Model CO2 as Greenhouse Gas: Tips & Results

[sylas]

Andre, the critical point is this. The no-feedback response to a forcing ... ANY forcing ... is NOT calculated by considering the atmosphere to be frozen in place. Neither does it work by presuming that the convective energy flux is fixed. It is about how the surface and the atmosphere together changes in temperature in response to additional energy, while holding the composition of the atmosphere and the surface properties fixed.

The whole of this account falls apart right at the start where you treat convection as something separate from the basic way in which an atmosphere establishes a temperature profile in the first place!

YOU said the IPCC ignored this process. I say that's absurd; because it's already built into first level of response of the atmosphere to a forcing.

If you think this convection result is a feedback, what on Earth do you think a non-feedback response looks like? You've been talking about "Planck response" before. Do you know how such thing is calculated? Do you even mean the same thing by that phrase as in a textbook on atmospheric physics?

Is convection / advection feedback or not?

It is not.

Feedback is...

Well, convection is an output of the Earth surface warming up, which is the output of the sun warming the surface. So convection is output. It's effect is to take energy/heat away that is in direct contact with the Earth surface. This air is replaces with cooler air which in turn does return less radiation to the surface, and hence is 'returned to its input in order to regulate its further output'. And since the sign (cool air) is opposite to the original input (warming sun) it is negative feedback.

It is fundamental to how the atmosphere establishes ANY temperature profile, and so is part of the non-feedback Plank response. Convection works faster than radiation in an unstable atmosphere which is hotter at low altitudes so as to give an unstable lapse rate.

It would make no sense at all to try and figure a temperature response of the atmosphere by radiation before convection comes into play. That's why convection is not considered feedback. It's part of the base response.

That's always how it is calculated. Nothing else would make sense. Everytime you see mention of "Plank response" or "no feedback response" in climate science, you are looking at a response of the atmosphere including the normal lapse rate.

Cheers -- sylas

 

[Richard111]

Gosh, lots to absorb. Having once held a PPL I have experienced some lively dynamics in the atmosphere, especially in the tropics.
To push the topic slightly and help my level of understanding; what happens if we move both columns over the sea? Upwelling radiation will be identical, returning radiation identical, but NO DELAY IN WARMING OF THE SEA SURFACE as longwave radiation only penetrates a few tens of microns of the sea surface. What happens with convection now?
I understand there is a slight increase in evaporation which should lead to a small cooling effect at the surface.

I have read somewhere that the total daily average of global sea surface evaporation is a about 2mm. I make that 2 litres or should we say 2kg of water vapour every 24 hours into each column. Good thing it soon reaches dew point and comes back down as rain.

edit: whoops... NO DELAY IN COOLING OF THE SEA SURFACE

 

[Andre]

YOU said the IPCC ignored this process. I say that's absurd;

Thanks for the nice strawman example. I have not said that. What I said exactly is:

Hence the upper levels hardly cool at night as the only cooling mechanism is ... greenhouse effect, radiation out. And at those levels, with strongly reduced water vapor, radiation escapes to outer space much easier. This effect appears to be neglected in the IPCC endorsed literature and if you don't account for it in the models, you're basically stuck to the GIGO principle.

Evidently, I'm talking of the convection energy/heat conveyer belt, nett transporting the heat energy one way, up.

because it's already built into first level of response of the atmosphere to a forcing...That's why convection is not considered feedback. It's part of the base response...

So we have a contraction here. If convection is a response to a forcing, then it is an output and any effects of the output of that system on the input/forcing thereof is feedback by definition, regardless of the time constant/delay, be it anywhere between micro seconds and millenia

Anyway it appears, that the way that Ray approaches convection, is slightly different than it used to be in the time that I learned it for my air glider license some 40 years go and how I teached it for my student pilots some decades ago or so, and the way it was toughed by eminance grise Richard Lindzen explaining...

..why simple radiative models with convective adjustment prove inadequate - qualitatively or quantitatively..

So, if I understand it correctly, in the radiative models, convection is already incorporated in the radiative models, adjusting the local lapse rate abarration. Nothing more. While I maintain that convection may play a prominent role in the energy / heat transport from the Earth surface to out radiation into space. The obvious difference in both approaches is whether or not radiation provides some quantitative (negative) feedback on variation in the input (forcing function) of the total Earth - surface / atmosphere system.

So if we can continue in a fallacy free fashion and consider this as the two hypothesis approach (1 radiation, 2 convection) and see what the evidence does, supporting either or both one or the other.

Before doing that, we need to complete the convection hypothesis with the water cycle / latent heat factor. Due to the large evaporation / condensation heat of 2.27 MJ/kg the transport of latent heat from evaporating at the surface and condensation in higher level into clouds, significant amounts of energy are transported aloft, that can radiate out more easily at higher altitude than at surface level. So, whenever the forcing function causes the surface to heat up more, the convection/latent heat conveyer belt will speed up (Claussius Clappeyron), transporting more energy aloft, increasing the out radiation, while reducing the heating at the surface. An additional factor is that more clouds reflect more short wave energy back to space and also reduce the surface heating.

Remember this is a (albout decades old) hypothesis that requires a prediction for either support or falsification, not the declaration that it is not science.

Also again, remember that one piece of not unimportant evidence of the energy transport associated with convection and latent heat, is in the raging power of tropical storms and tornadoes. Energy that has to orginate from somewhere.

So we need a prediction, in this case I predict that a study will be published in the near future about direct measurements of outgoing long- and short wave radiation, that would fully support the occurance of negative feedback on the variation of forcing functions for surface heating and convection, which is most prominent in the tropix.

And then we have not begun to consider the effect of advection when warmer air masses lifts up when colliding with colder air masses, which is certainly not a instantenous process.

 

[sylas]

Thanks for the nice strawman example. I have not said that. What I said exactly is:

Evidently, I'm talking of the convection energy/heat conveyer belt, nett transporting the heat energy one way, up.

So am I. Honestly. This is not a strawman; we really and truly are talking about the same thing, and it is not ignored by the IPCC or in perfectly conventional atmospheric physics. I'll emphasize this next paragraph as my main point:

Convection is the primary process by which the atmosphere maintains a temperature profile, and that means convection is part and parcel of the basic no-feedback Planck response of temperature to change in energy input.​

The Planck response is a new equilibrium temperature in response to an energy change, where the only thing that changes is temperature... and, of course, the energy fluxes associated directly with temperature. You can't have more temperature without having also changes in the associated energy flux -- the whole point is for temperature to increase to the point where energy balance is restored again.

In an atmosphere, we speak of the temperature at a given altitude, or pressure level. But the temperatures in an atmosphere at a given level don't just change by cooling or heating up air held in place. In an atmosphere, changes of temperature at a given level occur to a large extent by movement of air in and out of that level. Therefore this is already part of the basic Planck response, right there. The new equilibrium state includes a new flux of energy, both by movement of air (convection) and by radiation, so as to maintain energy balance. That's what equilibrium means.

For a planet, like Earth, with a gas that condenses in the atmosphere, there's a latent heat component as well, which shows up by using a moist adiabat rather than a dry adiabat for the temperature profile maintained by convection.

An aside on respect

Given the remark on strawmen, I'm feeling the need to back up and make an entirely secondary point. There's nothing personal about this. I'm inclined to like you, Andre. My reaction to you as a person has been positive. I don't want that to change, and I'd like you to work with me on that, please.

But it makes no difference to how I deal with differences over a topic in physics.

If I am wrong about something here, I am wrong honestly. I assume it is the same for you. I won't dishonour you, or me, by trying to raise strawmen or ignore your point. I am doing my level best in all honesty to give your account the fair and honest assessment you deserve, as a debating colleague serious about dealing with the physics. But I don't give automatic respect to the ideas themselves; and I don't mix up my respect for you as a person with a presumption that that your views on atmospheric physics must be legitimate scientific objections.

If you think I've missed some fundamental point, you can rely on this personal regard and try to put me back on track; as I am doing here also. I may not accept the validity of your corrections, but I accept your participation in good faith.

In all honesty -- you are badly mistaken here. You are describing something which is not ignored in conventional atmospheric physics at all. The physics of convection in your account seems perfectly adequate, and I not saying you misunderstand convection itself. The problem is where your description relates convection to greenhouse effects, and the notion that there's some convection related feedback that is ignored by the IPCC, or which stands as a useful new insight for atmospheric physics. This particular notion is of a kind that normally speaking would not be permitted in the forums, except for one curious fact -- the notion is not actually from you personally, but really does appear in the scientific literature.

There's a whole secondary debate about how on Earth this paper got published at all; but it's secondary, and not really appropriate here. Here we should stick to the physics. My own rule of thumb with respect to scientific literature is roughly in two parts:

• Sticking to ideas expressed in scientific literature is a good way to weed out a lot of ideas that lack any actual scientific merit.
• The scientific literature is not perfect; and sometimes even really fundamental errors slip through, that should have been picked up in review, but somehow were not. Hence there's no automatic presumption of merit to an idea that has got through the initial hurdle of publication.

Regardless of where it appears, the account you have given of how convection and greenhouse and feedback are related is incorrect, and it should be recognized as wrong by anyone who has dealt with that part of atmospheric physics that deals with the details of radiative and convection energy transfers in the atmosphere in the light of energy balance with the Sun. The guts of the error is to think that convection gives an additional feedback over and above the normal Planck response. It most definitely does not. This is a specific criticism of the idea on its own scientific merits -- or lack thereof -- and not a mere presumption because I am predisposed to reject the conclusion. The actual argument itself is fatally flawed on its own merits... and the flaw shows up originally in your reference, which is worse than useless for understanding atmospheric physics.

Convection and the adiabat

Anyway it appears, that the way that Ray approaches convection, is slightly different than it used to be in the time that I learned it for my air glider license some 40 years go and how I teached it for my student pilots some decades ago or so, and the way it was toughed by eminance grise Richard Lindzen explaining...


So, if I understand it correctly, in the radiative models, convection is already incorporated in the radiative models, adjusting the local lapse rate abarration. Nothing more. While I maintain that convection may play a prominent role in the energy / heat transport from the Earth surface to out radiation into space. The obvious difference in both approaches is whether or not radiation provides some quantitative (negative) feedback on variation in the input (forcing function) of the total Earth - surface / atmosphere system.

These two are the same thing. Convection maintains the lapse rate by transporting energy upwards in response to other changes. When you say "nothing more than adjusting lapse rate", this is the same as saying "nothing more than increasing the convective energy flux to balance the additional radiative cooling".

The error here is failing to recognize that the basic Planck response ALREADY includes the effects of convection to transport energy upwards so as to maintain an essentially adiabatic lapse rate in response to radiative cooling. If you try to propose a feedback adjustment to Planck response based on convection, you are double counting the process. If you try to isolate a response with convective energy transport held fixed, then you doing something frankly bizarre and certainly nothing like what anyone else in atmospheric physics calls the Planck response. I am completely positive that nothing in what you learned from basic atmospheric physics ever did anything like that.

The case of an optically thin atmosphere, also mentioned with a diagram in [post=2299540]msg #125[/post], is a good starting point. If the only mechanism of energy transport was radiation (in an optically thin atmosphere with only a very small amount of thermal radiation being absorbed [added in edit]) then the atmosphere would tend to be isothermal, at about 85% of the absolute surface temperature; with an effectively infinite lapse rate at the temperature discontinuity at the surface. Convection causes air heated at the surface to rise, and the process continues until you have a lapse rate up to the level where you intersect the "skin temperature" that would apply without convection.

In this atmosphere, there is a negligible net convective energy flux upwards, by the first law of thermodynamics, because there's nowhere for the energy to go. The atmosphere radiates very little of the energy; it is nearly all coming from the surface. There's turbulence and so on as the atmosphere repeatedly adjusts to variation, like the diurnal cycle for example; but the net energy flow upwards is the radiation from the surface that goes straight out to space, with very little net convection energy flux on top of that.

Now add a large pulse of some greenhouse gases to this atmosphere. Suddenly, it is no longer optically thin. There's a significant amount of absorption and emission of radiant energy going on. This is going to raise temperatures, but as the atmosphere changes in temperature, convection continues to work, as always, to maintain the lapse rate. When it has come back to an energy balance with the Sun again, you have some radiant energy going out into space from the atmosphere, which is cooler than the surface. The total energy to space is the same as before, so the surface has heated up... and hence (because the lapse rate is still the same) the atmosphere is warmed as well, by the same amount. The tropopause will be at a higher altitude than before.

At this point, there can be a net flux of convection going upwards, all the time, because now there IS a way for this energy to be lost. There's a continuous net flux of energy from the surface to the atmosphere, partially convective, partially radiative. We expect the new atmosphere to be a bit more turbulent than before, with the potential for a net convective flux that was not there before.

The crucial thing is... this IS the no-feedback response. This is how an atmosphere heats up. When you see a calculation of no-feedback responses, this is how it is done; it's about how the atmosphere heats up in response to energy, and convection is built into atmospheric heating because anything else would be unstable. There is no additional feedback from convection that is any different from the basic way in which the Planck response is calculated.

Calculating the equilibrium response to interactions of the atmosphere with radiation

You mention Ray's treatment of convection. Let's make that more concrete with some extracts. The text is "Principles of Planetary Climate", by R.T. Pierrehumbert.

I'll give a couple of extracts, in a blue font. These are largely from chapter 4, which is very technical and demanding. It relies on more basic material covered in chapters 2 and 3. I've tried to show where convection appears in the energy balance work as a kind of demonstration that it is just wrong to propose an additional feedback on top of what is already done for finding the base equilibrium temperature response to forcing.

Page 155:

The main reason for dealing with radiative transfer in the atmosphere is that one needs to know the amount of energy deposited in or withdrawn from a layer of atmosphere by radiation. This is the radiative heating rate (with negative heating representing a cooling). It is obtained by taking the derivative of the net flux, which gives the difference between the energy entering and leaving a thin layer. The heating rate per unit optical thickness, per unit frequency, is thus

[... equation 4.14 ...]​

This must be integrated over all frequencies to yield the net heating rate. For making inferences about climate, one ordinarily requires the heating rate per unit mass rather than the heating rate per unit optical depth. This is easily obtained using the definition of optical depth, specifically,

[... equation 4.15 ...]​

When integrated over frequency this heating rate has units W/kg. One can convert into a temperature tendency K/s by dividing this value by the specific heat c_p.[/color]​

The crucial point to note here is that the radiative energy transfers are not balanced. There IS heating and cooling going on, and hence there has to be another process involved to balance up the energy. There is. In the lower part of the atmosphere, the troposphere, this is (mainly) convection and latent heat for any condensable gases in the atmosphere; and the effect of radiation transfer is a cooling effect on the atmosphere.

Page 163-164:

An examination of the radiative heating rate profile for the all-troposphere case provides much insight into the processes which determine where the troposphere leaves off and where a stratosphere will form. We’ll assume that I_-,∞ = 0 and that the turbulent heat transfer at the ground is efficient enough that T_sa = T_g. Consider first the optically thin limit, for which the grey gas version of Eq. 4.28 is

[... equation 4.36 ...]​

assuming the stated boundary conditions. Since the radiative heating rate is nonzero, the temperature profile will not be in a steady state unless some other source of heating and cooling is provided to cancel the radiative heating. According to Eq. 4.36, the atmosphere is cooling at low altitudes, where T > T_g/2^1/4, i.e. where the local temperature is greater than the skin temperature. The cooling will make the atmosphere’s potential temperature lower than the ground temperature, which allows the air in contact with the ground to be positively buoyant. The resulting convection brings heat to the radiatively cooled layer, allowing a steady state to be maintained if the convection is vigorous enough. However, in the upper atmosphere, where T > T_g/2^1/4, the atmosphere is being heated by upwelling infrared radiation, and there is no obvious way that convection could provide the cooling needed to make this region a steady state. Instead, the atmosphere in this region is expected to warm until a stratosphere in pure radiative equilibrium forms. Indeed, the tropopause as estimated by the boundary between the region of net heating and net cooling is located at the point where T(p) equals the skin temperature; this is precisely the same result as we obtained in the steady state model of the tropopause for an optically thin atmosphere, as discussed in Section 3.6.[/color]​

The crucial point here is to note that the troposphere is where you have convection, and you have convection to maintain a balance of energy. This region of the atmosphere has a net cooling effect from radiation, and net heating from convection.

It's important to note that "heating" and "cooling" here does not refer to temperature changing, but to the sign of the energy flux into and out of a given level of the atmosphere. The equilibrium response being calculated here has energy in balance, which means that radiative cooling is balanced by convective heating, in the troposphere.

It's also important to note that "warming" in the sense of climate attaining a new higher equilibrium temperature works side by side with "cooling" in the sense that a greenhouse effect involves the atmosphere shedding energy (cooling) by radiating in the infrared. The "warming" in "global warming" is a comparison of the different temperatures in two different equilibrium states. The "cooling" is a reference to the net flux of radiant energy when the new higher temperature equilibrium is attained.

Here are two other quick extracts making rather basic points about atmospheric greenhouse effects:
Page 122:

It is very important to recognize that greenhouse warming relies on the decrease of atmospheric temperature with height, which is generally due to the adiabatic profile established by convection. The greenhouse effect works by allowing a planet to radiate at a temperature colder than the surface, but for this to be possible, there must be some cold air aloft for the greenhouse gas to work with.[/color]​

Page 256:

... One calculates the adiabat T_ad(p) corresponding to the ground temperature T_g and surface pressure p_s. Then at each timestep, wherever T(p) < T_ad(p), the temperature is instantaneously reset to T_ad. The rationale for doing this is that convection is a much faster process than radiative relaxation, and that wherever the temperature is below the adiabatic temperature, air parcels starting at the ground have enough buoyancy to reach that level, mixing air all along the way. ...[/color]​

That is: when you are calculating the equilibrium response, you use the adiabat, because that is what convection gives you.

Cheers -- sylas

 

[sylas]

I started this post some time ago, as a reply to Richard's thought experiment, but Andre beat me to it and since then we've had an interesting and useful debate over convection as feedback. This reply goes back to Richard's original post, which outlines a nice case for consideration.

Thanks for that. The Gas Laws rule.
Having no formal education in this subject I must rely on intuition and common sense (I hope) and reading blogs.

Another good option is reading some introductory texts on atmospheric physics. There's a good text on available online that I have recommended and used in this thread, and which I have been using offline to learn more about the subject. It's quite demanding, but you can get a lot just from the early chapters. It is "Principles of Planetary Climate", by R.T. Pierrehumbert at the Uni of Chicago.

The title of this thread, "Can You Model CO2 as a Greenhouse Gas", caught my attention because of a thought experiment I have been musing on. Imagine a column of air on a one square meter base. Accept the assumption that the sides of the column are impervious, no energy in or out. We know the mass of the column, about 10,333kg, the mass of contained CO2 about 4.13kg (0.04% say), and for water vapour we can choose any value from zero to say 4% and assume the base temperature and dry lapse rate is selected to ensure no physical water droplets will be in the column. We may need to limit our attention to some defined height of the column, say 300mb level or so.

Minor correction here: CO2 is about 0.04% by volume; so you have to scale by 44/29 (the molecular weight of CO2 and the average molecular weight of air) to get pretty close to 6 kilograms.

Removing the water vapour makes two significant differences from Earth's real atmosphere, even in clear sky conditions. First, the infrared absorption is much reduced. Second, the lapse rate is increased. But this is still a great example for clarifying the relevant physics.

Having defined the properties of our column (heh!), we consider the nature and properties of the base. We are free to choose water, land, grass whatever. Initially I have chosen a "greybody" with a surface temperature of 15C. Now the pips begin to squeak.
Ignoring convection from the surface and assuming the only "greenhouse" gas present is CO2 we should be able to surmise how much radiated energy is intercepted by the CO2, how much is transferred to the surrounding air molecules and how much is reradiated up and back down.

Definitive information on how long any CO2 molecule can remain in its energised state seems hard to come by. It would seem that at high densities, low altitude, where molecular spacing is closer, transfer by conduction is more likely. At higher altitudes the molecule may radiate a photon before encountering an air molecule. At this point my confusion index starts rocketing. Does the molecule radiate an equivalent photon? Or will the "new" photon be at a different wavelength/frequency? Anyway, to my thinking, (assuming there is no such thing as a free lunch) the "rate" of radiation will be less as the atmosphere cools with altitude.

Nice thought experiment. It is, of course, greatly simplified; and that's a good way to get at the physics of the situation.

It turns out that how much radiation gets absorbed is less important than WHERE it gets absorbed and where it gets emitted. The time in the "energized" state is not actually a useful quantity here.

With thermal absorption, the major effect is vibration of the molecules. This can last quite a while; generally long enough that the molecule collides with another molecule, and transfers kinetic energy. In brief; what happens when infrared radiation is absorbed is that the gas heats up. And, similarly; by virtue of having a temperature, the gas will radiate thermal radiation, in the same wavelengths that are absorbed.

So really, what you need is the temperature of the gas, and also the mean path length of a photon -- which is frequency dependent. CO2 is pretty much opaque at certain wavelengths, and pretty much transparent at others. The surface is very close to a blackbody, as far as thermal radiation is concerned. Surface emissivity is up around 0.98 or so, for most surfaces, at thermal wavelengths. So at 15C, you can pretty much use the blackbody emission spectrum, with 390 W/m² of energy and peak wavelengths around 3.7 microns. Differences in the surface are more significant for shortwave reflection (albedo) than for longwave emission.

The wavelengths where CO2 is opaque are the same wavelengths that get emitted as thermal radiation. This is Kirchoff's law; at any wavelength, the emissivity of a material is the same as its absorptivity.

Here are the absorption spectra for major gases (http://www.iitap.iastate.edu/gccourse/forcing/spectrum.html )
http://www.iitap.iastate.edu/gccourse/forcing/images/image7.gif

(Caution. The reference used here is some course notes from about 1997. I'm using it because the absorption spectra are shown nicely and clearly, and they are still accurate. But if you dig into the reference, some of the more detailed calculations are a bit out of date, and use CO2 forcing values that are about 20% too large. They were revised downwards from about 6.3 to 5.35 W/m²/Ln(CO2), as described in Myhre et al 1998. My posts in physicsforums have consistently used the more recent value for estimating forcings, as have IPCC reports from the third assessment onwards. In this post, we are using a simple zero-moisture example, which is going to be less accurate still, but hopefully useful for looking at the techniques.)​

I think I'll stop here. We know how much is being radiated up from the surface, we know CO2 can absorb at 2.7, 4.3 and 15 micrometers (µm), (I understand that this equates to about 8% of the available outgoing radiation). We do not, at this moment, know exactly how much is converted in heating the surrounding atmosphere. The remaining energy can be radiated isotropically such that about half will return to the surface.

So my present understanding is that under ideal conditions any surface radiation can expect something less than 4% of its output back again due entirely to CO2 thus slowing down the cooling of the surface by that amount.

The thing is that the backradiation you get coming down depends on the temperature of the gas rather than how much energy was absorbed to heat it up.

Here's how I'd do the analysis. To keep it really simple, I'll assume a dry well mixed atmosphere, and a surface radiating like a blackbody, with a fixed adiabatic lapse rate and the bottom of the atmosphere thermally coupled to the surface, so they are the same temperature.

Rather than talk about a percentage of radiation absorbed in total, I would consider how far radiation goes before being absorbed. We can do this by subdividing the atmosphere into a series of "slabs", each of which absorbs a very small amount of radiation. The total effect can then be obtained by an integration step, and from this you can infer fluxes of energy all up and down the column.

I had actually started this, but I'm short of time at present so I'll leave it here temporarily; and return to the thread when I have a bit more time.

Cheers -- sylas

 

[Andre]

...So we need a prediction, in this case I predict that a study will be published in the near future about direct measurements of outgoing long- and short wave radiation, that would fully support the occurance of negative feedback on the variation of forcing functions for surface heating and convection, which is most prominent in the tropix.

http://www.agu.org/contents/journals/ViewPapersInPress.do?journalCode=GL

Full paper http://www.leif.org/EOS/2009GL039628-pip.pdf

Abstract

Climate feedbacks are estimated from fluctuations in the outgoing radiation budget from the latest version of Earth Radiation Budget Experiment (ERBE) nonscanner data. It appears, for the entire tropics, the observed outgoing radiation fluxes increase with the increase in sea surface temperatures (SSTs). The observed behavior of radiation fluxes implies negative feedback processes associated with relatively low climate sensitivity. This is the opposite of the behavior of 11 atmospheric models forced by the same SSTs. Therefore, the models display much higher climate sensitivity than is inferred from ERBE, though it is difficult to pin down such high sensitivities with any precision.

Results also show, the feedback in ERBE is mostly from shortwave radiation while the feedback in the models is mostly from longwave radiation. Although such a test does not distinguish the mechanisms, this is important since the inconsistency of climate feedbacks constitutes a very fundamental problem in climate prediction.

 

[sylas]

http://www.agu.org/contents/journals/ViewPapersInPress.do?journalCode=GL

Full paper http://www.leif.org/EOS/2009GL039628-pip.pdf

It should be noted that Lindzen is a completely different kettle of fish from Chilingar, and that this paper has nothing to do with Chilingar's errors, or with the mix up over convection and feedbacks.

This paper is not about the greenhouse effect, or CO2, but rather about climate sensitivity.

One of the difficulties in this whole topic area is focus. I'd really like to avoid having every thread that mentions CO2 go off into every aspect of climate or a fight over AGW and global warming.

This thread, as I understand it, is about the basics of a greenhouse effect, and how physically CO2 in particular works to give additional energy and heating on a planet.

The point about feedbacks is that they are OTHER factors that apply for anything altering energy balances to help moderate or enhance the temperature.

Convection is not a feedback; it is a part of the basic Planck response to temperature change. Then, on top of the base response, there are changes in cloud cover and humidity in particular (the major source of the feedbacks being considered in the above paper) and they are NOT part of the Planck response; they give new secondary forcings in their own right on top of any primary forcing that initiates a change.

Cheers -- sylas

 

[lisab]

Sylas, forgive the dumb question, but I've looked at my statistical mechanics text and even my p-chem text...I don't see "Plank's response" there. Googling isn't much help.

So, can you clarify that term?

It should be noted that Lindzen is a completely different kettle of fish from Chilingar, and that this paper has nothing to do with Chilingar's errors, or with the mix up over convection and feedbacks.

This paper is not about the greenhouse effect, or CO2, but rather about climate sensitivity.

One of the difficulties in this whole topic area is focus. I'd really like to avoid having every thread that mentions CO2 go off into every aspect of climate or a fight over AGW and global warming.

This thread, as I understand it, is about the basics of a greenhouse effect, and how physically CO2 in particular works to give additional energy and heating on a planet.

The point about feedbacks is that they are OTHER factors that apply for anything altering energy balances to help moderate or enhance the temperature.

Convection is not a feedback; it is a part of the basic Planck response to temperature change. Then, on top of the base response, there are changes in cloud cover and humidity in particular (the major source of the feedbacks being considered in the above paper) and they are NOT part of the Planck response; they give new secondary forcings in their own right on top of any primary forcing that initiates a change.

Cheers -- sylas

 

[sylas]

Sylas, forgive the dumb question, but I've looked at my statistical mechanics text and even my p-chem text...I don't see "Plank's response" there. Googling isn't much help.

So, can you clarify that term?

The Planck radiation from a body is its thermal radiation. The Planck response is how much temperature changes for a given emission of energy if nothing else changes other than the temperature.

You can see the term used properly in Lindzen's paper, that Andre has cited.

Here are a couple of simple examples.

Planck response of a blackbody radiator

A blackbody at thermal equilibrium radiates energy Q at a temperature T according to the Stefan-Boltzman law, which is itself derived from the Planck radiation law.

Q = \sigma T^4​

The Planck response is dT/dQ.

\begin{align*}<br /> \frac{dQ}{dT} &amp; = 4 \sigma T^3 = \frac{4Q}{T} \\<br /> \frac{dT}{dQ} &amp; = \frac{T}{4Q}<br /> \end{align*}​

Planck response of a graybody radiator

A graybody radiates energy at each frequency that is a fixed fraction (the emissivity ε) of the blackbody radiation.

\begin{align*}<br /> Q &amp; = \epsilon \sigma T^4 \\<br /> \frac{dQ}{dT} &amp; = 4 \epsilon \sigma T^3 = \frac{4Q}{T} \\<br /> \frac{dT}{dQ} &amp; = \frac{T}{4Q}<br /> \end{align*}​

Planck response for a coloured body

In general, a radiator emits better at some frequencies rather than others, and in this case the Planck response will diverge somewhat from the simple case above, but because the Planck spectrum has a strong peak at the main emission window, the above approximation works pretty well in most cases.

Not the Planck response

In a complex system like the Earth, various things happen as temperature changes, which in turn alter how it interacts with energy. One of the major changes on Earth is that cloud cover and specific humidity will alter, and these have further knock on effects. This is called "feedback", and becomes part of a more complicated response than the simple Planck response in which it is only the temperature that changes.

Cheers -- sylas

 

[lisab]

Thanks! I'm new to this so I don't know the lingo.

 

[chriscolose]

Hi lisab,

Just to repeat everything sylas said (although maybe a bit different),

In general the fundamental constraint on Earth's climate (and all such climates for the rocky planets) can be thought of as a balance between the incoming energy from the sun, and the outgoing energy that the Earth emits to space (at the top of the atmosphere). In other words, the sun is the way the Earth gets virtually all of its energy, and it has to lose that energy somehow (otherwise the Earth would just keep heating up over time and would become too hot very early in its history). The way it loses that energy is through thermal radiation (which you can't see, but you can feel off of objects).

As a rather simplistic explanation, when you change the CO2 concentration in the atmosphere, it turns out that you greatly inhibit the efficiency at which the Earth loses that heat, while not changing the incoming energy to any significant degree. So after instantaneously changing the atmosphere through more CO2 an observer looking out from space at the Earth would see slightly less infrared radiation escaping (assuming that observer could see in the infrared), and to be even more specific, certain wavelengths in the infrared where CO2 is a strong absorber. So the temporary effect is for the Earth to be taking in more solar energy then it is losing infrared heat, and the Earth has a goal to get back in radiative balance. So to get that infrared-solar balance back, the temperature must rise. It turns out that the outgoing energy of the Earth is very strongly dependent on temperature (and for an ideal blackbody, only dependent upon temperature). So as the Earth warms it's going to lose more infrared heat, which is how it comes back to balance.

To get a bit more complicated, assume we have two Earth's that have completely identical climates. Now pretend we have the ability to change the CO2 content in the atmosphere without changing anything else (e.g., no changes in cloud cover, no changes in ice extent, no changes in humidity, etc). This means that the only thing that responds to warmer temperatures is the actual outgoing energy of the Earth. This is the Planck response, which for a doubling of CO2 is about a 1 C rise in global temperature.

Now on our second planet we let things go like they actually would. The reality is that when you force the climate to change through more CO2, through more solar energy, or whatever, you are certainly going to expect changes in ice extent, changes in cloud cover, changes in humidity, etc. Some of these responses are well understood and others are not. These responses also have their own effects on the energy balance of the planet by doing either or both of the following: absorbing or reflecting solar energy, or absorbing (or letting through) more outgoing infrared radiation. As one of the easier examples, ice is a very good reflector of incoming sunlight. So as you reduce ice cover in a warmer world, you not only get the effect of the CO2, but you also get the effect of reducing how much solar light gets reflected back away (i.e., more of it gets absorbed by the underlying ocean or land) and this will amplify the response. Thus the situation becomes more complicated than the Planck response because you not only have to figure out how much CO2 is going to warm the planet (and how much the outgoing energy is going to change with higher temperatures) but you also need to figure out how changes in temperature itself will alter the energy balance through other feedback mechanisms.

Hope that makes some sense

 

[Andre]

It should be noted that Lindzen is a completely different kettle of fish from Chilingar, and that this paper has nothing to do with Chilingar's errors, or with the mix up over convection and feedbacks.

But what both of them say, ultimately boils down to about the same effects so why is Chilingar so wrong?

This paper is not about the greenhouse effect, or CO2, but rather about climate sensitivity.

Fair enough, I'll make another thread.

Convection is not a feedback; it is a part of the basic Planck response to temperature change.

So I demonstrated that the mechanism of convection is exactly in conformity with the definition of feedback. Then it was argued that convection was a too fast process considering other processes, that it was not considered as feedback, instead it was assumed to be directly a part of the 'planck response', which basically restored the normal thermal gradient by transporting energy upwards in response to other changes. Then I contended that the time constant is not a part of the definition of feedback, be it mili-seconds or millenia. Furthermore I said that the horizontal version of convection, advection (lifting of warm air in frontal systems), very common in moderate lattitudes, is a process of days and weeks. So I wonder what is new now that the convection feedback can be denied once more, without further elaboration.

Actually in that process "Convection maintains the lapse rate by transporting energy upwards in response to other changes." Compare this with the normal effect of negative feedback: "maintaining". As negative feedback increases the stability of processes, like maintaining the lapse rate.


Mind that if convection is a part of the 'planck response' in general, it seems to me that there would be no difference in heat transport in two atmospheric area's, which are almost identical, except that one is static and the other is subject to advection processes.

So I tried find answers in the http://geosci.uchicago.edu/~rtp1/ClimateBook/ClimateVol1.pdf is not even mentioned.

 

[sylas]

But what both of them say, ultimately boils down to about the same effects so why is Chilingar so wrong?

They are not even close to the same thing; they are completely different. I'm rather baffled by this -- what's the same about it? Lindzen is talking about all the usual things atmospheric physicists talk about with feedbacks -- humidity and cloud, mainly. It's not remotely the same.

Lindzen does understand the greenhouse effect and how it works, and uses much the same basic no-feedback response as everyone else -- about 0.25 K/(Wm^-2) -- a simple approximation from the Stefan-Boltzman relation. He uses the same CO2 forcing as everyone else. Given current conditions, doubling CO2 gives an additional 3.7 W/m² of forcing... more energy. This is actually one of the most straightforward forcings involved in climate, and by now very well understood indeed. Lindzen's paper is NOT about greenhouse effects, or trying to rewrite the elementary thermodynamics involved in a greenhouse forcing -- which is what Chilingar does. The paper you cited is actually about sensitivity and feedback. A new thread on sensitivity might be interesting.

Another thread that would be useful, I think, is a kind of tutorial introduction to basic thermodynamics of how a greenhouse effect works at all, as a self contained thread entirely independent of sensitivity considerations, and using really basic science that ought to be a common basis for all these discussions; and is certainly taken for granted by someone like Lindzen, who actually IS a climate scientist.

The no-feedback response is about the new equilibrium temperature when nothing changes except temperature: and that means temperature all up and down the atmosphere. The atmosphere is a fluid, in constant motion, and that motion is part of the equilibrium, in which solar input is balanced by the net flow of energy from convection and latent heat and radiation. The no-feedback response is when nothing changes in compositions or cover. You keep cloud and humidity and all that fixed. You only alter the temperatures to the new equilibrium temperature profile that restores an energy balance; and so of course that has a different flux of energy -- that's what it means to restore balance. The new no-feedback energy flux for the new temperature profile involves all the ways heat energy gets transported.

Every time you see "Planck response", from any atmospheric scientist -- including Lindzen -- this is what it means.

So I demonstrated that the mechanism of convection is exactly in conformity with the definition of feedback.

Not in relation to the feedback of temperature and net energy, you didn't.

THAT'S what atmospheric physics means with respect to feedback.

You can, of course, try to use different variables, representing the different kinds of heat flow -- but it is just absurd to say this is something omitted from conventional climate science. The no-feedback response of temperature to restore energy balance is calculated using the basic convective-radiative equilibrium condition of the troposphere. There's nothing being ignored here.

Then it was argued that convection was a too fast process considering other processes, that it was not considered as feedback, instead it was assumed to be directly a part of the 'planck response', which basically restored the normal thermal gradient by transporting energy upwards in response to other changes. Then I contended that the time constant is not a part of the definition of feedback, be it mili-seconds or millenia. Furthermore I said that the horizontal version of convection, advection (lifting of warm air in frontal systems), very common in moderate lattitudes, is a process of days and weeks. So I wonder what is new now that the convection feedback can be denied once more, without further elaboration.

I tried to clarify this before. It's not actually the time that is the defining quality here; and my first attempt to explain this seems to have been misleading. My apologies. The rapid relaxation time for convection DOES show up in the way we calculate the new energy balance (see the fourth and last of my extracts from the climate text) but it is not actually the defining quality.

The essential thing about convection is quite simply that it is part of the energy balance equation. The no-feedback response is, by definition, the new temperature that will restore energy balance, and that will involve new values for convection as one of the basic energy transports in that new temperature profile, along with radiation.

So I tried find answers in the http://geosci.uchicago.edu/~rtp1/ClimateBook/ClimateVol1.pdf is not even mentioned.

The benefit of a basic text like this is not so much for finding keywords and phrases, but learning more about the underlying physics as background to these kinds of discussions. But it's a good idea to refer to other texts as well, and that might help. If you have a basic text on atmospheric physics you'd rather use, that should be fine. I'd be interested in a recommendation if you have one.

This is not advanced or disputed science. This is a foundation for understanding better some of the arguments that go on. The essential thing is simply that we find a text dealing with how the Earth -- or any other planet -- sheds the heat energy from the Sun back out into space.

This is not a meteorology text. There's very little on horizontal fluid motions; and this is also explained in the introduction. In the usage of this text, advection is convection... or perhaps better, the major part of convection. The text does not attempt to break convection down into parts. The "Plank response" is also called the "no-feedback response", but in this text there doesn't seem to be a special term for it. The concept of feedback as it applies in atmospheric physics is covered in several places, but most of the text is about calculating how energy flows for a given atmosphere and temperature. Invert that relation, and you've got the basics of what is called Planck response -- the temperature that gives a particular energy emission, for a fixed atmosphere and solar input.

Cheers -- sylas

 

[Andre]

Not in relation to the feedback of temperature and net energy, you didn't.

care to reread this?

...Ah let's try some ideas, especially with wet convection, involving latent heat. So, as a wet surface heats up, water evaporates (latent energy -which reduces the temperature increase). Conduction and radiation heat up the lower layer(s) of the troposhere, causing the well discussed convection. Heat- and latent energy -water vapor- are now transported up. Due to expansion the updraft cools adiabatically and water condenses forming clouds and releasing the latent heat again. Clouds are good radiators as they radiate on all water IR- frequencies. So this energy is radiated outwards in al directions as it would have done on the Earth surface without convection. But the difference is that energy -on water frequencies) emitted upwards will find less water vapor molecules because the upper levels are much drier than the surface levels. Evidently, the CO2 frequency bands are also less relevant here. Consequently the energy emitted by clouds (tops), on water frequencies, has more chance to escape into space than energy emitted by the surface in all bands including the CO2 frequencies.

Now if the greenhouse gas concentration was to increase then the heating of the lower atmosphere by radiation was also to be increased, this would enhance the convection rate, transporting more energy upwards, where more energy can radiate into space. Consequently it appears that convection acts as a negative feedback on greenhouse gas variation

Or in wrap up

system input: Sun heats surface
system output: surface heats lower atmosphere layes
system output: evaporation tempers the heating

other system effects:
lower hot air rising; convection causing...
colder air replacing the convecting air and
adiabatic cooling causing
cloud forming causing
higher albedo and ...
Less sun heating the surface (see input, affecting it negatively, feedback loop closed)

 

[Andre]

To amplifly to negative feedback effect, it should be noted that moist convection would reduce the temperature.

Mosit convection is most common around the equator, (hadley cell), the opposite effect, decending air is in the desert zones.

Now compare the average yearly temperatures of a equatorial station (Brazzaville) with a Sahara desert station (N'Guigmi):

http://data.giss.nasa.gov/work/gistemp/STATIONS//tmp.112644500000.1.1/station.gif

http://data.giss.nasa.gov/work/gistemp/STATIONS//tmp.133610490003.1.1/station.gif

Why is the desert warmer than the equator?

 

[sylas]

care to reread this?

I HAVE been reading your posts.

This isn't a case of my not reading you. It is a case of me trying to explain where what you have written goes wrong. You might disagree with me... but it isn't because I'm not reading.

The problem is that you are NOT describing feedbacks to the basic equilibrium temperature response. You are instead trying to break down the equilibrium response into bits and pieces and looking for interactions between those parts that you can call feedbacks ... but these are not feedbacks to the main energy balance relation.

You have a planet. You add some extra energy. It heats up, in order to shed that extra energy. The relation we are interested in is how much temperature brings the planet back to an equilibrium balance.

Now. By definition of the system in question, this is a relation between temperature and a total energy output that brings the system back to a balance. Lindzen, for example, cites that relation as 0.25 K / (Wm^-2), for the no-feedback Planck response. He uses a rather crude estimate, which gets into the right ball park, by simply treating Earth as a blackbody radiator.

Observe the units. It's a relation between temperature, and energy flux. But more than this... it is the TOTAL energy flux, including all the ways energy is transported, so as to balance up the Earth's total energy budget.

With this elementary understanding of the system in question, let's look at your list of processes. I'll put your text in blue, and add my comments afterwards indented.

system input: Sun heats surface[/color]

OK... we can say that the system input is a temperature, and let the output be the total radiated energy. That works.​

system output: surface heats lower atmosphere layes[/color]

Wrong. The system output is total energy emitted. Your proposed output is no longer talking about the energy balance system, but some kind of subdivision before you get to the output for basic no-feedback response. I am not persuaded at this point that you actually have a good handle on how to calculate a net temperature response. Maybe; its not easy. But in any case, like any physics or maths problem, it can be approached in different ways, which -- if they are physically sensible -- are different paths to the same answer. But crucially, you are not giving the system output here at all. You are merely decomposing the non-feedback response. The actual system output is the energy back to space to balance what is received from the Sun.​

system output: evaporation tempers the heating[/color]

This is not the output either. It seems to be a reference to one of the fluxes of energy that is involved in energy balance... the latent heat flux. But the system output is the total energy output into space.​

other system effects:
lower hot air rising; convection causing...
colder air replacing the convecting air and
adiabatic cooling causing[/color]

This, combined with the latent heat flux and the radiant heat flux, is a summary of the energy fluxes involved within the atmosphere. But they are not "feedbacks" in the sense of something responding to the temperature to give an additional forcing to the energy balance. The no-feedback response includes all these energy fluxes as part of the total transport of energy for the system in question, relating temperature to a total for energy emitted.​

cloud forming causing
higher albedo and ...
Less sun heating the surface (see input, affecting it negatively)[/color]

Cloud is a feedback. In the non-feedback response, you simply assume cloud cover remains unchanged. If temperature leads to changes in cloud cover, this then has an further effect on energy balance by absorbing or reflecting radiation. This is a feedback to the basic Planck temperature response, because it is not altering the energy balance simply by increasing the energy flux directly from temperature. It's actually modifying the composition of the system.​

The problem here is that by failing to define the system clearly, with the proper input and output as temperature and net energy emitted, you just end up talking at cross purposes. That's why I asked you previously to define what YOU understand as a non-feedback response.

I think that would still be a useful question for you to think about or even answer. What do you think of as "Planck response"? You've used the term before; what does it mean to you?

Here's my account. The conventional Planck response with zero feedback is still involving a total energy emission; latent heat, convection and radiation. There is nothing ignored here. You can calculate it more carefully than just treating Earth as a blackbody radiator, but the calculation of Planck response still works by holding fixed the composition of the Earth's atmosphere and surface. The cloud, the humidity, the surface and the atmospheric composition all remain fixed, and you simply calculate the new total energy flux when the whole system relaxes to a new equilibrium temperature.

To this we then add the feedbacks, which means temperature having an effect on the surface, or the atmospheric composition, or clouds, or anything else other than the simple direct change in total energy flux from temperature directly.

This is not merely a matter of two different ways of talking about the system. We got into this with a claim (based on Chilingar) that there's something going on with convection which is being inadequately addressed in conventional science, or else an effect of convection which has not been properly considered. That's poppycock.

Cheers -- sylas

 

[sylas]

Why is the desert warmer than the equator?

Probably... mainly because the planet is warming up, and land heats substantially faster than ocean, due to the ocean's massive heat capacity. Hence land anomalies have increased substantially faster than ocean anomalies. Brazzaville is much closer to the ocean.

But note that you can't conclude much at all by looking at individual sites like this. There's a lot of regional variations for all kinds of reasons. There's no reason for anyone to suspect a simple uniform change of temperature with latitudes.

Still, the difference between land and ocean is one of the most straightforward contributions that is likely to apply in this case.

Cheers -- sylas

 

[Andre]

Or in other words, it's all system and there are no feedbacks, it's all accounted for in the complete system of atmospheric/surface response.

Or are we interested in surface warming? with a primair warming source and a primair warming process directly dependent on the variables in the modified Stefan Bolzman equation, and all other processes are modifications of open and closed loop feedbacks?

But let's agree on the same goal post and let's not try to move them too much.

Notice that Brazzaville was around average 25 degrees C before 1980 and around average 25.5 degrees after 1990 while N'Guigmi was around 28C before 1980 and around 29C after 1990. But it's open source, do try the other equatorial stations versus Sahara stations around 10-12 North.

For instance much more to the East and still equatorial, Bangassou...

http://data.giss.nasa.gov/work/gistemp/STATIONS//tmp.109646560000.1.1/station.gif

against the much closer to the Atlantic desert station of Kenieba

http://data.giss.nasa.gov/work/gistemp/STATIONS//tmp.127612850002.1.1/station.gif

 

[sylas]

Or in other words, it's all system and there are no feedbacks, it's all accounted for in the complete system of atmospheric/surface response.

I specifically identified some feedbacks for you; so of course there are feedbacks.

What is not feedback is the response that follows from temperature directly. It's useful to separate out the Planck response, which is the extra energy flow (total) that arises directly from the new temperature; and the impact on energy that arises indirectly as temperature driven changes to the composition of the system; cloud, surface, humidity, etc,

I think it would be useful for you to describe in your own words what you mean by the Planck response. You've spoken of it previously in these threads. Are you satisfied with the meaning of the term I've tried to give, or do you mean something else? Do you recognize any value in trying to describe a response WITHOUT feedbacks? What does that mean to you?

Or are we interested in surface warming? with a primair warming source and a primair warming process directly dependent on the variables in the modified Stefan Bolzman equation, and all other processes are modifications of open and closed loop feedbacks?

It's not "or". It's "and". Of course we are interested in surface warming, with a primacy warming source. This temperature response to additional energy can be decomposed into a basic non-feedback response, and the feedbacks that can amplify or damp the base response.

It makes little sense to speak of "feedback" unless you are clear about the basic no-feedback relation to which the feedback applies.

Notice that Brazzaville ...

I really think this is off topic. We are meant to be discussing whether you can model CO2 as a greenhouse gas.

Looking at individual stations is not doing anything useful here. Regional variations arise for all kinds of reasons; and this involves much more than the simple basic thermodynamics of the energy balance for a planet. You can look at reasons for this kind of regional variation, but it has little to do with greenhouse effects, which tend to be pretty well mixed through the atmosphere. It tends to be a lot about redistributions of heat around the planet and weather patterns.

There are some really fundamental problems here in very basic thermodynamics that need to be cleaned up first, or else there's no hope of doing anything sensible with more complicated details; and one possible sticking point is simply the effect of thermal absorption in the atmosphere -- the impact of greenhouse effects at all!

Methods for calculating Planck response are relevant to this question, because the Planck response is all about how the Earth sheds energy out to space with a given atmosphere and surface.

Cheers -- sylas

 

[Andre]

So we have a complex process which can be boiled down to

http://pespmc1.vub.ac.be/feedback.html

Now if you take out only feedback processes selectively assuming that others are implicitely in the process, you may not end up with the most reliable representation of reality. Considering convection implicitely in the process disregards exactly what the four stations show in different positions of the hadley cel, a different temperature reaction due to different conditions in the atmosphere, moist convection in the tropix versus dry subsidence in the desert area. Yet that Planck response would probably not see the difference, would it?

So again the primary process is the integral of the momentary, local Stefan Boltzman grey body response on the insolation, all other processes are secundair and cannot be neglected regardless if the reaction time is a millisecond or a millenium.

 

[sylas]

So we have a complex process which can be boiled down to

http://pespmc1.vub.ac.be/feedback.html

The whole idea of any analysis is to make a useful abstraction of reality. We know that climate is complex, and involves many processes. But the whole complex assemblage still has to satisfy the laws of thermodynamics.

There are all kinds of things we could choose to consider as "outputs"; but it turns out that you discover useful things about a complex system by picking energy as an output worth examining.

Look at your citation to Lindzen. He speaks of the Planck response as 0.25 K/(Wm^-2).

Where are the multiple outputs there? Note that an energy balance analysis is not a complete description of every aspect of climate. It's an abstraction... but it's a USEFUL abstraction, because no amount of "complexity" can violate the basic laws of physics like conservation of energy. That is why it is productive to consider total energy as a particular variable (or "output") of interest.

It's the same thing in principle as the idea of the motion of the center of mass of some assemblage of objects. The whole system can be as complex as you like; but the "center of mass" turns out to be a useful abstraction for summarizing one aspect of the whole interaction.

Same with thermodynamics.

Now if you take out only feedback processes selectively assuming that others are implicitely in the process, you may not end up with the most reliable representation of reality.

Rubbish. This is a way to look at some overall abstractions of the collective behaviour for which we have fundamental laws of physics that are as reliable as all get out.

Also, what is this "take out" and "selective" and "assume"? The feedbacks and the base response they modify are useful ways to do an analysis. They aren't the only way to do things, but they are one good way to structure an analysis of a complex system.

Being able to define what we mean by a non-feedback response doesn't mean ignoring feedbacks. We don't, in fact, have a "reliable" representation in the sense of knowing the magnitude of climate feedback. The best we can do is to constrain the magnitude of feedback in the system, and this has been done with many studies, both theoretical and empirical.

But in order to express the magnitudes of these substantial uncertainties, we need to define what feedback magnitudes even mean. We need to define a base response, as a useful theoretical baseline.

YOU have been talking about feedback yourself, and (previously at least!) you've seemed to recognize that that in a feedback analysis you identify an input and output of interest, and you structure your description of the whole complex system into various feedbacks and a base response.

Are you saying that the whole notion of feedback analysis ITSELF is not "reliable"? (I would have thought not, but at this point it as well to ask!)

Or are you proposing some different way of picking your inputs and outputs? This is why I asked you to describe what YOU mean by "Planck response". If you are going to having anything useful to say about "feedbacks" beyond "it's all horribly complicated and we don't have a reliable representation", then you'll need to address this question eventually.

What is base response that feedbacks are modifying, in your analysis?

Considering convection implicitely in the process disregards exactly what the four stations show in different positions of the hadley cel, a different temperature reaction due to different conditions in the atmosphere, moist convection in the tropix versus dry subsidence in the desert area. Yet that Planck response would probably not see the difference, would it?

Convection is NOT considered "implicitly". It's explicit, and one of the primary ways that energy gets transported up from the surface into the atmosphere. The convective energy fluxes are crucial and explicitly quantified if you do a detailed calculation of non-feedback response to temperature.

Added in edit: In answer to your question, you calculate a Planck response for a given composition, including moisture. The moisture content of air IS very important in calculation of Planck response, and you will get different regional Planck responses if there are regional differences in humidity.

To again the primary process is the integral of the momentary, local Stefan Boltzman grey body response on the insolation, all other processes are secundair and cannot be neglected regardless if the reaction time is a millisecond or a millenium.

No; that is not how you calculate a Planck response.

Lindzen does something a bit like it, but without even integrating. He abstracts the whole surface and atmosphere as a single grey body radiator to estimate Planck response, which is crude, but gets into the right ball park; but it's definitely only a rough approximation.

The first step is quite simply to define what you are trying to calculate. I've given the definition of Planck response which is use in climate science; and I am not aware of any other definition. I am still unsure what you mean by this term, though I have now asked several times.

Given the definition of Planck response, you then have a basis for calculating it. In this calculation NOTHING can be "neglected" unless you can justify that the omission has a negligible quantified impact on the quantity we have defined.

The correct way to actually calculate the Planck response is to identify the composition of the system you are interested -- primarily, the albedo and the composition of the atmosphere, and then to calculate all the energy fluxes for a given temperature. This follows from the definition.

This means, in general, at least three major integrations. You have to integrate over every line of the electromagnetic spectrum; because it definitely isn't a grey body. You have to integrate over the surface, from poles to equator. You have to integrate over every altitude up the atmospheric column. It's a big calculation, but it can be done. Approximations are necessary, but as with any such analysis, you don't pick integration steps out of a hat... you have to justify the choice of step sizes by explicit consideration of the associated numeric errors.

The text I have cited is mostly structured around how that calculation works and the relevant thermodynamics used.

I've been thinking for some time that it would be a useful thread to actually break apart that calculation a bit and see more about how it works; particularly the radiative-convective equilibrium. The main idea would be to help interested readers get a bit more literate in the underlying science, without trying to resolve all the open questions or arguments over warming. I'd try to stick with basic thermodynamic foundations that should not be controversial to someone with a bit of interest in physics.

Here's a very useful and widely cited reference on climate feedbacks, which includes some helpful definitions and explanations of the terms being used (link to 3.2Mb pdf):

• Bony, S., et al (2006) "ftp://eos.atmos.washington.edu/pub/breth/papers/2006/Bony_etal_feedbacks.pdf"[/URL], in [i]Journal of Climate[/i], Vol 19, 1 Aug 2006, pp 3445-3482.[/list]
  
  Here's an extract from Appendix A, on definitions (page 3475)
  [indent][i]The Planck feedback parameter λ[sub]P[/sub] is negative (an increase in temperature enhances the LW emission to space and thus reduces R) and its typical value for the earth’s atmosphere, estimated from GCM calculations[sup]A1[/sup] (Colman 2003; Soden and Held 2006), is about -3.2 W m[sup]-2[/sup]K[sup]-1[/sup] (a value of -3.8 W m[sup]-2[/sup]K[sup]-1[/sup] is obtained by defining λ[sub]P[/sub] simply as -4σT[sup]3[/sup], by equating the global mean OLR to σT[sup]4[/sup] and by assuming an emission temperature of 255 K).[/i][/indent]
  
  Note that the value -3.8 they mention is obtained using Lindzen's method; though Lindzen quotes the inverse with units K/(Wm[sup]-2[/sup]); and limits it to the tropics. The more thorough method that actually looks at the energy fluxes of Planck response gives -3.2. Note that the parameter λ[sub]P[/sub] used here is what we've been calling Planck response. The alternative terminologies are briefly mentioned on page 3475:
  [indent][i] Since the feedback parameter is the sum of the Planck response (or Planck feedback parameter) and of all other feedbacks,...[/i][/indent]
  
  Cheers -- sylas

 

[vanesch]

Convection is the primary process by which the atmosphere maintains a temperature profile, and that means convection is part and parcel of the basic no-feedback Planck response of temperature to change in energy input.​

The Planck response is a new equilibrium temperature in response to an energy change, where the only thing that changes is temperature... and, of course, the energy fluxes associated directly with temperature. You can't have more temperature without having also changes in the associated energy flux -- the whole point is for temperature to increase to the point where energy balance is restored again.

In an atmosphere, we speak of the temperature at a given altitude, or pressure level. But the temperatures in an atmosphere at a given level don't just change by cooling or heating up air held in place. In an atmosphere, changes of temperature at a given level occur to a large extent by movement of air in and out of that level. Therefore this is already part of the basic Planck response, right there. The new equilibrium state includes a new flux of energy, both by movement of air (convection) and by radiation, so as to maintain energy balance. That's what equilibrium means.

For a planet, like Earth, with a gas that condenses in the atmosphere, there's a latent heat component as well, which shows up by using a moist adiabat rather than a dry adiabat for the temperature profile maintained by convection.

[...]

The error here is failing to recognize that the basic Planck response ALREADY includes the effects of convection to transport energy upwards so as to maintain an essentially adiabatic lapse rate in response to radiative cooling. If you try to propose a feedback adjustment to Planck response based on convection, you are double counting the process. If you try to isolate a response with convective energy transport held fixed, then you doing something frankly bizarre and certainly nothing like what anyone else in atmospheric physics calls the Planck response. I am completely positive that nothing in what you learned from basic atmospheric physics ever did anything like that.

This is interesting. In fact, I have to say that I erroneously thought, like Andre, that convection was added in after first considering a "frozen-in-place" atmosphere in which only the radiative transport was considered - as through layers of a solid, say. But it is in fact nothing more than a convention of what is considered "the basic response" and now that I've read the first few chapters in Pierrehumbert, it is clear that this approach (first "freeze in" the atmosphere as it is now, do the radiative transport, and then add a correction for convection) is essentially meaningless, because convection can completely alter what you would obtain by radiative transport alone - and even simplifies the problem.

In fact, the exercise is only interesting, because you would find that, with the atmosphere "frozen in", and an increased amount of greenhouse gas, the lower layers would heat up *more* than without the extra greenhouse gas (you increase "the heat resistance"). The top of the atmosphere would cool down, because if it has to radiate away the same amount of energy (= the solar influx that is not reflected by albedo which is constant) and now has a higher emissivity (because "blacker"), its temperature has to be lower. So the "frozen-in" atmosphere, with added greenhouse gas, would become hotter below, and cooler on top.

That would mean that the "frozen in" atmosphere has a sharper temperature profile, and hence that it would, if "unfrozen", *enhance* convection. If there is (strong) convection, then we know that the temperature profile follows the adiabat (wet or dry, accordingly). This convection (and that was my earlier discussion point in the thread) will restore the adiabat, and hence have a gentler temperature profile than the "frozen-in" atmosphere (with purely radiative transport), which means that the lower layers will be less hot and the higher layers will be less cool than in the case of "frozen-in".

So "switching on" convection does "temper" the greenhouse effect as compared to the (admittingly very artificial) "frozen-in-radiative-transport-only" atmosphere, but when all is said and done, as there is convection, the adiabat is restored.

Nevertheless, the exercise is not completely void of interest, because it shows that convection is maintained by adding greenhouse gasses. If it were different (if we had obtained that higher layers heated up, and lower layers cooled down), we might have been confronted with the problem that convection might stop, and in that case, we can't a priori say what is the temperature profile as we will have a stratified atmosphere.

In other words, as you say, assuming the adiabat means that convection did all the heat transport it could do upward.

If you assume the adiabat as a temperature profile, convection has already "tempered" the greenhouse effect to its maximum extent possible.
(of course, we still have to have the *right* adiabat, which is probably somewhat tricky when there are condensibles).

However, as the temperature profile of the atmosphere enters crucially into the radiation transport itself, it is indeed an almost useless exercise to do the "frozen-in" radiation transport exercise, because it would lead us to a harder problem (we would have to find out the temperature profile) than is the case with the adiabat (where the temperature profile is given). Hence no need to "freeze-in" the atmosphere, and then add convection afterwards as a "feedback" effect. Better take directly the adiabat, which includes already the convective contribution (and we know that it is there), and do the radiative transport directly with this profile.

Is that about right ?

 

[Andre]

That would probably right with a dry adiabatic convection, however this is a minor element in the complete element. In most instances convection and advection lead to cloud forming, changing the whole equation.

Commonly feedback of clouds is included in the theory but all clouds form due to cooling of that part of the artmosphere below the "dewpoint". Convection and advection are the major causes of that. Therefore these are a part of the many feedbacks

More later.

 

[sylas]

This is interesting. In fact, I have to say that I erroneously thought, like Andre, that convection was added in after first considering a "frozen-in-place" atmosphere in which only the radiative transport was considered - as through layers of a solid, say.

Yes, you've got it ... it definitely is not done like that.

You can do some interesting calculations with a "frozen-in-place" atmosphere, and this actually works as a model for the stratosphere, where there isn't a lot of convection. It's also a good practice example for simple problems in thermodynamics of radiation in a gas. But it's no good at all as a realistic representation of the troposphere, even as a Planck response. The troposphere is not in a radiative equilibrium, and so you really have to look at convection as a part of the energy flux for a stable equilibrium at any given temperature.

So what could you "freeze in place"?

You could perhaps try freezing in place the net flux of energy from convection, but there are very good reasons for not doing that either. Convection involves a rapid response to any instability from a lapse rate that is too high (when air temperature falls with altitude more rapidly than the adiabat) so it makes much more sense to freeze the lapse rate in place -- and this corresponds to letting convection continue to do its thing of transporting energy in the atmosphere as part of the net energy flux upwards as the Earth sheds its heat in a stable thermodynamic equilibrium with the solar input.

The paper I cited above (Bony et al, 2006) describes it simply as Planck response being when the atmosphere has a uniform increase in temperature at all altitudes. This corresponds to a fixed lapse rate. In practice, the actual calculations don't just pick a uniform temperature increase; but rather stick with adiabatic lapse rate (for the given humidity; also fixed), which is pretty much the same thing; and in the calculation there is a corresponding additional flow of energy into every level of the troposphere to compensate for cooling by radiant emission.

In Pierrehumbert, page 256 section 4.8 describes how this works, as part of calculation of the radiative-convective equilibrium in a real gas atmosphere. I quoted an extract of this page as the very tail end of [post=2311418]msg #154[/post].

But it is in fact nothing more than a convention of what is considered "the basic response" and now that I've read the first few chapters in Pierrehumbert, it is clear that this approach (first "freeze in" the atmosphere as it is now, do the radiative transport, and then add a correction for convection) is essentially meaningless, because convection can completely alter what you would obtain by radiative transport alone - and even simplifies the problem.

It is "convention"; but a pretty obvious one when you dig into it. Once you decide that we are interested in temperature and total energy flux, the rest follows pretty inevitably from the physics of the dynamic thermal equilibrium in the atmosphere for that energy flux.

Not that this is necessarily obvious. You can do a fair bit of thermodynamics without getting into the particular problem of atmospheric thermal equilibrium, so this isn't simply general knowledge. I've been learning about it over the last year or so, as useful background to following climate discussions.

[snip discussion...]So "switching on" convection does "temper" the greenhouse effect as compared to the (admittingly very artificial) "frozen-in-radiative-transport-only" atmosphere, but when all is said and done, as there is convection, the adiabat is restored.

Right!

There's an easy way to think of what you've just described. The "frozen in place" atmosphere corresponds to what is called "radiative equilibrium". In this case, essentially all the energy flux is from thermal radiation, and with minimal net impact from convection. Such an atmosphere will still have a lapse rate, but it follows from very different physics, and it can go negative if the atmosphere is able to absorb shortwave radiation... this is what happens in our stratosphere.

Usually the lapse rate in radiative equilibrium is weaker than the adiabatic (convective) lapse rate; and in such a case the atmosphere relaxes to the adiabatic lapse rate... and that is maintained up until the natural adiabatic lapse rate intersects with the natural radiative equilibirum. This gives you a basic physical theory of the tropopause. It is the point at which the atmosphere switches from convection to radiation as the primary factor establishing the lapse rate!

We often think of the tropopause as a local temperature minimum. But physically, you can have a more general theory of atmospheres that works across many different planets if you use the dynamic definition. See section 4.8 of Pierrehumbert, on "Tropopause height for real gas atmospheres"; though of course all the goes before helps lead up to this. (I'm currently working on understanding chapter 4 in my off-line study.)

There's another way to think of the tropopause... it marks a change on the impact of greenhouse effects for heating or cooling. Greenhouse effects work to cool down the troposphere, and the convection works to counter this and heat it back up again, at equilibrium.

Nevertheless, the exercise is not completely void of interest, because it shows that convection is maintained by adding greenhouse gasses. If it were different (if we had obtained that higher layers heated up, and lower layers cooled down), we might have been confronted with the problem that convection might stop, and in that case, we can't a priori say what is the temperature profile as we will have a stratified atmosphere.

You can calculate a temperature profile for a stratified atmosphere (a stratosphere) by knowing its composition and interaction with radiation. You can use the "frozen-in-place" model in this case; also called "radiative equilibrium" in the text.

(of course, we still have to have the *right* adiabat, which is probably somewhat tricky when there are condensibles).

Indeed. And in fact water vapour has a negative feedback by reducing the adiabatic lapse rate. The moist adiabiat is much weaker than the dry adiabat. If it was not for the fact that water vapour is also a strong greenhouse absorber, then the water vapour feedback in climate would be strongly negative, from its effect as a condensable substance. This is called "lapse rate feedback", discussed also in Bony et al (2006) cited previously.

Is that about right ?

I believe so...

Cheers -- sylas

 

[Andre]

Indeed. And in fact water vapour has a negative feedback by reducing the adiabatic lapse rate. The moist adiabiat is much weaker than the dry adiabat. If it was not for the fact that water vapour is also a strong greenhouse absorber, then the water vapour feedback in climate would be strongly negative, from its effect as a condensable substance. This is called "lapse rate feedback", discussed also in Bony et al (2006) cited previously.

Now let's focus on that feedback as compiled in "ftp://eos.atmos.washington.edu/pub/breth/papers/2006/Bony_etal_feedbacks.pdf"[/URL] find

[quote] The observed behavior of radiation fluxes implies negative feedback processes associated with relatively low climate sensitivity. This is the opposite of the behavior of 11 atmospheric models forced by the same SSTs...Results also show, the feedback in ERBE is mostly from shortwave radiation while the feedback in the models is mostly from longwave radiation.[/quote]

Now short wave feedback would be associated with direct 'reflection' of insolations, hence by more cloud forming -which in the tropics- is basically caused by convection. Hence it appears that the compilations of Bony et al rely on those same assumptions that lead to the mismatch observed by Lindzen and Choi.

Also the comparison of regardless which ground stations in equatorial Africa -as shown- with average temperatures, hovering around some 24-25 degrees Celsius with the stations in the Sahara, hovering around some 27-29 degrees, would generate some questions. Why is the tropics with a much higher moisture rate (water vapor feedback) still colder than the Sahara with much less moisture?

Now obviously this observation is confined to the tropics. However that's also the area with the largest insolation factor and constitutes close to half of the Earth surface. So what about the other half?

Looking at fig 2 of Bony et al, showing the typical cloudiness on Earth:

[ATTACH=full]128772[/ATTACH]

The equatorial clouds are associated with the Hadley convection cell and obviously in Lindzen's area of interest the only significant variable that can be associated with the short wave reflection (negative) feedback as observed by Lindzen and Choi.

Now the big curly waving frontal clouds on both hemispheres are the ones associated with frontal advection caused by colliding air masses where the warmer masses are lifted above the colder masses and generate clouds in much the same way as basic convection does. As explained in their fig 5:

[ATTACH=full]128773[/ATTACH]

obviously it seemed justified to investigate if higher temperatures also generate more clouds (Clausius Clappeyron) and hence more reflection of shoft wave energy? In other words is there really much difference in feedback between tropics and moderate climate zones?

Maybe it's an idea to review the origine of the assumed positive feedback like for instance the interpretion of the ice cores and the single Pinatubo incident (Soden et al 2002).

 

[chriscolose]

It's hard for deserts to lose much heat at the ground because the evaporation term in the surface energy budget is small compared to the moist tropics. This has little to do with water vapor feedback. If you could make the Sahara moister, the surface would cool even if you increase CO2 a bit.

It also doesn't follow at all from Clausius-Clapeyron that global cloudiness (or low clouds in particular which control the albedo more than any other kind) will increase in a warmer world. I don't understand andre's objections at all to sylas...they're repetitive and rather ill-posed.

 

[Andre]

It's hard for deserts to lose much heat at the ground because the evaporation term in the surface energy budget is small compared to the moist tropics. This has little to do with water vapor feedback. If you could make the Sahara moister, the surface would cool even if you increase CO2 a bit.

So latent heat is more important than CO2. But it's not about that, it's about why the sahara is warmer than the tropics. And it's not about feedback to variation in CO2 It's about feedback in the direct heating processes. Let's try again then,

If we assume the constructions presented by Pierrehumbert, which already incorporates convection in the Planck 'response', then the temperature reponse is not affected by convection, if I understand it right.

So for the Sahara compared to the tropics:

1: Latent heat evaporation/ condensation negative (cloud) feedback is virtually non existent
2: Absolute Atmospheric moisture is much lower hence the positive water vapor feedback is much lower in the desert
3: albedo is unclear but the low forest albedo of the tropics is neutralized by the high albedo of the more abundant clouds

So if the current greenhouse feedback ideas attribute the strongest effect to water vapor compared to the latent heat/convection then the tropics should be much warmer than the desert, which aint so.

Also, once more, the results of Lindzen et al suggest that there is something wrong with the estimation of the total feedbacks being positive.

Now why is the Sahara several degrees warmer than the tropics? Two possible reasons.

A: the negative cloud feedback (caused by moist convection) is stronger than the positive water vapor feedback.

B: The subsiding/ decending air above the desert heat up dry adiabatically while it had been cooled previously in the tropics by moist adiabatic convection. This results in Foehn effect, the dry air heating up much more than the original moist air cooled. Wouldn't this put some question marks by the assumption that convection (positive or negative) merely corrects the lapse rate?

It also doesn't follow at all from Clausius-Clapeyron that global cloudiness (or low clouds in particular which control the albedo more than any other kind) will increase in a warmer world.

No, http://www.science.uwaterloo.ca/~cchieh/cact/c123/clausius.html assume an exponential relationship between temperature and evaporation. Now the key is in "exponential". So if the lapse rate temperatures reacts lineair to an higher surface temperature, the difference in maximum water vapor content is exponential and when in the convection cooling starts from a higher temperature the difference in absolute humidity is larger with the same adiabatic cooling rate and more water will condensate.

Another question would be the source of the additional energy required to evaporate all that water with higher temperature. Would the assumed 3-4 W/m2 for double CO2 also be enough to evaporate enough water to attain that positive water vapour feedback?

 

[sylas]

I'm struggling with this thread also. It's becoming a real mess; and every time I think there's some progress, it just mutates into a new set of confusions.

In my opinion, one difficulty is that we keep shifting from the topic.

The thread topic was modeling CO2 as a greenhouse gas. This ought to be a quite straightforward, in the sense that of all the aspects of climate science it stands out as one of the simplest. The physical thermodynamics is comparatively straightforward, and mostly goes back to the nineteenth century; though with quantum mechanics having since given a better understanding of matter/radiation interactions.

On the way, we've covered some useful issues. A number of quite basic misconceptions have come up along the way, which have been addressed pretty clearly; though not, unfortunately, with a clear recognition from all participants, so I am still uncertain about how much real progress has been made in terms of getting the underlying physics sorted.

One of my own personal aims is to acknowledge explicitly and with thanks anyone who can find errors in my own work. This happens from time to time. In this thread, Andre noted correctly that I had implied convection gives a transport of heat both up and down, whereas of course it really only works to transport heat upwards. I acknowledged this promptly in [post=2299540]msg #125[/post]. It was a genuine error that I should not have made, but I'd messed up way the atmospheric equilibrium is maintained. I actively seek out useful corrections like this, and as a result the whole matter of radiative-convective equilibrium is on a firmer footing; for me in particular.

There have been a number of other similarly fundamental points which have just been left hanging. I don't know how much of what I consider the underlying physical basis for talking about feedbacks and sensitivity and radiation and convection and so on I can rely on as common ground for looking at different hypotheses about their magnitudes or signs.

Here are a couple of examples:

Magnitude of the greenhouse effect on Earth -- 33 degrees

This is an example where the discussion suddenly stopped with no recognition of the arguments I presented.

It is a fairly basic result for the nature of the Earth's surface temperature that for the amount of solar energy absorbed, surface temperatures are about 33 degrees warmer than it would be without longwave interactions in the atmosphere. This follows directly from basic thermodynamics; and it is specifically a consequence of the greenhouse effect -- primarily from CO2 and H2O.

Andre disputed this conclusion for a while, on the basis that the model was "too simple". I gave what amounts to the formal proof of why the 33 degrees is a strong lower bound on the magnitude of the effect in [post=2296677]msg #96[/post]; complexity can only make this number larger. Andre never commented on this bound further; so I don't actually know if he understands or accepts the argument for 33 degrees. I'm not just trying to rub noses in errors here; I do really find it disconcerting when someone proposes a focus ("Now let's concentrate on that." -- Andre, [post=2296495]msg #95[/post]) and then never makes any further comment when I give a detailed response.

If Andre recognizes the validity of the argument there's no shame in that. We can take it as progress and move on to consider other matters where we differ. If he doesn't -- then what was wrong with it? What about that suggestion we concentrate on this?

So just for the sake of seeing where we stand on what is recognized as common ground and what is disputed... Andre, did you understand the mathematical argument for 33 degrees being a lower bound on the magnitude of the greenhouse effect for Earth? It's a widely used number, with a clear physical basis. Can we now take this as common ground? If not, do you have any comment on the demonstration of the lower bound using Holder's inequality?

Planck response

I have asked many times for Andre to spell out what HE understands by the term "Planck response". (Messages 156, 161, 166, 171). I can appreciate one might not answer every question that shows up, but I've singled this one out as basic, many times. I STILL don't know whether Andre is on the same page as I am, or the same page even as his own sources that use the term!

This is a basic question, because it goes to the heart of how you identify what is and is not a feedback. Andre has spoken of certain processes being "assumed" as non-feedback; or about being "selective", words that suggest he considers this all rather arbitrary. But it is not arbitrary at all, and it is not a way of ignoring anything or leaving it out. It is simply a way of structuring the analysis of a complex system, and what is or is not feedback is not a matter of assumption, but a consequence of what the base response actually means. I've given the definition as I understand it many times now, and in fact this is not something that ought to be a matter of "debate" at all, in my opinion. It's really a matter of base level comprehension of the technical terminology of atmospheric physics.

Andre wants to talk more about feedback. Fine... but do we know what this actually means yet? What is the nature of the base relation to which feedback applies? If we can at least have that as common ground, it’s a basis for making some kind of substantive engagement on other points where we may differ on what evidence implies for the magnitudes and signs of feedback.

Thread focus

Both the net magnitude of the greenhouse effect, and the radiative-convective equilibrium for a given atmospheric composition which is the basis of no feedback response, go directly to modeling of the greenhouse effect, which is the main thread focus as I understand it.

CO2 is modeled as a greenhouse gas on the basis of its interactions with infrared radiation. Calculation of the base no-feedback equilibrium response shows up clearly how temperatures relate to the way energy flows between space and the surface. Any feedback process occurs when temperature feeds back into changes on some variable that appears in calculation of Planck response. Albedo alters solar input. CO2 alters longwave absorption. Cloud alters both albedo and longwave absorption. Humidity alters lapse rate as well as longwave absorption. And so on. The base response everything works upon is called the Planck response, or the net flux of energy for a given atmosphere and surface and solar input.

Now, unfortunately in my opinion, the whole matter of cloud feedbacks and humidity feedbacks -- which is really a distinct topic -- seems to be raised as a new focus.

And yet, this has come up while we are still left hanging with basic thermodynamic fundamentals unresolved from earlier in the thread.

-----

On the tropical feedbacks

OK. This seems to be a new direction Andre would like to take the discussion.

I have no problem with considering the matter of humidity feedbacks. There's been a fair bit of work on this recently, both theoretical and empirical, and nearly all of this work indicates a strong positive feedback effect. There are still a small number of papers proposing that the feedback is very small or even negative; and looking at this can be useful. Andre has provided a couple of useful references for this. But it is a new focus, and frankly, given the issues with basic underlying thermodynamics seen in this thread, I have little hope that it will make any real progress.

When I do post on this subject, I'll do it as a new post entirely, where I don't worry about other subjects or previous threads of discussion.

I just want to go on record for now concerning what I see as a whole pile of loose ends and missed opportunities in this thread.

The main issues ought to be clear and we should have been able to come to a clear mutual statement of what is common ground with respect to basic physics relating to the thread topic, of how CO2 works to give a greenhouse effect and to impact Earth's surface temperature. We could have done that without needing to resolve the question of "warming", or the magnitude of changes to CO2 concentrations, or the magnitude of Earth's sensitivity in general to small changes in forcings. The topic is more fundamental than that, and had good potential for clearing up a lot of deep misunderstandings of the relevant physics that sometimes degrades popular discussion.

Cheers -- sylas

PS. For Andre... I have referred to you above in the third person, from time to time, and I don't mean any offense by that. Sometimes people have taken offense at this in the past, so I'm adding this postscript to disclaim any attempt to belittle you. The grammar reflects nothing more than whether I am writing as a summary intended for readers in general, where I refer to you as an important thread participant; or whether I am intending to write specifically to you as one person in a two-way dialogue... that's all. Best wishes as always -- sylas.

 

[Rickeroo]

Greetings all. Brand new here, I came across this site Googling for the 270/385/1500ppm warming experiment under controlled conditions. So far, it's the closest I've come to an answer, which is I'll see essentially no difference in the temperature between the two containers. This makes sense to me when dealing with a trace gas, and it's what I expected.

I've also been on other threads, pointed to "whole atmosphere" experiments and direct measurements, where there are no doubt infinite variables. Basic scientific procedure, at least to me, would be to eliminate your variables and isolate what you want to test. Without 2 controlled containers of atmosphere, there will be a difficulty in convincing some of CO2's warming ability.

Those few controlled experiments use 100% CO2 to get a 5 C difference in temperature. Hopefully they are using a barometer to make sure they aren't over-pressuring the CO2 box. While Venus is a hot box full of CO2, I am also aware that its atmospheric pressure is 90 times that of Earth.

That being said, I've correlated two common data sets and I'd like your thoughts.

The first is the yearly ppm increase in CO2 at Mauna Loa:

http://www.esrl.noaa.gov/gmd/ccgg/trends/

The second is the temperature anonomly in degrees:

http://data.giss.nasa.gov/gistemp/2005/2005cal_fig1.gif

I had to blow the diagram up to get the yearly detail. What I found was quite interesting.

- The year to year CO2 ppm increase at Mauna Loa differed by as much as 600%, 1992 vs 1998.

- I was able to predict which years were cool, versus which years were warm, simply by looking at the ppm increase for that year. Lower ppm increases were associated with lower temperature for that year.

The cool years:

1960

1964 (significant)

1982

1992 (significant)


And these years will be on the warm side:


1965 (definitely warmer than 1964)


1972 (defintely warmer than 71)


1977 (warmer than 76)


1987 (warmer than 86)

1993, though not warm, will be warmer than 92.

Of course, 1998 and 2005 were heavy with CO2, and their yearly temperature reflects it.

This correlation, which is just about perfect, occurs in an atmosphere of ever-increasing CO2. What does this mean?

 

[sylas]

Welcome aboard Rickeroo.

Greetings all. Brand new here, I came across this site Googling for the 270/385/1500ppm warming experiment under controlled conditions. So far, it's the closest I've come to an answer, which is I'll see essentially no difference in the temperature between the two containers. This makes sense to me when dealing with a trace gas, and it's what I expected.

The atmospheric greenhouse effect arises in ten kilometers of atmosphere, with a substantial temperature gradient (lapse rate). It's not just the absorption that matters; but the fact that the portions of the atmosphere for which thermal radiation can escape into space are much colder, because they are at a high altitude. Without this temperature gradient, there would be no greenhouse effect.

The 270/385/1500 ppm CO2 corresponds to about 4/6/23 kilograms of CO2 per square meter, in about 10 kilometers of atmosphere. The forcing from this is very well known, and follows from well measured properties of CO2 and radiation, along with some basic thermodynamics.

The experiments that give the physical basis for inferring CO2 "forcings" do not work by simply testing a ten kilometer gas cell with atmospheric compositions. What experiments give you are the emissivity/absorbtivity of the gas at different frequencies, along with all the basic physics of thermodynamics that gives you the way heat flows through different materials of all kinds.

You can calculate from well established thermodynamics what is called the "Planck response", which gives you a temperature required for Earth to radiate back out again the solar input, but only for given fixed conditions. That's enough to define the "greenhouse effect" by itself, which is actually one of the most straightforward aspects of climate there is, and not in any credible scientific doubt at all.

This is a long LONG way from a full understanding of climate! The real problem is that conditions are NOT fixed. The Earth is a complex system, and all the genuine uncertainties of climate and temperature are not with the basics of the greenhouse effect itself, but with the whole sensitivity of the Earth climate system. You can't just use Planck response by itself, because as temperature changes, you also get changes in surface cover, humidity, cloud, carbon cycle, and so on. These are called "feedbacks", because they are changes driven by temperature that impact the variables that in turn establish the basic temperature response.

I've also been on other threads, pointed to "whole atmosphere" experiments and direct measurements, where there are no doubt infinite variables. Basic scientific procedure, at least to me, would be to eliminate your variables and isolate what you want to test. Without 2 controlled containers of atmosphere, there will be a difficulty in convincing some of CO2's warming ability.

Frankly, it is pretty much impossible to convince some people, and I am personally fairly relaxed about that. My main interest here is to try and help give a better general understanding of the underlying physics.

Experiments are nice, but they work best in the context of a theory to be tested by the experiment. For example, if you set up a long 10 kilometer tunnel, with a big lamp at one end to represent the sun, and used that to try and infer impact of different densities of CO2, you'd get effects quite different from the atmosphere, because you don't have a gradient of pressure with an adiabatic lapse rate driven by convection. And how would you know that this is important? Mainly, by knowing the theory of the greenhouse effect that you are supposedly testing.

The real issue for most people, I think, is not the lack of experiment. There are heaps of experiments and measurements that demonstrate the simple fact of a powerful greenhouse effect on Earth, but to see their relevance, you have to first understand the physics that they are testing. One of the clearest direct measurements of our greenhouse effect, in my opinion, is simply the direct measurement of the huge flux of atmospheric infrared backradiation coming down to the surface from the sky, with the spectrum matching our major greenhouse gases.

Atmospheric thermodynamics are quite complicated, but it is well within the capacity of a decent physics student to learn the basics of how the greenhouse effect works. The physics behind it is truly not in any credible doubt at all.

That being said, I've correlated two common data sets and I'd like your thoughts.

My first thought, on this sentence in isolation, is that correlation is a weak basis for confidence. It can be very useful as a test of predictions from theory, but in my opinion you don't really understand a physical situation until you have a theory; which means a proposed explanation of how something occurs. Finding correlations can be suggestive in looking for theories, but until you have the theory that is consistent with the observed correlations, the correlation alone can only be suggestive.

- The year to year CO2 ppm increase at Mauna Loa differed by as much as 600%, 1992 vs 1998.

- I was able to predict which years were cool, versus which years were warm, simply by looking at the ppm increase for that year. Lower ppm increases were associated with lower temperature for that year.

This isn't because of a greenhouse effect.

You are looking at a rate of increase of CO2, and comparing that with temperature. You can't explain this correlation (if it holds up) by proposing that CO2 is driving temperature. If CO2 was all that mattered, then you would expect temperature to rise all the time as CO2 is rising, but when CO2 rises more gradually, temperature would rise more gradually. But you are looking at temperature that goes up and down, which means there's something more than CO2 going on here for temperature.

And of course, there is. There's a heck of a lot going on with climate, all the time, which gives all kinds of natural variation on short terms. The major factor the big temperature increase in 1998, for example, was a very strong El Nino in that year. 1992 was cool, mainly because of the big Pinatubo volcano eruption. And so on. These are not merely correlation based arguments; there are good physical theories which indicate why you get hotter years with El Nino, and cooler ones with a big volcanic eruption.

The changes in the rate of increase of CO2 from year to year do not lead to the big temperature swings that you are looking at. We know the forcing involved, and it's a fairly strong steady increase, but not something that has huge short term forcings to make temperature swing wildly between different years.

This correlation, which is just about perfect, occurs in an atmosphere of ever-increasing CO2. What does this mean?

I don't think the correlation is all that good. I measured it for myself with a spreadsheet just now, using the annual mean grown rate for CO2 from the Mauna Loa site, and the GISS data underlying the graph you linked, from the Global Land-Ocean Temperature Index. The correlation I got was 0.735.

I think the most likely reason for any such correlation, if it is a real effect, is an impact of temperature on the carbon cycle. There are enormous fluxes of carbon in and out of the atmosphere from vegetation, and temperature is likely to have an impact on that, rather than the other way around.

Note that I am guessing at a theory for the correlation. This is the first step in a genuinely scientific project. The next would be to try and test the theory, with an experiment that has the potential to falsify it. I have no idea what that might turn up. The point is... merely noting a correlation is not a sufficient basis for a good scientific theory.

Cheers -- sylas

 

[Rickeroo]

Thanks Sylas.

With the correlation being .735, would that be enough to say that the comparision is relevant, or at least something to be looked at further?

There are enormous fluxes of carbon in and out of the atmosphere from vegetation, and temperature is likely to have an impact on that, rather than the other way around.

Would that suggest that temperature leads carbon, at least to some degree?

 

[sylas]

Thanks Sylas.

With the correlation being .735, would that be enough to say that the comparision is relevant, or at least something to be looked at further?

It's enough to be suggestive; worth looking at. I would not be at all surprised to see some kind of temperature related effect, but as I said, it's a bit of a guess. I suspect that if we hunted through the literature on the carbon cycle we'd find some consideration of such effects. I don't know off the top of my head and I haven't take time to look.

Would that suggest that temperature leads carbon, at least to some degree?

Sure. If you look at the Mauna Loa data, you should see that the total atmospheric CO2 looks like a tilted sine wave. There is a very strong annual flux of CO2 in and out of the atmosphere every year, and then a steady continuous increase going on all the time as well.

In a way, the Earth "breathes". The effect is seasonal, and it arises mostly (I think) from changes in the way vegetation takes in and emits CO2 between winter and summer. The total flux of CO2 involved is huge.

However, it is not cummulative. The carbon taken into vegetation is released again later. There's a continual cycle of carbon between atmosphere, vegetation, soil and ocean, going on all the time. The human input is a bit different, because we are basically acting as a way for carbon from geological reserves (fossil fuels) to make its way into the carbon cycle, and this leads to a gradual increase in carbon in all parts of the carbon cycle: the atmosphere and ocean especially.

Basically, the carbon cycle consists of several "reservoirs" of carbon, each with a different total capacity, and with carbon fluxes between them. Here's a diagram, from an online textbook: http://www.uwsp.edu/geO/faculty/ritter/geog101/textbook/earth_system/biogeochemical_cycles.html .) The numbers are the capacities of the reservoirs, and the total amount of carbon moving between them annually.

What human emissions do is add 5.5 GigaTonnes per year into the atmosphere. Over a century, this has resulted in a large increase in total carbon in the atmosphere, ocean and terrestrial reservoirs. But at the same time, there is about 90 Gigatonnes per year going each way between ocean and atmosphere, and about 120 Gigatonnes per year each way between atmosphere and vegetation/soil on land. Temperature effects can shift the balance of the reservoirs a bit, enough to make the net atmospheric increase rise or fall a bit, and I would guess this is the main reason for the correlation you have observed.

Cheers -- sylas

 

[Rickeroo]

Thanks Sylas. Yes, the sine wave at Manua Loa makes perfect sense with the seasonal vegetation level, as does the net addition of CO2.

It also makes sense that temperature would have an effect of CO2 transfer or absorption, something to be looked at anyway.

My next task will be to correlate the temperature with the rise in CO2, and to correlate the melting ice with the rise in sea level.

 

[Richard111]

silas - thank you for your comments and advice way back in post #155.
First I must admit I am not the originator of that thought experiment. I read it on a blog first, but sadly did not record the url. Later I attempted to work out the mass of the atmosphere on 1sq/m at standard pressure and found my answer was about right. This chuffed me no end, so I attempted to calculate the mass of global CO2 and got it wrong.

Minor correction here: CO2 is about 0.04% by volume; so you have to scale by 44/29 (the molecular weight of CO2 and the average molecular weight of air) to get pretty close to 6 kilograms.

Your correction improved my calculation but not enough yet. I was attempting to calulate the global mass of CO2. There are 10^6 square meters in a square kilometer so we have 6 x 10^6 kg/km^2.
From this link: http://www.net-comber.com/worldarea.html I selected 510,072,200 km^2 as the total global surface area and arrive at 3.06 x 10^15 kg. (umm.. still something wrong)

Towards the end of your post you mention dealing with the atmosphere in slabs and integrating the temperature changes, well, its 55 years since anyone last attempted to teach me calculus, so I am a lost cause there.

A point for clarification; in our column of well mixed gasses, as we progress upwards with a constant lapse rate, we not only have less temperature, we also have less density so the total mass of CO2 per "slab" will also be less. Therefore I feel we must take into account reduced mass as well as temperature.

I would like to get hold of the following book but it is not available in my local library. Might be in the Uni library. I will have to wait until Amazon offers used copies at much reduced prices. A short critique at:

http://climatesci.org/2006/05/05/co2h2o/

Relative Roles of CO2 and Water Vapor in Radiative Forcing
Filed under: Climate Change Forcings & Feedbacks, Climate Change Metrics — Roger Pielke Sr. @ 6:09 am
In the second edition of our book

” Cotton, W.R. and R.A. Pielke, 2006: Human impacts on weather and climate, 2nd Edition, Cambridge University Press, New York, in press ”

 

[shredder666]

before I make my views on this, how would current rate of change in global temperature affect the rate of which major greenhouse gases such as water vapour from the ocean, methane in garbage dumps, and CO2 trapped in soils, are released?

 

[sylas]

Your correction improved my calculation but not enough yet. I was attempting to calulate the global mass of CO2. There are 10^6 square meters in a square kilometer so we have 6 x 10^6 kg/km^2.
From this link: http://www.net-comber.com/worldarea.html I selected 510,072,200 km^2 as the total global surface area and arrive at 3.06 x 10^15 kg. (umm.. still something wrong)

That sounds pretty much exactly correct. That's what is given in wikipedia's Carbon Dioxide article. (Wikipedia is an unsafe source for this forum, but it's okay as a confirmation of your calculation.)

Note that people often speak of the mass of carbon in the atmosphere, which would be 12/44 times the mass of carbon dioxide. This corresponds to about 8.2 * 10¹⁴ kg, or 820 Gigatonnes. The carbon cycle diagram I've just posted gives 750; but that may just be an older value, corresponding to about 355ppm CO2 in about 1990.

A point for clarification; in our column of well mixed gasses, as we progress upwards with a constant lapse rate, we not only have less temperature, we also have less density so the total mass of CO2 per "slab" will also be less. Therefore I feel we must take into account reduced mass as well as temperature.

Quite so. However, we often use pressure as the altitude co-ordinate, which doesn't have that problem. This makes all sorts of calculations more straightforward.

I would like to get hold of the following book but it is not available in my local library. Might be in the Uni library. I will have to wait until Amazon offers used copies at much reduced prices. A short critique at:

http://climatesci.org/2006/05/05/co2h2o/

I can confirm for you right away the main conclusion. H2O is easily the most important gas in our atmosphere for giving the greenhouse effect. I've noted this a couple of times in the thread. An increase in humidity has a very strong effect; much more than a similar increase in carbon dioxide.

This is, in fact, the reason why "water vapour feedback" is such an important part of the more complex question of climate sensitivity. The amount of water in the atmosphere is mostly a function of temperature.

Industry emits huge amounts of water vapour into the atmosphere. Ironically, many pictures trying to show a picture of pollution are actually showing discharges of water vapour. A picture of CO2 emissions is much more boring, because it is invisible.

But the effect of human H2O emissions is almost nil on atmospheric water vapour. Anything extra we add comes out again almost immediately, because the water cycle is so rapid. So you really can't hope to increase humidity just by adding water. The best way to increase the water content of the atmosphere is simply to heat things up somehow. That's why carbon dioxide, despite being a smaller part of the total greenhouse effect, is what is forcing the changes. The warming effect of carbon is amplified by the effects of additional water from this feedback. See our previous discussion on "Planck response" and feedback. There are a lot of other effects to consider as well. Water vapour will reduce the lapse rate, which is a negative feedback; and changes to cloud can reflect sunlight (negative feedback) and also absorb infrared even more strongly than gaseous vapour (positive feedback). It looks like we may be pulling apart some of the scientific literature on this question as the thread progresses.

Cheers -- sylas

 

[Richard111]

Water vapour will reduce the lapse rate, which is a negative feedback; and changes to cloud can reflect sunlight (negative feedback) and also absorb infrared even more strongly than gaseous vapour (positive feedback). It looks like we may be pulling apart some of the scientific literature on this question as the thread progresses.

When water vapour reaches dew point and starts to condense on whatever CCN's are available, there is a drop in air pressure. I believe this can result in fierce updraughts within large cumulus clouds. You can see this effect here in Pembrokeshire. The "Finger of God" extending upwards from the cloud tops. Very impressive and a warning to any aircraft to keep clear. I would assume a lot of energy would be transported upwards even while the cloud is accumulating energy from the sunlight above and longwave radiation from below. At night I assume the "feedbacks" will change due to lack of solar input.

I feel I must acquire more understanding of the "greenhouse effect" of water vapour and liquid water (clouds, fog) and ice crystals (cirrus clouds) in the atmosphere and the effect on positive/negative feedback. Possibly, then, an understanding of the "feedback" due to increasing CO2 will be more clear to me.

So back to my imaginary 1m^2 column of air and a dry adiabatic lapse rate of 3C per 1000 feet and assuming the air temparature has stabilised from about 2 meters above surface level I expect the air temperature at 10,000 feet (plus 6 feet or so) to be some 30C cooler.

Now consider the air in 1000 foot slabs/layers, each layer 3C cooler than the layer below and that a net transfer of heat will only flow from hotter to cooler. We must also bear in mind that each layer has less mass than the layer below. The flow of energy is upwards. It appears only the bottom layer of a 1000 feet or so seems to have any feedback to the surface even as the net flow is upwards. It has been established that increasing the water vapour content does not effect the dry adiabatic lapse rate therefore any increase in CO2 also has no effect in dry air.

Sea surface temperatures appear to range from a minimum of -2C to a maximum of about 33C. A much smaller variation than on land and also less inclined to change sharply over short time periods. Seeing that slightly more than 70% of the Earth's surface is water I thought this might be a good place to start. In my attempts to gain some knowledge about water I have been looking at the Water Absorption Spectrum page on Martin Chaplin's site.

I must confess I find this site very heavy going, but extremely interesting. I never knew water could take on so many different molecular configurations which seem to be responsive to different temperature regimes. Every change seems to have its own spectral response. Quite awesome.

On the above page is a graph titled The visible and UV spectra of liquid water

http://www1.lsbu.ac.uk/water/images/watopt.gif

You can see clearly how light and some UV can penetrate quite deeply into clear water. (I read somewhere that you can get sunburn under water and thought Huh!) The area of the graph I am trying to get to grips with is the IR region. From about 3µm to 100µm. Here penetration seems limited. If I read that correctly I fail to see how downwelling IR from any source can possibly provide any significant heating into water. From other literature (haven't found it on Chaplin's site) I read that IR reacts with surface molecules of water to increase the rate of production of water vapour. How this may be quantified I haven't clue.

So to satisfy my curiosity I will suspend a shielded IR source over a measured quantity of water and try to record any temperature change. The IR source, still to be obtained, will be a circular slab of steel or cast iron of about 2kg mass and the shield will be a small drum such that airflow past the source is minimal but heat radiated downward will have a clear path to the water surface. Should be interesting. I will post the result in due course.