Estimating the impact of CO2 on global mean temperature

[sylas]

In another thread (see [post=2155677]msg #17 of "Only dirty coal can save the Earth"[/post]), user Bystander suggested I lay out more carefully for critical examination the physical basis for the impact of CO₂ on climate, so that the assumptions can be seen clearly. This is an excellent idea, and here is my attempt.

This post is a calculation, cited to the literature as forum guidelines advise, quantifying the importance of CO₂ to climate. It aims to be transparent, so each step in the calculation is clear and assumptions can be identified.

There's nothing here that is scientifically original or unusual. I am not attempting to cover every factor impacting climate, nor a full account of every period of history. I show that the impact of CO₂ is physically bound to be significant in the present. There are other factors, both positive and negative, known and unknown. CO₂ is one of the simplest. Its impact is basic physics, and necessarily a major contributor in recent decades.

This thread is not intended to take up every point of climate on which there is disagreement. I am specifically addressing quantification of only one of the many factors involved, to show that CO₂ is necessarily significant in the present. I request that we maintain that primary focus, and leave detailed discussion of other aspects of climate science to other threads.

Any point here can be expanded upon as needed. I will defend it vigorously as elementary physics. In my view, and in the view of nearly all scientists working on climate, popular disbelief on this specific point is poorly founded, and supported with bad science. But I will engage disagreement on that with civility and respect, and questions are certainly welcome. I will deal with objections on their own merits, and not by belittling people good enough to come and join in the thread. Be welcome, be comfortable, and for some of you -- be challenged!

The short form

Here's the sequence in summary form.

$$\begin{array}{lcll} \textrm{Energy balance} & \approx 239 & W/m^2 & \textrm{(energy }\text{flux to, and from, space)} \\ \textrm{CO2 forcing} & 3.7 & W/m^2/2xCO_2 & \textrm{(change to balance per doubling)} \\ \textrm{Current CO2 increase} & 2 & ppm/year \\ \textrm{Current 2xCO2 per year} & \log_2(1+2/385) = 0.0075 & 2xCO_2 / year & \textrm{(using 385ppm for current levels)} \\ \textrm{Climate sensitivity} & 3 \pm 1.5 & K/2xCO_2 & \textrm{(Temperature response for forcings, to restore balance)} \\ \textrm{Temperature impact of CO2} & 0.022 \pm 0.011 & K/year & \textrm{Estimated contribution of CO2 to current warming trends}\end{array}$$

​

This is a very conventional way of evaluating climate impacts in science. What follows is a brief explanation of the steps, with citations for the numbers and methods.

(1) Energy balance

Earth absorbs about 239 W/m² from the Sun, accurate to within a couple of percent. This is a global annual average, corresponding to a mean for the whole planet on the time scale of a couple of years, in Watts per square meter.

All of that energy has to go somewhere. Almost all of it is radiated back into space as IR thermal radiation. A small excess can be taken up with Earth's own finite internal stores of energy. The largest available internal store is the heat capacity of the ocean, and recent research indicates that a bit under 1 W/m² is currently being absorbed into the ocean, as it gradually increases in temperature. This is large by comparison with what is normal for the Earth. The flux into the ocean is known with about 20% accuracy at best; a recent published estimate is 0.85 +/- 0.15 (Hansen et al 2005).

(2) The atmosphere

Earth's atmosphere is the reason why temperatures here are on average so much higher than on our airless moon.

The Earth's surface is radiating something like 390 W/m² as IR radiation from the surface, known to within a couple of percent. Most of that ends up heating up the atmosphere, which then sheds heat into space. On top of that, thermal convection and heat of evaporation transfers almost another 100 W/m² into the atmosphere from the surface. Most of the Earth's thermal radiation into space actually comes from the atmosphere.

The atmosphere radiates in all directions, and on average something like 330 W/m² ends up coming back down to the surface. The net flow of energy is thus something like 60 to 70 W/m² radiant energy leaving the surface, plus the additional 100 or so W/m² as convection and latent heat. This is what balances with the energy absorbed at the surface, from the Sun.

Here is a recently published diagram, summarizing how energy flows at present between space, the surface and the atmosphere (Trenberth et al, 2009).

(3) Forcing the energy balance.

Temperature is governed by the flow of energy. The Earth's temperature is at a level that keeps a balance between energy received from the Sun, and emitted back out to space. There can be a small imbalance while excess energy is absorbed or released into the oceans. As equilibrium is reached, the whole planet has average temperatures sufficient for energy balance between solar input and thermal emission back out to space.

If anything occurs to upset the balance, such as a change in the Sun, or in Earth's albedo (reflection), or the atmosphere's composition (absorption), then the temperature of the Earth will respond until the balance is restored. An increase in energy flowing inwards will heat up the surface; and an increase in energy flowing outwards will cool the surface.

A forcing is defined as a change to the energy balance at the top of the troposphere, before balance is restored by changing temperatures. This definition is useful, because the temperature response to different forcings depends mainly on the magnitude of the imbalance, and not so much on how it arises. (Ramaswamy et al 2001)

(4) The forcing of carbon dioxide

Many factors for climate are hard to quantify, but carbon dioxide is one of the most straightforward. The impact, as a forcing, can be calculated. The effect is approximately logarithmic, meaning that the magnitude of the forcing depends on the factor by which concentrations increase. Doubling of concentration is a wildly used benchmark.

The method of calculation is as follows. The atmosphere is treated as a column of gas, with a longwave input at one end (the surface), and short wave input at the other (the Sun). Using the known absorption spectra of gases in the atmosphere, the calculation proceeds line by line through the spectrum to calculate how much radiation is transmitted and absorbed and emitted all along the atmospheric column. The end result is a profile of radiative heating and radiation flux, with a power spectrum of longwave emission at the top, and another for backradiation and solar transmission at the bottom. The calculation is repeated for different gas concentrations, and different conditions. With higher concentrations, the backradiation increases and the emission at the top drops. The forcing, by definition, corresponds to the change in longwave emission at the tropopause.

A standard reference for the calculation is in Myhre et al, (1998). The forcing for any doubling of CO₂ is about 3.71 W/m², or 5.35 per natural log. This is known to high precision for well defined conditions, and to about 10% accuracy in general for the Earth. That is, doubling CO₂ in the atmosphere results in 3.7 W/m² less IR emission escaping into space… until the surface heats up sufficiently to restore the balance.

It is a common convention in the literature to use 2xCO₂ as the unit for forcing, rather than W/m². There are good reasons for this, which are a bit beyond the scope of the post. Essentially, it is because the energy imbalance definition has about five different forms when you really nail it down, whereas doubling of CO₂ concentrations is comparatively unambiguous, with a straightforward radiative impact.

(5) The temperature response to forcing -- climate sensitivity

If nothing changes other than temperature, the response is straightforward. Emissions in general are proportional to the fourth power of temperature, and on Earth, with surface temperatures of about 288K and emissions to space of about 240 W/m², we get about 0.3C for each W/m² forcing. You can nail that down more carefully, but it does correspond closely to what you get with a radiation transfer calculation. It's called the "Planck response" in climate science. Converting forcing units, this corresponds to about 1.1C per 2xCO₂.

In reality, when you change temperature, all kinds of other things start to change as well, someone of which then have an impact of their own on the energy balance. This is called "feedback", and it means the climate response may be amplified, or damped. Several important feedback mechanisms are known (humidity, cloud cover, lapse rate, ice-albedo) and there have been attempts to quantify them. There are both positive and negative feedbacks involved, though the positive feedbacks are stronger. That is, the climate response in reality is rather more than 1.1C per 2xCO₂.

For this post, I will use observational constraints. The question is, what is the temperature response in degrees per unit forcing? This is called climate sensitivity, and it can be estimated with data for known episodes of climate change.

One line of research has been study of the brief episode of global cooling that follows a large volcanic eruption. Wigley et al (2005) study a number of eruptions, and three in particular where the data allows estimates of climate sensitivity. The estimates in K/2xCO₂ (with 2σ confidence limits) are 2.83{1.28 .. 6.32}, 1.54 {0.3 .. 7.73} and 3.03 {1.79 .. 2.59}. Sensitivity has also been constrained by a number of other cases, most especially the last glacial maximum some 20,000 years ago. There is quite an extensive literature on this, and nearly all of it continues to estimate sensitivity in the range of about 2 to 4.5C. A useful survey and attempt to combine estimates is by Annan and Hargreaves (2006). A very few isolated papers propose lower sensitivities, but these invariably are flawed by plainly identified methodological problems.

Sensitivity is a genuinely open research question with considerable uncertainty. It is, by far, the largest source of uncertainty for the conclusions of this post. I submit that climate sensitivity is very likely in the range 2 to 4.5 K/2xCO₂, and highly unlikely to be less than 1.5.​

(6) The contribution of CO₂ to modern warming

We have quite a good idea of atmospheric CO₂ levels in recent history. Pre industrial levels are about 280 ppm, and current levels are about 386 ppm. The current rate of increase is about 2ppm/yr.

For any change in CO₂ concentrations from A to B, the corresponding change in temperature is going to be S*log₂(B/A), where 1.5 < S < 4.5 is the climate sensitivity. You can also use S_e*Ln(B/A), where 2.2 < S_e < 6.5, which will be easier for most calculators. Best estimate is S=3, or S_e = 4.3.

For example, over recent decades the rate of increase of CO₂ has been around about 2ppm/year, on top of about 385ppm. The corresponding contribution of CO₂ to rising temperature is about S_e*Ln(387/385), which is in the range 0.011 to 0.034 C/year, with a best estimate of 0.022 C/year.

Direct measurements indicate that globally averaged surface temperatures on Earth are increasing in recent decades, at around about 0.02 C/year. (Brohan et al, 2005)

There are many factors that must be involved in any credible account of the causes for rising temperatures. It is certainly not driven by CO₂ only. The calculations here demonstrate that CO₂ is necessarily an important factor, and may help show why it figures so prominently in the scientific literature for modern global climate.

References

In order to keep discussions grounded in legitimate science, the Earth forum requires sources to be peer reviewed. I have therefore given a more than usually thorough list of references for the numbers used in this post.

• Annan, J. D., and J. C. Hargreaves (2006), http://www.agu.org/pubs/crossref/2006/2005GL025259.shtml, in Geophys. Res. Lett., 33, L06704, doi:10.1029/2005GL025259. (Looks at several observational constraints on sensitivity.)
• P. Brohan, J.J. Kennedy, I. Harris, S.F.B. Tett and P.D. Jones, http://www.agu.org/pubs/crossref/2006/2005JD006548.shtml. J. Geophys. Res., Vol 111, D12106, doi:10.1029/2005JD006548. (Measurement of change to global mean surface temperatures, with the HadCRUT3 dataset.)
• Hansen, J. et. al. (15 authors) (2005) Earth's Energy Imbalance: Confirmation and Implications, in Science, Vol 308, no 5727, pp 1431-1435. (Measurement of a small energy imbalance, and of the net energy flux into the ocean.)
• Myhre et al., (1998) http://www.agu.org/pubs/crossref/1998/98GL01908.shtml, Geophysical Research Letters, Vol 25, No. 14, pp 2715-2718. (Calculation of the CO₂ forcing.)
• Ramaswamy, V. et. al. (2001) http://www.grida.no/publications/other/ipcc_tar/?src=/CLIMATE/IPCC_TAR/WG1/212.htm , in Climate Change 2001: The Scientific Basis (Houghton, J. T. et al. eds, Cambridge, U.K.: Cambridge University Press. (Defines radiative forcing.)
• Trenberth, K.E., Fasullo, J.T., and Kiehl, J. (2009) http://ams.allenpress.com/archive/1520-0477/90/3/pdf/i1520-0477-90-3-311.pdf , in Bulletin of the AMS, Vol 90, pp 311-323. (Basic reference for energy flows and energy balance on Earth.)
• Wigley, T. M. L., C. M. Ammann, B. D. Santer, and S. C. B. Raper (2005), Effect of climate sensitivity on the response to volcanic forcing, in J. Geophys. Res., Vol 110, D09107, doi:10.1029/2004JD005557. (Sensitivity estimated from volcanoes.)

 

[Saul]

Hi Sylas,

(4) The forcing of carbon dioxide

Many factors for climate are hard to quantify, but carbon dioxide is one of the most straightforward. The impact, as a forcing, can be calculated. The effect is approximately logarithmic, meaning that the magnitude of the forcing depends on the factor by which concentrations increase. Doubling of concentration is a wildly used benchmark.

The method of calculation is as follows. The atmosphere is treated as a column of gas, with a longwave input at one end (the surface), and short wave input at the other (the Sun). Using the known absorption spectra of gases in the atmosphere, the calculation proceeds line by line through the spectrum to calculate how much radiation is transmitted and absorbed and emitted all along the atmospheric column. The end result is a profile of radiative heating and radiation flux, with a power spectrum of longwave emission at the top, and another for backradiation and solar transmission at the bottom. The calculation is repeated for different gas concentrations, and different conditions. With higher concentrations, the backradiation increases and the emission at the top drops. The forcing, by definition, corresponds to the change in longwave emission at the tropopause.

A standard reference for the calculation is in Myhre et al, (1998). The forcing for any doubling of CO2 is about 3.71 W/m2, or 5.35 per natural log. This is known to high precision for well defined conditions, and to about 10% accuracy in general for the Earth. That is, doubling CO2 in the atmosphere results in 3.7 W/m2 less IR emission escaping into space… until the surface heats up sufficiently to restore the balance.

It is a common convention in the literature to use 2xCO2 as the unit for forcing, rather than W/m2. There are good reasons for this, which are a bit beyond the scope of the post. Essentially, it is because the energy imbalance definition has about five different forms when you really nail it down, whereas doubling of CO2 concentrations is comparatively unambiguous, with a straightforward radiative impact.

What you state above is a simplification of the physical problem.

It is not a fact that a doubling of CO2 will increase forcing by 3.71 W/m^2. Why do you make that statement?

The consensus is the lower atmosphere is saturated from the standpoint of direct heating effects of CO2. Adding more CO2 to the lower atmosphere will not result in higher surface temperatures.

The CO2 warming is hypothesized to occur in the upper atmosphere, however, upper atmosphere temperatures have not increased as predicted by the models. A second indication that something is incorrect (GCM vs physical system) is that in the past CO2 levels have been high when the planet was cold and low when the planet was warm.

It appears based on measurements that something is incorrect with the basic modeling assumptions.

 

[sylas]

Thanks for joining in Saul. I was getting worried that no-one would be interested! I'm going to disagree with you (as promised ) but I appreciate the chance to go over some common points of popular dispute!

It is not a fact that a doubling of CO₂ will increase forcing by 3.71 W/m^2. Why do you make that statement?

Because it is a fact. I gave the reference, and you can find plenty of others that give similar figures. It's a physical inevitability, given the nature of how CO₂ interacts with thermal radiation, and we can demonstrate it with some calculations of radiation transmission in the atmosphere. I'll give a pointer to a suitable calculator here, and a worked example for doubled CO₂.

The consensus is the lower atmosphere is saturated from the standpoint of direct heating effects of CO₂. Adding more CO₂ to the lower atmosphere will not result in higher surface temperatures.

That's incorrect. As far as the physics is concerned, adding CO₂ inevitably leads to higher surface temperatures. The real question is how much. We know the forcing pretty well, but the sensitivity is not as well known.

When you speak of "saturation", the proper use for this term is radiative transfer; not temperature. We say that a part of the spectrum is "saturated" if the atmosphere is opaque in that band; that is, if all the light in that band of the spectrum is completely absorbed.

There is a consensus -- or better, it is a basic fact of life -- that CO₂ interacts with some wavelengths of light and not others; and that for most of the spectrum, the atmosphere is either saturated at those wavelengths, or else it is a band where CO₂ is transparent.

If that was universally the case at every wavelength, independent of concentration, then it would be true that adding more CO₂ would have little effect. The impact of additional CO₂ arises because as you add CO₂, the bounds of the saturated bands shift a little bit. This is a general effect not limited to CO₂, or to Earth's atmosphere. There is additional absorption in the "wings" of the absorption region, and this is the major cause of the forcing.

There are a number of tools available for looking at absorption spectra. I'll use an http://geosci.uchicago.edu/~archer/cgimodels/radiation.html , made available at the University of Chicago, which uses much the same techniques as in the paper I cited above, (Myhre 1998). It is not as accurate as fully up to date calculations, but it is able to run quickly for an on-line server, and thus makes a useful pedagogical tool. You can try this yourself as well.

First of all, here is a spectrum for a 1976 US-standard atmosphere profile. CO₂ levels are set to 375 ppm, no rain is considered, a surface temperature is set to 288.20K (about 15C) and the result is the spectrum you see coming up from the surface with a detector at an altitude of 20km (roughly the tropopause).

I've labeled three parts of the spectrum in this diagram, to illustrate how it works.

• A saturated band. The spectrum here follows close to a 220K blackbody, which corresponds to radiation from the top of the atmosphere. The atmosphere is opaque at these wavelengths.
• A transparent band. The spectrum here follows a bit above the 280K blackbody, which corresponds to radiation from the surface. The atmosphere is transparent at these wavelengths.
• A partial absorption band. The absorption here is nearly all H₂O; CO₂ is transparent in this part of the spectrum. H₂O is not evenly distributed in the atmospheric column, which is part of the reason the spectrum fits neither surface temperatures nor top of the atmosphere temperatures. CO₂ is much easier to handle in the calculation.

In this spectrum, the total outgoing energy flow is 258.893 W/m².

Now here is the spectrum calculated for the same atmosphere, but with 750ppm CO₂. It looks almost the same; so I've done an overlay of the 375 spectrum on the right, to help compare. You can see that the 750 CO₂ spectrum absorbs a little bit more radiation along the wings of the main saturated band; enough to reduce the outgoing energy by a little over 1%. Outgoing radiation this time is 255.470 W/m². The difference is 3.423 W/m²; close to what is obtained by the more precise calculations used in the literature. You should get similar differences when you try other atmospheric profiles or conditions, and double the CO₂.

Since I am not an expert in my own right, you are well advised to check up further from more authoritative sources, such as any undergraduate level text on atmospheric radiation. They should all deal with the concept of absorption in the wings of a saturated band. The spectra I have provided here are calculations, based on solidly established physics. The details of line broadening and absorption in the wings of a saturated bands can be seen directly in experiments. The current spectrum of the Earth can be observed from space, and it conforms to the theoretical calculation for prevailing conditions.

The CO2 warming is hypothesized to occur in the upper atmosphere, however, upper atmosphere temperatures have not increased as predicted by the models. A second indication that something is incorrect (GCM vs physical system) is that in the past CO2 levels have been high when the planet was cold and low when the planet was warm.

It appears based on measurements that something is incorrect with the basic modeling assumptions.

The real situation is a lot more complex than that. In fact, the major difference between greenhouse warming and other potential factors is that the greenhouse effect leads to strong cooling in the upper atmosphere, above the tropopause.


I think you may mean that there is expected to be strong warming in the middle troposphere, which is quite true. But that's not a greenhouse effect; we expect that for just about any cause of warming.

(1) Stratospheric cooling

The upper atmosphere is mainly warmed from the Sun. It is far from the surface, and above the tropopause which marks the limit of convection. The thermal radiation which is absorbed by greenhouse gases is already absorbed lower down, and the only thermal radiation received in the stratosphere is either at wavelengths where there's no absorption anyway to let heating occur (the transparent bands), or else at wavelengths being emitted the top of the troposphere, which is actually colder than the stratosphere (the saturated bands). Hence greenhouse gases don't help for heating the atmosphere at high altitudes. On the other hand, greenhouse gases still can emit thermal radiation very effectively. So the stratosphere is warmed by solar input absorbed by ozone in particular, and then sheds its heat by thermal radiation from greenhouse gases. The more greenhouse gases, the more efficiently the upper atmosphere is cooled.

(2) Mid troposphere warming

The troposphere is heated mainly from the Earth's surface. Heat is transported into the troposphere by convection, by latent heat, and by thermal radiation. Any source of increased heat at the surface is going to increase the energy flowing up into the troposphere, whether it be from extra greenhouse gas concentrations, more solar absorption, or more solar input.

Any surface warming is expected to be magnified in the mid-troposphere. This is nothing particularly to do with greenhouse effects; it is mainly a consequence of a changing lapse rate. With additional surface warming, for any reason, the lapse rate (the rate at which temperatures falls with altitude) is expected to drop. The "dry adiabat" shows a much sharper fall in temperatures than the "moist adiabat", and a warmer surface means more moisture and a reduced lapse rate – especially around the tropics.

The dry adiabat is about 9.8C degrees per kilometer. The moist adiabat can be half that. Now imagine that the surface warms, for any reason at all, by 5 degrees. At the same time, due to additional specific humidity, the lapse rate drops from, say, 6 to 5 degrees per kilometer. Then one kilometer above the surface will have warmed by 6 degrees, and two kilometers above the surface by 7 degrees, and so on; up until the atmosphere dries out a bit and the lapse rate comes back to normal. Hence the strongest warming is expected in the middle and high troposphere, in the tropics, but below the tropopause.

(3) Comparison with observations

It's a lot harder to measure temperature trends in the atmosphere than at the surface. There are many sources of error in measurements, and a sure sign of second rate information is when there's no quantification of these uncertainities.

Up until a few years ago, there was a genuine puzzle about tropospheric temperature measurements. The expected rise in temperature was not apparent, and it should have been, even given the measurement difficulties. Although this has often been presented as a problem for greenhouse warming theory, in fact it is nothing to do with greenhouse; but is a problem with the expected physics of temperature and heat transport in general.

More recently, these problems have been somewhat resolved. There are still large uncertainties in the numbers, and there's still lots of room for more careful measurements and testing of our models for lapse rate and atmospheric heat transport. But some sources of error have been clearly identified which obscured the tropospheric warming, which does seem to be there as expected. There's quite a lot of literature on this now; although it remains an active focus of research. Mainly it is a problem of trying to narrow down measurements. The big take home message, however, is that the question is entirely distinct from greenhouse warming. It's about how the atmosphere responds to any warming at all.

A major recent paper on this is:

• Santer, B.D. et. al. (17 authors) (2008) http://www3.interscience.wiley.com/journal/121433727/abstract?CRETRY=1&SRETRY=0, in International Journal of Climatology, Vol 28, Iss 13, pp 1703-1722.

This remains unfinished business, but as matters stand today there is no major inconsistency between theory and observation.

The stratosphere, on the other hand, has been cooling nicely just as expected. There are multiple causes of this; a reduced ozone concentration and reduced solar absorption has a major role alongside increase greenhouse concentration and enhanced thermal emission. But basically the measurements of the stratosphere give a strong confirmation that it is an enhanced greenhouse effect that is the major cause of global temperature trends over recent decades.

As a minor point, you have the situation with the past backwards. Throughout the past, high CO₂ is found with high temperatures, and low CO₂ with low temperatures. There's a kind of two way effect involved here, which is the more usual source of popular confusion. Rising temperatures can raise CO₂ levels, and rising CO₂ can raise temperatures. They tend to feed on each other, so it's not always clear which one is in the driver's seat! But the relationship is a very clear positive correlation throughout Earth's history.

Cheers -- Sylas

 

[Saul]

Thanks for joining in Saul. I was getting worried that no-one would be interested! I'm going to disagree with you (as promised ) but I appreciate the chance to go over some common points of popular dispute!

Because it is a fact. I gave the reference, and you can find plenty of others that give similar figures. It's a physical inevitability, given the nature of how CO₂ interacts with thermal radiation, and we can demonstrate it with some calculations of radiation transmission in the atmosphere. I'll give a pointer to a suitable calculator here, and a worked example for doubled CO₂.

Saul: A scientific theory must make predictions. In the case of this problem AWG predictions (GCM) do not agree with observations.

1. Tropospheric temperatures have not warmed as predicted by the GCM.
2. In the past the planet has been warm when CO2 levels were low and cold when CO2 levels where high.
3. The planet's temperature has been drifting down in the last 7 years. There is data that indicates that the cooling trending is accelerating.

When theory does not agree with observations the theory is in crisis. Something is fundamentally incorrect with the base assumptions or with the model.

That's incorrect. As far as the physics is concerned, adding CO₂ inevitably leads to higher surface temperatures. The real question is how much. We know the forcing pretty well, but the sensitivity is not as well known.

When you speak of "saturation", the proper use for this term is radiative transfer; not temperature. We say that a part of the spectrum is "saturated" if the atmosphere is opaque in that band; that is, if all the light in that band of the spectrum is completely absorbed.

Saul: Let's agree for this problem, the word "saturate" means whether additional CO2 added to the atmosphere can cause a significant increase in planetary temperature. There are other mechanisms that transfer heat in the atmosphere such as convection, evaporation and condensing of water. The oceans cover roughly 70% of the planet's surface. Those equations and links that you provide ignore the other methods of transferring heat in the atmosphere. They are toy models.

Higher CO2 concentrations in the lower troposphere does not result increased warming. If it were possible to have higher concentrations of CO2 in the lower atmosphere and leave the CO2 concentration unchanged in the upper atmosphere there would according to the GCM be no warming of the planet. The GCM show the warming will occur in the upper atmosphere which then warms the planet due to long wave radiation being radiated back towards the planet's surface.

There is a consensus -- or better, it is a basic fact of life -- that CO₂ interacts with some wavelengths of light and not others; and that for most of the spectrum, the atmosphere is either saturated at those wavelengths, or else it is a band where CO₂ is transparent.

Saul: You are speaking about a theoretical model that does not match reality. When observations does not agree with theory, theories must change.

I do not dispute the toy model's equations. The results of a toy model cannot be used to prove the planet will warm 3C if CO2 levels double. The question is the magnitude of the warming.

The toy models without feedback show an increase that is slightly less than 1C for a doubling of CO2.

(1) Stratospheric cooling

The upper atmosphere is mainly warmed from the Sun. It is far from the surface, and above the tropopause which marks the limit of convection. The thermal radiation which is absorbed by greenhouse gases is already absorbed lower down, and the only thermal radiation received in the stratosphere is either at wavelengths where there's no absorption anyway to let heating occur (the transparent bands), or else at wavelengths being emitted the top of the troposphere, which is actually colder than the stratosphere (the saturated bands). Hence greenhouse gases don't help for heating the atmosphere at high altitudes. On the other hand, greenhouse gases still can emit thermal radiation very effectively. So the stratosphere is warmed by solar input absorbed by ozone in particular, and then sheds its heat by thermal radiation from greenhouse gases. The more greenhouse gases, the more efficiently the upper atmosphere is cooled.

There are other mechanisms that can cool the stratosphere, for example changes in planetary cloud cover. Ozone in the troposphere absorbs short wave radiation. Clouds reflect short wave radiation back into space providing a second chance for the ozone molecule to absorb the short wave radiation.

As the planet's surface did warm in the late part of the 20th century, if it did not warm due to the increased in CO2 there must be another explanation. There are paper (disputed) that a significant portion of the 20th century warming was due to changes in planetary cloud cover.

(2) Mid troposphere warming

The troposphere is heated mainly from the Earth's surface. Heat is transported into the troposphere by convection, by latent heat, and by thermal radiation. Any source of increased heat at the surface is going to increase the energy flowing up into the troposphere, whether it be from extra greenhouse gas concentrations, more solar absorption, or more solar input.

Any surface warming is expected to be magnified in the mid-troposphere. This is nothing particularly to do with greenhouse effects; it is mainly a consequence of a changing lapse rate. With additional surface warming, for any reason, the lapse rate (the rate at which temperatures falls with altitude) is expected to drop. The "dry adiabat" shows a much sharper fall in temperatures than the "moist adiabat", and a warmer surface means more moisture and a reduced lapse rate – especially around the tropics.

The dry adiabat is about 9.8C degrees per kilometer. The moist adiabat can be half that. Now imagine that the surface warms, for any reason at all, by 5 degrees. At the same time, due to additional specific humidity, the lapse rate drops from, say, 6 to 5 degrees per kilometer. Then one kilometer above the surface will have warmed by 6 degrees, and two kilometers above the surface by 7 degrees, and so on; up until the atmosphere dries out a bit and the lapse rate comes back to normal. Hence the strongest warming is expected in the middle and high troposphere, in the tropics, but below the tropopause.

(3) Comparison with observations

It's a lot harder to measure temperature trends in the atmosphere than at the surface. There are many sources of error in measurements, and a sure sign of second rate information is when there's no quantification of these uncertainties.

Up until a few years ago, there was a genuine puzzle about tropospheric temperature measurements. The expected rise in temperature was not apparent, and it should have been, even given the measurement difficulties. Although this has often been presented as a problem for greenhouse warming theory, in fact it is nothing to do with greenhouse; but is a problem with the expected physics of temperature and heat transport in general.

More recently, these problems have been somewhat resolved. There are still large uncertainties in the numbers, and there's still lots of room for more careful measurements and testing of our models for lapse rate and atmospheric heat transport. But some sources of error have been clearly identified which obscured the tropospheric warming, which does seem to be there as expected. There's quite a lot of literature on this now; although it remains an active focus of research. Mainly it is a problem of trying to narrow down measurements. The big take home message, however, is that the question is entirely distinct from greenhouse warming. It's about how the atmosphere responds to any warming at all.

A major recent paper on this is:

• Santer, B.D. et. al. (17 authors) (2008) http://www3.interscience.wiley.com/journal/121433727/abstract?CRETRY=1&SRETRY=0, in International Journal of Climatology, Vol 28, Iss 13, pp 1703-1722.

Santer et al have only used data up until 1999. It is asserted in a paper that was been submitted for publication that more the data 1999 to 2008 analyzed using the Santer et al's methodology does not support Santer et al's conclusions.

This remains unfinished business, but as matters stand today there is no major inconsistency between theory and observation.

The stratosphere, on the other hand, has been cooling nicely just as expected. There are multiple causes of this; a reduced ozone concentration and reduced solar absorption has a major role alongside increase greenhouse concentration and enhanced thermal emission. But basically the measurements of the stratosphere give a strong confirmation that it is an enhanced greenhouse effect that is the major cause of global temperature trends over recent decades.

As a minor point, you have the situation with the past backwards. Throughout the past, high CO₂ is found with high temperatures, and low CO₂ with low temperatures. There's a kind of two way effect involved here, which is the more usual source of popular confusion. Rising temperatures can raise CO₂ levels, and rising CO₂ can raise temperatures. They tend to feed on each other, so it's not always clear which one is in the driver's seat! But the relationship is a very clear positive correlation throughout Earth's history.

Saul: I will provide a paper that provides data and analysis that shows the planet has been cold when CO2 levels were high and hot when CO2 levels where low.

Your comment refers to the current glacial/interglacial cycle.

There is currently no accepted explanation as to what causes CO2 levels to increase and decrease in the current glacial/interglacial cycle. Temperature changes can explain at most 6.5 ppm of the 80 ppm to 90 ppm drop in CO2 levels between the interglacial and glacial cycle.

The curious large unexplained changes in atmospheric CO2 is a puzzle which some are now calling a paradox.

Also there is no explanation why CO2 levels decrease and increase on a long term basis.

Cheers -- Sylas

Sylas, do you agree with my comment that increased CO2 does not cause direct warming of the troposphere? i.e. The troposphere is "CO2 saturated". I believe there is direct observations that supports that assertion.

Regards
Saul

 

[Saul]

This is the paper I was referring to that shows that atmospheric CO2 levels have been high when the planet was cool and low when the planet was warm.

Atmospheric carbon dioxide levels for the last 500 million years by Daniel H. Rothman

http://www.pnas.org/content/99/7/4167.full.pdf+html

The last 500 million years of the strontium-isotope record are shown to correlate significantly with the concurrent record of isotopic fractionation between inorganic and organic carbon after the effects of recycled sediment are removed from the strontium signal. The correlation is shown to result from the common dependence of both signals on weathering and magmatic processes. Because the long-term evolution of carbon dioxide levels depends similarly on weathering and magmatism, the relative fluctuations of CO2 levels are inferred from the shared fluctuations of the isotopic records. The resulting CO2 signal exhibits no systematic correspondence with the geologic record of climatic variations at tectonic time scales...

...The most recent cool period corresponds to relatively low CO2 levels, as is widely expected (30). However, no correspondence between pCO2 and climate is evident in the remainder of the record, in part because the apparent 100 My cycle of the pCO2 record does not match the longer climatic cycle. The lack of correlation remains if one calculates the change in average global surface temperature resulting from changes in pCO2 and the solar constant using energy-balance arguments (7, 26). Superficially, this observation would seem to imply that pCO2 does not exert dominant control on Earth’s climate at time scales greater than about 10 My. A wealth of evidence, however, suggests that pCO2 exerts at least some control [see Crowley and Berner (30) for a recent review]. Fig. 4 cannot by itself refute this assumption. Instead, it simply shows that the ‘‘null hypothesis’’ that pCO2 and climate are unrelated cannot be rejected on the basis of this evidence alone.

 

[sylas]

Sylas, do you agree with my comment that increased CO2 does not cause direct warming of the troposphere? i.e. The troposphere is "CO2 saturated". I believe there is direct observations that supports that assertion.

I don't understand what you mean; and I think this is because you are a bit confused. You are using technical terms in ways that don't really make sense.

I explained what "saturated" means for you before. You usage here is distinctly odd.

Adding more CO2 will increase the amount of absorbed radiation. That's a fact. If you mean by "saturated" anything different, then you're just wrong. Check any text on the physics of radiation in the atmosphere. There are many available, and they will all explain about saturation, and absorption, and about how increased absorption in a saturated mixture will occur at the wings. The same physics is used for Sun's photosphere, for the Earth's atmosphere, for a laboratory gas cell.

What you mean by "direct" is also unclear. On the one hand, the greenhouse effect only works because the atmosphere is warmed indirectly, from the surface rather than directly from the Sun. So in that sense the greenhouse effect works by indirect heating of the atmosphere with solar energy that has been first absorbed at Earth's surface.

As for how heat is actually transferred into the atmosphere from the surface, there are three major mechanisms. From least to greatest importance of total magnitude, they are convection, latent heat and radiant transfer. The radiant transfer is by far the largest flux of energy into the atmosphere, and it works by direct absorption of thermal radiation, captured by gases opaque to those wavelengths. In this sense, the greenhouse warming of the lower troposphere is as direct as all get out.

If you mean something else you'll have to spell what you mean by direct and indirect heating. It might help if you could cite some kind of source. It's a requirement of this forum that you use peer reviewed sources; and this restriction has been imposed, I think, mainly because of the climate discussions, where there is one heck of a lot of bad information out there.

Now personally, I don't mind all that much. But it's still worth checking. If you are getting your information from a blog or a newspaper or private web pages, you should be able to see what sources they are using. If you can't, that's a bit of a red flag. But in any case, feel free to indicate what you mean with a citation to these observations you speak of.

Cheers -- Sylas

PS. Aha. Our posts crossed, and I see you have added a reference. Thanks -- that will help. I'll have a look and get back to you.

 

[Saul]

I don't understand what you mean; and I think this is because you are a bit confused. You are using technical terms in ways that don't really make sense.

I explained what "saturated" means for you before. You usage here is distinctly odd.

Adding more CO2 will increase the amount of absorbed radiation. That's a fact. If you mean by "saturated" anything different, then you're just wrong. Check any text on the physics of radiation in the atmosphere. There are many available, and they will all explain about saturation, and absorption, and about how increased absorption in a saturated mixture will occur at the wings. The same physics is used for Sun's photosphere, for the Earth's atmosphere, for a laboratory gas cell.

What you mean by "direct" is also unclear. On the one hand, the greenhouse effect only works because the atmosphere is warmed indirectly, from the surface rather than directly from the Sun. So in that sense the greenhouse effect works by indirect heating of the atmosphere with solar energy that has been first absorbed at Earth's surface.

As for how heat is actually transferred into the atmosphere from the surface, there are three major mechanisms. From least to greatest importance of total magnitude, they are convection, latent heat and radiant transfer. The radiant transfer is by far the largest flux of energy into the atmosphere, and it works by direct absorption of thermal radiation, captured by gases opaque to those wavelengths. In this sense, the greenhouse warming of the lower troposphere is as direct as all get out.

...(Saul. cut)

Saul: The greenhouse molecules can only absorb specific frequencies. At the surface of the planet there is considerable overlap of the absorption frequencies. As this paper states adding CO2 to the lower atmosphere has a minor effect on planetary temperature due the overlapping frequencies of the greenhouse gases.

http://www.warwickhughes.com/papers/barrett_ee05.pdf

Greenhouse molecules, their spectra and function in the atmosphere by Jack Barrett

The absorption values for the pre-industrial atmosphere add up to 86.9%, significantly lower than the combined value of 72.9%. This occurs because there is considerable overlap between the spectral bands of water vapour and those of the other GHGs. If the concentration of CO2 were to be doubled in the absence of the other GHGs the increase in absorption would be 1.5%. In the presence of the other GHGs the same doubling of concentration achieves an increase in absorption of only 0.5%, only one third of its effect if it were the only GHG present. Whether this overlap effect is properly built into models of the atmosphere gives rise to some scepticism.

Cheers -- Sylas

PS. Aha. Our posts crossed, and I see you have added a reference. Thanks -- that will help. I'll have a look and get back to you.

The total theoretical increase in planetary temperature due to a doubling of CO2, is less than 1C. The 3C warming that is often quoted comes about due to positive feedback rather than negative feedback.

As observations do not match AWG (CO2) theory (long term)/General Climatic Models (short term), something must be incorrect with the base assumptions or there must be other processes that have not been modeled correctly.

Regards -- Saul

 

[Xnn]

Saul;

About the Earth's climate over the last 500 Million years.

Rothman's figure 4 highlights periods when Earth's climate was relatively cooler and warmer. That does not mean any were absolutely cooler than the present climate

For example, he list a cool period some where between the Jurassic and Cretaceous periods.

While the Cretaceous was cooler than the Jurassic, it'd be a huge mistake to conclude that either of these periods or their boundry was cooler than our current climate. They were both significantly warmer on an absolute measure.

 

[Saul]

Saul;

About the Earth's climate over the last 500 Million years.

Rothman's figure 4 highlights periods when Earth's climate was relatively cooler and warmer. That does not mean any were absolutely cooler than the present climate

For example, he list a cool period some where between the Jurassic and Cretaceous periods.

While the Cretaceous was cooler than the Jurassic, it'd be a huge mistake to conclude that either of these periods or their boundry was cooler than our current climate. They were both significantly warmer on an absolute measure.

If you look at figure 4 there are 4 ice epochs on the planet. We are living during the fourth. For the other ice epochs CO2 was above 2000 ppm which seems to indicate that the effect of CO2 on the planet's temperature, saturates at some level.

http://www.pnas.org/content/99/7/4167.full.pdf+html

I have another paper that notes there is a lack of correlation of planetary temperature and CO2 levels during this ice epoch.

Fig. 4. Fluctuations of pCO2 for the last 500 My, normalized by the
estimate of pCO2 obtained from the most recent value of . The solid line is obtained from Eq. 12 by using 0
36‰. The lower and upper limits of the gray area surrounding the pCO2 curve result from 0
38 and 35‰, respectively. The gray bars at the top correspond to periods when Earth’s climate was relatively cool; the white spaces between them correspond to warm modes (18).

 

[sylas]

OK; I've had a look. Saul cites:

Rothman, D.H. (2002) http://www.pnas.org/content/99/7/4167.abstract, in PNAS, April 2, 2002 Vol. 99 No. 7, pp 4167-4171

There are several important points. First, and by far the most important:

(A) This is a complete non-sequitur

The existence or otherwise of other factors that may be driving climate cycles apparent over long times spans, of 10 million years or more, is completely irrelevant to deciding whether or not CO₂ has a significant impact over the scale of recent decades.

The scale difference here is enormous. What is short term for Rothman is anything less than 10 million years. That is enough to encompass the entire Quaternary period several times over, with all its many the cycles of ices ages and interglacials!

This cannot be emphasized sufficiently strongly. There are many different factors impacting in climate, and in principle there is nothing implausible at all about long term climate change, on the scale of 10 million years or more, being driven primarily by factors other than CO₂. This is in no conflict whatsoever with the immediate physical impact of changing CO₂ levels, which is about solid as anything ever gets in science. It is truly elementary thermodynamics.

If people get nothing else from the thread, they should at least get this. A long term climate trend driven by something other than CO₂ is no conflict at all with CO₂ being a factor to drive climate above and below this other putative larger and slower trend.

It's as if you were using the theory of continental drift, which moves whole ecosystems vast distances over the scale of millions of years, to refute the idea of biological factors being important for the short term shifts in the boundary of a rainforest.

(B) Rothman himself explicitly denies the application being attempted in this thread

This should kill the argument dead. Looking at the extract from Rothman's paper, which Saul provides, Saul appears to have missed the implication as far as this thread goes. Rotham states:

Superficially, this observation would seem to imply that pCO₂ does not exert dominant control on Earth's climate at time scales greater than about 10 My. A wealth of evidence, however, suggests that pCO₂ exerts at least some control [see Crowley and Berner (30) for a recent review]. Fig. 4 cannot by itself refute this assumption. Instead, it simply shows that the “null hypothesis” that pCO₂ and climate are unrelated cannot be rejected on the basis of this evidence alone.​

The last sentence there is a triple negative, which can be confusing! But the rest of the paragraph is better. Rothman suggests, on the basis of his reconstruction, that CO₂ is not dominant on long time scales. As I have noted, this is completely consistent with the obvious immediate physical impact of CO₂ on shorter time spans.

Rothman is explicit that there is a wealth of evidence for CO₂ having at least some impact, and that his diagram does NOT refute this.

The clumsy final sentence in Rothman's paragraph claims that the other evidence – which he does not actually discuss! – can't reject the null hypothsis of no impact. That is an odd remark. Of course the evidence goes powerfully against the null hypothesis. The implication of Rothman's reconstruction as shown in his figure 4 is rather that the relationship with CO₂ may be over ridden on longer time spans by other unidentified factors. In principle there is nothing wrong with that notion, and it is in no conflict at all with the plainly demonstrated physical impact of CO2 in recent decades.

Rothman says this explicitly when he notes that his reconstruction "cannot by itself refute this assumption". Quite so. So let's stop pretending that it does.

But Rothman's work on the longer term scales ALSO has problems of its own. This is the next point:

(C) Rothman's reconstruction is at variance with related work before and since.

A summary of different attempts by a number of authors to estimate CO₂ levels over this period is available in Bergman, et al (2004).

• Bergman, N.H. et al (2004) COPSE: A new model of biogeochemical cycling over Phanerozoic time, in American Journal of Science, Vol. 304, May 2004, P.397-437; doi:10.2475/ajs.304.5.397

I've taken the figure 4 from that paper, which compares a range of published work using proxies or other others, and superimposed upon it the figure 4 from Rothman (2002) in green, rescaled so that the diagrams align. I have also included the icehouse periods identified in Rothman's diagram. Rothman's work diverges sharply from all other reconstructions past about 300 Mya BP.

This is a much less important point than recognizing that putative unidentified long term climate drivers is completely consistent with the physical importance of CO₂ on short time spans. But I think it is worth noting for the record that Rothman's reconstruction has problems of its own with other lines of evidence. It's all open research, of course; and none of it is any problem whatever for the simple physics of CO₂ and radiative transfers.

(D)Saul actually confirms the approximate magnitude of CO₂ forcing

Above, Saul speaks of the theoretical CO₂ impact:

The total theoretical increase in planetary temperature due to a doubling of CO2, is less than 1C. The 3C warming that is often quoted comes about due to positive feedback rather than negative feedback.

Of course. This is precisely what my original post explains. What you here call "theoretical" is the "Planck response" that I describe at point #5 of my first post. Repeating what I said originally: If nothing changes other than temperature, the response is straightforward. […] It's called the "Planck response" in climate science. Converting forcing units, this corresponds to about 1.1C per 2xCO₂.

There's a minor flaw in Saul's numbers if he gets less than 1C, but more importantly, this is not a "theoretical" impact for Earth at all. It's the impact for some other place where nothing changes as temperatures alter. The real theoretical work for Earth suggests an impact of about 3C per doubling; but I didn't cite theory for that. I went straight to observational constraints.

That climate feedbacks are positive is a real observation. The evidence I cited in my original post is empirical, and there's a lot more than my chosen example. Observations of the real world indicate that climate response is somewhere from 1.5C to 4.5C.

More amusingly still – by acknowledging the no-feedback "Plank response", Saul is actually using the very 3.7 W/m² that he started out by denying, or something very close to it! He acknowledges that there really IS a theoretical basis for non-feedback response of about 1C. Great. That means a forcing of about 3.3 W/m²; which is how much extra emission you get from Earth with a 1C rise as the base Planck response. Snap.

(E) A few loose ends

On tropospheric warming:

Santer et al have only used data up until 1999. It is asserted in a paper that was been submitted for publication that more the data 1999 to 2008 analyzed using the Santer et al's methodology does not support Santer et al's conclusions.

This bland assertion comes with no reference. Santer et al is pointing out errors in an incorrect analysis of older data. The truth of the matter is that data from 1999 to 2008 all shows that the troposphere is warming more rapidly than the surface. The real disputes are over how much; but the claims that troposphere amplification is absent are flatly false.

The major sources of information for recent years are satellite data from Remote Sensing Systems (RSS) and University of Alabama, Huntsville (UAH). This is an interesting case, because for a long time the whole issue was confused by UAH claims for a lack of warming. This was shown in 2005 to be an algebraic error in their analysis. It was as basic as doing a subtraction when they should have added. The UAH team acknowledged the problem promptly, and now everyone agrees that measurements show tropospheric amplification. The remaining differences are over how much; but frankly the UAH team has something of a credibility problem now.

• Reference: Mears, A.C., and Wentz, F.J. (2005) The Effect of Diurnal Correction on Satellite-Derived Lower Tropospheric Temperature, in Science, 2 Sept 2005, Vol. 309. no. 5740, pp. 1548 – 1551. DOI: 10.1126/science.1114772

Bottom line. Expectations are that the troposphere should warm more than the surface. This is what is measured. There remain large uncertainties in the measurement, so that it is hard to give a strong constraint on theory, but as matters stand observations are consistent with theory. The theory involved is entirely independent of greenhouse effects, and deals with lapse rates and any source of temperature change.

As observations do not match AWG (CO2) theory (long term)/General Climatic Models (short term), something must be incorrect with the base assumptions or there must be other processes that have not been modeled correctly.

Since your premise is false, the rest doesn't follow. Observations do most definitely match the physically expected impact of CO₂, especially on the time scale being addressed in this thread.

There are of course other processes that are not yet being modeled well, and which apply over time spans of hundreds of millions of years. We even know some of them, and the modeling is very difficult. (Continental drift; changes in land cover, both with new species and different biogeographic distributions.) What's the albedo of a Cambrian swamp, for example? Be that as it may, pretty much all work in very long term climate trends recognizes the impact of greenhouse gases along with whatever else is going on.

There's no basis in any of this for disputing the elementary physics that means CO₂ is necessarily significant for trends over recent decades.

I have another paper that notes there is a lack of correlation of planetary temperature and CO2 levels during this ice epoch. […]

No, you don't. It's the exact same paper: Rothman (2002). This is a rather revealing screw up. You even quoted the text from the paper, which confirms its source, but if you are speaking of it as "another paper", it suggests aren't really working direct from papers you have to hand at all, but that you are picking up extracts found in some secondary source.

Secondary sources can often be useful – but take care, because in this particular subject area many of them are utterly atrocious. For really basic stuff like saturation and absorption you are best to simply go with established textbooks. The forcing from carbon dioxide is not some open research question, but elementary undergraduate physics.

Cheers -- Sylas

 

[Saul]

OK; I've had a look. Saul cites:

Rothman, D.H. (2002) http://www.pnas.org/content/99/7/4167.abstract, in PNAS, April 2, 2002 Vol. 99 No. 7, pp 4167-4171

There are several important points. First, and by far the most important:

(A) This is a complete non-sequitur

The existence or otherwise of other factors that may be driving climate cycles apparent over long times spans, of 10 million years or more, is completely irrelevant to deciding whether or not CO₂ has a significant impact over the scale of recent decades.

The scale difference here is enormous. What is short term for Rothman is anything less than 10 million years. That is enough to encompass the entire Quaternary period several times over, with all its many the cycles of ices ages and interglacials!

Saul: We are living in the an Ice Epoch. There have been four ice epochs. There are currently permanent ice caps on both poles. By convention this period is called the "interglacial" period to differentiate it from the glacial period when planetary temperature drops further.

What causes the glacial/interglacial cycle where during the "glacial period" the ice sheets increase covering Canada, the Northern US, and Northern Europe is not known. The driver is not CO2. CO2 lags the temperature changes in the glacial/interglacial cycle by around 800 years. Curiously, it appears the Southern Hemisphere also cools at the same time as the Northern Hemisphere cools.

This cannot be emphasized sufficiently strongly. There are many different factors impacting in climate, and in principle there is nothing implausible at all about long term climate change, on the scale of 10 million years or more, being driven primarily by factors other than CO₂. This is in no conflict whatsoever with the immediate physical impact of changing CO₂ levels, which is about solid as anything ever gets in science. It is truly elementary thermodynamics.

If people get nothing else from the thread, they should at least get this. A long term climate trend driven by something other than CO₂ is no conflict at all with CO₂ being a factor to drive climate above and below this other putative larger and slower trend.

It's as if you were using the theory of continental drift, which moves whole ecosystems vast distances over the scale of millions of years, to refute the idea of biological factors being important for the short term shifts in the boundary of a rainforest.

Sylas, Rothman is showing evidence of Ice Epochs where there is massive ice sheets. During the other periods there are no massive ice sheets. Planetary CO2 levels do not correlate with the Ice Epochs.

(B) Rothman himself explicitly denies the application being attempted in this thread

Saul: Rothman states what his data shows. There have been four ice Epochs on this planet. We are living in the fourth. During the three previous Ice Epochs planetary CO2 has high. There is no correlation of planetary temperature and CO2.

Rothman's finding is also supported by Hay et al. (See link at the end of this comment.)

The correlation is shown to result from the common dependence of both signals on weathering and magmatic processes. Because the long-term evolution of carbon dioxide levels depends similarly on weathering and magmatism, the relative fluctuations of CO2 levels are inferred from the shared fluctuations of the isotopic records. The resulting CO2 signal exhibits no systematic correspondence with the geologic record of climatic variations at tectonic time scales.

This should kill the argument dead. Looking at the extract from Rothman's paper, which Saul provides, Saul appears to have missed the implication as far as this thread goes. Rotham states:

Superficially, this observation would seem to imply that pCO₂ does not exert dominant control on Earth's climate at time scales greater than about 10 My. A wealth of evidence, however, suggests that pCO₂ exerts at least some control [see Crowley and Berner (30) for a recent review]. Fig. 4 cannot by itself refute this assumption. Instead, it simply shows that the “null hypothesis” that pCO₂ and climate are unrelated cannot be rejected on the basis of this evidence alone.​

The last sentence there is a triple negative, which can be confusing! But the rest of the paragraph is better. Rothman suggests, on the basis of his reconstruction, that CO₂ is not dominant on long time scales. As I have noted, this is completely consistent with the obvious immediate physical impact of CO₂ on shorter time spans.

Rothman is explicit that there is a wealth of evidence for CO₂ having at least some impact, and that his diagram does NOT refute this.

Saul: No one is contesting that CO2 has some influence on climate. The paleoclimatic data indicates however: 1) There is a massive unknown mechanism that is cooling the planet independent of CO2 levels and 2) the planet's response to an increase in CO2 appears to a saturate such that increased CO2 does not cause a significant increase in planetary temperature.

The clumsy final sentence in Rothman's paragraph claims that the other evidence – which he does not actually discuss! – can't reject the null hypothsis of no impact. That is an odd remark. Of course the evidence goes powerfully against the null hypothesis. The implication of Rothman's reconstruction as shown in his figure 4 is rather that the relationship with CO₂ may be over ridden on longer time spans by other unidentified factors. In principle there is nothing wrong with that notion, and it is in no conflict at all with the plainly demonstrated physical impact of CO2 in recent decades.

Rothman says this explicitly when he notes that his reconstruction "cannot by itself refute this assumption". Quite so. So let's stop pretending that it does.

But Rothman's work on the longer term scales ALSO has problems of its own. This is the next point:

(C) Rothman's reconstruction is at variance with related work before and since.

A summary of different attempts by a number of authors to estimate CO₂ levels over this period is available in Bergman, et al (2004).

• Bergman, N.H. et al (2004) COPSE: A new model of biogeochemical cycling over Phanerozoic time, in American Journal of Science, Vol. 304, May 2004, P.397-437; doi:10.2475/ajs.304.5.397

I've taken the figure 4 from that paper, which compares a range of published work using proxies or other others, and superimposed upon it the figure 4 from Rothman (2002) in green, rescaled so that the diagrams align. I have also included the icehouse periods identified in Rothman's diagram. Rothman's work diverges sharply from all other reconstructions past about 300 Mya BP.

This is a much less important point than recognizing that putative unidentified long term climate drivers is completely consistent with the physical importance of CO₂ on short time spans. But I think it is worth noting for the record that Rothman's reconstruction has problems of its own with other lines of evidence. It's all open research, of course; and none of it is any problem whatever for the simple physics of CO₂ and radiative transfers.

There are different proxies to determine CO2 levels. Rothman's conclusion that CO2 levels do not correlate with planetary temperature is not contested. (See the paper linked to a the end of the comment.

Bergman's data shows CO2 levels in excess of 1500 ppm during the ice epochs and in other cases the CO2 levels drop but there is no ice epoch. Bergman's data also shows the Ice Epochs do not correlate with CO2 levels.

(Saul - Cut. I will respond to the other scientific issue in with separate comments.

Cheers -- Sylas

Saul:
It should be noted the lack of correlation of CO2 level and planetary climate is supported by other data. The problem with the CO2 driver hypothesis is the there is no mechanism to reduce CO2 and detailed analysis indicates CO2 levels were high when the planet was cooling, in the Cenozoic.


The Late Cenozoic uplift – climate change paradox by William Hay , Emanuel Soeding, Robert DeConto, and Christopher N. Wold


http://www.geo.umass.edu/faculty/deconto/hayetal.pdf

Because the GEOCARB model (Berner 1994; Berner and Kothavala 2001) does not have adequate temporal resolution to predict the structure of the Cenozoic decline in atmospheric CO2 concentrations, many geologists tacitly assumed that atmospheric CO2 decreased in parallel with the δ18O curve for deep-sea benthic Foraminifera. Although isotopic data from Mesozoic pedogenic carbonates suggest much higher levels of atmospheric CO2, the younger record is ambiguous (Ekart et al. 1999). Cenozoic paleosols suggest both higher (900–1,000 ppmv) and lower (270–210 ppmv) atmospheric CO2 concentrations. This view of a general decline in atmospheric CO2 throughout the Cenozoic has been challenged by recent studies. Pagani et al. (1999a, 1999b) estimated Miocene atmospheric CO2 concentrations from gp (magnitude of the carbon isotope discrimination
during photosynthesis) values based on δ13C in diunsaturated alkenones and the shells of shallow-dwelling planktonic Foraminifera from DSDP and ODP sites in the Atlantic, Indian and Pacific oceans. They concluded that atmospheric pCO2 levels were below 280 ppmv during most of the Miocene. They also found no feature comparable to the sharp Middle Miocene increase in δ18O interpreted as a major cooling step in the Antarctic. Similar results have been reported for the earlier Cenozoic by Pearson and Palmer (1999, 2000a, 2000b), based on interpretations of atmospheric CO2 concentrations from estimates of oceanic pH using δ11B of foraminiferal calcite...

On the basis of leaf stomatal indices in Ginko and Metasequoia, Royer et al. (2001) have concluded that atmospheric CO2 levels were between 300 and 450 ppmv during the Paleocene, Eocene and Middle Miocene, except for a brief excursion near the Paleocene– Eocene boundary. Veizer et al. (2000) found no direct relationship between the Phanerozoic δ18O record and the occurrence of glacial episodes documented by geological data, suggesting that the two phenomena are not coupled. Kump (2000) noted that this calls into question the currently accepted relationship between atmospheric CO2 levels and climate.

Regards -- Saul

 

[Saul]

OK; I've had a look.

Saul - Cut Sylas' comments a, b, c, and d. Left item E On tropospheric warming.

(E) A few loose ends

On tropospheric warming:

This bland assertion comes with no reference. Santer et al is pointing out errors in an incorrect analysis of older data. The truth of the matter is that data from 1999 to 2008 all shows that the troposphere is warming more rapidly than the surface. The real disputes are over how much; but the claims that troposphere amplification is absent are flatly false.

The major sources of information for recent years are satellite data from Remote Sensing Systems (RSS) and University of Alabama, Huntsville (UAH). This is an interesting case, because for a long time the whole issue was confused by UAH claims for a lack of warming. This was shown in 2005 to be an algebraic error in their analysis. It was as basic as doing a subtraction when they should have added. The UAH team acknowledged the problem promptly, and now everyone agrees that measurements show tropospheric amplification. The remaining differences are over how much; but frankly the UAH team has something of a credibility problem now.

• Reference: Mears, A.C., and Wentz, F.J. (2005) The Effect of Diurnal Correction on Satellite-Derived Lower Tropospheric Temperature, in Science, 2 Sept 2005, Vol. 309. no. 5740, pp. 1548 – 1551. DOI: 10.1126/science.1114772

Bottom line. Expectations are that the troposphere should warm more than the surface. This is what is measured. There remain large uncertainties in the measurement, so that it is hard to give a strong constraint on theory, but as matters stand observations are consistent with theory. The theory involved is entirely independent of greenhouse effects, and deals with lapse rates and any source of temperature change.


Since your premise is false, the rest doesn't follow. Observations do most definitely match the physically expected impact of CO₂, especially on the time scale being addressed in this thread.

There are of course other processes that are not yet being modeled well, and which apply over time spans of hundreds of millions of years. We even know some of them, and the modeling is very difficult. (Continental drift; changes in land cover, both with new species and different biogeographic distributions.) What's the albedo of a Cambrian swamp, for example? Be that as it may, pretty much all work in very long term climate trends recognizes the impact of greenhouse gases along with whatever else is going on.

There's no basis in any of this for disputing the elementary physics that means CO₂ is necessarily significant for trends over recent decades.

No, you don't. It's the exact same paper: Rothman (2002). This is a rather revealing screw up. You even quoted the text from the paper, which confirms its source, but if you are speaking of it as "another paper", it suggests aren't really working direct from papers you have to hand at all, but that you are picking up extracts found in some secondary source.

Secondary sources can often be useful – but take care, because in this particular subject area many of them are utterly atrocious. For really basic stuff like saturation and absorption you are best to simply go with established textbooks. The forcing from carbon dioxide is not some open research question, but elementary undergraduate physics.

Cheers -- Sylas

The troposphere does based on Radiosonde measurements (weather balloons) show cooling rather than warming. This paper asserts that weather balloon measurements in the 1980's had a warm basis. Even with that warm bias removed, however, there does appear to cooling of the upper troposphere/lower stratosphere.

Cooling of the upper troposphere/lower stratosphere would as I noted be consistent with the twentieth century warming being due to reduction in cloud cover rather than the GWG (CO2) warming of the stratosphere. There are published papers that show there is reduction in planetary cloud cover during this period.


http://homepage.univie.ac.at/leopold.haimberger/i1520-0442-21-18-4587.pdf

Toward Elimination of the Warm Bias in Historic Radiosonde Temperature Records—Some New Results from a Comprehensive Intercomparison of Upper-Air Data

The apparent cooling trend in observed global mean temperature series from radiosonde records relative to Microwave Sounding Unit (MSU) radiances has been a long-standing problem in upper-air climatology. It is very likely caused by a warm bias of radiosonde temperatures in the 1980s, which has been reduced over time with better instrumentation and correction software. The warm bias in the MSU-equivalent lower stratospheric (LS) layer is estimated as 0.6 +/- 3 K in the global mean and as 1.0 +/- 0.3 K in the tropical (20°S–20°N) mean. These estimates are based on comparisons of unadjusted radiosonde data, not only with MSU data but also with background forecast (BG) temperature time series from the 40-yr European Centre for Medium-Range Weather Forecasts (ECMWF) Re-Analysis (ERA-40) and with two new homogenized radiosonde datasets…

Refer to the figure 7 in the paper even with the removal of the warm bias the upper troposphere/lower stratosphere cools rather than warms.

FIG. 7. Same as Fig. 6, but for the tropics (20°S–20°N). The figure has been derived from 42 stations in the tropics with 26 out of 28 yr of data since 1979. Difference of dark blue curve in the 1980s minus the 2000s yields RAOBCORE bias estimates (approx 1 K in the LS; approx 0.6 K in the TS). RAOBCORE trends for the LS and TS layers are 0.34 K (decade)1 and 0.02 K (decade)1, respectively.

This is the paper I said that has been submitted that disputes Santer et al's conclusion that troposphere is warming. The paper uses the same data source as Santer et al and Santer et al's analysis methodology to determine the upper troposphere is not warming.

http://www.climateaudit.org/?p=4991

A debate exists over whether tropical troposphere temperature trends in climate models are inconsistent with observations (Karl et al. 2006, IPCC (2007), Douglass et al 2007, Santer et al 2008). Most recently, Santer et al (2008, herein S08) asserted that the Douglass et al statistical methodology was flawed and that a correct methodology showed there is no statistically significant difference between the model ensemble mean trend and either RSS or UAH satellite observations. However this result was based on data ending in 1999. Using data up to the end of 2007 (as available to S08) or to the end of 2008 and applying exactly the same methodology as S08 results in a statistically significant difference between the ensemble mean trend and UAH observations and approaching statistical significance for the RSS T2 data. The claim by S08 to have achieved a “partial resolution” of the discrepancy between observations and the model ensemble mean trend is unwarranted.

There was also a paper that asserted troposphere wind speed indicates that the troposphere was warming which if correct would imply that the past weather balloon data was not correct.

This new paper that has been submitted shows, however, that recent stratospheric wind data indicates that the troposphere temperatures are basically unchanged.

http://www.climatesci.org/publications/pdf/R-342.pdf

Recent work has concluded that there has been significant warming in the tropical upper troposphere using the thermal wind equation to diagnose temperature trends from observed winds; a result which diverges from all other observational data. In our paper we examine evidence for this conclusion from a variety of directions and find that evidence for a significant tropical tropospheric warming is weak. In support of this conclusion we provide evidence that, for the period 1979-2007, except for the highest latitudes in the Northern Hemisphere, both the thermal wind, as estimated by the zonal averaged 200 hPa wind and the tropospheric layer-averaged temperature, are consistent with each other, and show no statistically significant trends.

Sylas,
Because there is evidence that shows that there are periods of millions of years when there is not correlation of planetary temperature and planetary CO2 levels (i.e. The planet is cooling when CO2 levels are high and the planet is warming when CO2 levels are low) I believe, the current evidence that shows the stratosphere is not warming deserves consideration. (i.e. One observation and analysis supports the other.)

There are recent papers that have been published that try to explain why (Assuming the stratosphere did not warm and increases in CO2 do not warm the planet.) there is this discrepancy.

 

[sylas]

An aside… When you quote extracts from my post, you should not add your own comments nested inside the quote tags. If you have a comment on my text, then give a quoted extract, and nothing else, inside the quote tags. Then add your own comment afterwards, outside the tags. Repeat as necessary. It's very easy to miss stuff improperly added within what is quoted; it just looks like part of what you are quoting. It's important to get this right, out of fairness to the person you are quoting and also for proper emphasis and clarity of your contributions.

You also don't need to quote every part of my posts in your reply. If you do the quoting correctly, you also don't need to add your own name to indicate what you are writing.

It should be noted the lack of correlation of CO₂ level and planetary climate is supported by other data. The problem with the CO₂ driver hypothesis is the there is no mechanism to reduce CO₂ and detailed analysis indicates CO₂ levels were high when the planet was cooling, in the Cenozoic.

(For other readers, as useful background: the Cenozoic is the last 65 million years up to the present. The Quaternary is the tail end of the Cenozoic, or the last 1.8 million years.)

You have started another thread: [thread=308288]CO2 Variance in the Atmosphere[/thread], in which you talk about mechanisms for CO₂ variation. It's an interesting topic, and a useful initial post. That's the right place to talk about mechanisms. Not here.

For the record, however, it is flatly false to say there are "no mechanisms" for reducing atmospheric CO₂. Your own sources discuss some known mechanisms.

Over the only periods when we have direct data on atmospheric CO₂ levels, using ice core data, there is an unambiguous record of cycles of rising and falling CO₂ levels aligned with the ice ages of the Quaternary period. The mechanisms involved are not all understood, but the correlation with temperature is positive and unambiguous. If you calculate forcings using numbers from my first post, it's fairly easy to see that CO₂ change is not sufficient to explain ice age temperatures. There are other factors involved here, and the trigger is probably orbital changes. The observed fall in CO₂ is important part of quantifying causes for falling temperature; but it's only a part, and it's not the trigger.

That's going to remain true even if the correlation doesn't hold over longer time spans. There's nothing whatsoever wrong in principle with the idea that there are other factors working over long time spans that drive major swings in climate, so that the correlation is lost or even reversed on longer scales.

This is all irrelevant to the thread. There's no claim being made that CO₂ is the only factor to impact climate, and I have no problem in principle with other factors than CO₂ being a major factor in the Cenozoic. We already know for sure that other factors are important in the Quaternary – and that is a period where there is definitely a secondary greenhouse impact involved, of a magnitude that is calculated using the kinds of physics set out in my very first post!

You also quote a paper by Hay et al (2002).

• Hay, W. et al. (2002) The Late Cenozoic uplift - climate change paradox, in International Journal of Earth Sciences, Volume 91, Issue 5, pp. 746-774

There's nothing at all in that paper to deny or contradict the basic thermodynamics of the greenhouse effect, including the role of carbon dioxide, for driving temperature change. Indeed, it seems to be taken for granted – as it should! The real argument is that some other factor must have been involved in the late Cenozoic, because the CO₂ levels are not changing enough or in the right direction to be the major cause at that time.

I've got no comment on the merits of the paper. It presents what it identifies as "an outrageous hypothesis", so I guess it is a bit of an exercise of thinking outside the box, and that's a good thing for science. You quote Hay et al as follows:

On the basis of leaf stomatal indices in Ginko and Metasequoia, Royer et al. (2001) have concluded that atmospheric CO2 levels were between 300 and 450 ppmv during the Paleocene, Eocene and Middle Miocene, except for a brief excursion near the Paleocene– Eocene boundary. Veizer et al. (2000) found no direct relationship between the Phanerozoic δ18O record and the occurrence of glacial episodes documented by geological data, suggesting that the two phenomena are not coupled. Kump (2000) noted that this calls into question the currently accepted relationship between atmospheric CO2 levels and climate.

This is specifically saying that CO₂ levels can't do the job because they were too low. If he's right, this does indeed call into the question currently accepted trends on CO₂ levels at this time; and that would be surprising. But in principle, it's a perfectly credible notion to be considered on its own merits. It means, in context, that there's something else that is more important for climate over this period.

For some solid good sense on this whole issue, go one step further and read Kump (2000) directly. I've put it aside for reference in case you decide to take up the issue of climate and deep time in a more appropriate thread.

None of this, by any stretch of the imagination, is in the slightest conflict with the simple physics of how carbon dioxide impacts temperature! You are quoting papers proposing other factors at work in deep time. It's an irrelevant distraction for this thread – and it is also not fair to the scientists being quoted, who are not denying at all the physics of greenhouse and carbon dioxide.

The topic for this thread: estimating CO₂ impact NOW

My original post on this was crystal clear. I am talking about RIGHT NOW, where any notions of different factors driving climate can be tested with data right under our noses.

I'm not opposed to illuminating this question with data from prehistory, as long as there is some credible basis for linking observations to the specific thread topic. Just saying that CO2 is not always the major factor is a distraction. Take it to another thread if you are interested in deep time. It could be a great topic in its own right.

Your next post does look at the present, with atmospheric trends; I'll respond to that next.

Cheers -- Sylas

 

[Xnn]

If you look at figure 4 there are 4 ice epochs on the planet. We are living during the fourth. For the other ice epochs CO2 was above 2000 ppm which seems to indicate that the effect of CO2 on the planet's temperature, saturates at some level.

http://www.pnas.org/content/99/7/4167.full.pdf+html

Actually, the word "ice" does not appear anywhere in Rothman's paper.

This is not to suggest that no glaciers existed on Earth between 600 to 25 Ma. It is generally understood that 500ppm CO2 will not exert enough of a warming to melt all glaciers at higher elevations in the polar regions. This is one reason why East Antarctica is not expected to melt; it is simply at too high of an elevation. On the other hand, CO2 levels impact the elevation/latitude necessary to support ice sheets and this is why there has been such a retreat in the amount of ice at both poles.


Sylas;

Thanks for pointing out the inconsistencies in Rothmans reconstruction.

 

[Skyhunter]

The troposphere does based on Radiosonde measurements (weather balloons) show cooling rather than warming. This paper asserts that weather balloon measurements in the 1980's had a warm basis. Even with that warm bias removed, however, there does appear to cooling of the upper troposphere/lower stratosphere.

Cooling of the upper troposphere/lower stratosphere would as I noted be consistent with the twentieth century warming being due to reduction in cloud cover rather than the GWG (CO2) warming of the stratosphere. There are published papers that show there is reduction in planetary cloud cover during this period.


http://homepage.univie.ac.at/leopold.haimberger/i1520-0442-21-18-4587.pdf

Toward Elimination of the Warm Bias in Historic Radiosonde Temperature Records—Some New Results from a Comprehensive Intercomparison of Upper-Air Data
Refer to the figure 7 in the paper even with the removal of the warm bias the upper troposphere/lower stratosphere cools rather than warms.

You are not understanding what you are reading here.

The older radiosondes were not well shielded and were warmed by direct radiative flux from the Sun. This resulted in a warm bias in the radiosonde data especially during the 1980s. What Hamburg et al have done in their reanalysis is correct for the bias.

From the abstract:

Both of the new adjusted radiosonde time series are in better agreement with satellite data than comparable published radiosonde datasets, not only for zonal means but also at most single stations. A robust warming maximum of 0.2–0.3K (10 yr)^-1 for the 1979–2006 period in the tropical upper troposphere could be found in both homogenized radiosonde datasets.

By correcting for the warm bias in the 80's the radiosonde data is more in agreement with the satellites, models, and theory.

This is the paper I said that has been submitted that disputes Santer et al's conclusion that troposphere is warming. The paper uses the same data source as Santer et al and Santer et al's analysis methodology to determine the upper troposphere is not warming.

http://www.climateaudit.org/?p=4991
There was also a paper that asserted troposphere wind speed indicates that the troposphere was warming which if correct would imply that the past weather balloon data was not correct.

This new paper that has been submitted shows, however, that recent stratospheric wind data indicates that the troposphere temperatures are basically unchanged.

http://www.climatesci.org/publications/pdf/R-342.pdf

This paper has not been published in a science journal. It is against the forum rules to discuss it.

There are recent papers that have been published that try to explain why (Assuming the stratosphere did not warm and increases in CO2 do not warm the planet.) there is this discrepancy.

The stratosphere is expected to cool as CO2 increases.

The atmosphere cools from the surface outward until it meets the inversion layer known as the tropopause. Then it warms as you move outward. The reason is that below the tropopause the atmosphere is primarily heated by the Earth and above it is heated by the Sun. There is little mixing between the layers so very little convective heat transfer. Most of the heat loss from the stratosphere is in the form of electromagnetic radiation.

Now enter the world of quantum mechanics. CO₂ is a linear molecule with a zero dipole. O-C-O Each atom has a magnetic field and the center of the fields are equidistant from one another. When excited the molecule vibrates and has 3 primary quantum vibrational modes. http://chemmac1.usc.edu/bruno/java/Vibrate.html" is an animation of the quantum vibrational modes of the CO₂ molecule. The bending mode is the one that is significant for stratospheric cooling.

As the molecule vibrates the magnetic fields of the molecule are in motion and interacting with one another. This creates a rapidly shifting dipole moment causing the molecule to emit a photon in a random direction. When this happens in the stratosphere there is little probability that the photon will be reabsorbed in the stratosphere. Stratospheric cooling is the expected result of increased concentrations of carbon dioxide.

 

[Gokul43201]

It should be noted the lack of correlation of CO2 level and planetary climate is supported by other data. The problem with the CO2 driver hypothesis is the there is no mechanism to reduce CO2 and detailed analysis indicates CO2 levels were high when the planet was cooling, in the Cenozoic.

Even with your repeatedly raising this point, you fail to even once acknowledge the fundamental objection raised by Sylas, that you are implicitly assuming an infinite bandwidth for all feedbacks. The timescales of various mass and energy flow processes gives good reason to expect the gain from any feedback mechanism to be frequency dependent, to the extent that it seems reckless (at least to me) to assume/assert that the gain over timescales > 10MY should hold at timescales < 0.1MY

Now enter the world of quantum mechanics. CO₂ is a linear molecule with a zero dipole. O-C-O Each atom has a magnetic field and the center of the fields are equidistant from one another. When excited the molecule vibrates and has 3 primary quantum vibrational modes. http://chemmac1.usc.edu/bruno/java/Vibrate.html" is an animation of the quantum vibrational modes of the CO₂ molecule. The bending mode is the one that is significant for stratospheric cooling.

As the molecule vibrates the magnetic fields of the molecule are in motion and interacting with one another. This creates a rapidly shifting dipole moment causing the molecule to emit a photon in a random direction.

The dipole moment of a molecule (in the context of scattering, as is the case here) refers to the electric dipole moment and has nothing to do with magnetic fields (nor is there even such a thing in general, as the center of a magnetic field). You should probably have left out the QM part of this discussion as it is partly erroneous, serves little or no explanatory purpose (at least to me), and if at all, only raises more questions.

 

[Skyhunter]

The dipole moment of a molecule (in the context of scattering, as is the case here) refers to the electric dipole moment and has nothing to do with magnetic fields (nor is there even such a thing in general, as the center of a magnetic field). You should probably have left out the QM part of this discussion as it is partly erroneous, serves little or no explanatory purpose (at least to me), and if at all, only raises more questions.

I am not a physicist, and was relying on my inadequate memory and understanding of something I read a few years ago. Although my explanation and terminology may be lacking, CO2's quantum vibrations are the primary reason for stratospheric cooling since the bending vibration causes the molecule to radiate.

 

[Gokul43201]

Actually, all three vibrational modes make the molecule radiate. The bending and asymmetric stretch are IR active modes (they involve a change in the molecule's dipole moment) while the symmetric stretch is a Raman active mode (it involves a change in the polarizability of the molecule). I think (this is, at best a semi-educated guess) what's special about the bending mode is that it falls plumb in the middle of the black body spectrum for temperatures around 270K.

 

[Skyhunter]

Actually, all three vibrational modes make the molecule radiate. The bending and asymmetric stretch are IR active modes (they involve a change in the molecule's dipole moment) while the symmetric stretch is a Raman active mode (it involves a change in the polarizability of the molecule). I think (this is, at best a semi-educated guess) what's special about the bending mode is that it falls plumb in the middle of the black body spectrum for temperatures around 270K.

The symmetric stretch does not radiate and the asymmetric stretch is at to high a frequency to be excited by atmospheric temperature collisions.

 

[Gokul43201]

The symmetric stratch does not radiate...

Not true. This is a Raman active mode, so the radiation need not be in the IR band (look up Stokes and anti-Stokes lines in the Raman spectrum of a molecule).

But anyway, this is an unnecessary digression from the point of this thread, so I shall not pursue it any further. Besides, I have a long post to make in another thread...and too little time.

 

[Saul]

The stratosphere is expected to cool as CO2 increases.

The atmosphere cools from the surface outward until it meets the inversion layer known as the tropopause. Then it warms as you move outward. The reason is that below the tropopause the atmosphere is primarily heated by the Earth and above it is heated by the Sun. There is little mixing between the layers so very little convective heat transfer. Most of the heat loss from the stratosphere is in the form of electromagnetic radiation.

I have heard multiple explanations for the cooling of the stratosphere.

This is a new explanation. Are you suggesting the sun or the planet does not emit the frequencies that the CO2 molecule absorbs and emits?

My explanation for stratosphere cooling is that there has a reduction in low level clouds over the oceans, during the last part of the 20th century. Clouds reflect sunlight back into space. Sunlight that passes through O2 and O3 twice causes a greater increase in temperature than sunlight that only passes through it once. If there is less reflected sunlight, the stratosphere cools and the planet will warm.

There is published data that shows there was a reduction in planetary cloud cover in the 20th century.

 

[Saul]

(D)Saul actually confirms the approximate magnitude of CO₂ forcing

Above, Saul speaks of the theoretical CO₂ impact:

Of course. This is precisely what my original post explains. What you here call "theoretical" is the "Planck response" that I describe at point #5 of my first post. Repeating what I said originally: If nothing changes other than temperature, the response is straightforward. […] It's called the "Planck response" in climate science. Converting forcing units, this corresponds to about 1.1C per 2xCO₂.

There's a minor flaw in Saul's numbers if he gets less than 1C, but more importantly, this is not a "theoretical" impact for Earth at all. It's the impact for some other place where nothing changes as temperatures alter. The real theoretical work for Earth suggests an impact of about 3C per doubling; but I didn't cite theory for that. I went straight to observational constraints.

That climate feedbacks are positive is a real observation. The evidence I cited in my original post is empirical, and there's a lot more than my chosen example. Observations of the real world indicate that climate response is somewhere from 1.5C to 4.5C.

More amusingly still – by acknowledging the no-feedback "Plank response", Saul is actually using the very 3.7 W/m² that he started out by denying, or something very close to it! He acknowledges that there really IS a theoretical basis for non-feedback response of about 1C. Great. That means a forcing of about 3.3 W/m²; which is how much extra emission you get from Earth with a 1C rise as the base Planck response. Snap.

(Saul - cut)

Cheers -- Sylas

Sylas,

If I understand the 1.5C to 4.5C hypothesis. A doubling of CO2 will cause the planet to increase in temperature by 1.1C and "observations of the real world" shows the doubling of CO2 will cause a total net warming of the planet of somewhere of 1.5C to 4.5C.

The 1.1C temperature increase is per the equation logarithmic. What is the current expected warming for the increase from 280 ppm to 390 ppm?

http://co2now.org/

We both agree an increase in CO2 from 280 ppm to 560 ppm without feedback will cause some increase in planetary temperature. The difference in our position is the magnitude of the change. I am saying based on an analysis of the paleoclimatic data (planet's response to past CO2 changes) the rise in planetary temperature to a doubling of CO2 would be around 0.7C including feedbacks.

As you note and others note, CO2 can only absorb specific frequencies. The other frequencies pass out through the atmosphere. Is there anything that could possibly be incorrect with the 1.1C to 4.5C hypothesis?

Based on current temperature changes is it possible to falsify the 1.1C to 4.5C hypothesis?

If the planet were to suddenly cool would that disprove the 1.5C to 4.5C hypothesis? Or asking the same question another way, what is the maximum the planet can cool within the limits of the 1.1C to 4.5C hypothesis?

Regards --- Saul

 

[sylas]

Picking up two replies in one here:

I have heard multiple explanations for the cooling of the stratosphere.

There are two major reasons for stratospheric cooling. Reduced solar absorption, from declines in ozone concentration, and increased thermal emission, from increases in greenhouse gas concentrations.

Each of these has multiple secondary explanations, but because there is so little convection, it all comes down changing the interaction with radiation.

This is a new explanation. Are you suggesting the sun or the planet does not emit the frequencies that the CO2 absorbs and emits?

Think about this one for a minute: it's key. The frequencies that CO₂ can absorb and emit will not get from the surface to the stratosphere. They'll be absorbed, and emitted, all the way up the atmospheric column.

What eventually gets into the stratosphere, at these wavelengths, is emission from somewhere around the tropopause, which is the very coldest part of the atmosphere. You can see this in the calculated spectra I gave back in [post=2165483]msg #3[/post], and the same thing is seen in direct satellite measurements.

Increased interaction with thermal radiation, therefore, works up here as a cooling effect. You can think of greenhouse gases a bit like a large radiating surface, which can be used for passive heating or passive cooling, depending on whether you are hotter, or colder, than the background radiation around the radiator. This is not a new explanation at all.

My explanation for stratosphere cooling is that there has a reduction in low level clouds over the oceans, during the last part of the 20th century. Clouds reflect sunlight back into space. Sunlight that passes through O2 and O3 twice causes a greater increase in temperature than sunlight that only passes through it once. If there is less reflected sunlight, the stratosphere cools and the planet will warm.

That's an interesting notion, but I don't think it can work. Reflection from cloud is important in the troposphere, but is not likely to make much difference in the stratosphere. The UV radiation that is most effective for stratospheric heating is largely used up before it gets into the troposphere, and so the spectrum of reflected radiation is strongly depleted in precisely the wavelengths that would otherwise give stratospheric heating.

It's a neat idea, but in this forum you'll have to show that it has been seriously argued in the peer reviewed literature to take it any further.

There is published data that shows there was a reduction in planetary cloud cover in the 20th century.

Most of the literature indicates a trend of decreasing cover and a net cooling effect. As best I can determine, the following reference corresponds to what most scientists working on this subject see in the available data.

• Norris, J.R. (2005) http://www.agu.org/pubs/crossref/2005/2004JD005600.shtml, in J. Geophys. Res., 110, D08206, doi:10.1029/2004JD005600.

(Added in edit. Norris identifies a trend of decreasing cover, and associates it with a negative forcing. The impact of cloud is not limited to albedo. Cloud also interacts strongly with long wave radiation. More detail in the paper for anyone interested.)​

Sorting out the cloud effects is important for improving estimates of total forcing, as this is one of the largest uncertainties. (IPCC 4AR gives the range of possible forcings for cloud albedo over the 20th c. as -1.8 to -0.3 W/m², with a low level of understanding. The negative means it is a cooling effect.)

Resolving this uncertainty does nothing to alter the physical contribution from carbon dioxide or other well mixed greenhouse gases, which is known to a much greater accuracy.

If I understand the 1.5C to 4.5C hypothesis. A doubling of CO2 will cause the planet to increase in temperature by 1.1C and "observations of the real world" shows the doubling of CO2 will cause a total net warming of the planet of somewhere of 1.5C to 4.5C.

A doubling of CO₂ would give an equilibrium response of around 1.5C to 4.5C. That's what you get both with theory and with observation. Your "1.1C" figure there is the Planck response, which is what you would use for a planet with no atmosphere, given an equivalent forcing of 3.7 W/m². It's not a number you can use sensibly for the Earth.

The 1.1C temperature increase is per the equation logarithmic. What is the current expected warming for the increase from 280 ppm to 390 ppm?

Equilibrium response is log2(390/280)*3 (+/- 50%). That is about 1.4C, plus or minus 0.7.

This is not a prediction, because it does not include any other forcings, either positive or negative; neither does it consider the equilibrium response lag, of at least a couple of decades. It's not a climate model. It's the estimate, from well established physics, for the CO₂ impact all by itself. My aim not to explain climate, but simply to demonstrate that this CO₂ impact is necessarily a significant aspect of changing climates in the present;

We both agree an increase in CO2 from 280 ppm to 560 ppm without feedback will cause some increase in planetary temperature. The difference in our position is the magnitude of the change. I am saying based on an analysis of the paleoclimatic data (planet's response to past CO2 changes) the rise in planetary temperature to a doubling of CO2 would be around 0.7C including feedbacks.

Um, not trying to be rude here, but that makes no sense. You've just been arguing that paleoclimate is actually driven by some other forcing, which is plausible. But with that assumption, we've got no basis at all for estimating a CO₂ response from the data.

Unless you've got some way to estimate total forcings -- all of them, not just CO₂ -- you can't possibly estimate response or sensitivity. The large uncertainty in the carbon cycle beyond the Quaternary period, and the lack of data on albedo, makes it pretty much useless as a way of estimating sensitivity. Within the Quaternary we do have more of a handle on forcings, sufficient to get into the ball park for an estimate of sensitivity during the last glacial maximum.

As you note and others note, CO2 can only absorb specific frequencies. The other frequencies pass out through the atmosphere. Is there anything that could possibly be incorrect with the 1.1C to 4.5C hypothesis?

Not really.

Of course, no discovery in science is ever completely beyond question, there is always the possibility of something wrong with given estimates. What science actually does is keep trying to constrain the estimates and narrow the uncertainity.

These estimates ARE being narrowed. The 1.5C lower bound (not 1.1) is looking increasingly unlikely, and in the light of ongoing work, a more realistic lower bound is about 2C. The upper bound turns out to be harder to constrain, and this is a mathematical consequence of the non-linear nature of feedback response. (Ref: Roe and Baker, 2007)

• Roe, G.H., and Baker, M.B. (2007) http://www.sciencemag.org/cgi/content/full/sci;318/5850/629, in Science, 26 Oct 2007, Vol. 318. no. 5850, pp. 629 – 632, DOI: 10.1126/science.1144735

So a good summary for the current state of knowledge is that climate sensitivity is from 2 to 4.5 K/2xCO₂. Future developments may narrow that down, but as Roe and Baker point out, a substantial level of uncertainty is probably inevitable. Furthermore, sensitivity is not really a single precise number available to be discovered. It will vary somewhat with time and circumstance, and with the nature of forcings applied.

The range of 2 to 4.5 K/2xCO₂ for sensitivity is well supported, and there's little prospect of response being outside that range. There's also not likely to be a whole lot of narrowing of that range in the future, either. Climate is never going to simple to predict.

If the planet were to suddenly cool would that disprove the 1.5C to 4.5C hypothesis? Or asking the same question another way, what is the maximum the planet can cool within the limits of the 1.1C to 4.5C hypothesis?

That depends entirely on the forcings. A planet can only cool with a negative forcing, no matter what sensitivity you are using. If we were to see a sustained cooling effect, that would indicate a sustained negative forcing was at work – and so it would have to be something other than carbon dioxide.

I think I can see where this is going; so I'll just note for the record. The planet has lots of natural short term variability. From year to year there are changes going on that can give unpredictable differences, and this is effectively noise in the temperature records. The last 35 years has seen a strong trend of increasing global temperature, but for a randomly chosen sequence of ten consecutive years somewhere along that 35 year period, you might have anything from twice the trend to no trend at all, within about 2σ bounds of variation.

Such natural variations don't falsify anything, unless you know the specific forcings involved; and usually, we don't. But with a big volcano, you can estimate forcings and hence use information on the cooling response to constrain sensitivity. (Described with references in msg #1.)

The available data has already falsified the idea that sensitivity is less than 1.5, with 3σ confidence. If a new line of evidence is inconsistent with that, then we have one of the fun times in science when you know someone has to be wrong, but can't be sure who.

At present, however, the 2 to 4.5 estimate basically represents the range of what is credible, given all the data we can possibly throw at it.

Cheers -- Sylas

 

[Skyhunter]

I have heard multiple explanations for the cooling of the stratosphere.

This is a new explanation. Are you suggesting the sun or the planet does not emit the frequencies that the CO2 absorbs and emits?

No.
Here is a chart of atmospheric transmission
http://www.globalwarmingart.com/images/thumb/7/7c/Atmospheric_Transmission.png/495px-Atmospheric_Transmission.png

My explanation for stratosphere cooling is that there has a reduction in low level clouds over the oceans, during the last part of the 20th century. Clouds reflect sunlight back into space. Sunlight that passes through O2 and O3 twice causes a greater increase in temperature than sunlight that only passes through it once. If there is less reflected sunlight, the stratosphere cools and the planet will warm.

The stratosphere has been cooling since the 1950's consistent with the theory.

http://www.atmosphere.mpg.de/enid/2__Ozone/-_Cooling_nd.html"is a good site that explains it.

There is published data that shows there was a reduction in planetary cloud cover in the 20th century.

No there is not. At least nothing conclusive.

 

[mheslep]

...A doubling of CO₂ would give an equilibrium response of around 1.5C to 4.5C. That's what you get both with theory and with observation. Your "1.1C" figure there is the Planck response, which is what you would use for a planet with no atmosphere, given an equivalent forcing of 3.7 W/m². It's not a number you can use sensibly for the Earth.

Sylas - could you clarify: by 'equilibrium response' do you mean to include all the additional feedback mechanisms brought on by CO2 warming? I had thought that CO2 alone, without feedbacks would have a response of 1-2C at most..

 

[sylas]

Sylas - could you clarify: by 'equilibrium response' do you mean to include all the additional feedback mechanisms brought on by CO2 warming? I had thought that CO2 alone, without feedbacks would have a response of 1-2C at most..

I am estimating the impact of CO₂ for the Earth, which means real response of the planet, as best we can measure it.

I have not used any theoretical analysis attempting to quantify and model feedbacks. I used direct empirical estimates of sensitivity. There are a number of such empirical sensitivity estimates; I used Wigley 2005 in the the original post, as part of step 5 in the sequence. This estimates sensitivity from response to volcanic eruptions.

Equilibrium response simply means the total response after the planet comes back into energy balance. The original post mentions a 0.85 W/m² flux into the ocean. The equilibrium response is basically what you have once the ocean has warmed up enough to absorb that excess.

Where feedback shows up is in the attempt to actually model how the Earth's climate behaves, or to break down all the various physical aspects of the climate system. It's a complex interacting system, and the notion of feedback is a useful abstraction for mutual interactions. It is a very hand diagnostic for climate models. But it has nothing particularly to do with carbon dioxide warming. ANY change of temperature, for any reason, brings about a slew of changes and interaction within the Earth, and which bear upon Earth's climate sensitivity to any forcing.

I didn't worry about any of that in the original argument. I just went for empirical bounds on sensitivity.

It makes no sense to use a theoretical no-feedback Planck response, as would apply on the Moon, for example, in estimating how the Earth responds to anything.

Cheers -- sylas

 

[Saul]

I am estimating the impact of CO₂ for the Earth, which means real response of the planet, as best we can measure it.

I have not used any theoretical analysis attempting to quantify and model feedbacks. I used direct empirical estimates of sensitivity. There are a number of such empirical sensitivity estimates; I used Wigley 2005 in the the original post, as part of step 5 in the sequence. This estimates sensitivity from response to volcanic eruptions.

Equilibrium response simply means the total response after the planet comes back into energy balance. The original post mentions a 0.85 W/m² flux into the ocean. The equilibrium response is basically what you have once the ocean has warmed up enough to absorb that excess.

Where feedback shows up is in the attempt to actually model how the Earth's climate behaves, or to break down all the various physical aspects of the climate system. It's a complex interacting system, and the notion of feedback is a useful abstraction for mutual interactions. It is a very hand diagnostic for climate models. But it has nothing particularly to do with carbon dioxide warming. ANY change of temperature, for any reason, brings about a slew of changes and interaction within the Earth, and which bear upon Earth's climate sensitivity to any forcing.

I didn't worry about any of that in the original argument. I just went for empirical bounds on sensitivity.

It makes no sense to use a theoretical no-feedback Planck response, as would apply on the Moon, for example, in estimating how the Earth responds to anything.

Cheers -- sylas

Sylas,

You appear to be unaware of the current planetary temperature data. The planet is cooling.

How can that be possible with the 1.5C to 4.5C hypothesis. What is causing the planet to cool?

As I said, I am not arguing that CO2 cannot cause some warming. The scientific evidence does not appear however to support the 1.5C to 4.5C planetary warming for a doubling of CO2.

This is a recent paper that shows the planet was a "... resultant equilibrium climate sensitivity, 0.30 ± 0.14 K/(W m−2), corresponds to an equilibrium temperature increase for doubled CO2 of 1.1 ± 0.5 K. "

http://www.agu.org/pubs/crossref/2007/2007JD008746.shtml

Heat Capacity, Time Constant, and Sensitivity of Earth’s Climate System by Stephen E. Schwartz

ABSTRACT. The equilibrium sensitivity of Earth's climate is determined as the quotient of the relaxation time constant of the system and the pertinent global heat capacity. The heat capacity of the global ocean, obtained from regression of ocean heat content versus global mean surface temperature, GMST, is 14 ± 6 W a m−2 K−1, equivalent to 110 m of ocean water; other sinks raise the effective planetary heat capacity to 17 ± 7 W a m−2 K−1 (all uncertainties are 1-sigma estimates). The time constant pertinent to changes in GMST is determined from autocorrelation of that quantity over 1880–2004 to be 5 ± 1 a. The resultant equilibrium climate sensitivity, 0.30 ± 0.14 K/(W m−2), corresponds to an equilibrium temperature increase for doubled CO2 of 1.1 ± 0.5 K. The short time constant implies that GMST is in near equilibrium with applied forcings and hence that net climate forcing over the twentieth century can be obtained from the observed temperature increase over this period, 0.57 ± 0.08 K, as 1.9 ± 0.9 W m−2. For this forcing considered the sum of radiative forcing by incremental greenhouse gases, 2.2 ± 0.3 W m−2, and other forcings, other forcing agents, mainly incremental tropospheric aerosols, are inferred to have exerted only a slight forcing over the twentieth century of −0.3 ± 1.0 W m−2.

 

[sylas]

You appear to be unaware of the current planetary temperature data. The planet is cooling.

On the contrary; I am very familiar with the temperature record indeed, and I could see this response coming a mile away. In anticipation of this objection, I have already included this paragraph in an earlier post:

I think I can see where this is going; so I'll just note for the record. The planet has lots of natural short term variability. From year to year there are changes going on that can give unpredictable differences, and this is effectively noise in the temperature records. The last 35 years has seen a strong trend of increasing global temperature, but for a randomly chosen sequence of ten consecutive years somewhere along that 35 year period, you might have anything from twice the trend to no trend at all, within about 2σ bounds of variation.

2008 was a cold year, in the short term. It's the coldest year of the 21st century. 2008 was a hot year, in a long term. It's hotter than any other year of the 20th century, excepting only the big El Nino years of 1998/1997. That's according to either GISS or HadCRUT, and I'm sure you'll get the same in any reputable global temperature record.

Over the last ten years, the trend is for increasing temperatures. Depending on the dataset, this might be right back to the long term trend again, or else something rather below the trend but still positive. The difference is well within regression trend measurement errors over such a short time span.

If you look at even shorter windows of time, like the last eight years, only then you can find a negative trend. Short term variation in the trend is a normal consequence of the natural variation going on, all the time, and it is statistically insupportable to present this as an end to the long term trend of increasing temperatures.

In the original post of this forum, I cited the HadCRUT3v temperature record, with a reference to Brohan et al (2005). You can find the data yourself at Met Office Hadley Centre observations datasets (UK). I've had all this data on spreadsheets for a long time now, and it works much better than GISS data if you want to find recent cooling. Using the last 35 years, from 1974 to 2008 inclusive, this is what we find in HadCRUT3v (my calculations, annual data, linear regression):

• 35 year trend = 0.169 C/decade, +/- 0.033 (95% conf)
• 10 year trend = 0.107C/decade, +/- 0.212 (95% conf)
• 8 year trend = -0.111C/decade, +/- 0.178 (95% conf)
• variation in 10 year trend over the last 35 years: mean 0.169 C/decade, with 2σ bounds ranging from -0.054 to 0.392.
• variation in 8 year trend over last 35 years: mean 0.175 C/decade, with 2σ bounds ranging from -0.179 to 0.529.

That's the basis for my comment above: for a randomly chosen sequence of ten consecutive years somewhere along that 35 year period, you might have anything from twice the trend to no trend at all, within about 2σ bounds of variation. Eight year trends have even more variation, of course; and the last eight years are not even a 2σ outlier.

Natural variation like this doesn't happen by accident, and you can reasonably ask about causes. There is no indication whatsoever of some long term effect being involved, or of an end to the long term trend of increasing temperature that comes from a continuously increasing greenhouse forcing.

How can that be possible with the 1.5C to 4.5C hypothesis. What is causing the planet to cool?

As I said before; cooling means negative forcing, not different sensitivity!

Recently, it seems to be mostly due to the shift from a strong El Nino in 1998 to a strong La Nina in 2008. This is a natural cycle, with strong short term impact over a decade or two, but no long term cumulative effect. In my own personal opinion, which has no authority, I think there might also be a smaller contribution from the extended solar minimum. I've been trying to pull out a signal for the solar cycle from the datasets. It seems to be there, but subtle.

As I said, I am not arguing that CO2 cannot cause some warming. The scientific evidence does not appear however to support the 1.5C to 4.5C planetary warming for a doubling of CO2.

This is a recent paper that shows the planet was a "... resultant equilibrium climate sensitivity, 0.30 ± 0.14 K/(W m−2), corresponds to an equilibrium temperature increase for doubled CO2 of 1.1 ± 0.5 K. "

This is a much more interesting point. You've cited peer reviewed literature, as the forum requires. The citation is:

• Schwartz, S.E. (2007) http://www.agu.org/pubs/crossref/2007/2007JD008746.shtml, in J. Geophys. Res., 112, D24S05, doi:10.1029/2007JD008746.

Schwartz argues for a sensitivity value a long way below what I have proposed. He's suggesting 1.1C per 2xCO₂, which is pretty close to a no-feedback response. This is an example of just what I said above in #23: The available data has already falsified the idea that sensitivity is less than 1.5, with 3σ confidence. If a new line of evidence is inconsistent with that, then we have one of the fun times in science when you know someone has to be wrong, but can't be sure who.

I don't think this one is actually all that much of a puzzle. Schwartz is an extreme and isolated outlier in this whole area, and it didn't take long at all to see where he went wrong, and have this pointed out in the literature as well. Schwartz disagrees, of course; but that's the nature of the game in science. There was a fairly energetic set of responses and replies in the journal following Schwartz' original paper.

• Foster, G., J. D. Annan, G. A. Schmidt, and M. E. Mann (2008), http://www.agu.org/pubs/crossref/2008/2007JD009373.shtml, in J. Geophys. Res., 113, D15102, doi:10.1029/2007JD009373.
• Knutti, R., S. Krähenmann, D. J. Frame, and M. R. Allen (2008), http://www.agu.org/pubs/crossref/2008/2007JD009473.shtml, in J. Geophys. Res., 113, D15103, doi:10.1029/2007JD009473.
• Scafetta, N. (2008), http://www.agu.org/pubs/crossref/2008/2007JD009586.shtml, in J. Geophys. Res., 113, D15104, doi:10.1029/2007JD009586.
• Schwartz, S. E. (2008), http://www.agu.org/pubs/crossref/2008/2008JD009872.shtml, in J. Geophys. Res., 113, D15105, doi:10.1029/2008JD009872.

All this exchange is in the issue of 2 August, 2008. Fun reading.

There's plenty of precedent in science for someone to come up with new discoveries that bring about a sea change in their field. There's even more precedent for someone to come up with ideas that don't pan out, and that end up as falsified models. In the former case, the ground breaking new developments are often initially ignored or even dismissed as confusing or having the wrong result. The latter case, the failed new proposals are often taken apart quite thoroughly, showing exactly where they went wrong. This looks more like the latter case, from my perspective as an onlooker.

For this proposal to become accepted as a valid new result, Schwartz, or those who follow in his footsteps, will need to come up with credible explanations for all the data used to support estimates where his own value is well outside the range. Furthermore, this makes the directly observed temperature rise of recent decades a mystery.

Implications of unusually low sensitity

The calculation of a 3.7 W/m² forcing for 2xCO₂ is basic physics, with high confidence and low uncertainity.

A sensitivity of 1.1 K/2xCO₂ proposed by Schwartz, applied to the CO₂ change from 325 to 385 over the last 35 years, gives a CO₂ contribution of log2(385/325)*1.1, or about 0.27 degrees. The actual rise over this period is about 0.6 degrees.

So even with this surprisingly low sensitivity, the CO₂ contribution remains sufficiently large to be an important part of the whole equation… but no longer dominant. In this case, there would have to be some other very large positive forcing, and that would be a puzzle nearly as bad as explaining away all the data showing that sensitivity is actually something from 2 to 4 times larger than Schwartz proposes.

Cheers -- Sylas

 

[Bystander]

(snip)...suggested I lay out more carefully for critical examination the physical basis for the impact of CO₂ on climate, so that the assumptions can be seen clearly.(snip)

The suggestion also included something about "one assumption at a time."

This post is a calculation, cited to the literature as forum guidelines advise, quantifying the importance of CO₂ to climate. It aims to be transparent, so each step in the calculation is clear and assumptions can be identified.

(snip)
(1) Energy balance

Earth absorbs about 239 W/m² from the Sun, accurate to within a couple of percent. This is a global annual average, corresponding to a mean for the whole planet on the time scale of a couple of years, in Watts per square meter.

Satellite mapping of nightside? Complete? Statistical? Ongoing? How many years?

(snip)Almost all of it is radiated back into space as IR thermal radiation.

Or, more is radiated than is absorbed. Assymmetric, rotating, tattle-tale gray bodies are seldom at a steady state in radiative heat transfer between stars and the CMB --- and, the time scale for this revelation is what?

A small excess can be taken up with Earth's own finite internal stores of energy. The largest available internal store is the heat capacity of the ocean, and recent research indicates that a bit under 1 W/m² is currently being absorbed into the ocean, as it gradually increases in temperature. This is large by comparison with what is normal for the Earth. The flux into the ocean is known with about 20% accuracy at best; a recent published estimate is 0.85 +/- 0.15 (Hansen et al 2005).

"Small," large, "excess," deficit --- time scale? Size of heat reservoir is ONE factor --- response time is the other --- digging one watt signals from fluxes of several hundred watts that vary day to night by 50-75W/m² is "tall walkin' " when the uncertainty in overall average flux is "a couple per cent" of "~239 over a couple years."

(2) The atmosphere
(snip)
The Earth's surface is radiating something like 390 W/m² as IR radiation from the surface, known to within a couple of percent. Most of that ends up heating up the atmosphere, which then sheds heat into space. On top of that, thermal convection and heat of evaporation transfers almost another 100 W/m² into the atmosphere from the surface. Most of the Earth's thermal radiation into space actually comes from the atmosphere.

The atmosphere radiates in all directions, and on average something like 330 W/m² ends up coming back down to the surface. The net flow of energy is thus something like 60 to 70 W/m² radiant energy leaving the surface, plus the additional 100 or so W/m² as convection and latent heat. This is what balances with the energy absorbed at the surface, from the Sun.

100 W/m² for latent heat and convection? That's in base 2?

 

[sylas]

Earth absorbs about 239 W/m2 from the Sun, accurate to within a couple of percent. This is a global annual average, corresponding to a mean for the whole planet on the time scale of a couple of years, in Watts per square meter.

Satellite mapping of nightside? Complete? Statistical? Ongoing? How many years?

It's primarily basic physics. The value for Earth's absorbed energy falls out immediately as a consequence of knowing the solar constant, and Earth's albedo, and conservation of energy.

The value is (1-A)*S/4, where A is albedo, S is solar constant. The albedo is the reflected fraction, which is not absorbed. The division by 4 is for the ratio of surface area of a sphere to its cross section. Using A = 0.3, S = 1366, you get 239. The inaccuracy here is mostly in albedo, and works out to a couple of percent, as I stated above.

That's all the accuracy required for our purposes here. The more important point is simply to follow the idea of energy balance, and how changes in the balance are related to temperature.

If people would like to go into more detail on particular numbers, then the references I've given explain in a lot more detail. Satellite measurement is the main source of more accurate values in recent years. The major satellite projects involved have a home page on the NASA websites, and give background for things like mapping or coverage. See especially http://asd-www.larc.nasa.gov/erbe/ASDerbe.html.

Added in edit. As for time scales; the original post noted that these are global annual averages. The major reference, Trenberth et al 2009, gives values appropriate to March 2000 through May 2004.

The main point for this thread is to get the concept of "forcing", which is a change in the energy balance. The particular forcing and impact of CO₂ is unaltered by increasing precision in the measurements of absorbed solar radiation.

(snip)Almost all of it is radiated back into space as IR thermal radiation.

Or, more is radiated than is absorbed. Assymmetric, rotating, tattle-tale gray bodies are seldom at a steady state in radiative heat transfer between stars and the CMB --- and, the time scale for this revelation is what?

The energy balance numbers are annual averages. On this yearly timescale, almost all the energy absorbed is radiated back into space, because the Earth has no internal source of energy that comes anywhere close to the solar energy input.

For Earth, there is at present a small imbalance of energy being absorbed, measured with limited accuracy but less than 1 W/m². This was described in my post, with references.

Most objects in the solar system (rocky planets, asteroids, moons, etc) are definitely in a steady state balance of radiative heat transfer, with solar input effectively equal to reflection plus grey-body thermal emission out, even on very short time spans. The impact of stars and CMB is completely negligible. Some bodies, like Jupiter for example, have an internal heat source, and radiate more than they receive from the Sun. In this case, you can't use the same energy balance considerations that apply on Earth or most other solar system bodies.

"Small," large, "excess," deficit --- time scale? Size of heat reservoir is ONE factor --- response time is the other --- digging one watt signals from fluxes of several hundred watts that vary day to night by 50-75W/m² is "tall walkin' " when the uncertainty in overall average flux is "a couple per cent" of "~239 over a couple years."

Time scale for fluxes here is a global annual average, as noted in the original post. The small excess I mentioned here, with Earth currently absorbing a bit more than it emits, was explicitly quantified in my post, with explicit confidence limits. I cited a reference estimating 0.85 W/m², plus or minus 0.15. The time span for this excess is stated in the abstract (Hansen et al, 2005) as the last ten years. In the main text, this is clarified as 1993-2003.

For any imbalance between energy absorbed and emitted, there is a corresponding change in a heat reservoir. The only way response time bears upon this specific imbalance is precisely the response lag due to heating or cooling of a reservoir.

There's nothing controversial or unusual about working out energy budgets, and it's done with a lot more accuracy than just trying to calculate flux at every point of time, and take an average. Some fluxes can be known very precisely known indeed, even though the day/night variation is in hundreds of W/m². Nett shortwave insolation, for example. Others are less well constrained; but simply pointing out the day/night variation says very little about the uncertainties involved. The uncertainties don't primarily arise from the large diurnal variation.

If you are simply expressing a general skepticism that scientists could possibly estimate such things, then I can only shrug and point out that scientists in all kinds of fields make estimates of hard to measure quantities, with explicit confidence limits.

For the specific quantity mentioned here, of the small excess of absorbed energy over emitted energy, I just want to emphasize these points.

• It's highly uncertain. I cited one reference, giving 20% uncertainty bounds, but strictly speaking that is the bound given their method, and there's plenty of scope for other systematic problems that could mean the real value is outside those bounds.
• On my own behalf, when I am not citing others, I only tend to say that it is small, meaning under 1 W/m². That's also explicit in the first post. I'm presenting an upper bound on how far the Earth might be out of balance. To within a fraction of a percent the Earth is in net energy balance, as we should expect from conservation of energy.
• I don't put a whole lot of confidence on the value from one paper. It's an important paper and reputable work, but it's still appropriate to call this an open research question; so I use their value when appropriate and with suitable caution. Frankly, I suspect the value is too high. I could have used other lower published estimates, but this one works well as an upper bound on the imbalance.
• It makes no difference to the calculated impact of CO₂ given in this thread. I don't use this small and uncertain imbalance anywhere in the calculation. It is mentioned only because the concept of overall energy balance is so important, and hence it is appropriate to quantify the extent to which there may be an imbalance.

100 W/m² for latent heat and convection? That's in base 2?

Uh, no. It's good old base 10, and the value is quite definitely around that magnitude. The largest part of the contribution is from latent heat, and this is also the simplest to measure. You just need to know the annual total precipitation.

The approximate magnitude of this energy flux has been known since 1917 at least, where it is used in the first attempts to work out an energy balance of this kind.

You also highlighted some statements in red, which stand as very trivial basic background in no doubt whatsoever. That's at the level of high school or first year introduction to the atmosphere.

Cheers -- sylas

 

[joelupchurch]

I came across a new paper that suggests a climate sensitivity for CO₂ of somewhere close to one.

Douglass, D.H., J.R. Christy, 2009: Limits on CO2 climate forcing from recent temperature data of Earth. Energy & Environment, 20, 178-189 (Invited paper, reviewed by Editor.) You can download the pdf for free from Dr. Christy's web page http://www.nsstc.uah.edu/atmos/christy_pubs.html" .

The gist of the paper was the CO₂ warming should be consistent across all climate zones because CO₂ is well mixed and if some zones experience more warming than others, then it is for other reasons.

Here is the abstract:

The global atmospheric temperature anomalies of Earth reached a maximum in
1998 which has not been exceeded during the subsequent 10 years. The global
anomalies are calculated from the average of climate effects occurring in the
tropical and the extratropical latitude bands. El Niño/La Niña effects in the tropical
band are shown to explain the 1998 maximum while variations in the background
of the global anomalies largely come from climate effects in the northern
extratropics. These effects do not have the signature associated with CO2 climate
forcing. However, the data show a small underlying positive trend that is
consistent with CO2 climate forcing with no-feedback.​

 

[sylas]

I came across a new paper that suggests a climate sensitivity for CO₂ of somewhere close to one.

Douglass, D.H., J.R. Christy, 2009: Limits on CO2 climate forcing from recent temperature data of Earth. Energy & Environment, 20, 178-189 (Invited paper, reviewed by Editor.) You can download the pdf for free from Dr. Christy's web page http://www.nsstc.uah.edu/atmos/christy_pubs.html" .

Let's stop right there. That's not a credible source. It's not peer reviewed, and it is not in a recognized science journal. Here's the relevant part of the physicsforums guidelines

One of the main goals of PF is to help students learn the current status of physics as practiced by the scientific community; accordingly, Physicsforums.com strives to maintain high standards of academic integrity. There are many open questions in physics, and we welcome discussion on those subjects provided the discussion remains intellectually sound. It is against our Posting Guidelines to discuss, in most of the PF forums or in blogs, new or non-mainstream theories or ideas that have not been published in professional peer-reviewed journals or are not part of current professional mainstream scientific discussion. …​

This is applied especially stringently in the Earth science forum, mainly because there is so much low grade unscientific rubbish in the popular debate. The problem of course, is that "climate skeptics" generally perceive that the scientific community is "biased" against their work. But the only bias is the same that applies in all science: a bias against bad science and sloppy methodology. There's still ample scope for robust scientific debate on the many wide open research questions of climatology.

I'm not going to go into the merits of this article in any depth. Doing so would just give the incorrect impression that this is an actual scientific debate. It isn't. The authors are legitimate scientists in their own right, who do publish in real scientific literature; but on this topic they are both extreme isolated outliers within the scientific community, and they are both known for making error-filled arguments for insupportable nonsense, and public pronouncements that appear deliberately misleading. Their paper here shows both.

The journal "Energy and Environment" is notorious in this whole area. It has no impact rating, and almost no scientific visibility. It does not have the technical standards of a real science journal. It was, in fact, set up as an alternative publication venue for stuff on climate especially that would get tossed out of a real science journal.

Furthermore, there wasn't even an attempt at a fake peer review here. This paper is listed as an "invited paper" and "editor reviewed"; and the editor in this case is Sonja Boehmer-Christiansen; an English academic in political sciences who is a so-called "climate skeptic" with no technical expertise in the subject matter.

Most of the stuff in the quoted abstract is true enough, and pretty trivial. I have already described the guts of what is true in the abstract, back in [post=2171973]msg#28[/post]:

Recently, it seems to be mostly due to the shift from a strong El Nino in 1998 to a strong La Nina in 2008. This is a natural cycle, with strong short term impact over a decade or two, but no long term cumulative effect. In my own personal opinion, which has no authority, I think there might also be a smaller contribution from the extended solar minimum. I've been trying to pull out a signal for the solar cycle from the datasets. It seems to be there, but subtle.

These are additional non-cumulative natural variations that exist on top of any other long term trends. The ENSO cycle is not carbon related; although there is a potential for patterns of ENSO oscillation to alter as climate shifts in response to increasing global temperatures.

Why would Christy and Douglass bother to say ENSO is not carbon related? It could be just a comment that there are non-carbon related sources of temperature variation (duh!) but my cynical mind says that these guys are playing the usual misdirection card. I think they WANT to sow confusion, and are quite happy if a naïve reader picks up the impression that this must be some discovery or evidence against the larger greenhouse driven trends. It's no such thing, of course!

For a more sensible discussion of other factors bearing upon regional climate trends, see the thread [thread=306202]Only dirty coal can save the Earth[/thread], which addresses the causes of the exceptionally strong Arctic warming, and describes solid scientific research linking that Arctic warming to aerosols.

Where the abstract tips from trivial into junk is with final sentence, which speaks of a "small underlying positive trend" consistent with CO2 and no-feedback. They are effectively using a 1.1 K/2xCO₂ no-feedback sensitivity value, and everyone else uses 3 +/- 1.5 K/2xCO₂.

Effectively, they are saying that yes, CO2 is probably the major cause of long time climate trends; but that the trend involved is very small. In other words -- their argument hinges on assuming that the Earth isn't actually warming as much as other measurements indicate. This is papered over with a lot of other stuff within the article, some of which won't stand up too well to close examination, but ultimately it’s the temperature data that is IMHO the major reason they get ridiculous sensitivity numbers.

Christy uses his UAH_LT data for estimating the prevailing rate of temperature increase, limited to the tropics. This gives warming rates substantially less than almost any other source of data. There's a lot more to say on this, and that actually is a real scientific debate now underway… which Christy is losing. The warming rates I have used in this thread continue to be the better guide; and all the serious empirical work on estimating sensitivity continues to try and improve the bounds we now know from many different lines of evidence to be somewhere in the range 1.5 to 4.5 K/2xCO₂.

Cheers -- sylas

 

[chriscolose]

It is not particularly true that the forcing from CO2 is globally uniform, even though CO2 concentrations are essentially well-mixed. It is actually maximum in the tropics owing to the temperature contrast between the surface and troposphere, and even in the tropics is non-uniform because of dependence on humidity and clouds. The total response (accounting for feedbacks) is especially not uniform, which is why a fundamental characteristic of hothouse and coldhouse climates is in the changes pole-to-equator temperature gradient (being small in a warm climate, and much larger in a cold climate).

The sensitivity to a doubling of CO2 appears to be well within range of 2 to 4.5 C (IPCC 2007) and paleoclimate evidence does not support a very low sensitivity, such as a neutral feedback scenario. Many studies have examined this subject from a observational, paleo, and modelling standpoint and the results of a few decades of research really point to the IPCC range, so that is probably what policy decision should be based on. The high end is harder to constrain owing to the fact that feedbacks look like a converging power series (so you get asymptotic behavior as the gain factor approaches unity) but data doesn't really support a very high sensitivity (>5 C or so) either. A decade of flatline or cooling does not contradict any of this (and does not put any constraints on sensitivity), and is well expected to occur owing to the natural variability of the climate system, as discussed in a recent paper
http://www.agu.org/pubs/crossref/2009/2009GL037810.shtml

 

[mheslep]

Let's stop right there. That's not a credible source. It's not peer reviewed, and it is not in a recognized science journal. Here's the relevant part of the physicsforums guidelines

This is applied especially stringently in the Earth science forum,

Perhaps a disclaimer should have been made that this paper is not peer reviewed, but it seems to me that the last clause in the guidelines:

... or are not part of current professional mainstream scientific discussion.…​

would at least allow discussion on relevant writings from a scientist like Christy who has published so much peer reviewed material elsewhere in the field.

 

[sylas]

Perhaps a disclaimer should have been made that this paper is not peer reviewed, but it seems to me that the last clause in the guidelines:

... or are not part of current professional mainstream scientific discussion.…​

would at least allow discussion on relevant writings from a scientist like Christy who has published so much peer reviewed material elsewhere in the field.

John Christy is an active working climatologist, who has views that are strongly at variance with the great majority of his scientific peers. He publishes regularly in the real scientific literature. His ideas in the professional literature are actively engaged by other scientists, on their own merits.

You can introduce his ideas just fine with properly published work, and there are a number of advantages to doing it this way. Please make sure it is actually relevant to the specific topic of this thread, or if you want to explore some other issue, then consider a new thread for it.

I'm working on a longer response, which does look at Christy's published scientific work, and shows how it might and might not relate to this thread, and to the E&E paper.

Cheers -- sylas

PS. Caution: a lot of what Christy wrote prior to 2005 was flatly wrong, due to a basic algebraic error in his group's analysis, involving an incorrectly reversed minus sign, of all things. Everyone involved acknowledges this, and the scientific debate has moved on. This old error is now water under the bridge, but it certainly means that the older papers are well and truly out of date. Christy continues to argue for appropriately revised notions in the literature; and IMO he's losing that debate. But there's real engagement and scope to look into it, either here, or in another thread if that is more appropriate.