirmation bias — are stupid, or that they don’t use the underlying physics in their models or computations. Sure, you might discover something that they’ve forgotten. And they could be wrong about what processes ARE dominant lots of ways, because nonlinear models can often “work” around multiple clusters of “reasonable” parameters describing e.g. water vapour feedback especially when we perhaps lack sufficient data to properly constrain the models. But a study of the actual physics of the greenhouse effect being used in the models will suffice to show that your idea, while correct, is hardly original, nor is it an effect being omitted altogether from the GHE-warming arguments.

Leonard Weinstein

No! Radiation absorption and re-radiation, along with convection and evapotranporation and condensation transfer the heat absorbed by the Sun to a sufficient altitude to radiate to space. If the lapse rate and albedo do not change, the only way greenhouse gases increase the temperature is by raising the average altitude of outgoing radiation to space. It increases temperature at all altitudes in the Troposphere by shifting the entire temperature profile a small amount.

Paul Linsay

“Now it’s not impossible that since (a) in addition to radiation, heat is transferred from the Earth’s surface to greenhouse gases via conduction, and (b) convection currents (i) circulate the heated greenhouse gases to higher altitudes where energy transfer to space can take place and (ii) return cooler greenhouse gases to the Earth’s surface, that the process of heat transfer away from the Earth’s surface via greenhouse gases is more efficient than simple radiation from the Earth’s surface. ”
Yes, someone finally gets it. A simple calculation shows that a parcel of air that is 1K warmer than its surroundings and rising at 1 m/sec, walking speed, carries energy aloft at a rate of 1kW/m^2. That energy eventually rises to the top of the atmosphere where it is radiated away by H2O and CO2. Increasing CO2 will increase the top of the atmoshpere radiative cooling ability proportionatly while its heating effects at the ground only increase logarithmicly with concentration.

Gary Palmgren

This is a good argument but is limited by thinking of the atmosphere as one entity. A discussion about heat transport in the atmosphere very much needs to include the very different troposphere and stratosphere.
In the troposphere heat is mostly conducted by convection and the associated water evaporation and condensation. Temperature drops with altitude so warm air will rise, carrying massive amounts of heat as long as the adiabatic cooling does not reduce the temperature below that of the surrounding air. Condensing water vapor releases heat to keep a rising bubble of air warmer than otherwise and we get thunderstorms and other weather. At the tropopause, adiabatic cooling has reduced the temperature to about -55°C. Above the tropopause is the stratosphere.
The defining characteristic of the stratosphere is that temperature rises with altitude. Vertical convection effectively ceases. This is where heat transport by radiation finally becomes dominant. The dew point is very low throughout the stratosphere as it is controlled by the dew point at the tropopause.
A discussion on the effect of IR absorbing gases such as water vapor and CO2 needs to be broken down into two parts. What is the effect on convection in the troposphere? What is the effect on radiation in the stratosphere?
At a high enough altitude (40-50 km), CO2 is effectively cooling as it can emit radiation directly into deep space. Water vapor concentration is less than CO2 when the dew point reaches -55°C.
The effect on adding CO2 in the troposphere is going to be determined by how it interacts with water vapor and changes the convection. The shear complexity of this suggests CO2 will have no effect as the atmosphere will adapt to keep entropy generation at a maximum. More paths for energy transport will reduce the effect of a constraints on a given path so complexity ensures maximum entropy generation. The CAGW believers tell us that CO2 is a constraint on radiation transport of heat but this totally ignores the dominant convection and the likely hood that CO2 will simultaneously increase convection for no net effect.

Dave in Delaware

Convective heat transfer is enhanced by combined radiant + convective transfer.
For Convection, heat transfer from a surface to the air above it, the rate of heat transfer is dependent on the velocity of the air. For Natural Convection (no wind) the transfer rate is lower than for Forced Convection (windy) conditions, all other things being equal. There is a ‘boundary condition’ just above the surface where the air is barely moving, even on a windy day, and that ‘boundary’ limits the rate of convective heat transfer from the surface.
Introduce GHGs to the air above. Energy is also radiated from the surface, some photons are captured by the GHGs, and we are told, they are almost immediately ‘thermalized’, which is to say, the radiant energy is converted to thermal energy in the air. The warmer air directly above the surface now rises, increasing the Natural convection. The overall energy transfer, convection + radiant, is enhanced by the addition of GHGs. Its like the GHGs help the heat ‘jump’ the boundary layer at the surface. Rather than blocking the heat transfer, the GHGs actually would appear to speed up the rate of transfer, helping to move energy more quickly to the upper atmosphere, where it radiates out to space.

David Chamness

Reed, you’re post is very close to what I’ve been saying for years – but keep getting told I’m an idiot for saying it. CO2 both absorbs and radiates IR. I may be wrong, but I think it goes like this: if an unexcited molecule of CO2 is hit by a photon of the correct frequency, it will absorb the energy and move into an excited state. It will then re-release that photon in some random direction when moving back to an unexcited state.
Now, my problem is this: most of the atmosphere cannot absorb radiation from longwave IR. So as far as I can tell, it’s impossible for the re-radiated IR to heat the atmosphere directly. It can bounce around from CO2 molecule to CO2 molecule for 10 years and never “heat” the nitrogen and oxygen that makes up 98% of the air.
The question I have is this: is there another way to excite the CO2 molecule into radiating? Will conduction heat the CO2 molecule enough to cause it to radiate? If so, then would not the warm Earth impart heat to nitrogen and oxygen by conduction, which then can transfer that heat to CO2 via conduction, which will cause the CO2 molecule to fire off a photon of longwave IR on it’s own, without having absorbed one from the surface of the Earth?
Since more than 90% of the atmosphere is non-radiative in longwave infrared and does not absorb energy via radiation, most of the heat in the atmosphere is transfered there via conduction. It seems to me that a warm atmosphere will cause CO2 and H20 molecules to radiate away the atmospheric heat to space at increasingly efficient rates. The warmer the air gets, the faster those two gases radiate, regardless of what the surface of the Earth is doing, right?

David L. Hagen

Reed Coray
Re: “Energy that would normally escape into space is absorbed by these molecules, thus heating the atmosphere.”
That popular description would best be improved by noting that:
“Energy that would normally be radiated directly into space is BOTH absorbed AND then RERADIATED by greenhouse gas molecules IN ALL DIRECTIONS. This changes the atmospheric “lapse rate”.”
For a quantitative thermodynamically sound model of the atmospheric lapse rate see:
Robert E. Essenhigh (2006) Prediction of the Standard Atmosphere Profiles of Temperature, Pressure, and Density with Height for the Lower Atmosphere by Solution of the (S−S) Integral Equations of Transfer and Evaluation of the Potential for Profile Perturbation by Combustion Emissions.
As a “preprint” see his similar Paper No.03F-44: Western States Section Combustion Institute Meeting: Fall (October) 2003
Re: “Tyndall’s physical laboratory experiments found no evidence for any significant absorption of heat by nitrogen and oxygen in the longwave spectrum.”
While small, the greenhouse contributions of O2 and 2 are NOT negligible. The greenhouse effects of O2 and are quantified by Hopfner et al.(2012) The natural greenhouse effect of atmospheric oxygen (O2) and nitrogen (N2) Geophysical Research Letters in press 2012GL051409

We have found that on global average under clear-sky conditions the OLR is reduced due to O2 by 0.11 Wm2 and due to N2 by 0.17 Wm2. Together this amounts to 15% of the OLR-reduction caused by CH4 at present atmospheric concentrations. Over Antarctica the combined effect of O2 and N2 increases on average to about 38% of CH4 with single values reaching up to 80%.

Willie Soon discussed this fact that O2 and N2 not negligible for greenhouse gas emissions. ICCC 7, 2012 ~ 11:30 Tuesday 22 May 2012. See also other Heartland ICCC 7 presentations


This post is utter nonsense. Why use the analogy of an air cooled engine when an air cooled radiator works by conduction/convection and the ‘whole earth’ temperature is close to a solid sphere in a vacuum problem.
Anyone who thinks they have understood this post should realise that deductive reasoning in this style can tie you up in an ugly mess. Garbage in, garbage out.
It may well be that the classical greenhouse gas model is mostly wrong – this post is completely orthogonal to that question.

Mike McMillan

Not buying it.
Adding a second blanket at night increases thermal mass and thus makes me cooler?