Guest Post by Willis Eschenbach
I’ve been ruminating on the continuing misunderstanding of my position that a governor is fundamentally different from simple feedback. People say things like “A governor is just a kind of feedback”. Well, yes, that’s true, and it is also true that a human being is “just a bag full of organic chemicals, minerals, and water“ … but if you want to analyze humans, that’s far from sufficient.
Similarly, a governor is a kind of feedback … but if you want to analyze it, you can’t assume it’s just simple feedback.
For starters, let me offer my definition of a governor. A “governor” is a control system that uses both positive and negative feedback to maintain some system variable near a “set point” value. A common example in our daily lives is the cruise control on your car. It increases or reduces fuel flow (positive and negative feedback) to maintain the vehicle speed near some pre-set value.
I want to illustrate why you can’t analyze a governed system the same way you analyze an ungoverned system. Suppose I have a car, and I start out from a standstill on level ground. As I gradually add more and more gas by slowly pushing down on the gas pedal, the vehicle speeds up. I record the speed and the instantaneous fuel flow, and I get a graph like Figure 1. As is the usual custom, I’ve graphed the independent variable (fuel use) on the horizontal X axis, and the dependent variable (speed) on the vertical Y axis.
Figure 1. An illustrative graph of automobile fuel use versus speed, with no cruise control. Yeah, yeah, I know the shape of the curve won’t be exactly like that … which is why this is an “illustrative example”. [UPDATE: an alert reader has pointed out that the speed is of course in miles per hour (MPH) and not miles per gallon (MPG) … I can’t be bothered to redo all the graphs, so please just make the appropriate mental substitution.)
Figure 1 shows that the vehicle speed is some kind of function of fuel use. It’s not a straight line because as you speed up, the air drag on the car slows it down. And the drag increases by something like the cube of the speed. So as we add more and more fuel, the effect of each added unit of fuel decreases. You can see that at sixty miles per hour, the car uses 2 gallons of gas per hour.
The takeaway point from this analysis is that in terms of causation, we can see that a change in fuel flow causes a change in speed.
Now, let’s get the car going sixty miles an hour on level ground, turn on the cruise control, and once again measure the speed and the gas. Then, suppose the car goes uphill. It will start to slow down. As soon as it does, the cruise control will react to the reduction in speed by increasing the gas flow to keep the speed near to 60 mph. And when the hill steepens, speed goes down a bit more, and in response the governor further increases the fuel use to keep the speed up.
Then we crest the hill and start down the other side. The car starts to increase speed, and when it gets going faster than 60 mph, in response the cruise control will reduce the fuel flow.
This leads to a very curious situation—vehicle speed goes up as the fuel use goes down, and speed goes down as the fuel use goes up. Figure 2 shows an illustrative graph of the situation .
Figure 2. An illustrative graph of fuel use versus speed, with a cruise control set for 60 MPH.
I’m sure you can see the problem. If we analyze this situation in the exact same way as we analyzed the situation without the governor, we come to the ludicrous conclusion that if we increase the fuel use, it will reduce the speed of the car.
So how can we understand this change? The key to analyzing this system is to understand that the governor (cruise control) reverses the causation in the system. Without the governor, as mentioned above, a change in fuel flow causes a change in speed.
On the other hand, with the cruise control engaged the exact opposite is true—a change in speed causes a change in fuel flow. And as a result of this reversal of causation, our previous analysis method is useless because it incorrectly assumes that a change in fuel flow causes a change in speed.
But wait, it gets worse. Suppose we start up the car from a standstill as in Figure 1, but then when we get up to 20 mph we set the cruise control for sixty mph. The car will continue to accelerate as in Figure 1, but only until the car gets up to sixty mph. When it gets there, it doesn’t speed up any more. Instead, it takes up the pattern shown in Figure 2. So the complete graph of the run looks like Figure 3:
Figure 3. An illustrative graph of fuel use versus speed, with a cruise control that is set to 60 MPH just after the car starts moving.
Here we see that in the blue part of the graph, the change in fuel use causes a change in speed … but once it gets up to the set point speed of 60 mph, the causation reverses, and now the red line shows that a change in speed causes a change in fuel use. Note that this reversal does not require any change in the governor settings. When the situation is far from the set point, the causation goes fuel –> speed. But once it reaches equilibrium, causation reverses such that speed –> fuel.
How does this relate to the climate? Well, the underlying climate paradigm is that the forcing controls the temperature, such that a change in forcing causes a change in temperature.
On the other hand, I’ve proposed that there is a natural governing system regulating the temperature of the climate, a major part of which works as follows.
In the tropics, when it is warm, clouds form earlier and reflect away the sun to cut down the solar forcing. And when the tropics are cool, the clouds form later or not at all, which greatly increases the solar forcing.
What that means is that a change in temperature causes a change in forcing.
And that is why I say that the current method of analyzing the climate is totally incorrect, because it assumes the causation is going the opposite direction from what is actually occurring.
In closing, I can do no better than to post up this marvelous cartoon by Josh from five years ago showing my “Thundercloud” governor at work …
Best wishes to all,
w.
As Usual: If you disagree with someone, please quote the exact words you object to so that everyone can know both who you are addressing and just what it is that you disagree with.
Any natural regulatory system has bounds on the variations it can control, and there are events that could alter or destroy the regulation.
A quote from Willis, (the above sentence) which is one of the most important things ever to be said when it comes to the climate. It is spot on and the heart of the problem.
Hasn’t happened in 4 billion years, so probably not a concern. I’d be more worried about ebola.
Question:
Does the current obsession with “selfies” produce positive or negative feedback loops?
My answer would be “positive” — they tend to amplify narcissism — thus making selfies NOT governors.
Haven’t you heard? Selfies are passé. The latest narcissistic obsession are “dronies”: a selfie from a drone.
Don’t ask me to explain…
This whole argument breaks down into one of perspective. It reminds me of several years ago they were interviewing a geologists on television. While standing somewhere near the middle of Wyoming, he said: ” The equator use to run right through here” as he pointed at the ground beneath his feet. ** Wrong *** The place where he was standing use to be located on the equator. It is all a matter of perspective.
Willis is merely looking at the system from the standpoint of how much fuel is necessary, more/less equals positive/negative feedback. The engineers are looking at the system from the set point (i.e., intended speed). When the speed varies either plus or minus, it takes a negative feedback to return it to the set point. A positive feedback increases the deviation from set point.
Actually, Will’s point of view is probably more easily understood by the non-engineers in the crowd. It is also just as valid from a systems point of view. You can tell by the reactions of the poor control engineers that they have had he “negatives” of “positive feedback” drilled into them so hard, and for so long, just the mention of the words causes a “valvie”.
No, these are accepted, well-defined terms with a precise meaning. It is not a case of ‘ oh I think I’ll redefine for the layman’ , the use was wrong. W has corrected the article. If you also imagined it meant something else, you have once again learnt something from visiting WUWT.
How does this relate to the climate? Well, the underlying climate paradigm is that the forcing controls the temperature, such that a change in forcing causes a change in temperature.
On the other hand, I’ve proposed that there is a natural governing system regulating the temperature of the climate, a major part of which works as follows.
Willis further says which again is the heart of the problem for why nevertheless the climate changes.
Why why why????
It deals with a general epistemic problem in the sciences. Specifically that of dealing with assigning causes and analyzing them. This is just a specific case of ‘correlation is not causation’ with an illustration of how and why that can be.
I teach Physics and Environmental Science. I will use this article as well as the Comments this year in class.
Also, climate is, of course, very complex. To make things even more complicated, there are undoubtedly mechanisms that not only vary with temperature in the intensity of their effects, but there are also some mechanisms that only work over certain temperature ranges, thus turning on and off as factors in climate. Nothing can be assumed to be a constant factor, even the Sun, as ingenuously assumed by the IPCC.
For starters, let me offer my definition of a governor. A “governor” is a control system that uses both positive and negative feedback to maintain some system variable near a “set point” value. A common example in our daily lives is the cruise control on your car. It increases or reduces fuel flow (positive and negative feedback) to maintain the vehicle speed near some pre-set value.
There are “self-organized” systems, “evolved” systems (random variation and natural selection, as with biological organisms; also human success with “trial and error”), and “designed systems”. Your argument incorporates the ideas of the Reverend William Paley’s “argument from design” about the existence of God, and the modern arguments about “intelligent design”. That is, it does so if it is an “argument” instead of a bald “assertion”. Designed systems are explicitly teleological, with parts and attributes selected in advance to achieve a goal or improve upon the system as it is. Evolved systems have been modified with at least some randomness (independence of the goal or function), and selected post hoc when they are better than their predecessors. Self-organized systems are observed to operate with describable characteristics or within bounds because of the properties of locally-acting processes; they are also systems with “emergent properties”, but have not been explicitly designed or selected in order to achieve those describable characteristics and bounds.
Random variation and natural selection can be observed in every population of living things that has yet been studied. Intelligent design can be observed in the design and manufacture of automobiles and aircraft (though they are not completely deterministic — some ideas work better than others and are selected for more development), and self-organization can be observed in a lot of chemical systems such as “reaction-diffusion” reactions (and in some “deterministic” computational systems such as high dimensional non-linear dissipative systems, where the equations are not selected to achieve any particular behavior in phase space.)
When I questioned your use of the word “thermostat” I was not questioning your analysis of how it works (that may have been separate, though I think your analysis of clouds is basically sound), I was questioning the teleology implied by calling it a “thermostat”. What evidence have you that any of the properties of the system were designed or selected to produce any of the regularities of the observed system? Did some agent place the Earth within a range of distances from the sun to produce life (contrary to Mars and Venus), either by design or by trial an error (the “Goldilocks position” after the inadequacies of Mars and Venus). Did that agent create oxygen with the capacity to make ozone to protect us from too much sunlight and also to permit our biological energy processes? Did the agent create metals so that the process of random variation and natural variation could evolve enzymes like hemoglobin and chlorophyll? Some people answer those questions in the affirmative: Christians, Deists and more. But to paraphrase a famous aphorism, the invocation of the Agent adds nothing to the understanding of how the system works.
I appreciate and understand the engineering wonks and pedants getting their knickers up about ‘positive feedback’ and ‘governors’ precisely because Willis is using some of their terms. This despite that Willis could have, and did, call his terms anything and define them for the purposes of his argument.
What I do not appreciate are people in a deep search for problems that they can answer with their pet religious views. People such as yourself. When Willis moves on from descriptives to origin myths, please have my every encouragement to jump in and postulate your own. Indeed, feel free to rebut any origin myth that Willis or anyone else raises at that time.
Until then, be happy that your religious zealotry has been duly noted; and put a sock in it.
Matt, I generally value what you say, but you lost me here. What is your point, exactly? You are free to question the teleology of calling what is going on a “thermostat”, but I don’t understand just what word you might want to replace it with.
It seems that your objection is that you think there could not be a natural thermostat … despite the fact that there are natural heat engines.
So I’m just not getting what you’re trying to say. Could I ask you to boil it down into the “elevator speech” version, and make your point clearer?
Like I say, I ask because your thoughts are usually interesting.
w.
Willis: You are free to question the teleology of calling what is going on a “thermostat”, but I don’t understand just what word you might want to replace it with.
“clouds”
using a thermostat to explain the workings of the cloud feedbacks has the same conceptual problem as using “greenhouse” and “blanket” as analogies of the CO2 effects. Understanding greenhouses and blankets does not actually illuminate the discussion of how the CO2 produces its effects, and causes some people endless confusion. (e.g. [Adding an extra blanket can not “cause” additional body warming because the blanket is cooler than the body, so adding CO2 can not cause additional Earth surface warming .])
The thermostat and the cruise control have another important property in their workings that is absent in the clouds: somebody can “set” and “reset” the thermostat or cruise control. It isn’t at all clear that anyone can set the clouds, and the idea that we can set the clouds by controlling the CO2 is at the heart of the policy debate. If we could set the clouds by controlling the CO2, then the clouds might be like a thermostat. But your basic argument is that the cloud system is self-organized, and can not be set. At least I think that is your basic argument.
Thresholds, thresholds and thresholds are a big factor in a non linear, chaotic, random system such as the climatic system is. This if/when reached puts the governor on the climate system to the test.
For those who missed my statements on individual threads, I agree totally with those who have said that a generator uses just negative feedback … my bad, written on six hours sleep in two days. I have changed the two incorrect statements in the head post to reflect the correct terminology, and they have my thanks for pointing it out.
Let me add that this doesn’t change either the other statements or my conclusions in the post.
Regards to everyone, this is the beauty of the web. My mistakes don’t last long without being revealed … not always with the kindest of intentions and words, but that is the way of life, and I’m happy to have been corrected.
w.
Hi Willlis
This post brought to mind Le Chatelier’s Principle I was exposed to in chemistry classes. I am sure you can make an interesting post discussing Le Chatelier’s Principal and climate “change”…
kalya22,
This does indeed *resemble* Le Chatelier’s Principle but it is not a result of Le Chatelier’s Principle. That rule only applies to systems at thermodynamic equilibrium. Systems not at equilibrium might or might not respond as expected from Le Chatelier’s Principle, so it provides no predictive power in such cases.
Mike M. (period) August 2, 2015 at 10:33 am
Mike, what evidence do you have that the earth is NOT at thermodynamic equilibrium? It only heats or cools a tiny amount over centuries (± 0.3°C over the 20th century). How is that not thermodynamic equilibrium?
w.
Willis,
The temperature is not the same everywhere, therefore not at thermodynamic equilibrium.
The earth is in something approximating a steady state, if you don’t get too picky. Some people call that “equilibrium”. But that is certainly not “thermodynamic equilibrium”, which has a very specific meaning.
Brian,
“With swings of +/- 5 degrees C, with the associated glaciation and retreat in the long term (800K yrs) it’s hard to characterize it as “steady”.
I tend to agree, which is why I said “if you don’t get too picky”. And, of course, on time scales under a year, it is not at all steady.
Brian G Valentine August 2, 2015 at 12:19 pm
Thanks, Brian. You are correct. While the earth is thermodynamically stable, it is not in thermodynamic equilibrium.
However, let me add that that I don’t think that Le Chatelier’s Principle only applies at thermodynamic equilibrium as you said above. Le Chatelier said:
For example, the ocean is a buffered system which is (generally speaking) at chemical equilibrium. but far from thermal equilibrium. Le Chatelier’s principle operates there regarding the chemical species of the ocean.
As to whether the global climate system obeys Le Chatelier’s Principle, I’d say likely not … but instead, it obeys the dictates of the Constructal Law, which often amounts to the same thing. You might enjoy my posts on the subject, here and here.
Best regards,
w.
Willis,
“For example, the ocean is a buffered system which is (generally speaking) at chemical equilibrium. but far from thermal equilibrium. Le Chatelier’s principle operates there regarding the chemical species of the ocean.”
Actually, the ocean is not at chemical equilibrium. For example the pH varies from place to place and with depth.
In the real world, thermodynamic equilibrium only occurs on limited scales of time and place and within some limited range of accuracy. I might have liquid acetone at dry ice temperature in a Dewar flask and have it be usefully close to equilibrium for the duration of some experiment, but if I come back in a week I will likely find it to be at room temperature. Similarly, I might be able to treat some finite chunk of the ocean as being in thermal and chemical equilibrium even though the ocean as a whole is not. An example is measuring temperature; that is something that can technically be only done in a system at equilibrium.
I never heard of the Constructal Law before. At first glance it sounds both fascinating and crackpot; I will have to set aside the latter impression until I learn more. I seriously doubt that it has any meaningful connection to thermodynamics, but that does mean it is not useful.
But I would think that the Constructal Law is not useful if you can not predict a priori what the correct quantity is to extremize. You might suppose that a flowing river tries to maximize length, but I am pretty sure that it tries to minimize length (or maximize slope). Meanders occur when the slope is small as a result of a positive feedback loop that drives the flow away from the optimum, until it snaps back. The point here is not which view is correct; it is that you can’t get a reliable result until you know what extreme is “sought”. Maybe when I learn more, I will learn how to answer such a question.
A question for Willis:
Your proposed governor mechanism would tend to maintain a given “level state” of the climate system, as I understand your position. You point out that this is consistent with the evidence to within a decent tolerance (ie a very small percentage variation in average temperature pretty much throughout Earth’s history). My question is, what in your opinion determines what that level is in the long run? I apologize if you’ve been over this before, I can’t recall reading anything you’ve written to that effect.
Good question, Andrew. I doubt that there is a simple answer. The climate temperature control system is quite complicated. It includes many emergent climate phenomena, including inter alia clouds, thunderstorms, El Nino/La Nina pump, and dust devils. Some complex combination of these determines the global average temperature but I don’t think disentangling them is possible at this time.
w.
Andrew_FL says “My question is, what in your opinion determines what that level is in the long run?”
Pardon my jumping in here… The phase transitions of water are the boundaries over which the control system can operate. Exceed either boundary and it can “lock up” as for instance “snowball Earth”. The Earth remarkably has all three water phases (vapor, liquid and ice) concurrently and climate regulates, and is regulated by, the proportions of these phases. Weather is heat trying to reach the ice from the tropics.
Earth appears to have more than one stable state; the long glacial periods (100 thousand years or so) and the interglacials (ten thousand years or so). It also has a snowball state (very bad for life) and a hypothesized-by-warmists Venus state exactly the opposite of snowball.
I very much appreciate Leo Smith’s contributions here to the discussion. I have been thinking for some time that the real issue is whether or not climate has overall negative or positive climate feedback systems. The warming crowd have fixated on one potential variable in CO2 and decided that it is positive to warming and there are no offsetting negative variables This seems to me to be extremely unlikely given that earth has had a fairly stable environment for most of 3 billion years. This in spite of significant variation in CO2 levels. Additionally, Leo Smith mentions the tremendous amounts of heat transported upwards ( and outwards) by water vapour. This heat transport mechanism is a powerful and pure negative feedback on temperature, aided by the ability of warmer air to hold and therefore transport greater amounts of heat as temperatures rise. It seems to me that climate change has become an intellectual industry that consumes a fantastic amount of human effort in what I fear is a very wasteful pursuit of a solution to a problem that doesn’t exist. Can it be that hard to actually measure earth’s heat input versus output from space?
John Harmsworth,
“The warming crowd have fixated on one potential variable in CO2 and decided that it is positive to warming and there are no offsetting negative variables”.
That is simply not the case. All climate models, even the simplest as in my post above, include the “Planck feedback” (which I called “Planck response” above). If you change the radiative balance (say, with a change in solar output, or a big volcanic eruption, or by changing CO2), the temperature changes in response. That causes IR emission to change, via the Stephan-Boltzmann law. And that restores the radiative balance, at a slightly lower or higher T. So there is negative feedback that makes the climate system stable, in the sense that it does not run away.
The confusion comes from the fact that there are knock-on effects, such as changes in water vapor and clouds. Some people reserve the term “feedback” for those additional effects, excluding the primary Planck response. If you use that terminology, then a “positive feedback” means an overall response that is less stabilizing than the Planck response alone and a “negative feedback” means more stabilizing than the Planck response. Personally, I find that very confusing.
Congrats to Willis for knowing when to back down. The difference between positive and negative feedback isn’t trivial semantics; it’s the difference between systems that are inherently unstable and inherently self-stabilizing. Big, huge, important difference.
If you’re interested in a discussion of terms I suggest Otto Mayer, “The Origins of Feedback Control”, MIT Press, 1970. He opens with a discusssion of the feedback theory of Watt’s centrifugal governor. Both feedback and governor are in one sentence since there’s no real distinction between them.
His definition of feedback is
1. The purpose of a feedback control system is to carry out commands; the system maintains the controlled variable equal to the command signal in spite of external disturbances.
2. The system operates as a closed loop with negative feedback.
3. The system includes a sensing element and a comparator, at least one of which can be distinguished as a physically separate element.
In terms of the earth’s climate the input is obviously sunlight but exactly what is the output? Temperatures, clouds, precipitation, wind, local or global values,…, exactly what?
The “feedbacks” are obviously albedo and radiation to space since they subtract from the net solar input of energy to the oceans and atmosphere. The top of the atmosphere might be considered the comparator since the “difference” of input and output energy occurs there. But in a standard feedback loop the difference is with respect to the set point, the desired value of a variable, e.g., the speed. What’s that here? Is there some desired net energy input to the climate system with the interals just natural fluctuations?
The use of feedback terminology in climate science is a complete misuse of the concept. What they actually have is a system of coupled nonlinear partial differential equations with nonlinear couplings. Good luck with solving that.
Perhaps then you’d appreciate this cartoon.
http://www.maxphoton.com/straight-talk-on-art/
paullinsay August 2, 2015 at 1:05 pm
Thank, Paul, but “No real distinction between them”??? The governor is what he is calling the “feedback control system”. A “feedback control system” is most assuredly different from “feedback”. A governor is a control system which USES feedback, which means the control system and the feedback are not the same.
The “comparator” is provided by the tropical cloud system, which forms clouds when the local temperature (more properly ∆T, but in any location that averages out) exceeds a certain value. As a result, more clouds form when it is warm, and less clouds form when it is cool, and the “set point” is the local temperature at which the clouds form. See my post “The Thermostat Hypothesis” for a discussion of the issues.
Thanks,
w.
I say tomato, you say tamato, … ;-). It should be Mayr.
I think that your mechanism, which I’m not disputing, is subsumed in albedo and radiation as a combination of both, though black cloud tops absorb sunlight rather than reflect it.
The real issue is that the climate system is too complicated to be described by a simple one variable “feedback” system. There is no single control knob in CO2 that adjusts the global temperature, which is a meaningless unphysical concept anyway.
Nobody has mentioned what I believe is the real problem with Willis’ example yet so here’s my take on it. I’m also puzzled as to why nobody else has mentioned it yet because I know there are some experts in control system theory who watch this blog. Perhaps others commented on this above but I missed it?
SYSTEM IDENTIFICATION.
Go read up on the topic of system identification in control system theory as well as what the “state” of a control system is. For example, here:
https://en.wikipedia.org/wiki/System_identification
or
http://controltheory.org/
and many other places on the internet as well as any undergradutate text in control theory.
In Willis’ example, the system input is the desired speed. System output is the actual speed. The “plant” in this case consists of the engine computer, engine, transmission and speedometer. The input to the plant is a desired throttle setting — which is determined by the cruise control based on the difference between desired and actual speeds (that’s the feedback part). Fuel flow is an internal system variable that is indirectly driven by the engine computer.
Target Speed ==> Difference ==> Throttle ==> Engine Computer ==> Fuel Flow ==> Car ==> Speedometer
^ |
|———————————————–<– Feedback —-<———————————–|
When you look at fuel flow, it does not make a lot of sense in the example because you have started with the idea that fuel flow directly determines speed — and it does not. You have not correctly identified/modeled the system. Instead, if you make a proper model of the system:
1) Plant takes a throttle setting input.
2) Plant output is actual speed.
3) Speed is compared to desired speed to compute a new throttle setting (feedback).
Then it all makes sense. Now you can analyze the relationship between fuel flow and speed if you realize there is some internal transfer function that relates speed to fuel flow. This transfer function includes the concept that the car has a "state" which among other important variables includes the current speed or kinetic energy. Failing to include kinetic energy as part of the "state" will yield an inaccurate model of the plant and you are back to a state of confusion.
Here's a grossly oversmplified example of an equivalent analogy in the climate system. Say that temperature is what's being regulated — that means there's a desired temperature and an actual temperature. For the climate, desired temperature would be an inherent part of the overall system, not something that could necessarily be changed at will. The "plant" here would take water vapor as an input and through various processes yield an actual climate temperature. The feedback system would vary the amount of water vapor as a function of the difference between actual and desired temperatures. So, to translate Willis' example to this, it would be like trying to analyze temperature as a function of water vapor. What you would wind up with is how the climate's temperature responds to changes in water vapor. However if the climate system is not properly identified (e.g. not all of the input variables have been discovered), then your analysis will fail.
In summary, I don't know if the difference between a governor and a control system (whatever that might be as I'm confused on the point) is relevant here. Proper system modeling seems more relevant to me.
Another attempt a the little pictorial:
Target Speed ==> Difference ==> Throttle ==> EC ==> Fuel Flow ==> Car ==> Speed
^ |
| |
|———————————————–<– Feedback —-<—————–|
wxobserver August 2, 2015 at 5:40 pm
Fuel flow definitely determines speed … when cruise control is not turned on. Which was the condition in the first example, and which was why I mentioned it.
When the governor is engaged, causation reverses, and speed determines fuel flow.
Are you disputing either of those statements?
I ask because I read your links on “System Identification” and I don’t see anywhere that I misidentified any systems. So where is it that my example is incorrect?
w.
Thanks for the clarification Willis. Now I see that some of this is about semantics and not physics/control systems.
I was looking at this from the point of view of the open loop plant (cruise control off). From that perspective, fuel flow is a determining factor in speed but does not actually set the speed. Your example of a headwind illustrates this point.
Once the control loop is closed (cruise control is on), then you can still analyze behavior as before but only IF you still use the same definition of “plant”. In other words, the speed will still respond to fuel flow as before, but you can no longer directly set fuel flow as the cruise control is doing that.
Here’s where the semantics comes in. Once the loop is closed, it’s all circular. You can define any point in the loop as an “input” or “forcing” and any other point as an output. That’s what you’ve done in your example. I’ll try to say this again in different words.
Input is throttle setting, output is speed and fuel flow is an intermediate value set by the engine computer. From my viewpoint this is true whether the cruise control is on or not. With cruise off, it is my foot that sets the throttle (at my whim). With cruise on, the throttle is set by feedback from the speed. In both cases, the same relationship exists between fuel flow and resulting speed.
If I’m correctly understanding your viewpoint (correct me if I’m wrong), with cruise on you have changed the definition of the system input to be the speed and the system output is now fuel flow. You are free to do this of course because the loop is closed and you may pick your inputs and outputs. Realize though that you have changed the definition of “plant” in doing so.
For a different twist, you could define the input to be fuel flow and the output is throttle setting. Because the loop is closed, changes in fuel flow eventually change the speed, which changes the throttle setting requested by cruise control. You wound indeed find a relationship here but it would be quite odd and hard to understand.
To recap, I think what you’ve done is to define the “plant” as fuel flow ==> speed when cruise controls is off. When cruise is on, you’ve changed the “plant” to be speed ==> fuel flow. The second plant definition of course does not work when cruise is off because there is no relationship between speed and fuel flow in that case (well, there is but it’s the human driver acting as feedback in that case).
I’ve tried to show that you can pick any two points in a closed loop system and declare them to be input/output pairs but in doing so you are really changing the definition of the “plant”. As a result, you wind up with an apples to oranges comparison. I hope this makes sense…?
After thinking about this a bit more I realized another good way to look at this. Yes, causation is reversed, but you are the one who reversed it by redefining what the inputs and outputs are. The relationship between how fuel flow affects vehicle speed is still unchanged even with cruise on.
Positive Feedback
aka: The Star Spangled Banner
by Jimi Hendrix (the Ungoverned)
Gosh, … while we’re here …
Voodoo Child
by Jimi Hendrix
Considering the ocean and clouds in the equatorial Pacific, which is the control variable? As Eschenbach notes, you get a different slope for each:
Current science has CO2 as the control variable but if it’s a minor control other factors could cancel some or all of it out. Temperature seems more likely to the be control variable signaling each molecule to do a certain thing. Water to assume a certain form, rise or sink. I would not assign every last thing to temperature as all the other factors do impact it but it’s probably the predominate factor. That gives us a controlled stable situation as in the Temp/Cloud graph. Eschenbach’s figure 3 is informative. The blue line is the climate scientists view of things, and the red line is the pause that may have been caused by the governor that is not part of their understandings. The slope changed signs at the intersection and the CO2 is not doing what they expected. The blue line is on the way to equilibrium and once there the slope changes perhaps revealing CO2s true effect. CO2 is supposed to work longer but at the pause equilibrium, temperature showed its power. Theory sort of says CO2 can hijack temperature’s power and become the control variable. It could just be a minor player.
A little amused by this “fuel flow” business, Willis. Surely a cruise control is maintaining constant road speed regardless of engine load. You might argue that’s just splitting hairs but some of us have manual shift cars with cruise that allow you to shift gear (up or down) without resetting. You still need to dip the clutch though 🙂
A simpler example is a Watt type centrifugal governor on a steam engine or lawn mower. Constant RPM regardless of load.
A governor mechanism is inverse feedback. In the case of a cruise control, as road speed tends to increase above a set point, output power is decreased and visa-versa. And crucially, it’s proportional.
Don’t much like this positive/negative terminology, it just confuses the issue. Maybe that’s at the root of your ruminations; terms like forcing imply positive or negative exclusively and do not speak of proportionality.
Maybe saying a governor is an example of inverse proportional feedback is the way to go.
Feedback for either a single control loop of an entire dynamic system can be defined as follows:
1) If one unit of forcing results in one unit of response, feedback is “zero”.
2) If one unit of forcing results in more than one unit of response, feedback is “positive”
3) If one unit of forcing results in less than one unit of response, feedback is “negative”
A speed governor typically consists of several control loops. Some of those control loops employ positive feedback and some employ negative feedback. The “system” compares actual speed with desired speed and produces an output designed to produce one of three effects:
1) Hold existing speed.
2) Increase speed.
3) Reduce speed.
The amount of “system feedback” in a speed governing system is called “speed droop” in the speed governor ‘bidness. It may be varied from “zero” to various negative feedback values, the units of which are given in “percent”. In no case is speed governor “system” feedback “positive” (except for very brief transitions). Such a “positive feedback” system would go unstable at the first variation in “actual speed”.
A speed governor set with “zero droop” is called an “open governor” in the electric utility business. That governor will do whatever it takes (within the mechanical limits of the controlled prime mover) to match actual speed with desired speed. With a speed governor set at 5% droop, a fully loaded electric generator (prime mover at full throttle) running at desired speed will settle out at 105% of desired speed when that load is suddenly removed (load rejection).
AGW theory postulates that one unit of CO2 forcing results in more than one unit of climate system temperature response. That would be, by definition, “positive system feedback” to CO2 forcing. AGW theory makes that preposterous claim by postulating that attendant variations other atmospheric greenhouse constituents, primarily water vapor, cause CO2 forcing to be multiplied.
The question I’ve never been able to answer is the following:
If climate system feedback to temperature changes induced by atmospheric CO2 variations were positive, how could we all still be walking the face of the earth?
Don’t know where you got your definitions.
Most people would consider your # 1 to be more a definition of ” gain ” rather than feedback, or more generally of ” Transfer Function “.
I got my definitions in the course of designing control systems for some 40 years, including systems involving speed-governors. You confuse signal input versus signal output (gain) with system forcing and response functions. I do take your point, however, in that the speed-governor with a set-point is not the best analogy for what is meant by positive and negative system feedback in general. Let me try another approach using the definitions I laid out:
1) If a straightforward greenhouse calculation showed that a 5% increase in some gas constituent should yield a 1% rise in temperature and if measured results confirmed that 1% rise, then the conclusion would be that system feedback to temperature forcing by that gas was “zero”.
2) If temperature rose more than 1%, the conclusion would be that system feedback was positive.
3) If temperature rose less than the calculated 1%, system feedback was “negative”.
I’ll say again, the heart of AGW theory and the source of all the hysteria is the erroneous claim that atmospheric system feedback to an increase in CO2 greenhouse temperature forcing is “positive”.
Recall that Olavi Karner has demonstrated negative feedback in several atmospheric data series.
http://www.aai.ee/~olavi/
For instance:
http://www.aai.ee/~olavi/ISPM-app4f.pdf
http://www.aai.ee/~olavi/EE2007-ok.pdf
http://www.aai.ee/~olavi/cejpokfin.pdf
http://www.aai.ee/~olavi/2001JD002024u.pdf
I realize that this stuff is highly non-comprehensible for for non-experts on the matter, but with the large amount of experts on feedback here, it’s sad that all this never took off to refute the positive feedback alarmists.
“In the tropics, when it is warm, clouds form earlier and reflect away the sun to cut down the solar forcing. And when the tropics are cool, the clouds form later or not at all, which greatly increases the solar forcing.
What that means is that a change in temperature causes a change in forcing.”
This depends, as so often does, on definitions.
I suggest that the forcing in this case above is the suns input to earth, and does not change in this example.
Clouds form at a height and time dependant upon temperature, humidity, air pressure and other variables I am unaware of.
They also rain on us for the same reasons as above.
We know the genius of water, ice and water vapour, how is is able to transport a large amount of energy, and how very small changes in temperature/height/humidity causes clear sky, cloudy sky and rainy skies.
I would suggest that physics of clouds will hold the majority of information of how our climate works.
Could we ever understand ‘Global climate’? I suspect not in the near future.
Hi Willis,
As others have pointed out, a governor has a set point which is independtly determined. A simple negative feedback system hs no such set point. There are strong negative feedbacks in the climate system, with T^4 Stefan-Boltzman being the most obvious. The response of tropical cloud cover to changes in sea surface temperature is another. The formation of thunder storms or dust devils may be others. We know that negative feedbacks, in total, dominate the system, because the Earth’s temperature does not “run away”. But none of those negative feedbacks is a governor; they have no set point. They are temperature (and humidity) based atmospheric responses to local conditions.
If solar output were to change by a few percent (up or down) the Earth would either warm or cool on average, it would not remain at the same temperature. There would be a multitude of non-linear responses, of course, but the temperature would go in the expected direction: more solar energy, higher temperature, less solar energy, lower temperature. It is why we see seasonal temperature changes. The change in average temperature for a specific change in solar output would tell us the net sensitivity to external solar forcing. It seems to me that is pretty clear. Do you agree?
If we were to dump a huge quantity of sulfur hexafluoride (a strong and very long lived infrared absorber) into the atmosphere, then the loss of infrared energy from the Earth’s surface would be somewhat reduced, so we could reasonably expect that this would increase the average surface temperature. Just as with a change in solar intensity, there would be a multitude of non-linear responses, but the change in surface temperature, on average, would tell us the sensitivity of Earth to ‘forcing’ by sulfur hexafluoride.
The basic argument in climate science is about the relative importance of negative and positive feedbacks, and the resulting net sensitivity to ‘forcing’ by CO2, methane, etc, not about whether Earth has a set point temperature; the Geological record shows us there is no set point. I think there is strong evidence that the overall sensitivity to infrared forcing is relatively weak, and almost certainly far lower than GCMs suggest. But at the same time, I think it is incorrect (or at least very, very confused) to describe Earth’s climate as a governed system; it is just a complicated system with multiple negative and positive feedbacks.
Steve Fitzpatrick August 3, 2015 at 5:38 am
Thanks, Steve. Looking at the TAO buoy data, it’s clear that the strength of the climate response is greater in warmer conditions than in cooler. This argues that whether we understand how it happens, there is a set point. Otherwise, we’d see the same response at all temperatures.
SOURCE: Daily Albedo Cycle
You continue:
From our knowledge of the life cycles of stars, it is generally agreed that the luminosity of the sun has increased by about 5% over the last half billion years. This is the time that life has been popular on the planet.
However, we have absolutely no evidence of any such change in the temperature of the planet. Instead, over the time that the sun has warmed by about 5%, instead of warming by a percent or so as you suggest, the earth seems to have cooled by a percent or so.
As a result, I fear that your claim flies in the face of the evidence.
Next, if there is no “set point”, then why does the temperature return to the same level after something like a volcanic eruption?
And finally, if there is no “set point”, then why did the global temperature vary by only ± 0.3°C over the entire 20th century?
Best regards,
w.
Steve Fitzpatrick:
While agreeing that there’s no “set-point temperature,” I hasten to point out that there are no genuine “feedbacks,” negative or positive, operating in the climate system. “Climate science” has egregiously misappropriated that word from rigorous system analysis to mean a mélange of complicated responses and effects that have nothing to do with the output signal of the system being fed back as input without any delay. Even more egregiously, the frequency-dependent transfer function of linear feedback systems has been oversimplified to the static gain 1 / (1 +/ f) and applied to a system that is demonstrably nonlinear. As the widely scattered empirical results show, the attendant estimates of static “climate sensitivity” are physically quite meaningless.
Moderator:
Please correct the formula to read (1 +/- f) in the denominator.
I think there is more then one climate governor or regulator. One is terrestrial bound as Willis has suggested(tropical governor for example) , which is subject to terrestrial changes such as land/ocean arrangements, volcanic activity, ice dynamics and clouds to name a few.
The other climate regulator I think is extra -terrestrial due to Milankovitch Cycles, which is subject to solar variability, extra terrestrial impacts to name a few.
The dynamics of how all this relates then determines how effective the climate governors are which can vary from being able to maintain a stable climate lasting 1000’s of years to one that can change from an inter-glacial state to a glacial state over a rather short time period, but nevertheless will still be range bound to a degree in that the climate will not keep going in a warm/cold direction indefinitely.
The big unknown are climatic thresholds due to all of the above which are very likely to occur due to the fact the climatic system of the earth is chaotic, random and non linear and when it changes it does so in step fashion rather then in a slow gradual fashion.
All of that is evidence of thresholds in the climatic system which is not being addresses enough in my opinion.
Why? Probably because the essence of it is so elusive and beyond the capabilities of climate science to address in any real proper comprehensive manner. What confirms that statement is the fact there is still no real explanation as to why the climate at times changes abruptly and goes from a glacial to inter- glacial state, while at other times is relatively stable.
Till this is solve we will remain in climatic darkness as far as being able to predict when /how the climate may change going forward.
I expect that we will know the cause of the next step change due to all of the monitoring now in place. Lets hope the data is not massaged too much to prevent a good analysis…..
This time I only made it to this sentence, “Well, yes, that’s true, and it is also true that a human being is “just a bag full of organic chemicals, minerals, and water.“ Like poison ivy, my resistance is fades over time.
The terms you are discussing have mathematical definitions. If you don’t understand the math then you can argue by analogy all day long and you still will not understand. Do us all a favor and learn the math.
Dinostratus August 3, 2015 at 4:46 pm Edit
Dino, it appears that there is some unknown “mathematical definition” which you think I’ve not agreed with, or don’t know about, or something.
But since all you quoted was my statement about humans, along with a bunch of handwaving and insults, I don’t have a clue which “mathematical definition” you might be talking about, what you might think said definition is, or how I may have transgressed in whatever I said.
As I have requested many, many times,
QUOTE THE WORDS YOU DISAGREE WITH AND TELL US WHY YOU DISAGREE WITH THEM!
Is that so hard to do? Well, I guess if you don’t actually have a point and your intention is to just throw mud at me, it might be …
w.
I disagree with the part where you use words and arguments by analogy rather than mathematical definitions. Write the mathematical definitions. If you can’t, just say so.
Lots of people don’t know control theory, even simple linear control theory. There’s nothing shameful about that.
Dinostratus August 4, 2015 at 8:25 pm Edit
That’s your answer to a request for a direct quote?
Sorry, Dino, but I was serious. As I said:
As you have not done that, I fear I still don’t know what you are referring to.
In any case, if you wish to provide the math, please do so, as I have no interest in doing so. Me, I’m interested in the concepts. I’m not trying to explain how to do the math. I’m trying to explain the concepts to a lay audience. And the problem with writing for a lay audience is that as a rule of thumb, every number I put in the post loses me one reader, and every equation I put in there loses me ten readers. So I generally err on the side of explanation rather than math.
So if you wish to bring up the math, please go ahead, and I’m happy to discuss it. However, I don’t understand what the math would show that is different than my explanation … but hey, if you can disprove my work mathematically I’m more than interested.
Up to you,
w.