Guest Post by Willis Eschenbach
Stuck in the Nadi Airport in Fiji for the day, waiting for my flight to Brisbane. So … I read. I found a most clearly written and fascinating article in Quanta Magazine entitled “‘Next-Level’ Chaos Traces the True Limit of Predictability“.
Basically, it seems to be Godel’s Theorem meet the Turing Halting problem.
And it seems to me that their conclusions mean that future climate states are indeed fundamentally unpredictable, but NOT for the reason usually posited, that we cannot specify the starting conditions. It’s been referred to as the “butterfly problem”, that in a chaotic system a tiny change in starting conditions can lead to huge changes in outcomes.
That issue you can kinda get around by doing a bunch of runs with different starting conditions, and by using an ensemble of a whole bunch of models … although, of course, you have no way of knowing if the starting conditions or the ensemble members have fully explored the parameter space.
But this kind of unpredictability is a brand new one. They’ve shown that even if we knew the position and motion of all the molecules in the ocean, the future state of ocean currents still isn’t predictable.
Me, I’m curious about other folks’ take on what this finding means to climate science.
Best to all from F1J1,
w.
… reminds me of an old joke. Werner Heisenberg is speeding in his car when he’s pulled over by a policeman, who asks him “Do you know how fast you were going?”
“No,” said Werner, “but I know where I was!”
PS—I’m at my mate’s house in Oz, and it’s my first time going online with SpaceX Starlink. Here’s the Ookla Speedtest results … not exactly blazing fast, but it works very well.
The San Jose (California) Alum Rock School District convinced me that I am bad at math, or at least gave me a phobia. Given that, knowing climate fits chaotic math models would imply doing predictions is a wicked problem. Just how wicked is the question.
an impossible problem- not worth wasting time on
Agreed, Joseph, up to a point.
Until emphasis lessens on greenhouse gases, until the whole earth system is measured and researched, there is dangerous harm in making policies like “net zero carbon” because the full picture is not understood.
Two simple, critical examples needing better understanding are temperature effects related clouds and clarification of precisely what has caused measured warming since the little ice age – clarifications that require both elevation of mechanisms that best explain observations and less emphasis on mechanisms that do not.
So maybe more research is needed but it needs to focus on gaps in knowledge. Geoff S
Good point, Geoff.
The tragedy of climate modeling is that it has torpedoed climate physics for 40 years.
The climate models keep the chaotic effect small making ensembles a reasonable approximation method. But in reality the chaotic effect in weather, hence in average weather which we call climate, is very large. Weather and climate are intrinsically unpredictable. The math us clear on this.
Wojick ==> “Weather and climate are intrinsically unpredictable. The math is clear on this.”
Yes, absolutely correct. Both the Math and the Reality are clear on this.
IPCC TAR Chapter 14 Paragraph 14.2.2.2 Balancing the need for finer scales
and the need for ensembles
In climate research and modelling, we should recognise that we are
dealing with a coupled non-linear chaotic system, and therefore that
the long-term prediction of future climate states is not possible.
And there’s this illustration Kip Hansen put up here on WUWT sometime ago:
Steve,
Yes, Kip shows a lovely example with double pendulums.
Who would seek better understanding by taking an average of the positions over time?
Not me. Not you?
Yet, that is the “solution” used by climate modellers. Do many runs, then average them. These people are supposed to be mathematicians?
Geoff S
Sort of like the average number in the New York City telephone book
And how about using one or two temperatures a day to determine the “average” daily temperature and them averaging them again over months or even years to determine the temperature of Earth?
Especially when temperature doesn’t even adequately measure all the factors that enter into “climate”.
The problem is that they’re mathematicians, Geoff, and not physicists.
This is a wonderfully clear visual example of Pat Frank’s work on the propagation of errors in climate models. If a system as simple as a double pendulum is unpredictable, how much worse is it for the Earth’s climate?
I was waiting to groan at the dad-joke punch line if you wrote “Nein”, Willis.
🙂
All dry there now I hope, Willis.
Yes, it’s lovely today.
w.
Oz is a big place. My district is basically in drought. No real rain in many, many months. If you feel the urge, please post down all the excess to NE Vicdanistan. We’ll put it to use.
I am simultaneously proud, amused, and chagrined that I got all the references. I was going to replay my LPs of “The Eternal Golden Braid” but it appears my record player is broken.
I thought you designed your record player so it wouldn’t be broken by the EGB LP.
This time for sure! 😉
Recently asked the online AI/ML := AL systems to produce coding for some of the systems, like MU, that are subjects in GEB. Fun, and very much easier than either mind/hand or roll-your-own code.
I’ve been wondering for the last year about a question that I don’t have the resources to answer, and at 77 am not likely to acquire the necessary knowledge. To state it briefly, if a butterfly’s wings can beget a hurricane [chaos theory], it seems inherently true that of the effects of windmills bleeding energy out of the atmosphere could result in changing weather/climate. I would only be surprised if the answer is No. It seems obvious that we currently affect the system by some measurable amounts, but a future glut of windmills everywhere will be much more impactful. Is it possible that the question is too hard to quantify, or maybe the answer is too embarrassing to the climate alarmists that gatekeep the science?
The butterfly can only make the chaotic system do something it was already able to do. There would have to be a hurricane waiting for one more windmill to set it off.
The windmill intendeds to convert wind energy to electricity. That would be the reverse of a butterfly, where the butterfly intends to convert chemical energy from food into mechanical energy of wings, iow converting chemical energy into wind energy.
The point of contrasting windmill and butterfly is to say that it might not be possible for “one more windmill” to add anything to wind energy. The trigger would have to involve subtracting energy.
I ascribed intention to the windmill and the butterfly to make the sentences rhyme in structure.
I don’t think Mr. Capron is suggesting that a windmill can trigger a hurricane, rather, I think his question relates to the removal of energy from the wind and what that might do to the weather. Whether that has any affect on the climate is another question.
It has an effect there are countless published papers on it. US midwest because of the number of wind turbines has been extensively studied starting back in 2021
https://www.nature.com/articles/s41598-021-02089-2
I was wondering that if small things can spiral out of control, then what about big thing.
Unless I’m missing something fundamental, wind turbines have to cause some kind of climate change, however minuscule it might be. They’re also man-made, if that word hasn’t been banned yet. I wonder how much CC, and is it more or less than CO2 going from 320 – 640ppm. Wind turbines have been around a long time (my father worked on the gearing design for the ones on the Altamont pass at least 50 years ago).
https://www.scientificamerican.com/article/wind-power-found-to-affect-local-climate/
Highlights
https://www.scientificamerican.com/article/wind-power-found-to-affect-local-climate/
https://www.usgs.gov/news/national-news-release/wind-turbine-wakes-can-impact-downwind-vegetation
Wow, thanks.
Wind turbines via the electrical interconnect transport atmospheric energy (in the form of electricity that ultimately becomes thermal energy) to remote locations. This redistribution of atmospheric energy is an effect.
“The butterfly can only make the chaotic system do something it was already able to do”
And of course, the atmosphere can create hurricanes, or not.
Lorenz titled his presentation “Predictability: Does the Flap of a Butterfly’s Wings in Brazil Set Off a Tornado in Texas.”, and later said he was being literal.
He could have just as easily replaced the butterfly’s wing flap energy with an equivalent amount of energy from a wind turbine. That’s chaos for you.
I think it was Pielke Sr. who poo-pooed the idea that actual butterflies can have any effect on anything whatsoever regarding wind flow or currents. Any turbulence caused by the flapping of any bf wings gets quickly drowned out by even a gentle breeze.
And that makes perfect sense to me.
It isn’t a question of the relative power of potential influences of the environment. The fundamental assumption of the BF scenario is that even something very weak can trigger a different pathway for everything — not unlike a local puff of wind at a shooter’s location causing a bullet to deviate from the aiming point and consequently miss the target completely. Although, in that analogy I think it would take more than a butterfly to make a noticeable difference in the impact point.
I disagree. The relative power is everything.
Jeff, so you disagree with Lorenz and Feynman?
Might I ask if you can link to reproducible experiments demonstrating both were wrong?
As Einstein said “No amount of experimentation can ever prove me right; a single experiment can prove me wrong.”
A single experiment will do.
The point is that in a chaotic system, very small changes in initial conditions result in very large divergences as time goes by. In truth, it’s comparing two scenarios, otherwise absolutely identical but for the wing-flapping of one butterfly.
Similarly, multi-body orbital dynamics are chaotic. A while back, scientists ran a simulation of the solar system that covered several million years. Then they ran it again, changing one bit in Pluto’s starting location. Several planets ended up on the other side of the Sun.
pblase,
Yes, people steadfastly refuse to accept that there is no minimum change to initial conditions which may result in completely unpredictable outcomes.
Lorenz said he was being literal. Feynman pointed out that the uncertainty of simultaneously establishing the momentum and position of a photon leads to the same conclusion. Maybe Pielke Sr. knows more than Feynman and Lorenz did, but I haven’t seen his experiments demonstrating they were wrong. Maybe he was speculating?
If you’ve got 5mph winds, and a butterfly flaps its wings, how will that in any way affect the 5mph winds?
Jeff, maybe you miss the point of what both Lorenz, Feynman and many others have pointed out. Whether you use chaos theory or current physics, nobody can say for certain whether a butterfly flapping its wings in Brazil has caused your “wind” to be 5mph, to be coming from a particular direction, or even to exist all.
That’s the thrust of what both Lorenz and Feynman wrote. If you can quote either of them even intimating otherwise, I would be surprised, but I would accept reality.
The point Jeff Alberts is making (I think) is that:
And
Are the same starting point. It’s too small a difference to set things going down a different path, from that point.
And that’s a reasonable argument.
However, I disagree with Jeff Alberts. Just because a difference is too small for us to conceivably measure does not mean it’s too small for any measurement system to be unable to detect it, ever.
A chaotic system the size of the planet, extending over years in time, is very sensitive to differences.
In principle, chaos theory means that there is no quantising of starting conditions. This is, again, too extreme. I disagree with it.
But the quantising must be on a quantum (Heisenberg) scale, not merely relative to macro-properties like windspeed. There is a phase change at that scale – it’s not linear.
Jeff ==> There is a lot of nonsense going about making a fake controversy about “The Butterfly” — the literalists vs. those who understand what Lorenz was actually saying.
Nonlinear chaotic systems, especially two or more coupled nonlinear chaotic systems, are highly sensitive to changes in initial conditions, making long term predictability impossible.
That’s the real meaning of The Butterfly Effect.
You probably already read my Chaos Series here.
William ==> Windmills take energy from the atmosphere. That is the change. That change is part of the “initial conditions” first for local weather processes, and those local weather processes add up to local climate and all those local climates add up to planetary climate.
All changes above a certain level (unknown but knowable – some changes disappear due to physics) eventually cause “climate change” — which is continual and unpredictable.
Read my Chaos Series here at WUWT.
“Here’s the Ookla Speedtest results … not exactly blazing, but it works very well.”
Upgraded from impossible at any price in 1925.
I wonder what will seem ho hum in 3025.
Some generation will be the last generation to die.
SpaceX is a much bigger winner for Musk than Tesla. It is really three companies. A near commercial monopoly on space launch because of the reusability of Falcon 9 boosters. Starlink enabled by Falcon 9. And NASA funded Starship.
Reusability of the first stage is more a PR stint than decisive advantage. Monopoly on commercial launches is subsidized indirectly by military government contracts.
NASA funded Starship is total waste of taxpayers money. But Musk is good at this, he is a genius of converting government money in his image of genius.
We have Starlink on our boat in the Whitsundays (offshore QLD). Typically 5 times faster than Willis recorded
Feynman addressed a climate relevant form of this in volume 2 of his famous Caltech ‘Lectures on Physics’ in chapters 40 “The flow of dry water” and 41 “The flow of wet water’. He was able to mathematically derive precisely under all conditions the flow of dry water. But NOT wet water when viscosity was added. This was when his mind turned to the Navier-Stokes equations (flow of wet water) where he did several years of experiments illustrated in chapter 41.
Let me state his first few and last sentences of a concluding paragraph to chapter 41. “We have written the equations of water flow. From experiment we find a set of concepts and approximations to use to discuss the solution (that we cannot calculate precisely)—vortex streets, turbulent wakes, boundary layers. When we have similar equations in a less familiar situation (uses sun surface as example), we try to solve them in a primitive, halting, and confused way….All of these are really in the equations; we just haven’t found the way to get them out.”
(Sidebar. ‘Primitive, halting, and confused’ is an apt description of climate models.)
Undecidability just means we never will. A big generalization of Goedel’s 1931 incompleteness theorem in number theory.
Another uncertainty is added in the case of radiation transport. If a CO2 molecule gets excited by capturing an IR photon, there is no way to predict whether that energy will be re-emitted as a photon or transferred to another molecule in a collision.
Statistically you can.
Now roll your dice and get the answer…….
“transferred to another molecule” by means of one of that molecule’s electrons absorbing a photon emitted by an electron of the CO2 molecule. No actual “collisions”.
By the way, excitation (whether atomic or molecular) requires certain minimum energy inputs per photon – expressed usually in eV – electron volts. In general, the excited entity will rid itself of the excitation energy within nanoseconds, if the source of exciting energy is removed.
It’s a little more complicated than GHE cultists might have you believe.
However, CO2 (like other matter) simply cools if an external energy source is absent. If it’s gaseous, there’s certainly plenty of energy about!
I downloaded the article by Prof. Moore, and was struck by the remarkable fact that a brilliant person, Prof. Moore, understands the insolubility of certain problems; yet is led to the certainty that decarbonization of civilization is critical to its survival, based on an insoluble problem’s “solution”.
They have EYES but will not see.
Maybe they close their eyes and see too much.
For what it’s worth. I think every problem has a solution. If a solution cannot be found It does not mean that it does not exist. One may simply have insufficient knowledge or imagination to see it. If there is truly no solution then the problem is simply a fact mischaracterised as a problem.
The real issue is that some problems only have probability predictions of what will occur. In other words you will never get a final answer of “x = a single value”. You’ll get something like, x1 at 5%, x2 at 5%, x3 at 5%, and maybe even a continuous probability function where there is an infinite number of possible solutions.
I once saw a demo of chaos using drops from a leaking faucet. The nearest you could predict another drop was to use a pretty broad range of time. No matter how much data you could gather would reduce that range so you couldn’t predict the next drop except by pure chance. My thought at the time was Heisenberg’s Uncertainty Principle.
Same thing applies to forecasting the “climate”. We don’t even measure the necessary factors to begin to make a projection, let alone a prediction. It is why the IPCC and AGW advocates refuse to call their answers predictions. Yet a projection requires a range that surrounds it. When was the last time you saw an ECS quoted along with a possible range?
Jim, here’s an interesting way of wasting time – measure the interval between drops for a while. Now, plot each pair as Cartesian x/y coordinates. Watch chaos emerge.
Or don’t bother. It seems that I haven’t kept up to date with some things. Universities now have many different experiments demonstrating chaos from a dripping faucet.
Bugger.
See also
‘ultra chaos’,
‘clean numerical solution’ (CNS),
author Shi-jun Liao, sometimes as Shijun Liao, sometimes as S liao, sometimes as L Shijun
Google Scholar search:
https://scholar.google.com/scholar?as_q=&as_epq=ultra+chaos&as_oq=&as_eq=&as_occt=title&as_sauthors=&as_publication=&as_ylo=&as_yhi=&hl=en&as_sdt=0%2C33
“future climate states are indeed fundamentally unpredictable”
true! determinism fails
Einstein said “God doesn’t play dice”- but it seems that he does. It’s all just too complicated. IMHO, the cosmos is infinitely large, infinitely complicated and eternal. Believing that trends will continue is simple minded.
The fundamental problem here is understanding that the most probable outcome does *not* always come to pass. Sometimes the least probable outcome happens.
We are always told that the climate models are deterministic following a strictly causal chain based on initial conditions, i.e. everything averages out and the most probable outcome always wins out. The universe just doesn’t “average” all the time.
For some reason, the title of this post is in George Carlin’s voice in my head, as is the following retort: “Hope so!”
since there is NO RELIABLE unmassaged data to run any simulation with its just a waste of time to attempt anything like predictions … its the ultimate GIGO … and the input has been garbage forever … you can’t even begin to “tune” the models because the data you are trying to fit your model too is garbage … basically its an intellectual circle j**k …
A very nice word salad 🙂
100%
It means that if we do NOT eliminate all CO2 from the atmosphere, then SMoD, (i.e. Sweet Meteor of Death,) will wipe us out.
Forecasting the future is the n-body problem.
Intractable with current mathematics.
There are a near infinite number of ways of asking this question but in each the answer remains the same.
It’s possible climate change includes macroscopic Quantum processes. There have been apparently successful attempts to use equations from Quantum Physics to describe rogue waves.
https://physics.aps.org/articles/v12/146
If rogue water waves can be described using quantum physics equations, why not rogue atmospheric waves?
I have no idea if this means something truly quantum is happening, maybe the equations just happen to be a good fit. But it certainly is an eye opener.
From your link –
“Obviously, the instanton model would have to be validated by experiments able to probe these configurations.”
So it’s not a theory at all, just an untested hypothesis. Just another speculation.
As an aside. Mariners reported Rogue Waves for centuries, probably since man first ventured onto the sea, and for almost as long scientists said they were talking nonsense. Just another reason to be sceptical of experts and scientists opinions.
Schrodingers cat shows that quantum descriptions hold for macroscopic and microscopic objects. The key is time.
Nature selects the future similar to an electric spark jumping a gap. This is the wavefunction collapse. What we experience as “now”. Prior to collapse the future was a field potential between the wires. The wavefunction.
At the microscopic level the spark appears random. At the macroscopic level the sparks average out giving the illusion of predictability.
For example the most common roll of a pair of dice is 7. This gives us the illusion the dice are predictable at a different scale.
“They’ve shown that even if we knew the position and motion of all the molecules in the ocean, the future state of ocean currents still isn’t predictable.”
I haven’t read the article, but it sounds like someone has “reinvented” chaos theory.
You don’t even need chaos. As Feynman pointed out, given the uncertainty principle, you arrive at precisely the same conclusion, using ordinary physics. The uncertainty principle just states that it is impossible to simultaneously establish the position and momentum of a particle. Follow that through, and you quickly discover why you cannot predict the future state of the atmosphere, the ocean, or for that matter crustal movements.
The future is unknowable to any greater precision than that given by a naive predictor (say a smart 12 year old).
> They’ve shown that even if we knew the position and motion of all the molecules in the ocean, the future state of ocean currents still isn’t predictable.
That goes without saying.
“the ocean” is not a closed system, It has an atmosphere on one side of it and a planet’s surface on the other. Then there are all life forms moving in it. All are adding/removing energy in various ways at various points.
My Skynet (sic) is going a bit faster than that, though I’m in a rural area, so not much competition for bandwidth. Generally get 250-300Mbps down and 15-25Mbps up. Heavy cloud cover can reduce this though.
Willis,
Your old Heisenberg joke is wrong, unfortunately.
Heisenberg noted that uncertainty limits calculation of values of some pairs of related parameters. The joke implies certainty instead of uncertainty with “I know where I was”. Correctly should be like “I do not know quite where I was”.
(Please pardon pleasant pedantry). Geoff S
Re ordinary chaos this approach is false: “That issue you can kinda get around by doing a bunch of runs with different starting conditions, and by using an ensemble of a whole bunch of models …”
The average of the multiple results is no more likely than any of the individual results. Ensembles work for non-chaotic systems where the initial conditions are not precisely known when the error has a normal distribution. In chaotic systems there is no such distribution.
David, you are correct. It’s a pity that many physicists, let alone “climate scientists” refuse to accept reality, even though the IPCC grudgingly admits that predicting future “climate states” is not possible.
Chaotic systems typically have attractors such that the dynamics is limited to a limited portion of the phase space. This allows a degree of prediction to large-scale quantities such as pressure and temperature. We are not interested in knowing position and velocity of each gas molecule.
“a degree of prediction”
Meaning the actual state is still unknowable. A six sided die allows a degree of prediction of the next roll?
But an attractor is only useful if it is stable. When the attractors are chaotic too, predictions of the next answer is impossible. I have never seen a paper or article describing stable atmospheric attractors that could help resolve future climate states accurately.
Mactoul, that’s not useful, is it? For example, the “limited phase space” for temperature has a lower limit of zero.
So you can confidently “predict” that a series of chaotically generated temperatures will be above absolute zero?
Pressure likewise, from zero (vacuum) to infinity. You may say that the past shows limits, but that’s the point. You can’t predict the future by looking at the past.
Chaotic systems have “strange” attractors.The attractor keeps moving – no equilibrium, no reversion to a mean, no meaningful anomalies.
Chaos may be internally generated, without the need for external input of any sort.
Predict the future of a chaotic system? Go ahead and do it, and I’ll believe you. Don’t try to pass off assumption or speculation as “prediction”, though. Oh, and don’t just assert that such and such “will happen”. The future hasn’t occurred yet – that’s why it’s called the future!
+100
That speed is brilliant compared to our National Broadband network (NBN) in country areas which in a recent upgrade spent $750 million of the taxpayers hard-earned for just 100 new customers. Starlink would have provided pretty good access for areas in our outback for a fraction of the cost used for subsidies to our very remote Australians.
If we knew the position and movement of every particle in the universe, we could always calculate the next step, in which case there would be no free will and we could look into the future…and into the past.
And there would be no anthropogenic global warming because we would be innocent pre-programmed, mindless automatons predetermined to release co2.
“If we knew the position and movement of every particle in the universe, we could always calculate the next step, in which case there would be no free will and we could look into the future…and into the past.”
Actually you can’t. Quantum mechanics is based on probability. The most probable outcome isn’t necessarily the one that always happens. There isn’t any way to calculate with 100% certainty the next role of the universe die. You might be able to calculate the most probable future but you’ll never know for sure if that will come to pass.
Addendum:
“You were doing 100mph.”
“Great! Now I have no idea where I am!”