Thursday, August 04, 2011 ... Deutsch/Español/Related posts from blogosphere

Can one sharply separate forcings and feedbacks?

Andrew Dessler is employed as a climate alarmist in Texas (at Texas A&M).

He recently expressed his opinion about the new paper by Spencer and Braswell (reactions in the media via Google News). According to Dessler, the "paper is not really intended for other scientists, since they do not take him seriously anymore (he’s been wrong too many times)."

Dessler himself who is never wrong (except when he speaks or writes) addresses his results to other scientists which is why he sent his reaction concerning Spencer's and Braswell's paper to Think Progress, a community server of militant Marxist guerillas.

I wonder whether he can see the irony. My guess is that he can't. Yesterday, alarmist Alexander Ač didn't manage to identify that a report about Bernanke in the pub came from the Onion and he presented it on his blog as real news. ;-)

These people are no longer capable to distinguish reality from fiction even in the most obvious cases. At any rate, Dessler's e-mail to Joe Romm also says the following:

To understand this paper, you have to understand the difference, between a “forcing” and a “feedback.” Forcings are imposed changes to, the climate, while feedbacks are processes that respond to changes in, the climate and amplify or ameliorate them. So the addition of carbon dioxide to the atmosphere by humans is a forcing—it is simply an imposition on the climate. Water vapor, on the other hand, is a feedback because the amount of water vapor is set by the surface temperature of the planet. As the planet warms, you get more water vapor in the atmosphere, and since water vapor is itself a greenhouse gas, this leads to additional warming.

The canonical way to think about clouds is that they are a feedback—as the climate warms, clouds will change in response and either amplify, (positive cloud feedback) or ameliorate (negative cloud feedback) the initial change.

What this new paper is arguing is that clouds are forcing the climate, rather than the more traditional way of thinking of them as a feedback. This is not, in fact, a new argument. Spencer’s 2010 JGR, paper as well as the new Lindzen and Choi 2011 paper both make this argument.

Overall, the argument made in all of these papers to support the conjecture that clouds are forcing the climate (rather than a feedback) is extremely weak.
Dessler's text gets extremely repetitive so I didn't copy the whole thing. OK, his main claim is that clouds are "just a feedback" and "not a forcing" and this is the key dogma that every paper about clouds has to obey.

Oh, really?

Bosses and slaves in the climate

First, what does it really mean? He means that there are the "real forcings" which are like the "bosses", the important effects, and they're the only truly active, independent players who command everything else. All other players in the system are just "assistants" to these bosses who "amplify" the orders by the bosses (or present their commands in a softer, more diplomatic tone - "negative feedbacks"), or they're just passive spectators.

Quasi-mathematically, he says that everyone has to reconcile his theory of the climate with the following multi-step template:
{\rm Emissions}_{CO_2} &= F_1({\rm Carbon\,\,tax})\\
{\rm Concentration}_{CO_2} &= F_2 ({\rm Emissions}_{CO_2})\\
{\rm Clouds} &= F_3 ({\rm Emissions}_{CO_2})\\
{\rm Temperature}&= F_4 ({\rm Concentration}_{CO_2})\\
and so on. So if you look at this hierarchical system of laws, clouds and everything else is ultimately a function of the carbon tax - the only "true forcing". But even if you forget about the carbon tax (which has no measurable impact on \(CO_2\) in the real world, unlike recessions and technological discoveries), this world view is based upon the belief that some quantities in Nature are "active" while others are "purely passive" and controlled or determined by the "active" ones.

That's how the dogmas of third-class scientists of Dessler's caliber are clearly built. But does Nature obey these dogmas?

In principle: it doesn't

In principle, when you're arbitrarily accurate, it surely doesn't. There are no fully passive players in the physical world. There is no exception. There are no observers that only react to the external changes but don't influence their environment. Dessler's reasoning is surely fundamentally flawed as a matter of fundamental physics.

The laws of classical physics may be expressed using the principle of least action:
\(\delta S = 0\)
The action \(S\) is a function of all degrees of freedom - positions of particles, shape of things that can change their shape, values of electromagnetic and other fields at each point, and so on.

If you vary the action, you obtain the so-called Euler-Lagrange equations: they're like equations similar to \(\vec F = m\vec a\) or Maxwell's equations for electromagnetism. It's important to realize that in all realistic situations, a dynamical degree of freedom \(\Phi\) enters more than one equation. If a degree of freedom \(\Phi\) is dynamical, and not e.g. a Lagrange multiplier, there must exist at least one term in the action that is at least bilinear in \(\Phi\) or its derivatives, e.g. \((\partial_t \Phi)^2\).

Consequently, it appears in the equations of motion obtained by varying \(\Phi\) itself. But unless \(\Phi\) is completely decoupled from everything else - a ghost that can't even react to other things - it also appears in the equations where another degree of freedom is (or other degrees freedom are) varied. It follows that \(\Phi\) influences the other degrees of freedom, too.

If \(\Phi\) is influenced by \(\Psi\), it must also be true that \(\Psi\) is influenced by \(\Phi\). Maybe I have used an unnecessarily complicated language to convince you about Newton's third law, the law about action and reaction.

So \(\Phi\) cannot be a passive player that doesn't influence anyone else. In particular, if you choose \(\Phi\) to remember some information about the cloud cover, it cannot be the case that \(\Phi\) is only a "passive respondent" that cannot influence other players in an independent way. It always can!

And we know damn well that clouds surely do influence the rest of the system. Tropical depressions have clouds in the middle. The clouds help to cool the interior - because they reflect the sunlight - which also helps to further reduce the pressure in the middle. Tropical depressions may later grow to tropical cyclones and hurricanes (which have a cloudless "eye" in the middle that helps to stabilize the system). This temporary instability wouldn't be the same if there weren't clouds - and it's likely that it wouldn't work at all. If clouds weren't important for a hurricane, you would probably get many hurricanes during sunny days which seem somewhat unusual. ;-)

The atmospheric-ocean interactions that lead to El Nino or La Nina conditions are affected by cloud cover, too, even if Dessler finds it inconvenient. But is there at least some approximation in which some degrees of freedom are "masters" ("forcings") while others are just "assistants" (feedbacks) or "slaves" (completely passive spectators)? Well, yes, there is.

Slow and fast degrees of freedom

In quantum physics, we know such things from the Born-Oppenheimer approximation. For example, the nuclei of the atoms are moving much more slowly (in the molecules) than the electrons. The momentum of the nuclei and the electrons are comparable - by the action and reaction principle (or momentum conservation, if you wish). But because the kinetic energy is \(E_K = P^2/2M\), it is clear that for the same momentum, the kinetic energy of the nuclei that have a much higher mass \(M\) may be neglected relatively to the kinetic energy of the lighter electrons.

So you may imagine that the nuclei are sitting at fixed positions and you calculate the behavior of the faster electrons relatively to these fixed positions. In this way, you obtain all the wave functions for the electrons. They also tell you that there is an effective potential energy for the positions of the nuclei, \(V(\vec R_1,\dots,\vec R_N)\). This effective potential energy is calculated as the total energy eigenvalue corresponding to the motion of all the electrons.

At the end, you may include the kinetic energy of the nuclei as well, add it to the potential energy you calculated from the electrons' wave functions, and neglect everything else about the electrons. In another language, the "fast" degrees of freedom of the electrons have been "integrated out": you no longer have to look into the detailed motion of the electrons. You have just extracted how it affects the slow degrees of freedom, namely the locations of the nuclei.

("Integrating out" in quantum field theory, using the logic of the renormalization group, is based on the same principles although the mathematical realization differs.)

An analogous "hierarchic" treatment exists in classical physics, too. At least when there are fast processes and slow processes. For example, the change of the concentration of \(CO_2\) in the atmosphere is a "slow" process - it takes roughly a century for the concentration to make a substantial part of its return to the equilibrium value for the current temperatures which is around 280 ppm.

Spongebob Squarepants in the Endless Summer, a sequel to An Inconvenient Truth (AIT) addressed to audiences that are 80% smarter than the average AIT viewers. The U.S. department of education will allow the students to choose to watch either Spongebob and Global Warming or a movie by Michael Moore, so that kids may get both sides. ;-)

On the other hand, the change of the concentration of water vapor is a "fast" process. It takes a much shorter time for water to evaporate from the ocean - or condense into the ocean - and reach same value of the concentration that is expected for the existing values of some "slow" parameters. In this sense, the water vapor is analogous to the electrons in the Born-Oppenheimer approximation: they also immediately adjust their wave functions to minimize the energy, given some "nearly constant" positions of the nuclei.

So the only justification of the dogmatic proposition that "clouds are not a forcing" is that they are always a "fast process". And indeed, you may find situations in which this approximation is relatively OK.

But you may also find situations and effects in which this approximation is totally wrong. If the Solar System moves through a spiral arm of the Milky Way so that the amount of cosmic rays is higher, these cosmic rays almost immediately influence the cloud cover but they may be doing the same thing for - well - one half of 140 million years (the period it takes for the Solar System to "bubble" through the Milky Way from one spiral arm to the next). Or any other period of time if there are some faster patterns in the cosmic rays, e.g. those influenced by the slow 400-year solar cycles.

So you could say that the cosmic rays are the forcing, and not the clouds themselves, but that's just about your choice of words. It's still true that from the terrestrial viewpoint, you will see a change in the cloud cover that looks independent of other internal quantities and that may (and probably did) change the Earth's temperature by many degrees if not a dozen of degrees.

You could say that the cosmic rays play a similar role of an "external agent" as humans do when they add \(CO_2\) into the atmosphere. And you would be right. But that doesn't mean that the cosmic rays are necessary for the clouds to be able to influence other degrees of freedom as if they were the "local bosses" for a certain "project". Of course that the clouds can play this role themselves.

While Dessler assumes that clouds are "so extremely fast" degrees of freedom that they're just ultimately slaves that should never be considered "independent degrees of freedom", even little children know that this is rubbish. It can take quite some time for clouds to disappear if you experience cloudy skies. The Sun above the clouds may be doing the right thing, everything may be in place to make those nasty clouds disappear but they're still there.

(We know it in Czechia. July 2011 was the coldest and rainiest July of this century so far. So if you had 100+ degrees Fahrenheit in Texas, please don't forget that there are also places outside Texas where your experience doesn't necessarily generalize.)

In fact, there are all kinds of lags associated with the clouds and some of them are comparable to months. Note that the typical lags in changes of temperature associated with the heat capacity of the upper ocean are of order 5 years. It simply takes years for the oceans to heat up or warm down by their contact with the adjacent atmosphere.

If I borrow a part of Figure 3 in Spencer-Braswell, it looks like this:

What is drawn is a regression coefficient between two quantities (some energy flux and temperature) in \({\rm W}/{\rm m}^2/{\rm K}\) as a function of some lag between them in months. When you understand some logic of this research, it's a pretty simple and fundamental a thing to try to draw. The green curve is extracted from the observations. The red and blue curves are calculated from the climate models. Within the accuracy we expect from the IPCC (and using clever approximations such as \(2035=2350\)), you see that the green, red, and blue curves almost exactly coincide.

You may ask: which of the IPCC climate models - those with a lower sensitivity (blue) or those with a higher sensitivity (red) - are better in describing this feature of reality? The answer is that it doesn't really matter. All IPCC climate models are extremely far from the reality (even though the lower-sensitivity models are a little bit better).

First of all, the models' maximum achieved slope is about 3-5 times lower than the observations: that shows that the IPCC models strongly overestimate how much tightly the atmosphere wants to store heat when you raise the temperature. When the Earth gets warmer, it simply emits more thermal radiation. The slope of this dependence is substantial and many times greater than the prediction of the IPCC models that essentially say that the Earth doesn't respond (and doesn't cool itself) when the temperature rises.

Second of all, the optimum lag where the curve is maximized - around 4 months according to the observed data - is something that the models totally disagree with. The contemporary IPCC-sponsored models - and the third-class scientists who defend them - have absolutely no clue about such quantities and patterns. They have never tested any of them. They have never asked the question. They're never being fired if they utterly fail in their job even though they should be severely punished if the system were working and if science mattered. So they don't even have to try.

What they do is to prepare some presentations indicating that the models "look OK" that the laymen who are ignorant about physics and mathematics are ready to buy. The assumption is that the laymen can't understand that if \(Z(t) = X(\Delta t)Y(t-\Delta t)\), one may try to determine the quantity \(X(\Delta t)\), encoding the relationship between \(Y\) and \(Z\), as a function of the delay \(\Delta t\). The number of people who are ignorant enough to be unable to understand these tests is high enough for the alarmist machine to run. Well, it hasn't ever been about the equations, anyway. So the likes of Dessler never do any genuine science. They never try to ask and answer meaningful questions. They're just defending some blatantly wrong dogmas that are convenient from various personal and collective perspectives.

Dessler is promising us that he will find some actual problems with the paper by Spencer and Braswell. He doesn't know of any at this moment but he's already sure he will find some problems - well at least, problems legitimate enough so that they can get through the "peer review" done by his soulmates (whose standards are near zero when the paper is helping to create panic).

So far, he has only realized - correctly - that the results of the paper are inconvenient and contradict the IPCC's perception of \(CO_2\) as the only relevant "master", or "forcing", in the climate they want to be discussed. The idea that the cloud cover at a given place could be a real number and an independent piece of information that is not a function of the \(CO_2\) concentration - or, heaven forbid, that clouds could even influence the temperature - is as scary for the likes of Dessler as the heliocentric system was for the medieval Catholic bigots.

Even if the Earth were not the center of the Universe and even if clouds had the potential to independently influence the temperatures beneath them, we surely can't allow this heresy to be loudly articulated, can we? This would lead to a complete collapse of the society as Dessler et al. visualize it.

People like Spencer, Braswell, Lindzen, Choi and others are working hard and they're rather badly treated for this work. This won't change until at least 95% of the pseudoscientific likes of Mr Dessler will be eliminated from the scientific community where they have absolutely no moral right to oxidate. What they're doing is nothing else than sabotage. They do everything they can for the genuine scientific research to be suppressed and de facto impossible. That's what they are apparently paid for.

You see that as soon as Spencer and Braswell plotted the observational data - and the model predictions - in a slightly original and unusual way, the alarmists were just stunned. They didn't know how to react - except for vacuous ad hominem attacks (like claims that no one surely takes them seriously because they're not a part of the alarmed community). In the same way, if you wrote a paper comparing the regional variability of the warming trend (the standard deviation that I analyzed in a few previous postings) with the model predictions, they would also be shocked. They have never tested their models whether they predict these rather basic quantities correctly. More seriously, they don't want the models to be tested because every new test means one more failure. And every new failure diminishes their anti-God called \(CO_2\).

There are many independent players and degrees of freedom in Nature - and in the climate system - and they influence each other in many ways. Some of these influences are more important, some of them are less important. They may be described by one equation or another - and the right one can't be clear a priori. The right theories about these influences must be selected by their agreement with the data, and not just the most obvious laymen's data such as the overall global temperature change, but also the agreement in many graphs that only skillful scientists such as Spencer, Braswell, Lindzen, and Choi may design and quantify. One example was reprinted above.

These four people are doing real climatology, real theory, real phenomenology, and/or real observations and experiments; Dessler and dozens of others are doing nothing else than propaganda, mudslinging, and defense of their politically motivated but scientifically unjustifiable dogmas.

And that's the memo.

P.S.: Chris Mooney has constructed a new and detailed proof that Spencer and Braswell are wrong. He borrowed a cartoon and rigorously demonstrated that Andrew Dessler is the same person as Albert Einstein. Because Albert Einstein is right, it follows that Andrew Dessler is also right. So Spencer and Braswell must be wrong because that's what Dessler says.

Such a brilliant and elegant proof. Why haven't I thought about it myself? Well, maybe I am not as smart as Chris Mooney.

Add to Digg this Add to reddit

snail feedback (1) :

reader netudiant said...

With all due respect for climate science, it seems to me that you are wasting your time on the topic.
It is like shooting fish in a barrel, the science presents a mass of targets.
However, the science and its shortcomings was never the issue, rather it is tax money and control. The skills needed in that fight are held in abundance by Dr Havel, but not by most scientists or physicist.

(function(i,s,o,g,r,a,m){i['GoogleAnalyticsObject']=r;i[r]=i[r]||function(){ (i[r].q=i[r].q||[]).push(arguments)},i[r].l=1*new Date();a=s.createElement(o), m=s.getElementsByTagName(o)[0];a.async=1;a.src=g;m.parentNode.insertBefore(a,m) })(window,document,'script','//','ga'); ga('create', 'UA-1828728-1', 'auto'); ga('send', 'pageview');