With apologies to my many readers who understand the basics of heat transfer in the atmosphere and really want to hear more about feedback, uncertainty, real science..
Clearing up basic misconceptions is also necessary. It turns out that many people read this blog and comment on it elsewhere and a common claim about climate science generally (and about this site) is that climate science (and this site) doesn’t understand/ignores convection.
The Anti-World Where Convection Is Misunderstood
Suppose – for a minute – that convection was a totally misunderstood subject. Suppose basic results from convective heat transfer were ridiculed and many dodgy papers were written that claimed that convection moved 1/10 of the heat from the surface or 100x the heat from the surface. Suppose as well that everyone was pretty much “on the money” on radiation because it was taught from kindergarten up.
It would be a strange world – although no stranger than the one we live in where many champions of convection decry the sad state of climate science because it ignores convection, and anyway doesn’t understand radiation..
In this strange world, people like myself would open up shop writing about convection, picking up on misconceptions from readers and other blogs, and generally trying to explain what convection was all about.
No doubt, in that strange world, commenters and bloggers would decry the resulting over-emphasis on convection..
First Misconception – Radiation Results are All Wrong Because Convection Dominates
There are three mechanisms of heat transfer:
- radiation
- convection
- conduction
Often in climate science, people add:
- latent heat
In more general heat transfer this last one is often included within convection, which is the movement of heat by mass transfer. Although sometimes in general heat transfer, heat transfer via “phase change” of a substance is separately treated – it’s not important where “the lines are drawn”.
Update note from Dec 9th – I leave my poorly worded introduction above so that readers comments make sense. But I should have written
In fact in atmospheric physics we almost always see the breakdown like this:
-radiation
-latent heat
-sensible heat
Latent heat being the movement of heat via evaporation – convection – condensation. Sensible heat being the movement of heat via convection with no phase change. Conduction is actually also included in sensible heat, but is negligible in atmospheric physics.
Therefore when convection is written about it is both sensible and latent heat. That is, heat transfer in the atmosphere is via either convection or via radiation.
End of update note
Let’s look at conduction, safe from criticism because it is largely irrelevant as a form of heat transfer within the atmosphere. Conduction is also the easiest to understand and closest to people’s everyday knowledge.
The basic equation of heat conduction is:
q =- kA . ΔT/Δx
where ΔT is the temperature difference, Δx is the thickness of the material, A is the area, k is the conductivity (the property of the material) and q is the heat flow. (See Heat Transfer Basics – Part Zero for more on this subject).
Notice the terms in this equation:
- the material property (k)
- the thickness of the material (Δx)
- the temperature difference (ΔT)
- the area (A)
Where are the convective and the radiative terms?
Interestingly, conduction is independent of convection and radiation. This is a very important point to understand – but it is also easy to misunderstand if you aren’t used to this concept.
It doesn’t mean that we can calculate a change in equilibrium condition – or a dynamic result – only using one mechanism of heat transfer.
Let’s suppose we have a problem where we know the temperature at time = 0 for two surfaces. We know the heating conditions at both surfaces (for example, zero heat input). We want to know how the temperature changes with time, and we want to know the final equilibrium condition.
The way this problem is solved is usually numerical. This means that we have to work out the heat flow from each mechanism (conduction, convection, radiation) for a small time step, calculate the resulting change in temperature, and then go through the next time step using the new temperatures.
For many people, this is probably a fuzzy concept and, unfortunately, I can’t think of an easy analogy that will crystallize it.
But what it means in simple terms is that each heat transfer mechanism works independently, but each affects the other mechanisms via the temperature change (if I come up with a useful analogy or example, I will post it as a comment).
So if, for example, convection has changed the temperature profile of the atmosphere to something that would not happen without convection – the calculation of conduction through the atmosphere is still:
q = -kA . ΔT/Δx
And likewise the more complex equation of radiative transfer (see Theory and Experiment – Atmospheric Radiation) will also rely on the temperature profile established from convection.
So – an ocean surface with an emissivity of 0.99 and a surface temperature of 15°C will still radiate 386 W/m², regardless of whether the convection + latent heat term = zero or 10 W/m² or 100 W/m² or 500 W/m².
Second Misconception – Atmospheric Physics Ignores Convection
This is a common claim. It’s simple to demonstrate that the claim is not true.
Let’s take a look at a few atmospheric physics text books.
From Elementary Climate Physics, Prof F.W. Taylor, Oxford University Press (2005):
From Handbook of Atmospheric Science, Hewitt & Jackson (2003):
From An introduction to atmospheric physics, David Andrews, Cambridge University Press (2000):
In fact, you will find some kind of derivation like this in almost every atmospheric physics textbook.
Also note that it is nothing new – from Atmospheres, by R.M. Goody & J.C.G. Walker (1972):
Both convection and radiation are important in heat transfer in the troposphere.
Lindzen (1990) said:
The surface of the earth does not cool primarily by infrared radiation. It cools mainly through evaporation. Most of the evaporated moisture ends up in convective clouds.. where the moisture condenses into rain..
..It is worth noting that, in the absence of convection, pure greenhouse warming would lead to a globally averaged surface temperature of 72°C given current conditions
Note the important point that convection acts to reduce the surface temperature. If radiation was the dominant mechanism for heat transfer the surface temperature would be much higher.
Convection lowers the surface temperature. However, it only acts to reduce the effect of the inappropriately-named “greenhouse” gases. And convection can’t move heat into space, only radiation can do that, which is why radiation is extremely important.
The idea that climate science ignores or misunderstands convection is a myth. This is something you can easily demonstrate for yourself by checking the articles that claim it.
Where is their proof?
Do they cite atmospheric physics textbooks? Do they cite formative papers that explained the temperature profile in the lower atmosphere?
No. Ignorance is bliss..
Third Misconception – Convection is the Explanation for the “33°C Greenhouse Effect”
Perhaps in a later article I might explain this in more detail. It is already covered to some extent in On Missing the Point by Chilingar et al (2008).
As a sample of the basic misunderstanding involved in this claim, take a look at Politics and the Greenhouse Effect by Hans Jelbring, which includes a section Atmospheric Temperature Distribution in a Gravitational Field by William C. Gilbert.
If you read the first section by Jelbring (ignoring the snipes) it is nothing different from what you find in an atmospheric physics textbook. No one in atmospheric physics disputes the adiabatic lapse rate, or its derivation, or its total lack of dependence on radiation.
Clearly, however, Jelbring hasn’t got very far in atmospheric physics text books, otherwise he would know that his statement (updated Dec 9th with longer quotation on request):
T is proportional to P and P is known from observation to decrease with increasing altitude. It follows that the average T has to decrease with altitude. This decrease from the surface to the average infrared emission altitude around 4000 m is 33 oC. It will be about the same even if we increase greenhouse gases by 100%.
– was very incomplete. How is it possible not to know the most important point about the inappropriately-named “greenhouse” effect with a PhD in Climatology? Or even no PhD and just a slight interest in the field?
What determines the average emission altitude?
The “opacity” of the atmosphere. See The Earth’s Energy Budget – Part Three. Clearly Jelbring doesn’t know about it, otherwise he would have brought it up – and explained his theory of how doubling CO2 doesn’t change the opacity of the atmosphere – or the average altitude of radiative cooling to space.
Gilbert adds in his section:
I was immediately amazed at the paltry level of scientific competence that I found, especially in the basic areas of heat and mass transfer. Even the relatively simple analysis of atmospheric temperature distributions were misunderstood completely.
Where is Gilbert’s evidence for his amazing claim?
Gilbert also derives the equation for the lapse rate and comments:
It is remarkable that this very simple derivation is totally ignored in the field of Climate Science simply because it refutes the radiation heat transfer model as the dominant cause of the GE. Hence, that community is relying on an inadequate model to blame CO2 and innocent citizens for global warming in order to generate funding and to gain attention. If this is what“science” has become today, I, as a scientist, am ashamed.
I’m amazed. Hopefully, everyone reading this article is amazed.
The derivation of the lapse rate is in every single atmospheric physics textbook. And no one believes that radiative heat transfer determines the lapse rate.
And the important point – the Climate Science 101 point – is that the altitude of the radiative cooling to space is affected by the concentration of “greenhouse” gases.
Actually understanding a subject is a pre-requisite for “debunking” it.
Conclusion
Many people read blog articles and comments on blog articles and then repeat them elsewhere.
That doesn’t make them true.
Science is about what can be tested.
What would be a worthwhile “debunking” is for someone to take a well-established atmospheric physics textbook and point out all the mistakes. If they can find any.
It would be more valuable than just “making stuff up”.
References
Elementary Climate Physics, Prof F.W. Taylor, Oxford University Press (2005)
Handbook of Atmospheric Science, Hewitt & Jackson, Blackwell (2003)
An introduction to atmospheric physics, David Andrews, Cambridge University Press (2000)
Atmospheres, R.M. Goody & J.C.G. Walker, Prentice-Hall (1972)
Some Coolness Regarding Global Warming, Lindzen, Bulletin of the American Meteorological Society (1990)
You’re right, of course, that atmospheric scientists are very conscious of convection.
Lay people often think by analogy with what they see on the scale of a fireplace, or room heating, for example. Here the temperature gradients are very large – in ° per metre, and the air can be considered neutrally stable, so convection cells are easily set up.
But on the scale of the atmosphere, there is convective stability at any lapse rate less than 0.01 K/m (<10 K/km, as it is most places, most times), absent condensation. That means there is an energy drain on natural convection, at a rate proportional to the extent the lapse rate is below 10 K/km.
That doesn’t mean convection is impossible. Large structures like Hadley cells work as heat engines, transferring air from hot to cold, which overcomes the stability energy drag. But it does limit its ability (when dry) to be a major mode of vertical transport. It usually needs some surface inhomogeneity (thermals) to get something moving.
Im unsure what to make of your comment Nick…
As i understand it, convection is the dominant means of energy transport in the lower troposphere, but with altitude, and reducing pressure, shortening path length, radiation becomes more and more dominant at energy transfer, until it completely takes over and removes it from the troposphere.
Now i think you are saying that dry convection isnt a very efficient mover of energy? Due to reduced path length? And the fact that with altitude, water isnt condensing releasing latent heat into the parcel, keeping the parcel warmer than its relative altitude?
Its really this sentence
” That means there is an energy drain on natural convection, at a rate proportional to the extent the lapse rate is below 10 K/km.”
That im not understanding.
Mike,
I’m just saying that there isn’t a huge amount of buoyancy-driven convection, and, yes, it isn’t transporting a big fraction of heat.
The “energy drain” refers to the convective stability. Suppose you have some air which has been heated, perhaps over a hot bit of ground, to 1C higher than ambient. Then it is buoyant and rises, cooling by 10K every km of rise as it goes.
If the lapse rate is only 5 K/km, then within 200 m it is back to ambient temperature, but has KE. So it keeps going, but cools more, becoming heavier than other air at that level. So it loses KE. Work is done to slow it down.
The same happens in reverse for cool air falling. Again it warms to ambient, and thereafter is slowed and loses KE.
If the lapse rate had been 9 K/km, then the rising air would have travelled 1 km before reaching ambient. That’s why the effect is proportional to the different between the actual lapse rate and the adiabat.
If the lapse rate had been 11 K/km, then warm air rising will actually get more buoyant as it rises. That is the convective instability regime. It is self-correcting, because the upward transport of heat reduces the lapse rate.
Nick Stokes @ December 8, 2010 at 5:45 am says
“I’m just saying that there isn’t a huge amount of buoyancy-driven convection, and, yes, it isn’t transporting a big fraction of heat.”
In SoD’s article, he takes this extract from Lindzin 1990
“..It is worth noting that, in the absence of convection, pure greenhouse warming would lead to a globally averaged surface temperature of 72°C given current conditions”
And again immediately above Lindzins quote, in the last page on display from “atmospheres” it is stated, that convection is responsible for lowering the lower atmosphere T by 60K.
The rate of convection, is going to be heavily reliant on the opacity at a given altitude. If the energy being radiated up from the surface is trapped in lower altitude air, it will warm until it does exceed the lapse rate, until it rises to a level where the energy can be moved via radiation… The obvious, very visual example would be a heat storm, when warm “humid” air rushes up to form a thunder head. The rate at which it rises is easily observable. Or the top of cumulus clouds as they bubble from the warm moist air hitting a colder layer.
I dont think you are correct in believing, that convection isnt transporting a big fraction of heat. Warm air, is generally moist air, air looses energy as it rises, performing work displacing the air above it, the more energy contained in a parcel of air, the higher it can convect.
Mike,
I’m happy with Trenberth’s (2008) numbers – 17 W/m2 thermals, 80 W/m2 evapotranspiration. Lindzen is probably lumping them both in as convection.
Yes, and why not? How else would you describe the physical process by which water vapour rises through its denser surroundings?
I regard the obscure wording of the NASA and the Kiehl Trenberth energy budget diagrams as a bit of a travesty.
Nick thermals = sensible heat transport, evapotranspiration = latent heat transport. Both are related to convective, not radiative, transport of heat energy.
Yes or no?
I am not sure I understand that bit :
“Convection lowers the surface temperature. However, it only acts to reduce the effect of the inappropriately-named “greenhouse” gases.”
My first intuition would be to think that convection will decrease surface temperature whether there are “GH” gases or not.
If the ground is heated up by short wave radiation and then some of that heat is transfered to the air directly above it (through conduction as the atmosphere would be mostly transparent to radiation), buoyancy of that parcel of air is increased and you have got convection going, don’t you?
Once that hot parcel of air is moved, it is replaced by colder air thus enhancing conduction and “pumping” heat away from the ground. I would be inclined to think that the result would be faster cooling of the ground and ground level atmosphere.
What I am not able to conceptualize yet is the equilibrium/regime state of that process. I guess if the atmosphere can not radiate, air brought to higher altitude through convection will not be able to cool down to restart the cycle.
Is that what you meant with that sentence? That in the case of a non-radiative atmoshere, convection won’t matter because what ultimately drives the cooling of the climate (toward space) is radiation and if the atmosphere does not radiate, convection can do whatever it wants to the temperature profile of the atmosphere, it will not affect the climate? Hence, convection can only play a role in counter-balancing the “GH” effect by bringing hot air to an altitude where it can more easily radiate to space? (allowing heat to “bypass” part of the optical thickness of the atmosphere)
(sorry for the lengthy and somewhat confused post, I guess I worked my way through the problem while writing. I’ll post it anyway in case I went completely wrong and someone can point it out.)
I think the simple point SoD was making is that the radiative+convection lapse rate (in the Troposphere) is steeper than would be the case for a radiative only Tropospheric lapse rate.
Thus convection ‘reduces’ the effect of GHGs.
At least that is my interpretation.
I think, its more that, ghg’s and convection both work to move energy out of the system. But the bottom line is, that the height that the energy is able to leave the atmosphere, will effect the T profile down through the troposphere. Just by the need to maintain the T differentials up through the various layers to enable the movement of energy up to this altitude.
So if you raise the height that radiation escapes, that new altitude will need to warm till it is radiating out the energy coming into the system, and the next layer down will need to warm enough that it can transport the energy to this layer, and so on all the way down to the surface.
From reading the above exchange between Mike and Nick one can see how the impression has got around that some climate scientists ignore or play down the role of convection in the atmosphere. As I understand it, Nick was excluding the buoyancy-driven transfer of water vapour from his usage of the term ‘convection’, while Mike was including it in his.
Doom has already in this post drawn attention to this potentially confusing variation in usage where he writes:
“Often in climate science, people add:
latent heat
In more general heat transfer this last one is often included within convection, which is the movement of heat by mass transfer.”
Doom adds that “…it’s not important where “the lines are drawn”.” It may not be important, but lack of clarity as to the assumed position of the lines can (and from the evidence in this thread does) sometimes cause confusion.
Indeed.
Either water vapour is part of the convection or S.o.D needs an extra category in his list of ways energy gets shifted off the surface of the planet:
4) Latent heat
Evapo-Transpiration
Or whatever.
The night/day differences in heating, along with latitude differences, planet rotation and even tilt and path variation of the planet cause large air movement. Add to this the buoyancy due to ground heating, and there is no possibility for a case of no convection. The points being made on what it means are a waste of time.
Which case for no convection? What points are the waste of time?
Sorry if I’m missing something blindingly obvious, but it’s not clear to me what you’re referring to.
Ned,
I was referring to the discussion around the statement:
“..It is worth noting that, in the absence of convection, pure greenhouse warming would lead to a globally averaged surface temperature of 72°C given current conditions”
OK, I’m still not completely clear on what you’re saying (Lindzen is wasting people’s time? Nick and Mike are wasting their time? All of the above?)
… but it’s not important. Don’t feel you need to waste your time answering this.
🙂
I’d be so bold as to go even further and say that if we didn’t have day/night differences, latitudes, planet rotation, tilt, path variation (or in other words we would be floating on 3 elephants standing on a turtle), we’d still be having convection due to gas being gas (mostly vacuum with occasional particle mixed in). It goes to places by itself very well.
Some of the confusion about the relative importance of radiative cooling, convection and latent heat comes from not defining whether we are talking about:
a) TOTAL Upward LW radiation originating at the surface (390 W/m^2 average from W = oT^4) OR
b) NET upward energy flux by LW radiation (390 up – 324 down = 66 W/m^2).
Given that 83% of the energy initially directed upwards gets absorbed and returned, radiative cooling is a very inefficient process near the surface of the earth. Of the 66 W/m^2 that escapes as net radiation, 40 W/m^2 goes through the atmospheric window directly to space. Since GHGs absorb and emit at the same wavelengths, once LW energy is absorbed by the atmosphere, that energy can not be emitted through the atmospheric window and must be emitted at wavelengths where the mean free path of photons is relatively short. So the best way “out” is to be absorbed by the ground and then emitted through the atmospheric window. The atmospheric window handles a net 40 W/m^2 from the surface, while another 24 W/m^2 eventually manages to escape via wavelengths absorbed by GHGs. Interestingly, the approximately 4 W/m^2 of radiative forcing calculated for 2X CO2 amounts to decreasing this 24 W/m^2 by about 1/6; a figure that is compatible with some estimates of the relative importance of CO2 in the greenhouse effect. However, the figure of 24 W/m^2 is what is left over after subtracting some very large numbers (390-326-40), so there should be a large amount of uncertainty in this residual 24 W/m^2.
The amount of energy transfered upward as latent heat (78 W/m^2) is slightly larger than the NET flux of energy by radiation (66 W/m^2). In a previous comment, I showed how a figure close 78 W/m^2 is easily calculated from the average annual precipitation, which is 1 m. (11/11/10 at 6:31 pm https://scienceofdoom.com/2010/11/05/the-three-body-problem/).
Doesn’t anyone know whether energy transfer by thermals was determined by observations or the bookkeeping needed to balance surface energy fluxes? (My guess is bookkeeping.) Since rain falls from relatively low altitudes, thermals may become more important than latent heat higher in the troposphere. Thunderstorms and deep convection appear to represent thermals taking over from latent heat as a major mechanism of upward energy transfer.
SOD,
Thanks for mentioning the document “Politics and the Greenhouse Effect” written by Hans Jelbring and myself about a year and a half ago. As you already know (because I told you in a previous thread), that document was written as a political document and was given to select Swedish parliamentarians prior to a government hearing in Sweden. It was not meant to be a scientific document. I’m not sure how it got published on the internet. But even though it is a political document (and pretty tame by recent standards) the science in it is correct.
In the document you quote me as saying: “If this is what “science” has become today, I, as a scientist, am ashamed.” That statement was true then and it is even truer today. But my shame is nothing compared to what you should be feeling right now. I was very surprised to see you take Dr. Jelbring’s quote out of context and twist its meaning. I have been following your blog for some time now and have found it to be based on sound science. That is why I was very surprised and disappointed when you quoted Dr. Jelbring as follows:
You sarcastically implied that Dr. Jelbring did not understand the role of optical depth with regards to the effective emission altitude. You stated that his knowledge was “incomplete”. But, as you know, in the original document the full statement is as follows:
No, Dr. Jelbring’s knowledge is not incomplete, but your quote is very incomplete and misleading. He was obviously talking about the temperature profile, not the physics of GHG emission. His statement is very correct. Greenhouse gases have nothing to do with the gravity induced lapse rate observed in planetary atmospheres. Your further insults to Dr. Jelbring are uncalled for and completely untrue. I advise your readers to read the document and decide for themselves whether or not Dr. Jelbring understands atmospheric physics. I would further suggest that you brush up on your atmospheric physics and read Dr. Jelbring’s excellent E&E paper The “Greenhouse Effect” as a Function of Mass. It can be found here: http://ruby.fgcu.edu/courses/twimberley/EnviroPhilo/FunctionOfMass.pdf. Why don’t you write a post based around this paper. It would be very, very interesting and informative. I must repeat your statement in this blog thread: “Actually understanding a subject is a pre-requisite for “debunking” it”.
In the meantime I think a retraction and correction to this post is in order.
As to your mockery of my statements in the paper, I am sorry that you do not appreciate a little sarcasm in a political document. You seem to have little problem using sarcasm in this blog which is supposed to be based on Science and hopefully is not intended to be political. But the science behind my statements is completely correct. The “Greenhouse Effect” is a result of the gravity induced temperature lapse rate. An isothermal atmosphere would not have a “Greenhouse Effect”. The climate scientists who continuously use the flawed radiation transfer paradigm and talk about “backradiation” as the cause of higher surface temperatures should also be ashamed. Backradiation is a secondary effect and has nothing to do with the surface temperature. The surface temperature is solely a function of the incoming SW radiation, the atmospheric lapse rate and the effective emission temperature resulting from that temperature gradient. I think even the Swedish parliamentarians understood that basic fact.
Williamcg,
Regarding the quote,
“This decrease from the surface to the average infrared emission altitude around 4000 m is 33 oC. It will be about the same even if we increase greenhouse gases by 100%.”
Which part(s) will be the same? The average emission altitude, the lapse rate, or the temperature difference between the surface and the average emission altitude?
It’s possible that different meanings exist between author and reader. As I understand it, the lapse rate would be the same, but the mean altitude of emission would be different. Hence, a different temperature at the surface.
Regarding your statement, “The “Greenhouse Effect” is a result of the gravity induced temperature lapse rate.” Gravity is a function of mass; the greenhouse effect is a function of absorption/emission. We are not significantly changing the mass of the atmosphere (that I am aware of, some, but not much), but we are changing its radiative properties. I don’t see how you can claim that changing the radiative properties of a body has no effect on the thermodynamic properties. If you have two planets, both with atmospheres of equal mass, but with different radiative properties, would you expect the surface temperature to be the same for both? Or, perhaps more accurately, would you expect the heat content of the atmosphere, land, and oceans to be the same for both?
williamcg,
The atmosphere would indeed have a lapse rate in the absence of IR active gases in the atmosphere. But the surface temperature would not be the same as it is now. Instead of a narrow atmospheric window, all surface radiation would go directly to space. The average temperature would be a lot lower because the ground would be seeing a brightness temperature near 0 K. Since the atmosphere couldn’t lose energy to space, convection wouldn’t matter either except to lower the difference in temperature between the equator and the poles. That actually raises the average temperature because of the T^4 dependence of emission.
You are correct that the greenhouse effect depends on the temperature decreasing with altitude. But the lapse rate alone does not determine the surface temperature. The lapse rate combined with a non-transparent atmosphere means radiation near the surface is higher than radiation at high altitude. Higher radiation equates to higher brightness temperature and higher surface temperature. The brightness temperature of the planet observed from space is ~255 K (~240 W/m2 emission) C. The average surface temperature is 288 K or a difference of 33 K Note that these numbers are measured, not modeled.
The brightness temperature of the atmosphere with greenhouse gases is about 275 K (324 W/m2). That’s rather a lot higher than 0 K. The net radiation is proportional to the difference between the fourth power of the temperatures of the two surfaces so a high atmospheric brightness temperature means a higher surface temperature for the same net radiation.
If you don’t believe the atmosphere has a brightness temperature, you can buy an IR thermometer and measure it. It’s probably too cold now, but it’s easy to do in the summer.
DeWitt,
I agree with everything you said in your first post above. It fits with my statement “The surface temperature is solely a function of the incoming SW radiation, the atmospheric lapse rate and the effective emission temperature resulting from that temperature gradient”. If the effective emission altitude is at the surface then the effective emission temperature would be about 255K.
I’m not sure what your next two posts are trying to tell me. I do not deny the existence of backradiation, but it is not the determinant of the surface temperature – it’s quantity is a result of the surface temperature.
DeWitt is better at this than I am, but from a simple perspective, your statement that “…but it is not the determinant of the surface temperature…” somewhat implies that the surface somehow ignores the energy that it is receiving from the atmosphere. If it absorbs energy that it would otherwise not, would it not rise to a higher temperature?
Back radiation is, of course, not the sole determinant of surface temperature. Neither is solar radiation. But back radiation is a significant component of the total energy balance at the surface, which is what determines the surface temperature.
Another question, regarding:
“…the effective emission temperature resulting from that temperature gradient.”
How does the T gradient affect the emission temperature?
Energy emitted in the long run is very close to energy absorbed. Given a surface area, Stefan-Boltzmann sets a temperature for emitting a set amount of energy. I don’t see where the gradient, or lapse rate fits in when determining the emission rate.
SOD: Your post is titled “Things climate science has totally missed”. The key word is “totally”. I could do a post on “Things climate science often overlooks when focusing attention on radiative forcing by GHG’s”.
For example, in your post on “Venusian Mysteries”, Figure 2 shows the temperature vs height profile for the atmosphere of Venus. A quick inspection of this figure shows that most of this profile is dominated by a fixed linear lapse rate controlled by buoyancy-driven convection, not the curved plot expected for equilibrium radiative transfer. (See Figure 2.9 in the linked CO2 Part 5.) However, the caption to Figure 2 says “Temperature profile calculated from the Venus radiative transfer model”. This almost certainly was a radiative-CONVECTIVE model. The proper terminology is used elsewhere in your post, but the lapse rate on Venus wasn’t mentioned – leaving the role of convection on Venus indeed a mystery! (The surface radiation of 16,100 W/m^2 is so impressive that it is easy to overlook back radiation of about 16,099 W/m^2 and the need for cooling by convection.)
The other thing that many overlook is the fact that a radiative forcing of 3.7 W/m^2 from 2X CO2 results in a direct temperature rise of only about 1 degC from well-understood, non-controversial physics. For example, in your excellent series of posts “CO2 – An Insignificant Trace Gas?”, it appears to have taken seven posts to finally admit that the direct effect of doubling CO2 ALONE isn’t a cause cause for alarm. Finally an answer: By itself, CO2 IS fairly insignificant! You’ve overlooked the need for conclusive evidence that it does become significant when feedbacks are included. :-)) Or maybe I’ve overlooked your evidence. Interestingly, the seventh post is subtitled: “The Boring Numbers”. I’ll bet these numbers would have been less boring if calculations had shown a direct temperature increase of 10 degK for 2X CO2 and that rise had been reduced to 1.5-4.5 degK by negative feedbacks.
However, these are minor complaints. The total science content at this site greater than at any other blog. You seem to have attracted a number of sharp readers who seem to correct these minor weaknesses. Best of all, you and they make an honest attempt to help commenters like me.
On the topic itself, I wonder if it is a perception problem. For instance General Circulation Models (GCM) have been around for some time. The name ‘circulation’, to me, implies that those that wrote them had some idea that the atmosphere circulates, and convection is just one aspect of circulation. However, convection is not a force in and of itself; it exists because of energy imbalances between locations in the system. So, it isn’t what is driving the changes we are seeing.
I would hazard a guess that those who really seek to understand the forces driving the changes tend to focus on those, but the perception of others can be that they are ignoring convection. Not to paint with too broad a brush, but there are also those who bitterly complain that the sun has a lot to do with the energy level of the earth and it is being ignored. I’m pretty confident that it is not; you have to shut your eyes pretty tight not to see any references to solar activity in the climate study realm. And the same would apply to convection, but there we are, trying to convince people to open their eyes.
Chris,
I believe most of your questions are answered in the document (SOD provided the link). But let me give you the whole paragraph that SOD partially quoted:
Thus the lapse rate is independent of the GHG radiative properties of the atmosphere. The dry adiabatic lapse rate is dT/dz = -g/Cp. The only way that the addition or removal of GHG gases will affect the dry adiabatic lapse rate is if it changes the heat capacity (Cp) of the atmospheric mass. Water vapor can have a small affect since its heat capacity is twice that of dry air, but its molecular weight is only 62% of dry air. CO2 is insignificant. Water vapor has a much larger effect on the lapse rate due to the latent heat released during condensation. But that is not a radiative effect.
The lapse rate is gravity induced in that gravity causes a pressure differential across the atmosphere based on the varying vertical mass above any given altitude. This, in turn, leads to a temperature differential based on PV = nRT. The greenhouse effect is a direct result of this temperature gradient since the emission temperature at the effective emission altitude must be such to allow a balance of energy fluxes between the incoming SW radiation and the outgoing LW radiation (~235 W/m2). Thus the surface temperature must be greater than the effective emission temperature (~33°C for earth). Backradiation is just a secondary phenomenon. It increases or decreases as a result of the surface temperature and the total heat content of the atmosphere. It follows the surface temperature; it does not control the surface temperature. The net radiative heat flux through the troposphere changes very little, if at all, with surface temperature. The net heat flux through the troposphere is primarily convective, not radiative.
Thanks, but no, you haven’t answered my questions.
You have stated a lot of basic information that is correct, PV = nRT and all that, but nobody I know is twisting the ideal gas law. What is happening is that you are assuming that SW plus backradiated LW is equal to SW alone and it is not.
Chris,
Who said SW plus backradiated LW equals SW alone? I said that incoming SW = outgoing LW. With the exception of the IR window from the surface, the surface and the atmosphere are basically trading photons with each other. They are trading a lot of photons, but that is not important. There is little or no net radiative heat flux occurring within the troposphere. The net heat flux from the surface to the TOA is primarily convective.
Williamcg,
Well, it’s possible I misunderstood you when you said:
“Backradiation … has nothing to do with the surface temperature.”
Which I took to mean that the surface temperature is not affected by backradiation, which would imply that the surface temperature is only affected by SW and not LW. What else did you mean?
Also, I’ve read various estimates of the mean emission altitude; they have varied from 6km to your estimate of 4km. All the estimates are well within the troposphere over most of the globe; so, I have a problem reconciling your estimate of the mean emission altitude with your statement that there is little or no radiative heat flux within the troposphere.
Chris,
The cause of any change in backradiation is a change in surface temperature. If the surface temperature increases, the LW emission increases and the atmosphere absorbs and emits more LW emission; thus backradiation increases. Backradiation is an effect of surface temperature not the cause.
When I say that there is very little net radiation flux within the troposphere, I am referring to the flux between the surface and the troposphere and troposphere to troposphere. The troposphere can roughly be assumed to be in local radiative equilibrium at any given point. Thus there is little net radiative flux. But the troposphere can emit directly to space at higher altitudes. But what is emitted to space cannot be absorbed or emitted by the surroundings so the LTE, no net flux situation continues.
Well, you’ve fleshed out Jelbring’s quote. But it just gets worse. He’s taken a mangled version of the Ideal Gas Law (PV=RT) and deduced that “Hence T = P/R, T is proportional to P and P is known from observation to decrease with increasing altitude.”
But it isn’t. You’ve given the law correctly – PV=nRT, where n is the number of moles of gas in V. So P/T=ρ (R/M) where M is molar mass. T is not proportional to P but to P divided by density, which is certainly not constant.
And he notes that the IGL is a natural law which “unscrupulous scientists can twist”.
williamcg:
As I have no idea how he can reach his conclusion it’s pretty difficult to know what “out of context” actually means. Your “complete quote” will do just as well and still makes exactly the same point. I also welcome readers to read the whole article and I will be delighted to update the article to quote the full statement that you would like.
Actually, I thought I was being charitable. The alternative is worse.
The standard theory is clear – increasing the concentration of “greenhouse” gases increases the opacity of the atmosphere which increases the altitude of effective emission to space.
As the temperature is lower the higher we go in the atmosphere this would result in a reduction in OLR. A reduction in OLR will increase in more heat being retained in the atmosphere until a hotter climate finally restores an OLR at the old value which balances absorbed solar radiation. Well that’s the theory.
By not even mentioning this fundamental element of atmospheric physics your readers – like the confused crowd where I last saw your article promoted – will (and do) reach the wrong conclusion.
We all agree that the lapse rate is a function of the thermodynamics of moist air and nothing to do with its radiative properties.
You can see that clearly explained in the atmospheric physics text books I have scanned a few pages from. You can see it any atmospheric physics textbook.
Where you have failed your readers is by not making clear what the climate science argument actually is.
And as you’ve repeated the same stuff here without actually dealing with the main point you probably don’t understand it. So I have a simple question to ask to confirm your thinking:
1. Do you believe that the effective emission altitude to space will NOT change regardless of the concentration of radiatively-active gases?
Note: main article now updated with “full quotation”.
I’m delighted that William Gilbert is here to defend his criticism of climate science.
I did ask him on Sep 27th if he had any changes to his article or wanted to see this article before it was published but didn’t get any response.
My next questions for William Gilbert are:
2. Do you believe that the references I have provided are the correct derivation of the dry and moist adiabatic lapse rate?
3. Where in atmospheric physics text books or papers on the radiative-convective models is your claim supported:
I disagree with Leonard Weinstein (December 8, 2010 at 1:27 pm) – although I guess that’s obvious because otherwise I would not have quoted the statement in the first place..
The point that Lindzen makes is a very useful one and not a waste of time at all.
Examining “what would the world be like if” scenarios helps to illuminate us students about the relative importance and role of various mechanisms.
Just like “what the world would be like if the atmosphere was transparent to longwave radiation” and many other useful questions.
Most atmospheric physics textbooks contain the derivation for the “grey atmosphere with no convection” and I found it very useful to study this problem. (Some explanation at Vanishing Nets).
For hr and others –
I believe I have (unintentionally) confused some/many/all with my poorly worded introduction and will revise it shortly.
In fact in atmospheric physics we almost always see the breakdown like this:
-radiation
-latent heat
-sensible heat
Latent heat being the movement of heat via evaporation – convection – condensation.
Sensible heat being the movement of heat via convection with no phase change.
Conduction is actually also included in sensible heat, but is negligible in atmospheric physics.
Almost every paper I have read, when writing about convection is clearly writing about latent heat and sensible heat. The exceptions are when the breakdown between them is the topic of discussion – which of course makes it even more explicit so I don’t believe in the world of atmospheric physics papers and textbooks there is confusion over this topic.
Main article now updated.
Thanks, Doom. Your clarification is much appreciated.
SOD,
Sorry, but I never saw your post on the other thread asking me about writing something about the Politics and the Greenhouse Effect document. Send me an email next time if you need a sure response to such a request.
But my answer then would be the same as now – why on earth would you want to discuss a political document put together for Swedish parliamentarians on what has been a very serious science blog? Are you nuts? The section that you excerpted for this thread was from a paragraph titled “The High School Approach”. And you criticize Dr. Jelbring for not fully explaining the effective emission altitude and temperature? Hell, these parliamentarians don’t even know what the words “emission” and “opacity mean. They probably think “emission” is some kind of gentle bodily function and “Opacity” is the latest cell phone! Get serious. I don’t know why you keep dragging this document out in the open. You first did it on the Venus thread and I explained to you that this was not a scientific paper. And now you have dragged it out again. Climategate must have really gotten to you.
But if you would like to discuss Dr. Jelbring’s E&E paper or my recent E&E paper, The Thermodynamic Relationship Between Surface Temperature and Water Vapour Concentration in the Troposphere, I would be happy to help you put something together. (And, yes, E&E is a British publication and I had to spell “vapor” that way). These are peer reviewed papers and would be a better fit for what has been a very good science blog. I have also written an article for a private skeptic blog on the “Climate Science Paradigm” which you may be interested in. It discusses the current climate science paradigm with respect to Thomas Kuhn’s book The Structure of Scientific Revolutions and why the current crop of climate scientists cannot see the alternate paradigms in atmospheric physics. We could also discuss the Miskolczi papers which you seem to avoid, at least thus far. But, no, I am not interested in trying to turn a political document into a scientific paper, even though the science part is correct but limited. That would be a waste of your, my and the readers time.
williamcg:
Well, my article is out now, so now is the time to answer the 3 questions I have asked.
In other news, I have read the paper: “The ‘Greenhouse effect’ as a function of atmospheric mass”, Jelbring, Energy & Environment (2003).
It has the same flaws as the political paper.
The author appears totally unaware of what the inappropriately-named “greenhouse” effect is actually claimed to be by climate science.
There is no comment in the paper about the effect of the opacity of the atmosphere on the OLR. No comment means “I didn’t know about it”.
This is also clear in the citations. Arrhenius (1896) as the last word in the radiative-convective model?
What about Manabe & Wetherald (1967) or Ramanathan & Coakley (1978)?
The substance of my new article would be the same as this one. Or the same as On Missing the Point by Chilingar et al (2008).
If you don’t even mention the main point then what else is there to say?
So back to my earlier Q1 – Do you believe that the effective emission altitude to space will NOT change regardless of the concentration of radiatively-active gases?
I may have missed it, but I haven’t seen mention in this thread of the relation of convective energy flux to other parameters. If one assumes this flux, in a first approximation, varies linearly with lapse-rate changes, then why is the lapse rate presumed a tropospheric constant? Convective flux is typically of order 100 W/m2 at the surface and markedly reduced at the tropopause.
Is the apparent lapse rate constancy a sign that small temperature gradient changes lead to large convective flux changes and, if so, why shouldn’t small lapse rate increases then compensate to some degree for reduced radiative fluxes due to increased IR absorption?
Scienceofdoom,
I think to a large degree that you, williamcg and I agree on the net cause of atmospheric greenhouse gas effects. It is some of the details and emphasis you disagree on. I think a comment may help here.
We all agree that the greenhouse gases absorb and also radiate LW radiation from the solar heated ground.
SOD uses the argument that this raises the altitude of out going radiation, which would then be colder (temperature drops with altitude due to the lapse rate), so the lower atmosphere and ground warm up until the temperature at the higher altitude is raised so that balance is restored.
I and williamcg use the argument that lapse rate is a gradient, not a specific set of temperatures. We agree there would be a raising of the altitude where the outgoing radiation balances the incoming (and thus sets the temperature at that altitude), but this directly increases the ground temperature by the increase in altitude times the lapse rate (due to adjustment in convective heat transfer).
I think both views get to the same end point. However, the lapse rate is not dependent on the radiation level for Earth, only the mass of the atmosphere, gravity, and specific heat of the atmosphere. Thus raising the outgoing location would, by definition of the lapse rate and temperature of the equilibrium level, raise the ground temperature.
What SOD states, is what he considers to be the transient response mechanism of an increase in greenhouse gases. What williamcg and I state is the steady state response. However, the increase due to human activity is relatively slow (0.5% CO2 increase per year), so steady state solutions are valid. However, the increase in back radiation in no way caused the increase in temperature, the convective adjustment maintaining the average lapse rate did it.
This is a cause vs effect argument, and both happen, but since lapse rare is not radiation dependent for Earth (on the average and at low altitudes), the back radiation has to be the effect, not cause.
A question regarding, “…the lapse rate is not dependent on the radiation level for Earth”
Is it?
I don’t know how to quantify this, but absorptivity and emissivity go hand it hand; so, if something has a higher absorption rate, it also has a higher emission rate. For instance, I read some material lately (“Cooling due to the greenhouse effect” http://www.atmosphere.mpg.de/enid/20c.html) that indicates that an increased emission rate contributes to stratospheric cooling. I’m thinking that molecules don’t really care whether they are located in the troposphere or stratosphere as far as their emissions are concerned; so, the troposphere air must also be loosing energy faster than it would with less CO2.
The original source of the LW is the surface; so, I’ll hazard a guess that in the lower troposphere the higher absorption rate forces a warming and above the mean emission altitude the higher emission rate forces a cooling. (And we are in reasonable agreement that the emission altitude is well within the typical tropospheric altitude.) Of course, convection will keep heat content fairly well mixed, but I would expect it to be less than 100% efficient at this. So, I’m thinking there might be a reason for the lapse rate to be slightly, just slightly, higher with a more radiative atmosphere.
Incidentally, the within the troposphere, the temperature and lapse rate are close to what PV=nRT would predict, but not exactly. So, there are other factors at work; radiative heat loss has to be one of them.
so, the troposphere air must also be loosing energy faster than it would with less CO2.
No, this is dependent on the path length, at low altitudes, where the air is warmer, it is emitting more than at higher lower temperatures, but due to the proximity of the molecules, they are absorbing more. So if you look at radiation, as an expanding sphere from a molecule, the density of the absorbing molecules will effect how far this energy travels before it is absorbed.
So with the denser atmosphere at lower altitudes, what a molecule emits is absorbed by its neighbors, and what its neighbors emit is absorbed by it. Reducing the loss of energy. However, when the path length shortens, as in the stratosphere, it means that the radiation, being emitted by a molecule is traveling a lot further before it is absorbed, by neighboring molecules, and as the sphere of radiation increases o course, the more dispersed is the energy/photons) it is receiving less and less of the energy it emits. But the real crux of the stratosphere, is that it is heated by O3, O2, and UV. Not LW from the ground. CO2, due to the path length, is a net emitter, it emitts more than it absorbs, taking energy from the stratosphere and radiating it away.
Where as adding co2, to the troposphere, will lengthen the path length, raising the altitude that energy can escape.
Dosnt really answer your Q though. 🙂
Mike,
Thanks, but no, that doesn’t really answer my question.
While I agree with most of what you wrote, I disagree with the notion that the mean path distance is spherical. Because the direction of emission is uniformly distributed about a sphere, it would be if the atmosphere were of uniform composition and density, but even in dry air, density changes with altitude. I believe the path distance is slightly egg-shaped, and more so as you increase the definition of where the boundary exists as a percentage of absorption from 0 to 99.99 percent. Another way of looking at this is that the hazard function of capture declines as the photon travels upward and increases as it travels downward.
The mean distance of emission of a photon to space is the 50 percent point with respect to outbound LW radiation. Correct?
So, if absorptivity == emisivity, and the 50% boundary for un-recaptured outbound photons is at altitude X. I believe that it would follow that there would be a warming effect below X and an cooling effect above X. Kind of like if you replaced one insulative material with one just slightly better. It’d just be hard to detect since I don’t think it would be very large to begin with and convection would tend to blur it. Though, I’d be curious to see if any there were any radiosonde data mining studies that looked for this.
Chris G
My spherical analogy, was just talking molecule to molecule in a parcel, to show “why” the path length changes with pressure.
We have an inversion from the tropopause up, i would think this would be a good place to look for your 50% boundary.
Not sure what you are talking about with regard to looking for looking for the 50% boundary near the tropopause. It’s been looked for already, and found, sort of, there’s just some quibbling about where exactly it is. It is a lot lower than the tropopause. For instance,
“The infrared radiation that cools
the Earth comes from an average height of about
5.5 km at present,…”
Click to access hadley%20center%20climate%20change%202005.pdf
and
“…to the average infrared emission altitude around 4000 m…”
from Jelbring paper under discussion.
Yes, I’m aware that the tropopause is not at any set altitude and it varies by day and night, season, and latitude. Regardless, the average emission heights above are in the ballpark of half the height of the tropopause; I kind of mentioned this in my earlier comments.
Leonard: I’m interested in the implications of your very clear explanation of cause-and-effect. To see if I understand correctly, I’ll restate your argument. Cause: Increasing GHG’s raise the altitude of average emission to space, so that altitude must warm, and that warmth is translated to the surface by the lapse rate. Effect: A warmer lower troposphere and surface – which cause increases in both upward and downward LWR near the surface of the earth.
This argument seems to imply that greenhouse gases are unimportant at altitudes where the lapse rate is constant. CO2 is well-mixed, but water vapor is not and water vapor feedback is critical to CAGW. SOD had a post several months ago about the effect of increasing water vapor at different altitudes. My instinctive (and probably incorrect) response was to assume that all changes in radiation below the average emission level would be countered by changes in convection, but this wasn’t what the papers SOD cited showed (including one by Roy Spencer). I presume these papers were using some sort of model that included radiation and convection.
Can anyone help with this dilemma?
Quondum,
There are only four things that can transfer energy from the solar heated surface:
1) Thermal conduction (which is significant right at the surface, but very small elsewhere)
2) Radiation (which has two components; a clear window direct to space, and a relayed radiation from absorption and re-radiation)
3) Dry convective transfer (which is due to a combination of buoyancy and winds)
4) Evaporation combined with convection and condensation at altitude (which is the largest effect for Earth).
Three and four combined are the dominant sources of heat transfer from Earth to the upper atmosphere. There are cases where the lower atmosphere will not be at the adiabatic lapse rate, such as at night, at the poles when it is low Sun, and with clouds, but the average global long term lapse rate will be the wet or dry adiabatic lapse rate just due to the mass of the atmosphere, gravity, and the specific heat of the atmospheric gases. Buoyancy from solar heated ground, lateral pressure and temperature variation due to night/day and latitude variation, planetary rotation and other factors (ocean currents) cause winds and convection to form which are sufficient to maintain the wet or dry adiabatic lapse rate on average. At higher altitude, more and more radiation can go through the remaining atmosphere, and at some point this energy loss makes the lapse rate decrease (exceeds the ability of convection to maintain it), until at the tropopause it goes to zero. Close to and above this there is no significant convection, so radiation dominates. However, in the stratosphere, UV absorption by Ozone heats the gas, and the temperature increase with altitude. This reverse lapse rate is very stable wrt convection.
Quondam,
Please excuse my misspelling your name in my previous entry. In answer to your question “Is the apparent lapse rate constancy a sign that small temperature gradient changes lead to large convective flux changes and, if so, why shouldn’t small lapse rate increases then compensate to some degree for reduced radiative fluxes due to increased IR absorption?” convection is strong enough to adjust for reduced radiative fluxes due to increased IR absorption. That is exactly why I take the position I do. Convection is the dominant energy transfer mode and maintains the adiabatic (dry or wet depending on location) lapse rate on the average.
Leonard,
We share the same view regarding the importance of convection in compensating for increased GHG radiative resistance. That convection can lessen thermal changes inferred from radiation-only arguments is a truism, but by how much? This is the reason for my comment seeking ideas of the relation of convective flux to thermal gradients. One irritation I have with skeptical science is that it’s quick to critique but doesn’t feel an obligation to work out a better solution.
I once went through a spreadsheet calculation using MODTRAN results for net radiative flux before and after CO2 doubling which assumed a local convective flux change proportional to a thermal gradient change. The proportionality constant had dimensions m2/s (times the heat capacity). It required values >50 to keep unperturbed lapse rates within a K/Km of the adiabat, with a net temperature increase across the troposphere of 1K from doubling. With values >1000, radiative changes were all but ‘shorted out’. It’s tempting to identify this constant with eddy diffusion (a velocity times a distance). This showed me that a compensation comparable to the perturbation was feasable, but I still lacked a handle for the limits of turbulent compensation.
Quondam,
In the Earth’s atmosphere, radiation forces convection. If we could freeze the atmosphere in place by some means so that all energy loss had to be by radiation, the lapse rate would be greater than the dry adiabatic lapse rate of 10 K/km. A lapse rate greater than the adiabatic rate is unstable to convection in the normal atmosphere so heat has to be transferred by convection to maintain the adiabatic lapse rate. In a transparent atmosphere there would be much less heat transfer by convection.
Sensible and latent heat live:
see here and paper explaining the calculation here
Sensible and latent heat:
correct link to near live-graph can be found here.
The calculations do not make sense if air temperature is negative, hence the gaps in the curves.
Sensible and latent heat:
dear scienceofdoom: please correct the previous link: it should start by http: and not http.
(http://meteo.lcd.lu/today_01.html)
Thanks and sorry about that…
SOD,
I guess our main point of contention is that term, “the main point”, that you keep throwing out. Atmospheric opacity may be your main point but I can think of a lot of other main points that may be more important than your main point. Jelbring covers a lot of main points in his paper that explains a lot of empirical observations of earth as well as Venus and Titan. I’m sorry that he failed to include your main point in his analysis. If I am not mistaken he intentionally left out electromagnetic radiation to show that basic fundamental physics explains the basic structure of the greenhouse effect (i.e., the difference in temperature between the surface and the effective emission altitude). After those main points are established, your main point can then be considered. But since you are stuck (along with many others) in your radiative transfer, thermal energy paradigm, you struggle to see what he is telling you.
Now for your question #1. “Do you believe that the effective emission altitude to space will NOT change regardless of the concentration of radiatively-active gases?”
I won’t bother to tell you what I believe since that is of no importance to the real world out there. The Church believed different things from Galileo and that did not do Science very much good at the time. I am not a practitioner of the Global Warming Orthodoxy. But I will tell you what I think I know (and that changes on a daily basis).
1. The hypothesis that increased GHG concentration will increase the optical depth and thus raise the effective emission altitude to a higher, cooler level is well thought out and makes sense.
2. The corollary hypothesis that the greater optical depth will lead to a higher surface temperature is also credible.
3. The 3rd corollary hypothesis that higher surface temperature will lead to higher water vapor concentrations in the atmosphere and further increase the optical depth is, I think, pushing things a little too far.
4. Unfortunately the massive empirical data compiled over a 61 year history published in a recent peer reviewed paper by Miskolczi, shows no change in optical depth despite rising CO2 levels. This does not support the first hypothesis and his empirical data has not been effectively challenged as far as I know.
5. The 3rd hypothesis is not supported by the Miskolczi paper. Nor is it supported by empirical data published in peer reviewed papers by Partridge and Soloman which show water vapor levels declining in both the upper troposphere and the stratosphere despite rising surface temperatures. My peer reviewed paper provides a possible mechanism as to why this is occurring.
It seems like your “the main point” is getting a little less main. Maybe, just maybe there are other main points out there that may better explain what is going on? Since your main point is based primarily on radiative heat transfer and thermal energy, maybe we should consider some other factors involved with atmospheric physics. Conduction and convection as heat transfer mechanisms may be a good start (as you are trying to do in this post). But since convection is primarily a mass transfer mechanism, maybe we should consider the role of mass in this whole thing as well (you radiation guys like things based on m2, but maybe we could also try kg). Then there is that energy thing. You and the radiative transfer advocates like thermal energy (it causes familiar things like “temperature”). But I hear there are other energy forms running around out there such as chemical potential energy (latent heat to you), gravitational potential energy and, my all time favorite, work energy. And surprise, convective mass transfer affects all three of these other strange energy forms. What do think atmospheric physics would look like if all these things (main points?) were considered in a proper, thorough manner? Maybe some more main points will come to light. Who knows?
I have some ideas that I would like to share with you and your readers. I need to find out which ones will hold water after a thorough critique. And this site has top quality commentators. But until you print a retraction of the personal attacks you made towards Hans Jelbring and me, I don’t think that will happen. I can see no justification for these attacks. Dr. Jelbring is just approaching the scientific issues from a different perspective than you. Why does that justify vilification? You have had no personal contact with him as far as I know. But you have had contact with me on this blog, so I will leave the retraction optional in my case.
Thanks, I’m done.
Chris G on December 10, 2010 at 2:41 am
You were talking about co2 before, 15 micron… now you are talking about average emission o energy… There is a difference, co2, from the graphs ive seen here and at other websites, seems to be emitting out from a T o 220k…. a bit higher than the average emission.
Mike,
You lost me.
CO2 molecules and any other molecules in the near vicinity will share energy through collisions and that sharing will result in the mixed gasses having pretty much the same temperature. No?
If so, then CO2 molecules will vary in temperature with the rest of the atmosphere from bottom to top. So, what is this about CO2 emitting only from 220K?
Yes, I’m talking about the total emissions from the atmosphere, and emissions from CO2 are part of that.
Going back to the line of reasoning I started, if the average altitude of emission to space is well below the tropopause, that would mean that there is significant radiative loss at altitudes lower than the tropopause. I’m not saying anything new with that, Leonard says the same thing with,
“At higher altitude, more and more radiation can go through the remaining atmosphere, and at some point this energy loss makes the lapse rate decrease (exceeds the ability of convection to maintain it), until at the tropopause it goes to zero.”
Maybe I’m belaboring a point that doesn’t need to be, but I get the feeling from some commenters that they don’t think radiative energy loss is significant until the tropopause.
I haven’t seen a counter to the idea that a higher density (or partial density) of a greenhouse gas means a higher rate of emission. I see that you brought up that more will be absorbed, and I’m not arguing with that, but absorption will never be 100% of what is emitted. So, more emissions means more energy loss.
If the rate of energy loss is higher, I’m not sure, but I’m thinking that might have an affect on the lapse rate. Ah, point of clarification, I’m talking about the normal or environmental lapse rate, which I know is variable. I’m just thinking that the mean rate would change with a change in radiative properties.
You know, I’m starting to think that my personnel discovery is likely an old hat to those well-versed in atmospheric science.
williamcg:
There are two reasons for my harsh comments in the article:
1. Regardless of whether you agree with what I call “the main point” – claiming that you have disproven a basic element of atmospheric physics without even mentioning this essential element is a major problem.
Proof by ignoring is not proof, it’s misleading.
After all, it is the central claim of how the “no feedback” temperature increase will be caused by increases in CO2 concentrations.
Some people disagree and explain why they think it is wrong, because it is customary when debunking a theory to attempt to describe it correctly.
You and Hans Jelbring have ignored it and said “see we’ve proved the ‘greenhouse’ effect is unaffected by increases in CO2”.
I can only describe this as misleading – I assumed it was unintentional and therefore due to not understanding the subject very well.
2. You have derived the same formula as everyone else and claimed that atmospheric physics totally ignores it. This is not true and falsified by the textbooks shown in the article.
You have claimed atmospheric physics has a “paltry level of scientific competence”.
You haven’t demonstrated this claim. The textbooks above – on your point of contention – prove the exact opposite.
Great scientists like S. Manabe, R.M. Goody, V. Ramanathan, R. Cess, R. Lindzen and 100s of others have done thorough and distinguished work over many decades.
If you describe their excellent work as “incompetent” with no evidence then you invite harsh criticism.
Back to your specifics:
Item 1 and 2 are the claim about “no feedback” increases in surface temperature due to increases in “greenhouse” gases.
I have not made any claims for item 3 in this article.
What is fascinating is that you agree that 1 & 2 are “credible”.
Therefore, your existing article is fatally flawed by omission. (And even if you didn’t believe that 1 & 2 were credible your existing article is fatally flawed by omission).
Chris G,
The average height of emission is a mathematical construct than, IMO, often leads to confusion. Optical depth is a function of wavelength so the altitude where the optical depth reduces to a value of 0.5 (= the altitude of peak emission ) also varies with wavelength. For weakly absorbing lines or for water vapor, the peak emission altitude can be quite low, less than 2 km above the surface. Water vapor pressure decreases with altitude much faster than for CO2 and other non-condensable gases. CO2, OTOH, has very strong absorption in the 15 micrometer band so the peak emission altitude is high, near the tropopause.
The emission at any given wavelength and temperature cannot exceed the emission of a black body at that same wavelength and temperature. Because the peak emission altitude for CO2 at 15 μm is near the tropopause where the temperature is about 220 K, the emission intensity is much lower than for emission from lower altitudes and correspondingly higher temperature.
Thanks, that makes perfect sense and I see where Mike was coming from better now. Though, I don’t see that this changes my conclusion.
One quibble, the peak of a distribution is not necessarily equal to the point marking 0.5 of whatever. For instance, if memory serves, in a chi-square distribution, the mean is a bit to the right of the peak.
SOD: A number of these discussions about the rising altitude of average emission keep returning my thoughts to how radiative forcing is calculated. The IPCC defines radiative forcing as the change in flux at the tropopause (and Myhre does his calculations for the true minimum). From this change in flux, we can calculate warming at the tropopause and then translate that warming to the surface via the fixed lapse rate. The problem is that there isn’t a fixed lapse rate. For part of the distance from the surface to the tropopause, there is a fixed lapse rate determined by convection and therefore by heat capacity and gravity. That lapse rate doesn’t change with GHG mixing ratio (until one wants to introduce the complication of changing water vapor and lapse rate feedback). Above the altitude where convection is no longer important, the shape of the temperature vs. altitude graph is presumably controlled by radiative equilibrium. The curved shape of that part of the graph DOES change with GHG mixing ratio (including water vapor). If it weren’t for the ultraviolet radiation absorbed by the stratosphere, there wouldn’t even be a minimum on the graph to call the tropopause. The altitude of the minimum is changed by the effect of GHGs on stratospheric temperatures – an effect that is greater in magnitude and opposite in sign to the change lower in the atmosphere and stratospheric temperature is even more complicated when ozone is thrown into the mix.
It would have been ridiculous for the IPCC to define radiative forcing at the surface, where radiative equilibrium doesn’t exist and temperature is controlled mostly by convection. However, defining forcing at the tropopause may be almost as bad – at least if one wants the calculated forcing to be relevant to warming at the surface. For purposes of defining and calculating radiative forcing, wouldn’t it have been better to pick the lowest altitude where the fixed lapse rate ends and temperature begins to be controlled by radiative equilibrium? At that altitude, we know precisely how warming (from W=oT4) will translate to the surface. Above that altitude, other factors besides fixed lapse rate have the potential influence the answer. For example, Santer et al showed that the height of the tropopause has rise over the last 30 years and that models showed 60% of that change was due to changes in ozone and 40% to changes in GHGs. In a comment, Pielke noted that the upper troposphere hadn’t warmed during that period and therefore all of the rise in the height of the tropopause could have been due to ozone. And Santer defined the tropopause as the height at which the magnitude of the lapse rate dropped below 2 degK/km, well below the true minimum.
Does anyone (DeWitt?) know what happens to the radiative forcing for 2X CO2, when it is calculated at the top of the fixed lapse rate rather than the tropopause?
(Technically, the difference between warming at the tropopause and warming at the surface can be lumped into one catch-all term, “lapse rate feedback”. We can say that 3.7 W/m^2 for 2X CO2 translates to 1 degK “before feedbacks”. However, the more feedbacks one needs to consider, the further one strays from basic physics.)
Frank,
The logic of calculating forcing at the tropopause includes a step you left out. Because the energy distribution in the stratosphere and above is controlled solely by radiation, equilibrium is achieved rapidly, on the order of a few months. So the CO2 is doubled, the stratosphere is allowed to equilibrate and then the forcing at the tropopause is calculated. I’m not sure what definition of the tropopause is used for this calculation. I’m doubt that it makes a significant difference.
I don’t think this is how it’s done. I wouldn’t do it that way. In fact, the forcing varies with latitude (~ 3X max/min with the max at the equator and the min at the poles) so you really have to use a GCM or at least a 2D radiative/convective model to calculate the warming.
Interesting post. I enjoyed it.
“But what it means in simple terms is that each heat transfer mechanism works independently, but each affects the other mechanisms via the temperature change (if I come up with a useful analogy or example, I will post it as a comment).”
I offer my example of a pan of boiling water on the stove. The temperature of the water (100 C) is determined by the detailed balance of heat input from the gas setting, conduction through the pan, radiation from the sides of the pan, radiation from the surface of the water, convection within the water and evaporation.
The question of which one “dominates” (as it is meant here) in determining the temperature is not a matter of which one is biggest (if the water is only barely boiling, evaporation is relatively small), or of whether one mode of heat transfer controls the values of the others (turning the gas tap to apply more heat can affect all the others, but none of the others can turn the gas tap), or even which ones are essential to the water boiling in the first place (obviously, if you turn the gas off completely, a lot of these considerations will swiftly disappear). An effect dominates to the extent that the final result (the water temperature) is controlled by one of the heat transfer modes. i.e. why is the temperature of boiling water 100 C?
I think this confusion is more a result of semantics than it is a fundamental difference on the physics – but I’m not sure.
“Second Misconception – Atmospheric Physics Ignores Convection”
Note that there is an important difference between atmospheric physics not discussing convection at all and it not discussing the importance of convection in the greenhouse effect. You have given a lot of examples to disprove the former, but the latter is what sceptics usually mean by it.
(Incidentally, it’s all very well relying on Ramanathan and Coakley to support your argument that climate scientists always discuss convection, but most people rely on the public outreach efforts of trusted science organisations for their understanding. These are what they are complaining about.)
I keep trying to argue that atmospheric physics is perfectly well aware of convection, and uses it in its calculations of the greenhouse effect, it’s just that it often neglects or misunderstands it in its public explanations of the effect. Not always, though.
My favourite example is the discussion in Soden & Held 2000, in the text below figure 1. You might like to add it to your list of examples.
Nullius in Verba:
It’s an excellent paper. I cited it – and recommended the paper – in Clouds and Water Vapor – Part Two.
The diagram for the benefit of other readers:
First of all, from the many claims I have read in blog articles (and comments on blog articles) almost all are not as you describe. They are claims that climate science ignores convection, climate science doesn’t understand convection, climate science is so focused on its radiative paradigm that it doesn’t understand the importance of convection..
All without attribution to an important paper or a textbook..
On your point: “..not discussing the importance of convection in the greenhouse effect..” – would you like me to scan in many pages from atmospheric physics textbooks and climate science papers with this laid out point by point?
It will mean a trip to the library but if that is what it takes I am more than happy. I know that climate science does get this right because it is in the many textbooks I have read.
“What sceptics usually mean by it..” – I call myself a sceptic, which means testing things that people claim.
Many people call themselves sceptics yet accept any cock and bull story that supports a preconceived point of view. I don’t believe that they can really be called sceptics.
According to St. Google:
“skeptic: someone who habitually doubts accepted belief.”
“Contemporary skepticism (or scepticism) is loosely used to denote any questioning attitude, or some degree of doubt regarding claims that are elsewhere taken for granted.”
That’s a broad claim that I find hard to believe – “These are what they are complaining about..”
In any case, I don’t have an interest in dissecting the work of marketing departments and public outreach groups.
When someone says “climate scientists don’t understand this subject” I take that to mean “climate scientists don’t understand this subject“.
If it turns out that the claimers have never read any climate scientists then I think it is reasonable – in fact, required – for me to point that out.
If people instead wrote ” the NASA marketing department doesn’t do a good job of explaining this” I would probably agree with them if I was interested in reading what the NASA marketing dept had to say.
This figure is an excellent representation of the failure to communicate between consensus and skeptic. Both plots have the same slope, e.g. 6.5K/Km. The consensus takes this slope as a fixed parameter derived from equilibrium properties – convective equilibrium. The skeptic considers this expression oxymoronic, asserts that stationary states are not equilibria and asks what about drawing a 2x line with the same x-axis intercept, but passing through the unperturbed tropopause point? (This avoids raising temperatures above the tropopause by the same amount as at the surface as depicted.) Does, perhaps, some rule guide nature to a compromise?
It’s my impression that the consensus takes the adiabat to be some sort of equilibrium property and a column of gas insulated from its environment would show a temperature difference between top and bottom. Were this the case, one could take two columns containing different gasses of different heat capacities and expect different temperatures at the two tops if they rested on the same thermal base. One could then connect the tops with a thermocouple and extract useful work – perpetual motion of the second kind.
This equilibrium notion carries into the perturbation analysis on the following page. It presumes the mean flux a constant determined by solar input, but includes no constraint that the system remain in a stationary state. Once agreement can be reached that the atmosphere is better described by a stationary state far from thermodynamic equilibrium, we may start converging towards nature’s solution.
Thank you.
“First of all, from the many claims I have read in blog articles (and comments on blog articles) almost all are not as you describe.”
Yes. I suspect in a lot of cases the meaning is taken to be obvious from context, and for a lot of others, it is because of the point I alluded to and will discuss again below: that most people engaged in the policy debate haven’t heard about the greenhouse effect from reading the technical literature. They haven’t been told it’s an urgent problem requiring immediate action to cut emissions, tax energy, and stick up ugly windmills all over pristine wilderness from reading the technical literature. They think they’ve been told how climate scientists think the greenhouse effect works, and it doesn’t include convection.
“would you like me to scan in many pages from atmospheric physics textbooks and climate science papers with this laid out point by point? It will mean a trip to the library but if that is what it takes I am more than happy.”
Personally, I think that would be helpful. But then, I’m biased.
I’m a little tetchy at the moment, because I’ve just spent two days arguing back and forth with fervent AGW believers in defence of textbook AGW science. Somebody gave the radiative-only version of the GHE, as if it was the whole story, and asked rhetorically how anyone could doubt the conclusion. I pointed out that they had omitted convection and feedbacks from their explanation, and that these needed to be covered before you could draw such a conclusion. And I was told, at great length by several articulate people, that this was nonsense, the greenhouse effect didn’t involve convection at all, and I was contradicting the expertise of all the world’s climate scientists in my efforts to prove that CO2 had no effect on temperature. Which, ironically, I wasn’t even attempting at the time.
So I hope you can understand my mild irritation at being told I’m complaining about a strawman.
“In any case, I don’t have an interest in dissecting the work of marketing departments and public outreach groups.”
You don’t have to. But you should at least acknowledge that they are the primary reason for much of the misinformation about “what climate scientists believe” you keep seeing, and not keep on blaming/criticising the people who have been misled by them.
While people might not think the NASAs of this world are necessarily right, they do at least expect them to represent their own scientific views accurately. If I say that NASA thinks CO2 acts like the glass on a greenhouse, scepticism is fully justified. But when NASA themselves say it, how far are you supposed to take this scepticism? You are supposed to believe that they misrepresented their own opinion? How are you supposed to argue against that?
“If it turns out that the claimers have never read any climate scientists then I think it is reasonable – in fact, required – for me to point that out.”
I agree. Having someone who clearly knows what they’re talking about saying in public that you can’t rely on the media journalists and the campaigners to tell you accurately what climate scientists are really saying, and that they’re frequently getting it wrong, that would be tremendously helpful. In fact, we need a whole load of scientists to stand up and say it.
And then to tell us what’s in those textbooks and papers.
(Which you’re doing an excellent job of doing, and please carry on. I’m 100% supportive of what you’re doing here.)
Dewitt: Thanks for your reply of 12/10 8:17. You are correct that I omitted allowing the stratosphere and tropopause to equilibrate before calculating radiative forcing. Myhre uses the true minimum as the location of the tropopause and explores how using different pressures (altitudes) for the height of the tropopause at different latitudes influences his calculation of radiative forcing. Different assumptions about the tropopause change the forcing by 10%. http://folk.uio.no/gunnarmy/paper/myhre_jgr97.pdf
Figure 4 of Collins et al (http://www.cgd.ucar.edu/cms/wcollins/papers/rtmip.pdf) shows that radiative forcing calculated at the surface is only 1 W/m2. Radiative forcing rises to a maximum at some altitude, is about 4 W/m2 at 200 mb, and is negative in some parts of the stratosphere. You may be overconfident to assume that: a) the radiative forcing for 2X CO2 is a clearly-defined 3.7 W/m2 and b) that the associated no-feedbacks climate sensitivity of 1.0 or 1.2 degK has any relevance to the changes that will ensue at the surface. The IPCC may have chosen the tropopause to calculate radiative forcing simply because forcing is largest at that altitude.
SOD and Nulluis: The Soden and Held scheme is excellent – to a first approximation. Going to the next step, however, readers need to recognize that the radiative forcing for 2X CO2 VARIES with altitude. Forcing is: only 1 W/m2 at the surface, 4 W/m2 at 200 mb, negative in the stratosphere, and has a maximum somewhere. (See above Collins reference.) The fixed lapse rate doesn’t extend all of the way from the surface to the tropopause. If radiative forcing were calculated at the top end of the fixed lapse rate (as this diagram suggests), the scheme would be much more accurate. If you believe in a large water vapor feedback, then the slope of the 2X line must be steeper, negating about half of the temperature rise at the surface.
Frank,
The LW forcing (figure 4, panel A in Collins, et.al.) is always positive, i.e. less LW radiation is emitted at high CO2 when looking toward the surface at any altitude. The increase in LW emission from CO2 in the stratosphere is not enough to make the total forcing negative unless you are looking layer-by-layer. That is not the case in figure 4. The SW (panel B) forcing is negative at 200 mb and the surface and slightly positive at TOM (top of model). Positive means more radiation than the base case because the ozone level is lower. Everywhere else, more incident radiation is absorbed by the increase in CO2 so less SW radiation reaches the surface.
It appears that GCM’s underestimate LW forcing compared to a line-by-line model.
Dewitt: If one goes up high enough (10 mb), the effect of 2X CO2 is to reduce temperature. If one were to calculate radiative forcing at this altitude, are you telling me that the forcing (down-up) will be positive?
I’m not sure what you mean by down-up. Which down and which up? Forcing is usually defined in terms of the surface temperature as in ΔTs = λRF, where λ is the climate sensitivity. Forcing is then the difference between upward LW radiation at the tropopause after stratospheric equilibration with and without doubled CO2. If you look only at upward LW with and without CO2, the difference is always positive except at the surface where it’s zero. Layer by layer forcing is more complicated because you have to include the reduction of SW radiation as well as the increase in LW radiation. So the stratosphere cools with additional CO2 because emission of LW increases and absorption of SW remains the same. But that’s supposed to have already happened before you calculate the climatic forcing.
In Judith Curry’s recent post CO2 no-feedback sensitivity there’s a lot of concentration on net downward radiation change at the surface. But I don’t see the point. The final equilibrium surface temperature will not be defined by the forcing only at the surface. The troposphere will warm and that will increase downward radiation at the surface regardless of the instantaneous value.
Dewitt: I’ve been wondering about the meaning of the IPCC’s calculated radiative forcing of 3.7 W/m2 for 2X CO2 and the related no-feedbacks temperature rise of 1.0 or 1.2 degK. My original point was that one can get radically different answers for the radiative forcing caused 2X CO2 by varying the altitude at which the forcing is calculated. Which altitude is most relevant to calculating a no-feedbacks warming relevant to surface temperature change? The IPCC could have picked the tropopause simply because radiative forcing is greatest there. I’ve suggested doing calculations at the lowest altitude where convection is unimportant and radiative equilibrium determines temperature. A constant lapse rate will transfer the same temperature change to the surface.
To show that an even wider range of radiative forcings are possible and emphasize the fact that radiative forcing must have a maximum somewhere, I added my assumption that radiative forcing is negative in the stratosphere where 2X CO2 causes the temperature to drop (more than it rises in the upper troposphere). Was I wrong? Obviously one can’t use the IPCC’s refinement of calculating forcing after allowing the stratosphere to equilibrate when one wants to talk about forcing in the stratosphere itself. I have to admit to sometimes being a little bit confused about whether forcing should refer to LW or LW+SW, and the convention for sign (which I appear to have gotten right). The definition below from the TAR seems closer to my intuitive understanding than what you wrote above (which seemed to focus only on the change in upward LW, but yours may be the only important change):
“the change in net (down minus up) irradiance (solar plus long-wave; in Wm-2) at the tropopause AFTER allowing for stratospheric temperatures to readjust to radiative equilibrium, but with surface and tropospheric temperatures and state held fixed at the unperturbed values”.
The concept of allowing the atmosphere above the altitude at which one calculates radiative forcing to equilibrate before doing the calculation is interesting. It suddenly dawns on me that this must be what Judith must do when she talks about calculating forcing at the surface.
Do climate scientists understand convection? I am sure they do. Do climate scientists ignore convection (and back radiation) when calculating the surface temperature of the Earth? Again, I´m sure they do.
Let me be more specific about this. Certainly they rely on the presence of convection to maintain a lapse rate which allows specifying the temperatures needed for the calculation of emissions at increasing altitudes. Can you find the amount of heat being conveyed by convection in Watts/m2 anywhere in the calculation? It seems almost certain you cannot. It is the same thing with back radiation. The sums only involve the up/outward radiation and do not involve the down/inward values at all.
In principle one could determine the surface temperature by finding the value that is needed to balance the energy flows at the surface but this seems to be difficult because there is no clear relationship between surface temperature and convective/latent heat loss. The back radiation is easily dealt with as it is directly linked to the surface temperature via lapse rate/radiative equations.
I should clarify that I am mostly talking about equilibrium 1D models rather than the calculations in the GCMs which are quite different.
It seems to me that there is not much room for debate about the very simple models that allow the use of Modtran/Hitran to calculate the relationship between surface temperature and OLR for varying amounts of CO2. They just do not need to have any stuff involving convection or back radiation to do their job.
Whether or not these simple models represent the real world is another question entirely.
SoD, I’m afraid you have missed the point. You paste in various text books that work out the lapse rate. But this tells you very little, and certainly does not show that climate scientists take into account convection. In order to say anything about the heating of the surface or any layer you need to work out the total heat flux into/out of that point. To do this you need to work out the radiative heat flux and the convective heat flux. You have done the radiative heat flux, but you have not quantified the convective heat flux. This is basically what Jorge is saying. Unfortunately, it is very difficult to quantify.
The fact that convection can’t transport heat to space is irrelevant. It can transport heat upwards very efficiently up from the surface through the atmosphere to where radiation can take over.
The more general point arose in your series of articles on CO2. While the radiative aspect was very thorough, the convective discussion (in part 5) was brief and muddled, as pointed out by Frank at the time.
Quondam – you didn’t miss it! There is no simple relationship between the convective heat flux and the lapse rate. That’s what makes it so difficult to measure. NiV’s example of a boiling pot is a good example – the temperature is a fairly uniform 100 degrees, regardless of how high you turn up the gas.
Frank: you say that “(t)he IPCC may have chosen the tropopause to calculate radiative forcing simply because forcing is largest at that altitude.”
This claim seems utterly unjustified to me. The tropopause marks the division between the region of our atmosphere that is well mixed by convection (due to the steady decline in temperature through the troposphere) and the higher levels where temperature doesn’t decline steadily with altitude and (as a result) convection becomes unimportant as a heat transport mechanism. So it’s a natural place to look for net forcing– unlike lower levels, at the tropopause heat transport is principally radiative, so the radiative climate forcing can be calculated there.
If you still want to claim that some kind of ‘cheating’ wzs involved in the choice to calculate the GHG forcing at the top of the convectively-mixed region of our atmosphere, you need to give a real explanation of where you think it ought to be calculated and why that’s the right place to do it for climatological purposes.
The IPCC (or Myhre at least) defines the tropopause as the location where the lapse rate is zero, so this does not mark actually the boundary between: a) the region where the atmosphere is well-mixed by convection and the variation in temperature with height is controlled by a constant lapse rate, and b) regions where energy flux is only by radiation and the variation in temperature with height is controlled by radiative equilibrium. I agree with you that radiative forcing should be calculated at this boundary, so that the resulting temperature change can be translated to the surface by a constant lapse rate. Unfortunately, in between the top of the constant lapse rate and the “zero lapse rate” tropopause, there is a variable lapse rate that changes with GHGs.
Furthermore, there is a highly variable relationship between the top of the convective domain and the location at which the lapse rate is zero. If it weren’t for the fact that the stratosphere absorbs ultraviolet light and is therefore warmer than parts of the troposphere, we wouldn’t have a tropopause at all. The height of the tropopause depends far more on changes in the stratosphere (ozone, cooling by 2X CO2) than on the troposphere. In polar regions in the winter, there is no tropopause, and convective transfer of heat is meridional, not vertical.
I would be happy if the IPCC had done what you think they do – calculate radiative forcing at the boundary between convective and non-convective domains. However, since they don’t calculate radiative forcing at the altitude(s) most predictive of surface temperature change, I’m suspicious. Is the difference substantial? Who knows? Myhre tried different methods for accounting for the variation in the altitude of the tropopause and found that radiative forcing varied by 10%. Radiative forcing at the surface is only 1 W/m2, 70% less than at the tropopause.
One could also ask why the IPCC chose to allow the stratosphere, but not the troposphere, to equilibrate before calculating radiative forcing. One could assume uniform warming of the troposphere and surface by a fixed lapse rate.
For me as a sceptic, this is a very disappointing post. It would have been an opportunity to discuss why the current creed of GCM’s are incapable of modelling convection accurately. Because, for IPCC scenarios it matters little that climate scientists *know* that there is convection. Their so-called experiments are model runs with large raster sizes, and the current models have flaws which make them mispredict things again and again and again… (I’m using “predict” colloquially here, like other people use the word ‘trick’.) So it is important to understand that knowing about convection does not mean one is able to model it with todays technology. And i would love to see the journalists who get so many training sessions about how to report about climate change explain the shortcomings of the models to the interested public. But that would be jeopardizing the consensus, right? Can’t have that.
Science of Doom has responded, but I will add…
Re: DirkH…
1. The Earths atmospheric temperature profile has been known for many decades and the reasons for it have also been known. Convection has been a part of that knowledge. Basic school level films of the 1950s and other info include it.
It is ridiculous to state that knowledge is ignored today.
2. Climate models have included convection since the 1960s, when it was realised adjustments were required.
Jorge:
Can you provide a reference?
As I said in the article “Science is about what can be tested.”
I would like to test your claim but you need to be more specific than “you are sure” that climate science does ignore something.
DirkH:
Before discussing whether or not a subject is done well it seemed important to discuss whether or not it is done at all.
Many people – and a specific example given in the article – claim it is not done at all.
People provide a calculation of the adiabatic lapse rate and claim “climate science ignores this”.
This claim is not true.
The other claim, again demonstrated in the specific example, is that convection explains the “greenhouse” effect.
This claim is also not true.
Are you disappointed because you disagree with the content of the article or disappointed because you would like me to cover another subject?
SOD,
I beleive the claim is that climate science does not fully understand either concept and thus ignores its implications with respect to atmospheric physics.
This was never said. But your statement does emphasize your continued misunderstanding of the physics being discussed. You are not alone, unfortunately.
williamcg: strong assertions– not much backup. SOD has presented multiple citations for the basic physics. You claim there are fundamental and egregious gaps in work by climate scientists on these issues. I think you owe us some real examples from the literature.
From among the many people who think that climate science does not use convection properly or fails to understand convection properly, can someone provide one or more specific examples:
e.g.
In book x the author calculates y on page z which is incorrect
In book x the author fails to calculate parameter y which is needed to solve value z
In paper x the author does not understand parameter/concept y and so incorrectly calculates/fails to calculate z
And also comment on whether you believe this is because convection is an insoluble problem so that climate science is doing it to the best knowledge available OR whether they just don’t understand the basics.
Clearly many people have strong beliefs but I need testable claims. If it’s “I read it on another blog” go and ask the article writer for evidence.
And for people who wonder whether climate science understands this subject but have refrained from claiming when they don’t know for sure, it is also ok to ask:
“I think that atmospheric convection works like – is this how it is used in climate science?“
Can you not look into this for yourself? There are plenty of examples. What is the aim of this blog? Is it
(a) To investigate and discuss openly and objectively various issues of the science of the climate, or
(b) To defend unquestioningly the IPCC dogma and attack skeptics?
See the comment from Dirk H.
SoD
I am sorry that you did not go on to read the rest of my comment.
I was quite clear that no calculation of convective heat transfer was made in some 1D radiative models. Convection is acknowledged by imposing an external temperature profile based on a lapse rate that is maintained by convection.
I suggest that exactly this approach was taken by Ramathan and Coakley 1978 and also taken by Kiel and Trenberth 1997.
As you state yourself in CO2 – An Insignificant Trace Gas? Part Five “Note that, like R&C, they assumed a temperature profile to carry out the calculations because convection dominates heat movement in the lower part of the atmosphere.”
It seems we all agree that there is no calculation of energy moved by convection. Convection is just mentioned to justify the use of an adiabatic lapse rate.
I would argue that convection is ignored in the calculation except that a bare unspecified minimum is needed to allow the lapse rate assumption.
Jorge – your claim is false.
From Ramanathan and Coakley, Reviews of Geophysics and Space Physics Vol. 16 p. 465 (1978) we have:
Eq. 6: the “Thermodynamic Energy Equation”, includes all the relevant energy terms, and averaged becomes Eq. 7 which explicitly includes a term ‘q_c’, the convective flux.
Eq. 9 then states explicit formulas for ‘q_c’ under different vertical temperature gradient conditions.
Did you even bother to read the papers that you then claimed made “no calculation of convective heat transfer” ???
I can see why Scienceofdoom sounds irritated these days. People, back up your statements with actual facts, read the papers you’re critiquing, don’t just make stuff up out of thin air. This is really disturbing.
Arthur, don’t be so rude. The formula (9) is useless because it requires you to be able to measure the difference between the temperature gradient and the lapse rate, which you can’t do, because this difference is virtually zero.
Arthur Smith says: “Did you even bother to read the papers that you then claimed made “no calculation of convective heat transfer”?” One could ask the same of Arthur.
Immediately after Eqns 6 and 7 mentioned by Arthur, R&C admit that “Equations 7 and 8 do not fully defined the problem… Here we will be concerned with a formulation for the convective flux. An exact treatment for q_c would require the solution of the equations of motion and continuity… This ambitious task has not been attempted… In general, q_c is accounted for by semi-empirical or empirical techniques”
Eqn 9 mentioned by Arthur is a semi-empirical technique. R&C conclude the paragraph discussing Eqn 9 by saying that it “may be inadequate for treating convective heat transport under realistic atmospheric conditions”. Eqn 9 appears problematic to me because it says convection is proportional to the amount the lapse rate exceeds the adiabatic lapse rate – an inherently non-equilibrium treatment – in a framework assuming radiative-convective equilibrium.
R&C next say that empirical techniques for dealing with convection do what Jorge says they do. R&C: “The empirical technique … does not treat the convective processes explicitly. Instead, the effects of convection are included implicitly by assuming that convection maintains a critical temperature lapse rate within the convective region.” Jorge: “It seems we all agree that there is no calculation of energy moved by convection. Convection is just mentioned to justify the use of an adiabatic lapse rate.”
A public link to R&C 1978 is:
Click to access Ramanathan_Coakley_Radiative_Convection_RevGeophys_%201978.pdf
As for K&T 1997, Section 4 states: “The remaining heat flux into the atmosphere from sensible heat is deduced as a residual from the condition of the global energy balance at the surface”. K&T do cite other estimates of sensible heat flux, but two of their citations are older textbooks while the non-peer reviewed third citation probably involves estimates from surface balance considerations. http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.168.831&rep=rep1&type=pdf
AOGCMs try to calculate sensible heat flux from first principles (but the grid cells in these models are much too large match the size of real convection). Radiative-convective models treat sensible heat flux as a fudge-factor needed to balance upward and downward energy flux. There isn’t anything inherently wrong with this process, because the fudge-factor is attributable to a real energy flux, convection. The problems arise from a natural tendency to avoid discussing the fudge factor, not ignorance of it.
Hi Arthur – nice to meet up again.
You are quite right that I have not read the full R&C paper as I don´t download such large files on my dialup line.
From the abstract I see that they seem to have done a lot of calculations of convection and radiation. It looks good and I will try to find time to download it.
I was obviously mislead by reading the CO2 article by SoD where he says:
“But how do we include convection? If we don’t include it our analysis will be wrong but solving for convection is a very different kind of problem, related to fluid dynamics..
What R&C did was to approach the numerical solution by saying that if the energy transfer from radiation at any point in their vertical profile resulted in a temperature gradient less than that from convection then use the known temperature profile at that point. And if it was greater than the temperature gradient from convection then we don’t have to think about convection in this “slice” of the atmosphere.
By the way, the terminology around how temperature falls with height through the atmosphere is called “the lapse rate” and it is about 6.5K/km.”
Naively, it seems, I took this to mean that the lapse rate (known temperature profile) was used in the calculations rather than any actual values for convection.
Looking at the abstract for R&C does not cause me to change my mind. They say:
” It is shown that the value of the critical lapse rate adopted in radiative-convective models for convective adjustment is significantly larger than the observed globally averaged tropospheric lapse rate.”
That seems clear enough. They are using the lapse rate in the models as a convective adjustment which does not have to use the fluid dynamics needed to compute real convective heat flow.
Perhaps, Arthur, you might care to quote my remarks in full the next time as I think “no calculation of convective heat transfer” is not the same as the point I was making that “no calculation of convective heat transfer was made in some 1D radiative models.”
My comments have been fairly nuanced and my claim about ignoring convection made clear precisely where and how I thought it was being done.
If a reasoned comment is going to be construed as an attack on ignorant climate scientists by a stupid person I may be in the wrong place to continue my attempts to understand atmospheric physics.
Jorge:
You have a number of points in your comments. Let me try and address some of them.
1. From your initial comment it appears that you think climate science is failing to use “best practice” heat transfer knowledge.
I say “it appears”. But perhaps you can confirm what you think.
After all this article is primarily addressed at the idea that climate science doesn’t understand the subject. (This is the point of view of William Gilbert, for example, although he hasn’t provided any evidence).
2. You are correct that knowledge of the convective heat transfer is not needed to calculate the upward and downward radiation at each height in the atmosphere – the radiative-convective models.
What is needed is knowledge of:
a) temperature at each height in the atmosphere – to calculate emission at each height.
b) concentration of various “greenhouse” gases to calculate absorption at each height
Calculating convective fluxes is a very difficult problem.
If the problem (of calculating radiative changes from “greenhouse” gas changes) can be solved using temperature, which is much easier to determine, and a knowledge of the convective lapse rate then is this the wrong approach?
Do you think it is?
3. Back radiation can be easily calculated in exactly the same way (e.g., see Theory and Experiment – Atmospheric Radiation) and often is. I have seen (older) papers where vertical profiles of upwards and downward fluxes are given.
I don’t see any evidence that it is “not used”. It is easily calculated.
4.
This is correct and is the essence of the convective heat transfer problem. Convective heat transfer is calculated in practice using empirical formulas (aka “bulk formulas”) which also require knowledge of surface wind speed. Because we include latent heat loss we also need to know surface humidity.
But if we can calculate all the radiative terms (which we can) then the surface convective term appears as the balancing item.
5.
If you are talking about the 1d radiative convective models then the question is what is their job? And how do people use them?
Their job isn’t to completely represent the real world. Their job is calculate the changing energy balance in the climate due to changing “greenhouse” gases.
GCMs are a different proposition as you said.
Back to your original first “claim” –Climate science ignores convection.
There are many papers which are full of convection. Convective formulas, convective flux calculations and measurements.
One important question is “where do we need to know convective heat fluxes and is it ignored when it should be used?”
Another important question is “is convective heat transfer not understood by climate science to best practice methods in the wider science community?”
It seems like many people are saying yes to both questions. Jorge, perhaps you are not, perhaps you are.
Many people are saying yes to one or both questions, no doubt more people will be forthcoming with some evidence.
To be honest I’m a bit confused how it is easy to calculate the radiative heat transfer and at the same time it’s difficult to calculate the convective heat flux. The problem on the easy part is that it’s easy to calculate it *if* we know the temperature profile, however by definition in the case of warming – the temperature profile has to change somehow which would change the radiative heat flux and it’s all gonna get messed up. If we did calculate it then couldn’t we find the convective heat flux by using A-B (we have the temperature profile and radiative heat flux – the rest ought to be convective?).
Another thing I’m slightly confused about is how DLR is dependent on the amount of absorbing gases in the atmosphere (directly at least).As far as I understand, as long as there are enough particles of that kind in the atmosphere that the atmosphere is opaque, the atmosphere can roughly be viewed as a kind of a blanket (a thin layer with 100% absorbtion). Changing the concentration would only change how high the blanket is from the surface (what changes is the effective emitting altitude – the same way as with outgoing radiation). If what is being said is that the effective height will go so much lower due to concentration rise that the temperature at that height is higher as well (due to it being colder further away from the surface), then it makes sense – but I would call this an indirect effect from rising the concentration of absorbing gases.
What I’m trying to say is if the temperature at effective emitting height at concentration A is the same as the temperature at effective emitting height at concentration B is the same, the difference in concentration of A and B is irrelevant.
PS. I’ve made quite a few unwritten assumptions here, but I hope my question still shines through in this a bit somehow.
Mait:
I am in the process of writing an article on convection basics, maybe that will answer some questions.
Why is one heat transfer calculation easier than another?
That’s just the way it is.
The reason in essence is because radiation and conduction rely on knowledge of temperatures and material properties.
Convection requires knowledge of fluid flow. Turbulent fluid flow is difficult.
If we want to calculate heat transfer by radiation for 1 hour ago we just need to know a few parameters that we have measured.
If we want to calculate heat transfer by radiation in January 1st, 2020 we will need to be able to predict these parameters which then relies on being able to predict how the surface/atmosphere will change.
Two different problems. Now vs the future.
For your comment ending:
You can call it an indirect effect but we still need to know it. If we want to know DLR we need to know all the values that affect it.
I don’t understand this one at all.
PaulM:
It is very difficult to quantify.
Does this mean that climate science doesn’t understand it or is ignoring it?
Are you simply saying that this is a challenging area of climate science?
Or that climate science doesn’t understand the subject the same as heat transfer specialists?
What exactly is muddled in Part Five?
The 1d radiative-convective models are able to correctly calculate the OLR (top of atmosphere), the DLR (surface) and the spectral characteristics of the atmospheric radiation upwards and downwards. See Theory and Experiment – Atmospheric Radiation.
Do you believe they are “making it up”?
If they aren’t “making it up” how do they get it correct when they haven’t calculated the convective heat flux?
This is the benefit of the 1d radiative-convective model, and the solution to the problem was a big advance in atmospheric physics in the 1960s.
It’s not irrelevant at all. It appears that you think I am saying that convection is irrelevant?
No. I am simply saying that radiation is important because only radiation determines the energy balance of the whole climate system.
I’m sure we all agree on this point. So I can’t see how it can possibly be considered irrelevant to climate science.
OK, we are getting some points of agreement. It’s good that you acknowledge that it is difficult to quantify, though I wonder why you didn’t mention this in the original post.
“Does this mean that climate science doesn’t understand it or is ignoring it?”
I wouldnt put it in either of those crude terms. I would say climate science tends to downplay both the importance of convection in heat transport, and the difficulty of evaluating it. (IPCC AR4 Chapter 1: ‘radiation’ appears 44 times. ‘convection’ appears twice, both in the context of the ocean).
I will put comments on Part 5 over there.
On the last section, you’ve lost me completely. I never said anything about the DLR calculations being wrong or made up. I guess we are talking past eachother 😦
It can be difficult discussing complicated technical issues in this format.
SoD
Thanks for taking the time to respond.
You raise a number of points about what I think on a number of issues which I will try to answer.
1. On the question of whether climate scientists understand convection I would simply say they are in the same boat as everyone else. The problem is a very hard one to solve. To the extent that anyone knows how to do calculations involving convection and latent heat, they do what they can.
2. This is a bit more difficult to answer. The thing is that I agree we need to know the temperatures and compositions at all altitudes to calculate OLR from the surface temperature. I also think that real OLR measurements in the right conditions do confirm calculations of this kind. This seems to put us in the position of being able to say, with some confidence, that if the surface temperature and temperatures and compositions at altitude were to be so and so then the OLR would be such and such.
A lapse rate assumption is obviously helpful as it means that if we know the surface temperature then we know the temperatures at altitude. The problem seems to be that the lapse rate we should use is not obvious as it involves all the convective/latent heat/moisture content troubles that we have already agreed are hard to solve.
I am not putting this very clearly but I am trying to make the point that the lapse rate assumption is fine as long as you know the lapse rate.
3&4. There is no disagreement here. I made this very point about the ease of calculating back radiation when talking about doing a surface energy balance.
5. This is a bit tricky. I would say that the 1D radiative models allow one to calculate the OLR given the current state of the atmosphere (compositions and temperatures) but not much else. The reality is that changing the CO2 content will alter the state of the atmosphere in many ways and I do not see that these models can tell you what that new state will be. Once the new state is established the models can again be used to calculate the OLR. Presumably this will be the same as the original if the new state also represents an equilibrium.
There were some other question in this part but I hope my answers can be inferred from what I have already written.
Jorge: How do you avoid the conclusion that, since OLR will, in the end, balance incoming radiation, atmospheric changes that reduce OLR will lead to an increase in net energy (and thus temperature) in the climate system? Are you claiming that changes in convection will somehow ‘make up’ for the effects of GHGs on OLR? Or are you just claiming that we can’t be certain that such compensating changes won’t occur?
If the latter, how much weight do you think scientists and policy makers should give to such purely speculative possibilities? Our best efforts to understand what actually is and will occur have shown substantial skill wrt recent temperature increases as well as perturbations such as volcanic eruptions, show no signs of any such compensating changes, and have led to a very strong consensus among climatologists. Moreover, global temperatures are actually rising, as I’m sure you’re aware. My view is that idle skeptical speculations are not a constructive basis either for scientific purposes or for policy making.
Hi Bryson
Well cripes. I do my best to try to explain my understanding of what these 1D models can do and you go on about idle sceptical speculations.
Suppose the earth had an equal temperature all over the surface and an absolutely fixed moisture content and no phase changes anywhere. Under those asumptions I think it is very likely that an increase of CO2 would simply raise the surface temperature and all temperatures above. This increase can obviously be calculated iteratively by running something like Modtran with new surface temperatures until the original OLR is restored.
As soon as we allow the possibility of changes in lapse rate, moisture content, phase changes and clouds when surface temperature alters we cannot just use Modtran as above. Now we have to also recalculate all the temperatures and compositions at altitude before we can compute the new OLR.
The same would be true if we try to use a 3D version of Modtran except the difficulties are compounded as there are now a whole lot of sideways movements that have to be dealt with as well.
All one can do with radiative models is find the OLR when the state of the planet is known. If the surface temperature changes, the state of the planet has to be redone before applying the new radiative computation.
As far as I know there is nothing controversial about this. It is simply not the job of radiative models to compute how the state of the atmosphere might change with a change in surface temperature.
The whole question of how changes in the state of the system can be determined is really the same question as determining the size and direction of whatever feedbacks there may be. This does still seem to be a matter of controversy.
The case of the simple 1D model with the assumption that changes in surface temperature only affect the higher temperatures (via the same lapse rate) and everything else remains the same gives the standard 1ºC for a CO2 doubling. Given the assumption I see no reason to doubt that figure.
After looking at your comment again about the role of convection I see I have not really answered it directly. I do not think that the existence of convection means that an increased value can negate or enhance the effects of increased CO2. This is with the proviso that an increase in convection does not change the lapse rate or any other relevant system parameter.
Figure 2 of Ramanathan 1981 has a graph that shows how radiative forcing for 2X CO2 varies with altitude. Newer radiation calculations would change the magnitude of this forcings, but probably not their trend with altitude. If one accepted my suggestion that radiative forcing should be calculated at the boundary of the regions of the atmosphere dominated by convection and radiation (rather than the tropopause), that boundary would need to be at about 5 km for a significant part of the global to produce a 50% decrease in calculated radiative forcing. This doesn’t seem reasonable. My suggestion appears more likely to produce reductions up to 25%.
http://journals.ametsoc.org/doi/pdf/10.1175/1520-0469%281981%29038%3C0918%3ATROOAI%3E2.0.CO%3B2
Does the radiative forcing for 2X CO2 increase with altitude because of the overlap between the absorption of CO2 and H2O and the exponential decrease in water vapor with altitude in the troposphere?
Jorge,
This was helpful, but I’m still not as clear on your position as I’d like to be. As SOD’s work here illustrates, we often do use very simplified models to capture particular processes associated with some phenomena of interest. I’ve always understood simple 1D models in those terms. So I’ve been puzzled by the level of concern raised here about the features of the real world that they set aside.
WRT ‘real’ mid-term climate sensitivity, I appreciate that feedbacks aren’t entirely settled (hard to imagine how they could be). Of course some important positive feedbacks are pretty well grounded (increased water vapour, albedo changes, CO2 and methane emissions from melting permafrost regions, etc.). Further, arguments from paleoclimatological evidence seem to support the view that there is substantial overall positive feedback (else the Milankovitch cycles are insufficient to account for interglacial periods). So it does take a fairly hard skeptical turn to throw up our hands rather than just pursue the physics as best we can (unless your point is really just about the limitations of the simpler models you’re discussing): Combined with the success of models in capturing many climatological patterns and responses to some perturbations along with the recent warming trend and some of its observable features (polar amplification inter alia), this and other evidence (about which I’m no expert) has been the basis for IPCC assessment reports, which are, at least by some published measures, on the conservative side of the consensus of climatologists. Nothing we know at this point about convection or feedback effects suggests this work is on the wrong track (and no empirically well-founded model yet available does either, to my knowledge).
Anyway, I gather you’re saying that the lapse rate may change, in ways that are related to feedbacks in the climate system. But I don’t understand, first, how that’s related to your remark about other relevant system parameters except insofar as they affect the lapse rate. And I’m not quite clear on the significance you assign to point you’re making about possible changes in lapse rate or other features of the climate system. Suppose that all the 1D models do is illustrate the radiative implications of increasing CO2 levels. Surely that’s still a legitimate use for them, not grounds for arguing that climate science has been badly done.
Of course if some have been led to think that such models tell the complete story, that’s a mistake, and something that good communication about the science shouldn’t encourage. But that would only show that (on occasion, at least) climate science has been badly communicated. To show that climate science has been badly done would require showing that there’s substantial confusion in the relevant scientific community about what these models do and don’t show.
In response to my question asking for specific examples, PaulM said:
Obviously from this and other comments many people perhaps think that b) is the aim of this blog.
In many ways, the blog original aims get somewhat derailed by the various audience comments, which I feel I need to respond to.
There is huge confusion about basic science. There are frequent claims about the poor understanding in climate science of the basics. These get repeated in blog articles and comments on blogs – as well as on Science of Doom.
It’s not possible to provide coherent complete summaries on the state of the science for one subject in every aspect for a few reasons:
1) There’s a massive amount I don’t know. I read papers and textbooks to find out stuff I don’t know and it takes a lot of time.
2) It would take 100s, maybe 1000s, of hours to produce a complete summary of a particular topic like convection.
3) Almost no one would read the resulting super-long article.
Therefore I write articles about subjects I find interesting and about subjects my audience finds interesting or challenging or controversial. And I write articles about subjects I find less interesting but where the subject has apparently widespread confused or false claims.
For those who think that the objective of this blog is to:
Don’t read it. Take Science of Doom off your blog reader and unsubscribe.
For those who are prepared to consider that I might be trying to discuss climate science and find out what is true, false or uncertain please keep reading.
I never expected to write as many articles about the Second Law of Thermodynamics. Readers comments have “inspired” them. They were not written to avoid a particular “more important subject” that some readers wished were discussed.
If you have a subject that you really want discussed then ask. I already have quite a list. But perhaps your request might move up the list. Perhaps not.
And before writing comments attacking people’s motives, take another look at About this Blog.
Bryson
I would prefer to stay away from the more general aspects of how well extremely complicated models perform. It is a big subject and I am still trying to solidify my understanding of simpler things that are related to basic physics that are within my comprehension.
I was muddled when talking about lapse rate and other system parameters. I should have said:
This is with the proviso that an increase in convection is not accompanied by a change in the lapse rate or any other relevant system parameter.
Sorry for the confusion.
Yes, I do think 1D radiative models are useful for relating OLR to the state of the system. It has taken me some time to realise that one cannot do the reverse and go from OLR to the state of the system with this simple type of model. I may have been fooling myself because this is exactly how the standard no feedback climate sensitivity is calculated. I can now see that this is a special case theoretical construct.
Yes, this is one of the key points, and it was made back in May by Frank, see CO2 Part Five,
Frank:
“…temperature is controlled by a variety of factors including latitude, convection and radiation and certainly can’t be calculated directly from radiative equilibrium.”
May 24, 2010 at 9:57 pm Frank
“You won’t know the temperature unless you take convection into account. “
Frank,
If I look only at the change in upward LW radiation with 2X CO2 using MODTRAN, I get a steady increase in the lower atmosphere, relatively constant in the region of the tropopause and reduction in the stratosphere (graph here). However, the difference is always positive. At the surface looking up, there would be an increase in LW radiation but a decrease in SW radiation so the net forcing at the surface is small. Looking at the graph, it seems logical to me that the chosen forcing is where it changes the least with pressure. That’s approximately the tropopause.
I would need a better model that included SW absorption and convective heat transfer to calculate layer by layer forcings. Although, the assumption that the lapse rate won’t change takes care of convective transfer and is a very good first approximation since we’re dealing with fairly small temperature changes compared to diurnal and seasonal changes.
Thanks. To see the negative forcing associated with stratospheric cooling, it may be necessary to look at the downward flux as well as the upward. For 2X CO2 to cool the stratosphere (which happens above about 50 mb according to the IPCC’s fingerprint), the stratosphere must be warmer (from absorbed UV) than the regions of the atmosphere with which it is exchanging photons. In the stratosphere, most of the radiative cooling and absorption presumably is at wavelengths absorbed and emitted by CO2 (but radiation flowing directly from the surface to space in the atmospheric window may be the largest the flux, but this doesn’t change with CO2 until the surface warms).
I think we should calculate radiative forcing (and a no-feedbacks temperature change) at the locations in the atmosphere that are most RELEVANT to the temperature change that will occur at the surface below – not at an altitude where forcing doesn’t change with pressure or where forcing is greatest. From my perspective, that altitude is at the boundary of the region dominated by convection and region dominated by radiation. From that altitude, a lapse rate controlled by g/Cp (not GHGs) will translate the temperature rise at the boundary to the surface. However, no one has made any favorable comment about this proposal, so perhaps I have “lost it” on this subject. However, I have made a little progress in getting others to realize that radiative forcing does change substantially with the altitude.
But that’s precisely where the forcing reaches a maximum and becomes fairly constant. If I add temperature to the plot, the breakpoints on the forcing and temperature curves happen at the same altitude or pressure. Above that altitude, the temperature and forcing are relatively constant until the temperature starts to increase because of SW absorption.
A nice outline of the case for both the basic CO2-driven radiative forcing and how various feedbacks add up to the (widely accepted) 3 degrees C (with a range rougly from 2 to 6 or so) as the result of doubling CO2 (from pre-industrial levels) i.e. about 560ppm can be found at RealClimate.
Whether you accept the argument as outlined or not, it’s a good place to start working through it.
Frank:
I hope to write an article with a bit more content on this subject as you have already asked some questions about it.
But in brief, you are correct, the value of “radiative forcing” depends on altitude.
I commented in response to a similar question on Judith Curry’s blog:
The point being that so long as you know your frame of reference for calculating the radiative forcing you don’t have a problem.
SOD: The only frame of reference that is really important is surface temperature. As long as there is a fixed relationship between temperature at the surface and temperate at the altitude where radiative forcing is calculated, everything will be fine. However, if we calculate radiative forcing in the stratosphere, where 2X CO2 is expected to reduce temperature, we probably will predict global cooling rather than global warming. Likewise, if we calculate radiative forcing at the surface, where radiative equilibrium doesn’t determine temperature, the temperature change we calculate from that forcing won’t mean anything.
One interesting question is why does the radiative forcing for 2X CO2 vary with altitude. I suspect the overlap between the absorption of water vapor and carbon dioxide may play some role. But radiative flux also depends on the temperature of the regions that are exchanging photons. One explanation I have heard is that 2X CO2 is supposed to cool the stratosphere because more CO2 molecules radiate energy away from the relatively warm stratosphere, but don’t absorb as much from the cooler tropopause.
SOD, re. distance to the centre of New York: I’ve used a similar approach in explaining STR to non-science students: each observer agrees with the others about the different results they get for various kinds of measurements (and, of course, about invariants). But I’m not as clear on how it applies in this case. I thought that the top of the troposphere made physical sense as a place to evaluate radiative forcing, since below that level convection moves heat energy around in the climate system, while at that level (radiation in – radiation out) determines the overall forcing, i.e. the net increase or decrease in energy in the climate system. Of course other figures for radiative forcing at various altitudes are perfectly sound physical facts. But they don’t tell us as much about the energy balance of the climate system, do they?
Bryson Brown:
I agree with you. But the tropopause has different definitions, and with those different definitions the value of radiative forcing, of course, changes.
In the end radiative forcing is a “headline figure” and the heating and cooling rates vertically through the atmosphere are the real information needed.
But it doesn’t change much however you define the tropopause. That’s not true anywhere else.
Reply to DeWitt 12/18/10 4:39pm: I really appreciate seeing the data you posted. (I think there may be an extra “i” right after “http://” I was able to see the plot after removing it.) I hope you won’t mind a few questions: 1) Where does the temperature vs altitude data come from? Is this some sort of a global average profile? 2) Are you still showing just the change in upward LW flux (not the net change/forcing)? Is this also some type of global average that makes sense to compare with the temperature profile?
It sure looks like the radiative forcing at 12-13 km – where the constant lapse rate ends – is essentially the same as at the tropopause, which might be anywhere from 13-18 km on your graph. If generally true, this would mean the there is no practical difference between the IPCC definition for radiative forcing (and the no-feedbacks temperature rise that could be calculated from it) and the alternative I prefer.
An edit and preview function would be nice. If the site could be made compatible with CA Assistant you would get all that. While you’re at it, how about TeX too.
The link to the graph works for me in Firefox.
The temperature profile is for mid-latitude summer. In this case it’s from David Archer’s MODTRAN site. If you elect to save text and then click on the link below the graphs, you get all sorts of data including the spectral data.
And it is the difference in LW up only. Near the tropopause, that is the net effective forcing as the change in SW is minimal and there is very little downward LW.
SoD,
I’ll try to explain the point I was trying to make with the emitted radiation earlier.
What I meant is that if we take a box which doesn’t emit any radiation (let’s call it a whitebox) and fill it with two different gas mixes which are kept at the same uniform temperature (probably by magic, but this is quite important simplification for the point I’m trying to make), the radiation incident on the walls of the box is the same regardless of the concentration of absorbing/emitting gases in the mix, as long as the opacity of the gas is low enough so the gas is opaque in smaller distances than the distance between the walls.
This would mean that the DLR on Venus for example wouldn’t change that much even if we halved the concentration of CO2 on the condition that we leave the temperature profile the same.
This was covered in the Venus discussion. The height of the atmosphere is a large factor in the magnitude of the greenhouse effect. Halving the concentration of CO2 while maintaining the same surface pressure does indeed make only a small difference in Venusian DLR and probably surface temperature at steady state. Similarly, doubling the surface pressure of the Earth by increasing the mass of nitrogen while maintaining the same partial pressure of CO2 would result in significant warming.
[…] Things Climate Science has Totally Missed? – Convection […]