Guest Post by Willis Eschenbach
Figure 1. The Experimental Setup
I keep reading statements in various places about how it is indisputable “simple physics” that if we increase the amount of atmospheric CO2, it will inevitably warm the planet. Here’s a typical example:
In the hyperbolic language that has infested the debate, researchers have been accused of everything from ditching the scientific method to participating in a vast conspiracy. But the basic concepts of the greenhouse effect is a matter of simple physics and chemistry, and have been part of the scientific dialog for roughly a century.
Here’s another:
The important thing is that we know how greenhouse gases affect climate. It has even been predicted hundred years ago by Arrhenius. It is simple physics.
Unfortunately, while the physics is simple, the climate is far from simple. It is one of the more complex systems that we have ever studied. The climate is a tera-watt scale planetary sized heat engine. It is driven by both terrestrial and extra-terrestrial forcings, a number of which are unknown, and many of which are poorly understood and/or difficult to measure. It is inherently chaotic and turbulent, two conditions for which we have few mathematical tools.
The climate is composed of six major subsystems — atmosphere, ocean, cryosphere, lithosphere, biosphere, and electrosphere. All of these subsystems are imperfectly understood. Each of these subsystems has its own known and unknown internal and external forcings, feedbacks, resonances, and cyclical variations. In addition, each subsystem affects all of the other subsystems through a variety of known and unknown forcings and feedbacks.
Then there is the problem of scale. Climate has crucially important processes at physical scales from the molecular to the planetary and at temporal scales from milliseconds to millennia.
As a result of this almost unimaginable complexity, simple physics is simply inadequate to predict the effect of a change in one of the hundreds and hundreds of things that affect the climate. I will give two examples of why “simple physics” doesn’t work with the climate — a river, and a block of steel. I’ll start with a thought experiment with the block of steel.
Suppose that I want to find out about how temperature affects solids. I take a 75 kg block of steel, and I put the bottom end of it in a bucket of hot water. I duct tape a thermometer to the top end in the best experimental fashion, and I start recording how the temperature changes with time. At first, nothing happens. So I wait. And soon, the temperature of the other end of the block of steel starts rising. Hey, simple physics, right?
To verify my results, I try the experiment with a block of copper. I get the same result, the end of the block that’s not in the hot water soon begins to warm up. I try it with a block of glass, same thing. My tentative conclusion is that simple physics says that if you heat one end of a solid, the other end will eventually heat up as well.
So I look around for a final test. Not seeing anything obvious, I have a flash of insight. I weigh about 75 kg. So I sit with my feet in the bucket of hot water, put the thermometer in my mouth, and wait for my head to heat up. This experimental setup is shown in Figure 1 above.
After all, simple physics is my guideline, I know what’s going to happen, I just have to wait.
And wait … and wait …
As our thought experiment shows, simple physics may simply not work when applied to a complex system. The problem is that there are feedback mechanisms that negate the effect of the hot water on my cold toes. My body has a preferential temperature which is not set by the external forcings.
For a more nuanced view of what is happening, let’s consider the second example, a river. Again, a thought experiment.
I take a sheet of plywood, and I cover it with some earth. I tilt it up so it slopes from one edge to the other. For our thought experiment, we’ll imagine that this is a hill that goes down to the ocean.
I place a steel ball at the top edge of the earth-covered plywood, and I watch what happens. It rolls, as simple physics predicts, straight down to the lower edge. I try it with a wooden ball, and get the same result. I figure maybe it’s because of the shape of the object.
So I make a small wooden sled, and put it on the plywood. Again, it slides straight down to the ocean. I try it with a miniature steel shed, same result. It goes directly downhill to the ocean as well. Simple physics, understood by Isaac Newton.
As a final test, I take a hose and I start running some water down from the top edge of my hill to make a model river. To my surprise, although the model river starts straight down the hill, it soon starts to wander. Before long, it has formed a meandering stream, which changes its course with time. Sections of the river form long loops, the channel changes, loops are cut off, new channels form, and after while we get something like this:
Figure 2. Meanders, oxbow bends, and oxbow lakes in a river system. Note the old channels where the river used to run.
The most amazing part is that the process never stops. No matter how long we run the river experiment, the channel continues to change. What’s going on here?
Well, the first thing that we can conclude is that, just as in our experiment with the steel block, simple physics simply doesn’t work in this situation. Simple physics says that things roll straight downhill, and clearly, that ain’t happening here … it is obvious we need better tools to analyze the flow of the river.
Are there mathematical tools that we can use to understand this system? Yes, but they are not simple. The breakthrough came in the 1990’s, with the discovery by Adrian Bejan of the Constructal Law. The Constructal Law applies to all flow systems which are far from equilibrium, like a river or the climate.
It turns out that these types of flow systems are not passive systems which can take up any configuration. Instead, they actively strive to maximize some aspect of the system. For the river, as for the climate, the system strives to maximize the sum of the energy moved and the energy lost through turbulence. See the discussion of these principles here, here, here, and here. There is also a website devoted to various applications of the Constructal Law here.
There are several conclusions that we can make from the application of the Constructal Law to flow systems:
1. Any flow system far from equilibrium is not free to take up any form as the climate models assume. Instead, it has a preferential state which it works actively to approach.
2. This preferential state, however, is never achieved. Instead, the system constantly overshoots and undershoots that state, and does not settle down to one final form. The system never stops modifying its internal aspects to move towards the preferential state.
3. The results of changes in such a flow system are often counterintuitive. For example, suppose we want to shorten the river. Simple physics says it should be easy. So we cut through an oxbow bend, and it makes the river shorter … but only for a little while. Soon the river readjusts, and some other part of the river becomes longer. The length of the river is actively maintained by the system. Contrary to our simplistic assumptions, the length of the river is not changed by our actions.
So that’s the problem with “simple physics” and the climate. For example, simple physics predicts a simple linear relationship between the climate forcings and the temperature. People seriously believe that a change of X in the forcings will lead inevitably to a chance of A * X in the temperature. This is called the “climate sensitivity”, and is a fundamental assumption in the climate models. The IPCC says that if CO2 doubles, we will get a rise of around 3C in the global temperature. However, there is absolutely no evidence to support that claim, only computer models. But the models assume this relationship, so they cannot be used to establish the relationship.
However, as rivers clearly show, there is no such simple relationship in a flow system far from equilibrium. We can’t cut through an oxbow to shorten the river, it just lengthens elsewhere to maintain the same total length. Instead of being affected by a change in the forcings, the system sets its own preferential operating conditions (e.g. length, temperature, etc.) based on the natural constraints and flow possibilities and other parameters of the system.
Final conclusion? Because climate is a flow system far from equilibrium, it is ruled by the Constructal Law. As a result, there is no physics-based reason to assume that increasing CO2 will make a large difference to the global temperature, and the Constructal Law gives us reason to think that it may make no difference at all. In any case, regardless of Arrhenius, the “simple physics” relationship between CO2 and global temperature is something that we cannot simply assume to be true.
Sponsored IT training links:
Download RH302 questions & answers with self paced 70-270 practice test to prepare and pass 646-985 exam.
equilibrium = science fiction.
RE: Mooloo (02:20:07) : “I understand that more CO2 might create a greater greenhouse effect. I think it’s untrue because CO2 is at saturation…”
The situation is something like asking a painter to come in day after day and paint a black stripe down the left two inches of your kitchen window. Each time he does this there will be a little slop-over and the painted area will gradually increase at an ever decreasing rate.
gtrip.
It is obvious that you have never studied the geology of rivers as the forces which control the meandering of streams are indeed quite complex.
The difference in the velocity between the inner and outer banks of the oxbow generates differential erosion of the banks causing the stream to meander. While my explanation is simple, modeling all the factors involved would not be. This includes modeling turbulent flow, eddys, erosion, and I’m sure many other factors of which I am unaware.
It is not analogous to a plywood board covered in sand. The scale of the phenomena is very significant to the effect generated.
Willis’s post is informative and his analogy is valid. Your simplistic attempt at finding fault is simply uninformed and ignorant of the factors involved.
jeez aka charles the moderator
BTW, I don’t like the G. P Bear posts either, but it is just a matter of taste.
gtrip (00:23:13)
gtrip, let me suggest that you do some research before being so quick to judge and condemn. Whether in sand or soil, rivers meander and wander.
I’m glad to hear that your guess is that without any obstructions the river would flow straight down the hill. Unfortunately, it is well known and accepted science that your guess is wrong. You might Google “stream table” for some information on the subject before deriding it all as “junk science”. Despite your claim, it has long been known that a river does not “plot its course by the path of least resistance”. Nature is often counter-intuitive, what seems obvious “simple physics” is often not so. Like the man said:
“The world is not just more complex than we imagine, it is more complex than we can imagine”
Attributed to J. B. S. Haldane
Take a look at the photo in Fig. 2., and tell me what “obstructions” could cause that succession of changes and oxbows and places where the river used to flow. Do you truly think that the river is making those endless changes, forming loops and cutting them off, shifting from place to place and never settling down in one channel even after thousands of years because of “obstructions”, or as a result of following the path of least resistance? If those were true, the river would pick one channel and be done with it … but the evidence of all the historical channels shows that that is not the case.
I also strongly suggest that you read the information I linked to about the Constructal Law. While you may think it is also “junk science”, Bejan is one of the 100 most cited scientists on the planet … and I doubt that’s because people think he is peddling junk science.
My best to you, good to see you thinking about these questions, keep an open mind,
w.
I also feel a bit uneasy about the analogy of the meandering river and climate.
Despite the impossibility to model the exact path by which water will flow down hill, one came safely presume that the water will finally end up at the same spot: a more or less well established end point that is defined by the position of the mouth of the river.
The climate modellers may argue that with a certain amount of CO2 increase the atmospheric response cannot be predicted in detail in terms of place, time and severity (in the same way we cannot predict the exact course of a meandering river) yet the final end point will be a temperature increase of say 3 degrees with a doubling of CO2.
Any analogue can thus be attacked or twisted around. I would rather stick to the direct arguments (for which there are many) for questioning the validity of climate models.
Phil. (22:39:51) :
Allan M (12:05:52) :
From Reid Bryson:
Q: Could you rank the things that have the most significant impact and where would you put carbon dioxide on the list?
A: Well let me give you one fact first. In the first 30 feet of the atmosphere, on the average, outward radiation from the Earth, which is what CO2 is supposed to affect, how much [of the reflected energy] is absorbed by water vapor? In the first 30 feet, 80 percent, okay?</em?
Complete nonsense, not even close (never mind the bizarre comment in brackets [of the reflected energy])
Well, you can cite your aeronautic friend: I was was just quoting from an interview with Dr. Reid Bryson, the “Father of Climatology,” and at the time, the most cited scientist in the literature. The fact of it being a transcript of an interview explains the square brackets.
I know some don’t like ‘arguments from authority,’ unless it suits their case, but, as I say, I wasn’t arguing, merely citing.
Incidentally, Bryson got seriously caught up in the global cooling scare in the ’70’s; you know, the one that never happened. And it seems that he learnt his lesson with regard to panic and certainty.
See: Wisconsin Energy Cooperative News; archives, may 2007.
cthulhu (14:38:10) : “No model has ever shown strong negative feedbacks, and the history of Earth’s climate defies it. If there were strong negative feedbacks in climate, the climate would barely change at all.“
Well, looking at the temperature history, it actually did ‘barely change at all’.
http://geocraft.com/WVFossils/PageMill_Images/image277.gif
To me, staying within 10C over hundreds of millions of years is a picture of abject stability. Will you admit that despite wild changes in CO2 in earth’s past there is zero long term corrrelation to temperature or will you dispute the data?
Something seems to be regulating temperature far more effectively than CO2 is able to force it and not only CO2 – continents moving all over the place, vast lava flows and volcano activity, ELE asteroid impacts, etc. Everyone reading this post knows that the current climate models used to scare everyone wouldn’t have a SNOW BALL’S CHANCE of duplicating the above temperature record WRT even CO2 alone and, I repeat, that renders them .. GARBAGE!
Mooloo (02:20:07) :
“These gases are colder” relates to what? There is more CO2 in the outer atmosphere, yes. But the outer atmosphere black box radiates out at the same rate regardless of what gas it is made up of. So what other mechanism is at work then?”
Gasses are not blackbodies and will only radiate/absorb certain spectra. O2 is not active in the infrared, so it does not radiate/absorb infrared light. CO2 does radiate/absorb infrared light, that is why it is called a greenhouse gas.
The temperature of the high CO2 is colder than the temperature of the surface because of the pressure difference (lower pressure means colder).
Thanks a lot Willis for taking so much time to explain the flow system climate.
Here you show the earth as a heat engine . Can I just compare it too a worlpool driven by solar energy? It starts in the tropics with heating up the surface and evaporation, then convection, turning latent heat into energy mass (liquid water) a long the top of the source of the streamflow. In big thunderstorms, the clouds shoot up and lift the upper boundery level called “tropopause” even higher. From there the stream (through advection) flows towards the pole where the tropopause is much deeper. Maybe we could compare the average width of the global stream with the average height of the tropopause. The major gradient could be defined as coming from the high tropopause “peaks” over the tropics to the low troposphere “basin” over the poles, where excess heat can easily escape. If the average tropopause is higher, global temperature as one of the functions of the global energy balance is higher too (see Scientists Find “Fingerprint” of Human Activities in Recent Tropopause Height Changes. Of course I got convinced – partly by your articles – that this is not true but what Roy Spencer claims here:
and
and
I have several times in my career translated serious numerical modeling code from Fortran to modern languages and thus had to deal with the issues of validating the results. In the real world mistakes cost money and sometimes lives. Most recently I translated some aerodynamics code for a New Space company. I spent weeks doing nothing but validating and checking to be sure the output was reasonably trustworthy for questions within the realm of interest. When Rand told me the CRU model code did not even handle numeric overflows I was speechless.
Let me explain. Computers represent numbers in binary. Any signed representation (ie one that handles plus and minus) will use some formatting trick to differentiate the two. The problem is, if a positive number gets incremented to be one bit too big… it may suddenly become a negative number. Regardless of what does happen, any calculation using the value after an overflow might as well be a random number generator. The results are totally, utterly worthless. There is not a chance in hell that the output will be meaningful.
There are ways of dealing with this sort of thing but I will not go into that sort of techno-detail here. My goal is simply to point out that if the statements I heard are true, I must cease to believe the validity of any output from CRU and CRU related models.
http://www.samizdata.net/blog/archives/2009/12/a_few_thoughts.html
ThinkingBeing: “Any “cooling” you see in a short time span is just the meandering of a system that is never nicely in equilibrium.”
Substituting the word “warming” for “cooling” in your statement:
“Any “warming” you see in a short time span is just the meandering of a system that is never nicely in equilibrium.”
Wouldn’t that be an equivalent statement?
For the cooling that occurred during the period c. 1945 – 1978, the AGW crowd claims that was anthropogenic, but this time it is due to nature. AGW proponents can’t have it both ways.
Compared to billions of years of earth climate, the mere 160 years or so that man has been directly measuring the earth climate system temperatures represents only a few drops of water in the pool. Prior to that, we have only anecdotal reports and proxy data. Even ice core temperature data are a type of proxy data, since they are not measured in “real time”. Proxy data are not only non-global in scope (i.e. sparse), they are unreliable since they have to be interpreted, interpolated, extrapolated (!), correlated, adjusted, and then adjusted again. If the proxy data don’t support the hypothesis, it gets discarded, at least by some (alleged) scientists.
And based on a few billion years of earth history, wouldn’t say 10,000 years qualify as a “short time span”?
I wish ThinkingBeing hadn’t been driven off. I am puzzled by fast feedbacks. It would be interesting to see how that calculation works. It would also be interesting to see if he really knows how the calculation works or if he’s “talking through his hat.”
Mr Ace says: The temperature of the high CO2 is colder than the temperature of the surface because of the pressure difference (lower pressure means colder).
You were doing well until you got to this point. You can have a low pressure gas with very high temperatures. It is done all the time in plasma physics.
Temperature is a measure of the average energy of the atoms/molecules in a gas. Density then determines total energy.
@climatepatrol
Thanks for that link. Earlier I was wondering “where are the main flows?” and in my ignorance I had no idea there was already lucid article about it!
http://wattsupwiththat.com/2009/06/14/the-thermostat-hypothesis/
Weather (and Climate) equals mass times the speed of light squared.
Sometimes, when the microscopes are focused on a speck on a gnat’s wing, its difficult to see the gnat, the piece of bark it is resting on, the limb the bark is attached to, the trunk the limb is attached to, the tree the trunk is attached to, the trees surrounding the tree we’re looking at, or the whole snow covered forrest. I admit to being a pygmy in a land of giants. I am always amazed at the levels of knowledge that generally pour forth on WUWT. Now that the majority of travelers on Spaceship Earth have been disillusioned by the abortion that was Copenhagen, who are we to follow? Who will pick up the standard of humanity and shout “Follow Me!”, another group of charlitans, or someone with the guts to say: “At this point in time we don’t know enough about the weather to say what’s happening, so don’t worry, we’ll get back to you when we really have something solid.”?
anna v (23:15:45) :
I was puzzled with your statement: how could any molecule not have an infrared spectrum. Tom Vogt, where are you when you are needed?
Because homonuclear diatomics (like O2) don’t have a dipole! The reference you gave to forbidden transitions caused by photodissociation giving rise to highly vibrationally excited O2 (including triplet states) in a near vacuum (~100km altitude) isn’t relevant to the atmosphere.
I’m selling “Space Blankets”. No refunds.
MikeF (22:33:56) :
I remember them (first portable PCs) we knew them as lugables. Osbornes and Compaqs as I recall. Round about the same time as the first 64K DRAM (that’s 65,536 bits) were available commercially. I think we were only just changing from worrying about being frozen to being fried. About 1981 or just after.
You’ve made me ancient just thinking about it!!
MikeF (00:38:07) :
Phil,
I am sorry that my choice of words have offended you. Please accept my sincere apology.
Thank you I appreciate it.
You must admit, however, that your listing of you friend’s accomplishments and other appeals to authority sounded very much like bragging that little kids do about their fathers.
It wasn’t intended as an appeal to authority just a counterpoint to Willis’s argument by weird analogy.
If you want people to take your references to your friend seriously you should use his full name and cite papers he had published.
Anyone interested in following up could easily do so based on what I’ve already said but as said above that wasn’t the point.
And stop assuming that you are the authority on computer programming, without proof that sounds really childish.
Actually that was Willis: “This is not unusual. People think that computers are able to do many things that they can’t do. As someone who has programmed computers for more than forty years, I can assure you that a model is no better than the understanding of its programmer.”
Regarding your “since” statement – he did design wings in 70s, right? So, what did he use then? Since it couldn’t have been a laptop, what was it? Was it a slide rule? Or was it a computer? Maybe computer with “super” added to it? As in “supercomputer”?
As I recall it was a ‘mini’ in the lab which progressed to a ‘super mini’ (a Stardent?), it certainly wasn’t a ‘Cray’. The most impressive demonstration was when giving a lecture he would set up a laptop to optimize a wing during a lecture and have the computations show up live on a screen and finish in less than an hour.
Person of your experience with computers must be aware that modern laptops are much more powerful than 70’s supercomputers, right?
Of course, the first computer I used was all of 4k, somewhat later I did a lot of work in the 70s with the PDP8, one of which is now apparently on display in the Smithsonian!
And why is it when I read your posts I reminded of a time when I was in kindergarden myself (pretty long ago, unfortunately)?
I’ve no idea.
Let’s do simple thought experiment. Assume that you want to make river’s life easier by taking some of it’s resistance out. Let’s cut a bypass through one of the existing loops. This should be less “resistive” to the river then going the long way, agree? Not only would it make river shorter, you would line the new canal with nice slippery concrete so it flows easier. Well, as it turns out, the river will create extra loop elsewhere and maintain its length. It looks like it is not “interested” in least resistance as much as in maintaining its “status quo”.
This would falsify your “least resistance” argument, no?
Ok, what I did was pretty simplistic, but it is lot less simplistic than what you did.
Now, I am willing to admit that my “proof” is not really bulletproof. It might be possible to come to conclusion that particular length that is being maintained IS a path of least resistance after calculating “resistance” very carefully (and I’d imagine that some people would just redefine “path of least resistance” as a path that river takes), but it would only confirm the main premise of Willis’s post – there is nothing simple in many processes that actually happen in nature and “simple physics” is not the way to describe them.
Mike M (03:39:56) :
cthulhu (14:38:10) : “No model has ever shown strong negative feedbacks, and the history of Earth’s climate defies it. If there were strong negative feedbacks in climate, the climate would barely change at all.“
He needs to define “barely”. Also, strong negative feedback can be manifested in both gain and bandwidth. Bandwidth determines how fast the loop will respond to changes, gain how well. E.g., a PID loop can have very high gain at low frequencies yet low bandwidth. It will compensate low frequency disturbances very well. High frequency, not so well. In addition, if the loop is nonlinear, the gain and bandwidth depend on the operating condition.
Hank Henry (05:50:38) :
It would also be interesting to see if he really knows how the calculation works or if he’s “talking through his hat.”
I vote “hat”. Check out Tamino’s blog, if you can stomach it. That’s where he gets all his information.
Willis Eschenbach (12:37:06) :,
I get the same kind of thing from warmists all the time. My response is similar to yours. And I have made some monumental errors that are in fact documented on the ‘net. When I was convinced of my error I said so – out loud and in many places.
BTW in one of the places where I was wrong I had the opportunity to erase the record. My side and theirs. I didn’t touch a letter let alone a word. Why? Because my integrity is worth way more than hiding the evidence of my fallibility.
In addition, if the loop is nonlinear, the gain and bandwidth depend on the operating condition.
We see this effect in operational amplifiers all the time. The bandwidth for small signals is much larger than the large signal bandwidth. Which is why spec sheets show the results (scope traces) for small signals and large signals.
And that is for supposedly linear amplifiers (constant gain).
Hi Willis – Couldn’t agree more: Climate the mother of all nonlinear systems; “simple physics,” a prescription for error. Some observations:
1. Your 1st example inspired – substitute CO2 input for heating, and you’re off and running. Question is what are the negative feedbacks that induce climatal homeostasis? Lindzen says clouds /iris; fertilization effect (increased primary production), another.
2. Impt. to distinguish between homeostasis ( your 1st example) and chaos. Both require nonlinearity, but homeostasis can obtain in equilibrium systems – think Michaelis-Menton kinetics. In the case of mammalian body temperature, sweating, increased peripheral blood flow, etc., responsible for so-called “thermo-neutral” region. Of course, if animal subjected to too much heating, these mechanisms swamped: core temperature rises; animal dies. Implication for climate is that too much CO2 input could similarly exceed regulatory capacity, in which case, see below, one expects abolition of intrinsic variability (see below). Question then becomes, “when does this happen?”
3. Boundedness of terrestrial climate – at least since snowball earth (assuming Proterozoic glaciation as extensive as it’s cracked up to be) – plus climatic variability on wide range of time scales, suggests climate determined by presence of both positive and negative feedback – necessary, but not sufficient condition for chaos.
4. Sensitivity to initial conditions a consequence of chaotic topology – chaotic sets organized about infinite numbers of non-stable periodic orbits (saddle cycles) – i.e., every point on a chaotic set arbitrarily close to such a cycle. Climatal periodicities ranging from annual to millennial cycles therefore a “fingerprint” of chaos. Likewise, shifting periodicities – for example, switch from 100 ky to 40 ky glaciation cycles.
5. The conventional interpretation of climate as an equilibrium system subject to “forcings” bogus. By a theorem of Takens, any observable of a chaotic system also chaotic. Climate sensitivity in principle computable from the state of the system. If the latter chaotic, so also is sensitivity.
Some of these ideas developed at length in “A Surfeit of Cycles” recently published in Energy and Envt – available at http://bill.srnr.arizona.edu/mss/Surfeit.pdf.
Best wishes.
First, it is possible to ignore constructal law and threat it as noise. But then, you have to know what kind of noise you will get. Then you know if you have to make a 1-year average or a 10-years, 100-years average in order to remove noise. At this point, it seems most alarmists believe a 10-years average is all you need, while skeptics believe a 1000-years average migth not be enough. My conclusion, the science is not there.
Second, it is interesting to notice that all greenhouse effect gases could cool the temperature as much as they could warm it. In fact, if our sun was a red dwarf, those gases would absorb a lot of infra-red rays coming from the star before they get to the ground. So the planet might cool because of greenhouse gases. In conclusion, maybe we should have the best data about the output of the sun at several frequencies, because the sun’s output is responsible for most of the climate.