Arguments
Lindzen Illusion #6: Importance of Greenhouse Gases
Posted on 10 May 2011 by dana1981
Throughout the Lindzen Illusions series, we've gone back and scrutinzed his 1989 MIT Tech Talk to evaluate how Lindzen's climate positions in 1989 compare to his beliefs today, and to observational reality. In the talk, Lindzen made a statement about greenhouse gases (GHGs), downplaying the importance of carbon dioxide (CO2):
"Water vapor is far and away the most important greenhouse gas, except for one form which isn't a greenhouse gas: clouds. Clouds themselves as liquid water are as important to the infrared budget as water vapor. Both swamp by orders of magnitude all the others. With CO2 one is talking about three watts per square meter at most, compared to a hundred or more watts per square meter for water."
Similarly, in a 1992 article, Lindzen said:
"Even if all other greenhouse gases (such as carbon dioxide and methane) were to disappear, we would still be left with over 98 percent of the current greenhouse effect."
It's worth noting that this is one of Lindzen's previous positions which has gone by the wayside. Lindzen still believes that water vapor and clouds play key roles as feedbacks (which is true, although he has the magnitude and sign of those feedbacks wrong, as we saw in Lindzen Illusion #4), but he no longer downplays the significance of CO2 in the greenhouse effect. However, "water vapor is the most powerful greenhouse gas" has become one of the more popular climate myths, and remains pervasive among "skeptics", so it's worth once again examining the validity of this argument.
Schmidt et al. (2010)
This issue was addressed by two recent papers from NASA GISS, Lacis et al. (2010) and Schmidt et al. (2010). Schmidt et al. examined the contributions of various GHGs to the Earth's greenhouse effect. Schmidt et al. concluded as follows:
"we find that water vapor is the dominant contributor (∼50% of the effect), followed by clouds (∼25%) and then CO2 with ∼20%."
Schmidt et al. estimate the global mean longwave radiative flux from CO2 at approximately 30 Watts per square meter (W/m2), clouds at 40 W/m2, and water vapor at 80 W/m2. This is a stark contrast to Lindzen's 1989 claims:
"With CO2 one is talking about three watts per square meter at most, compared to a hundred or more watts per square meter for water."
Lindzen claimed that water vapor and clouds contribute at least 30 times more to the greenhouse effect than CO2, but as Schmidt et el. show, the ratio is closer to a factor of four. On this issue, Lindzen was off by an order of magnitude because he underestimated the CO2 radiative flux by a factor of ten.
Similarly, Lindzen claimed without GHGs other than water vapor and clouds, "we would still be left with over 98 percent of the current greenhouse effect." In reality, water vapor and clouds account for approximately 75% of the greenhouse effect.
Lacis et al. (2010)
Lacis et al. answer the question of "the most important GHG" even more directly. In fact, the first sentence in the abstract of their paper reads:
"Ample physical evidence shows that carbon dioxide (CO2) is the single most important climate-relevant greenhouse gas in Earth’s atmosphere."
The authors note that unlike water vapor, CO2 does not condense and precipitate from the atmosphere at current climate temperatures. Unlike many other greenhouse gases such as CO2 which can be added to the atmosphere, the level of water vapor in the atmosphere is a function of temperature. If extra water is added to the atmosphere, it condenses and falls as rain or snow within a week or two. In short, water vapor is a feedback, not a forcing. NASA GISS summarizes the results of Lacis et al. (2010) as follows (emphasis added):
"Because carbon dioxide accounts for 80% of the non-condensing GHG forcing in the current climate atmosphere, atmospheric carbon dioxide therefore qualifies as the principal control knob that governs the temperature of Earth."
So when it comes to governing global temperature changes, Lindzen's statement is backwards. CO2 plays a much larger role than water vapor and clouds, which act to amplify the CO2-caused warming, but don't remain in the atmosphere long enough to drive global temperature changes themselves.
A Lindzen History
As noted above, this is one of Lindzen's 1989 climate beliefs which has gone by the wayside, but unfortunately, remains a fairly widespread belief amongst global warming "skeptics". Once a climate myth is born, it's very difficult to kill, as the Skeptical Science Arguments database shows.
In the next Lindzen Illusions installment, we'll examine the history of Lindzen's climate arguments - what he argued in 1989, what he argues today, what is necessary for his arguments to be correct, and how those requirements stack up against observational reality.
Every time they repeat an erroneous statement you can bet that some politician somewhere leaps on it as a reason not to act and God knows it is action that is becoming more and more urgent.
It isn't as though we have the church telling us that the earth is at the centre of the galaxy and that sort of stupidity that happened in Galileo's day. These people are free to act in any way their conscience leads them.
This is a useful factoid for those that counter denialists’ points in the ether of forums and comments sections.
So I can't understand why Lindzen is so far off the mark.
Does he still believe his "orders of magnitude" claim?
Has SkS invited the attention of Dr Linzen to the series of “Illusion” articles it is publishing and, more importantly, has it invited him to respond?
[34] We conclude that, given the uncertainties, that water
vapor is responsible for just over half, clouds around a
quarter and CO2 about a fifth of the present‐day total
greenhouse effect. Given that the attribution is closer to 20%
than 2%, it might make more intuitive sense that changes in
CO2 could be important for climate change. Nonetheless,
climate sensitivity can only be properly assessed from
examining changes in climate, not from the mean climatology
alone [Annan and Hargreaves, 2006].
What this paper shows is what may happen in the atmosphere given the slab role. What it does not show is what the actual contributions of h20 vapor, clouds etc will have.
This is a stab at sensitivity using a poor model to boot.
Camburn - we're not talking about climate sensitivity here. Try Lindzen Illusion #4.
This argument of Lindzen's is a red herring, which as a physicist he knows very well. Only non-condensible IR-absorbing gases such as CO2 and CH4 can force the temperature upward, resulting in a higher vapor pressure of water and thus more water vapor, which amplifies the effect.
“... the amount of water vapor is a function ONLY of the temperature ...”
“... There is no "natural" source of water vapor that can be suggested as an alternative source of the last century's warming, because there is an ample inventory of liquid water, just waiting for some OTHER forcing to send it up into the atmosphere. ...”
Water vapor, however, is - in comparison with the CO2 - differently "placed" in the atmosphere. And it is influenced by very different factors - such as circulation: the atmosphere-ocean.
Temperature determines the amount of water vapor in the atmosphere - I agree - but the same amount of water vapor in the atmosphere - depending on the relative humidity - that cause other RF.
A very simple model for the water vapour feedback on climate change, Ingram, 2010.:
„... the heat radiated by the climate system is a function primarily of relative humidity, not temperature.”
Its effect on a real accumulation of heat - in the climate system - can be so much greater than that resulting from the comparison the direct RF water vapor - of direct RF CO2. Water vapor - in fact-so it can be more important than the result of simple calculations.
“... atmospheric carbon dioxide therefore qualifies as the principal control knob that governs the temperature of Earth.”
... but only with water vapor. Its direct impact - doubling - to max 1.1-1.2 degrees C. And we do not know exactly “how it works” water vapor.
To quote again Ingram, 2010.::
“A substantial positive feedback [on increase CO2] on climate change is then to be expected, and an approximate doubling is well established numerically. However, a physical explanation for the size of this effect is conspicuously lacking.”
And widely Lindzen's views - precisely - are shown (I think) here.
To what concentration of CO2 does this refer?
For the past 20 years I've thought that we have about 240W/m2 to power the climate: H2O in its various forms contributing about 180W/m2 and 280ppm CO2 about 9W/m2. Going up to 560ppm should contribute an extra 3.5W/m2. The temperature of the Earth's surface is around 285K, so to a crude first approximation the temperature rise involving a doubling of the natural CO2 level will be 3.5 x 285 / 240 = 4.2C which is about what the IPCC claim.
30W/m2?? What is going on?
“For instance, one cannot simply take the attribution to CO2 of the total greenhouse effect (20% of 33°C) and project that onto a 2 × CO2 scenario. That would exaggerate the no‐feedback impact of the extra CO2 while ignoring the role of feedbacks that might change the water vapor and clouds.”
Oh yes, this is certainly true ...
Where did you get your 9W/m2 number from (for CO2)?
Just a cursory look shows that it's not a reasonable number. If a doubling will add 3.5 W/m2, we can estimate that if we were to halve CO2 down to 140 ppm, then the contribution would drop from 9 W/m2 to 5.5 W/m2. Another halving to 70 ppm would drop the CO2 contribution from 5.5 W/m2 to 2.0 W/m2. But at that point you can't even halve CO2 again, because the contribution goes negative.
In theory, we should be able to halve CO2 about eight times (log2280 = 8.129). That would give us 8 * 3.5 = 28 W/m2... which is pretty close to the 30 W/m2 attributed by Schmidt et al 2010.
Of course, that's probably not strictly correct (I imagine things must get complicated when levels of CO2 are fairly low)... but certainly 9 W/m2 at 280 ppm seems way off.
Also, 241 W/m2 is the inbound/outbound component of the system. The radiative component which warms the planet above 255˚K is 155 W/m2 (per Schmidt 2010) which is added to that 241 W/m2. That is the component which must be divided among water vapor, clouds and CO2.
Lindzen's opinion is nothing close to what you quoted from Schmidt et al 2010. His opinion is that CO2 will be completely canceled by negative feedbacks from clouds.
The difference between his stance, and Schmidt's objective comment, is vast.
While Schmidt's comment is "certainly true," as you say... Lindzen's actual position is flimsy at best, and is in no way comparable to, or supported by, Schmidt's statement.
Ref 3. Modeling Experiements:
(14). WE use the IPCC AR4 version of GISS ModelE to calculate the instanteous changes in radiative fluxes to changes in indivueal LW absorbers, while holding the climate (spatial and temporal distributinos of temperature, surface properties, etc) fixed.
This is a slab atmosphere. Thank you.
Also, I don't know what you mean as a sound bite as it is clear that by holding the climate fixed, this is the definition of a slab atmosphere.
How does that quote in any way imply that the model uses a "slab" atmosphere?
From Schmidt et al [2006] describing modelE:
Reducing degrees of freedom within a problem in order to simplify calculations (i.e to make them remotely doable) is a pretty standard practice in most fields, and one that does not equate to a "slab" atmosphere.
http://alanbetts.com/workspace/uploads/us-senate-roleofcloudsinclimate-7-1283380998.pdf
This hearing, chaired by Senator Al Gore; including prepared statements by:
Albert Arking, Alan K. Betts, Robert Cess, Robert Charlson, Michael Hall, Richard S. Lindzen, V. Ramanathan, William B. Rossow, Shelby Tilford, Col. Robert Townsend and Kevin Trenberth
as well as the lengthy discussion that followed, which clarified several issues.
This hearing is an interesting snapshot of scientific opinion on the climate change issue in 1991. Speakers were invited to present a range of opinions. I was invited to present a counter argument to the negative feedback mechanism (upper tropospheric drying by deep convective clouds in the tropics) proposed by Lindzen at that time.
My prepared statement ("The Role of Clouds in Climate Change") is P33-37; Lindzen's is P14-20. The general discussion starts on P50. The discussion of Lindzen's hypothesis is mostly P64-76. Lindzen formally withdrew his hypothesis during the discussion (on P68 - this was news to Al Gore). However I don't think this retraction was well-publicized; and Lindzen has continued to suggest other mechanisms for low climate sensitivity - which are discussed elsewhere on this site.
[DB] Hot-linked URL.
How does that quote indicate a slab atmosphere? "Slab" is typically taken to mean a single isothermal layer with uniform absorptivities over the LW and SW spectrums. GISS ModelE certainly isn't that.
Radiative forcing by definition is the change in radiative flux at the TOA considered whilst keeping the climate "fixed" (after stratospheric equilibration).
What you are quoting is a partial slab. When Schmidt stated that the climate paramiters were fixed....that is the definition of a slab atmosphere in entirety.
He used AR4 GISS E and fixed the climate. I don't understand the questions that have been raised as I stated what Schmidt stated about the model used. The model was restrained by the paramiters it was subjected to. Kinda like one slice of a CT scan isn't it?
If you believe it is correct, please provide a citation.
See my comment at 20 to understand why certain parameters are held constant.No. He presumably ran the model to arrive at a representative state at equilibrium, then held that state constant while computing the radiative fluxes of the various components.
He did not "restrain" the model by parameters it was "subjected to."
There is a huge difference. You misunderstand what was done.
http://www.gps.caltech.edu/classes/ese148a/lecture8.pdf
I've never heard of this definition -- it differs so very much from the "partial" definition used in simplified energy balance models. Perhaps you can dig up a reference?
Your post appeared after I posted so I will reply. From what I have read, and my understanding, a slab is considered a stationary fix of the atmosphere. There can be many cross sections of the slab. Hence my reference to a CT scan. By holding all climate paramiters stable, it is a fixed slab of atmosphere.
Anyways..That is my take.
Your link doesn't support your definition.
It was also refreshing to read Gore's statements and questions. He actually had a solid understanding of climate science and desire to learn from the expert testimony. A stark contrast to today's congressional climate hearings, where politicians are just trying to score points, and Republicans just try to jam as many myths as possible into their alotted time.
I see nothing unusual about the reference to a slab atmosphere in the document provided. It in no way supports your own defition.
Beyond this, GISS modelE consist of 20 layers, not one single "slab." Beyond this, that slab is further subdivided into a rather extensive grid:In this context, your dismissive "slab" comments are meaningless.
Separate from this, your confusion with a "slab" holding parameters constant is unfortunate, but the confusion is yours. Again, reconsider how the model works. It has a large number of variables, that work in concert while it is running, each affecting and being affected by others. But the output of the model is not necessarily the value being hunted in this particular scenario.
So the model is allowed to run and achieve an equilibrium state, and then with all other values fixed it is possible to compute the radiative fluxes, almost as if time were frozen for an instant, to be able to measure everything instantaneously, and in an abstract way defeat the Heisenberg Uncertainty Principle.
There is nothing whatsoever wrong with this approach. You are falsely applying sinister motives and techniques, as if they first set the model to provide the expected output, and then arrived at the desired result.
This is simply not the case.
Please read the total link. There is a two layer model of a slab atmosphere. There is nothing wrong with a slab atmosphere as a tool.
Schmidt gets results by using a slab atmosphere. It is not a dirty word. It only means that the composition of all variables are kept constant, which he certainly indicates in his paper. He then changes one variable to obtain a result. Is this a pre-concieved result? I wouldn't have a clue as I didn't run the model.
My initial response was to the usefullness of this run. Being it was run in a static mode, (Is that less offensive?).....the application to climate is minimal because any change in composition in the real world results in various changes in climate as a whole.
I would suggest describing the "fixed" parameters, assuming that any serious modelling will be using a 20+ layer atmospheric model, rather than just arguing over a side definition.
I didn't want to argue over the deffinition of slab at all.
I did look at the entire thing. The page that is confusing you is where it says "Earlier, we developed a ‘two layer’ model of a slab atmosphere:" What it means is that it treated each layer as a slab (i.e. it used the slab approach, but with two of them instead of one). This is quite common, and ultimate any model -- even one with 1,000 layers -- is in this way a "slab atmosphere."
In any event, as I already posted, Schmidt did not use a slab atmosphere. The model was horizontally gridded with either 90x35 = 3150 cells, or 180x70 = 12,600 cells. This in turn is divided into 20 layers, for either 63,000 or 252,000 individual, 3-dimensional cells.
Hardly a slab.Then why do you use it disparagingly?
No it doesn't, and no he didn't. Did you bother to read my explanation about this?
First, "slab" has nothing whatsoever to do with setting up static parameters.
Second, ou are very, very confused about how models are structured, how they are used, and what was done in this case.
Let me try again.
They ran the complex, global climate model, which has been under development for decades, and takes weeks to months on a very high powered computer to run for anywhere near reasonable time scales. It involves physics and physical interactions of complex atmospheric makeup, radiation, ocean, and what-not, in a grid containing 252,000 individual cells.
They ran this to achieve a state which represents the 1980 climate (or, more likely, several states spread out over the course of one solar year) and at points froze the simulation to allow them to compute -- over 252,000 cells (slabs, if you prefer) -- the radiative fluxes.
From the paper:Your presumption that the model was somehow initialized (other than starting parameters to simulate 1980 conditions) hand artificially held constant (other than for the purposes of being able to extract the necessary numbers and perform the desired calculations) is mistaken.
1) As pointed out by Sphaerica, RobertS, and KR, your definition of a slab model is false. Even allowing for a two slab model, or a twenty three slab model, a slab model only models radiative transfers. Further, a slab model treats the entire surface of the Earth as being isothermal, and homogenous, and does the same for each "slab" of atmosphere it incorporates.
IN contrast a Global Circulation Model models heat transfers by convection, advection and latent heat. It divides the world up into cells which have different surface properties. It models heat transfers by any means between each layer, and between adjacent cells at each layer. As a result, a GCM will automatically generate the major atmospheric circulations including Hadley cells, Ferel cells and Polar cells with associated trade winds, westerlies and doldrums. It will also generate internal representations of clouds and precipitation automatically.
Describing a GCM as a slab model shows complete ignorance of the subject, and renders your comment irrelevant.
2) Schmidt et al did not hold "all variables" constant. He held "... the climate (spatial and temporal distributions of temperature, surface properties, etc.) fixed". (My emphasis.) That means when varying CO2 levels, he did not allow water vapour content to fall. But the temperature was not held constant, rather it was held to the annual cycle of temperatures at each particular location and altitude. Likewise with water vapour content and cloud cover when they were not being explicitly varied. That is not a "static mode".
If your understanding of the experiment where correct, Schmidt et al would only have modelled the changes to radiative forcing at a particular temperature at a particular location. But as it stands, they modelled the change over a range of temperatures across all seasons and across all locations in the globe. As such, it usefully determines the radiative forcings of the atmospheres components across the normal range of conditions we face.
You are right, of course, that that exercise is not entirely useful. Far more useful is to allow temperatures to vary with changing atmospheric concentrations, and to allow water vapour and cloud levels to vary with temperature. Given your qualms about Schmidt et al, you will no doubt endorse Lacis et al who do exactly that. So, assuming you are truly a sceptic and not just a denier, you would agree that without CO2 in the atmosphere, the water vapour concentration of the atmosphere would plummet, and the Mean Global Surface Temperature drop by around 34 degrees C.
Slab models allow for no lateral energy transfers; and
Slab models do not allow for convective or latent heat transfers.
I know you are being generous to Camburn by allowing as much truth to his argument as you can without absurdity, but you have, I believe allowed too much to avoid confusion.
Further, Camburn's defence in terms of multiple slabs is not warranted. His original claim was of a slab atmosphere. Singular. Had he said "it's just a twenty slab atmosphere" he would still have been incorrect for reasons given above, but it would have punctured his rhetoric of suggesting absurd oversimplification. So he suggested slab (singular) for rhetorical effect, and only allows for multiple slabs later to pretend he is being reasonable.
I will not endorse Lacis etal in intirety. They are assuming that h20 vapor is only a positive feedback. I am not ready to assume that. Schmidt also assumed that.
Dressler showed that it could be negative, so that option must remain open as a consideration, which would change the results of both Lacis and Schmidt.
You are totally wrong in your assesment. A slab can have many cross sections. Think what you may....I don't really care.
I made an observation as to the usefullness of this paper in a chaotic climate, which is what we live in.
For some reason you don't like the word slab....which why you don't is beyond me.
Do a search for other posts I have made on this site, then formulate your opinion.
Thanks in advance.
It has nothing to do with liking or not liking; it's about correct terminology. As Tom so patiently explained, a slab atmosphere model can have more than one layer, but in any case, it treats each layer as isothermal and homogeneously mixed, with uniform broadband absorptivity over the entire SW and LW spectrums (typically a_SW=0, a_LW=1). All energy flow is purely radiative, and the planet is treated as a uniform radiating sphere. You cannot calculate the radiative contribution of each LW absorber properly using this method, and that's not what Schmidt et al. do.
Nowhere in your link does it state that any model of the climate system is a slab model if "the climate is held fixed" (as defined by Schmidt et al.) because a slab atmosphere is a specific [and overly simplistic] representation of atmospheric energy flow, nowhere near the complexity used in GCMs.
Anyway, this conversation is off topic and more than a little silly.
@40 As Dana points out, you are wrong about the water vapour feedback. What is more, the difference between the effective temperature of the Earth, and the surface temperature is known; and the consequent difference in power also. It follows that if one factor contributing to the difference between the two is less than is thought, then the other factors must be greater to compensate. If the water vapour feedback is negative, in other words, then CO2 forcing must be stronger.
dana1981@42. You are totally correct. I was wrong in my ref to Dressler and h2o vapor.
H20 vapor may not be always positive as a feedback. It depends where it is at.
http://www.springerlink.com/content/m2054qq6126802g8/
Or whether or not radiosonde data is of better quality than satellite data and model expectations, of which the authors of the paper you cite are clearly not convinced.
Stick to pizza ... acquire enough empty boxes and you might begin to gain insight into how GISS Model E and other GCMs work.
The authors, in the abstract, say this, for instance:
"it is important to establish what (if any) aspects of the observed trends survive detailed examination of the impact of past changes of radiosonde instrumentation and protocol within the various international networks."
if any ... like, we don't really believe this ourselves, we're just putting the radiosonde data out there for people to look at and, hey, don't blame us, we're putting caveat after caveat in our abstract and paper.
Camburn thinks it's significant, even though the authors think it probably isn't ("if any").