The artist as climate model expert

Guest essay by Charles Battig, M.D.

Artist_models

There are reams of publications devoted to the mechanics of computer modeling of global climate dynamics, and the predictive results derived therefrom. Mathematicians, experts in chaos theory, fluid dynamics professionals, climatologists, geologists, oceanographers, satellite-data analysts all toil away at constructing a computer model which will accurately embody these sciences in a faithful representation of the global energy interplay known as climate. No one has quite succeeded yet, even with ever faster and more complex computers. The actual observed climate behavior of our planet continues to confound the very best efforts of the very best computer modelers and their models to replicate it.

True climate denial is the real-world denial by mother earth of faulty ersatz computer representations and impersonations.

Science and the arts have different ways of viewing the world, i.e. reality. It is in the field of visual fine arts that an elusive truth of climate dynamics has already been identified and documented. Long ago, an artist portrayed what many in the field of computer modeling of climate have not noticed, or have refused to acknowledge. Yet, his message is often ignored by the computer modeler whose passion for his program blinds him to the truth embodied by this artist. Ardent climate modelers have proclaimed their allegiance to their computer model, rather than to reality.

This icon of visual truth hangs in the Los Angeles County Museum of Art, a few miles away from the University of California at Los Angeles (UCLA). Just recently, a group from UCLA issued a report exploring climate mitigation via a number of modalities. Had members of that report visited the nearby County Museum, perhaps they might have tempered their enthusiasm for the reported findings, all based on studies…studies being shorthand for computer simulations.

Ceci n’est pas une pipe,” or “This is not a pipe,” painted by René Magritte in 1928 shows a perfectly recognizable smoker’s pipe. Suppress that urge to declare that painter to be out of touch with reality. His painting is more in touch with reality than some climate modelers, for example Prof. Chris Folland, Hadley Centre for Climate Prediction and Research who illustrates such affection for his computer generated world that he proclaims: “The data doesn’t matter. We’re not basing our recommendations on the data. We’re basing them on the climate models.”

Such blind love of one’s own creation is immortalized in Greek mythology by the tale of Pygmalion. Falling in love with your own creation makes for interesting philosophical debates when it is a statue; when it is a climate model, it is science gone awry. It is no longer science but a fanciful detachment from reality.

Magritte’s visual message is that a copy or rendition of an object is not the original object itself. David Blakesley and Collin Brooke’s commentary on visual rhetoric notes that:

“Magritte highlights for our consideration the idea that an image of a pipe is not the same thing as the pipe itself (or the letters p-i-p-e). It is a representation of a pipe, once removed from its referent, the object to which it refers. He also forces us to consider our own reaction to the painting by suggesting that our compulsion to call the image a pipe reveals our predisposition to confuse the image with the thing it represents.”

Here then is the lesson offered by Magritte for the global climate modelers. The computer model is not reality; it is just your attempt to replicate reality. In the complex system known as global climate dynamics, the models are not reality on a computer chip. They are scenarios akin to the fabled three blind men describing an elephant. Stephen Few has used this tale as the starting point for his essay on perception and business intelligence. Each modeler may, or may not, capture a small piece of the climate puzzle correctly, but no one has gotten the whole picture reduced to a valid computer representation. The climate system is so complex and chaotic, that computer modelers have to settle for simplified constructs with open-ended assumptions (parameterizations). Like Magritte’s pipe, the final result is even further removed from reality.

Recently, the National Climate Assessment was released and claims to “detail(s) the multitude of ways climate change is already affecting and will increasingly affect the lives of Americans. Institutionalized computer forecasting masquerading as scientific fact does not make it fact.

Extending the reach of Federal control not only over us, but also over future global climate, the EPA has a new slogan: “Thirty per cent less by 2030.”

Perhaps the “30 by 2030” was chosen by a focus group because it has a rhythmical ring to it, or perhaps a computer projected it to have mass appeal.

Expectations of realistic research results and future climate states based on faulty underlying models are, well, just so many pipe dreams.


 

Charles Battig, M.D. , Piedmont Chapter president, VA-Scientists and Engineers for Energy and Environment (VA-SEEE). His website is www.climateis.com

0 0 votes
Article Rating

Discover more from Watts Up With That?

Subscribe to get the latest posts sent to your email.

64 Comments
Inline Feedbacks
View all comments
Frank Kotler
June 3, 2014 10:40 pm

James says:
June 3, 2014 at 12:42 pm

The models do include rain, wind and snow.
—————————————————
I don’t think so. Wind probably wouldn’t hurt – the thing’s got a fan in it anyway – but real rain and snow would mess up your super computer pretty badly. tI’s all ones and zeros – mostly zeros.

June 4, 2014 2:28 am

Mi Cro says:
June 3, 2014 at 11:58 am
As what most would consider an expert in computer modeling of electronics

Bob Pease was of the opinion that even models of electronic circuits were near useless. And that no matter what the model showed you had to build a prototype to check model results. And this in a very limited domain where things are very constrained. I do EE and agree with Bob..
OTOH – climate….

Reply to  M Simon
June 4, 2014 4:40 am

M Simon says:
June 4, 2014 at 2:28 am

Bob Pease was of the opinion that even models of electronic circuits were near useless. And that no matter what the model showed you had to build a prototype to check model results. And this in a very limited domain where things are very constrained. I do EE and agree with Bob..

I started using digital TTL gate level simulators and timing verifiers around 82, went to work for Valid in 84, then Viewlogic in 91, went into PLM in 97. Most of that time was spent getting demo circuits from guy’s like you and Bob, modeling them, and showing you (or your boss) how there was value. But a lot did depend on the kind of circuit you designed, the level of fidelity your circuit needed to show it indeed when prototyped (which was routine) would work. The early years while you could put layout parasitic’s into the simulation, if you had proper margins, or were careful during layout you’d have good success with your prototype. From when I first started till the time I left more and more of the real physical effects were available to be added into the designs. In parallel to this the Integrated Circuit guys would start with transistor level circuits, and in the case of digital would then characterize a gate level model. All this lead to FPGA’s, LSI, and VLSI chips, which had their own modeling issues. Bu the 90’s we had tools to analyze inner connect transmission line effects, basically a spice model of the inner connect with analog drivers and receivers, then you could splice the resultant effect into digital gate or behavioral simulators.
But no one I know of makes a mask set for an IC, or spins a 20 layer board without doing simulations first, the tighter the margins, the more physical effects have to be included. If your design is all in an FPGA, if you make a mistake, you fix it and reprogram it.
Basically, you still need to prototype, but hopefully you build a couple, not 8-10. And that saves a lot of time and money.
When I left, computers still struggled simulating the boot process of a computer to test the hardware and firmware. By this point in time complex chips when used for a system level simulation were based on a software model (that should have been validated against a physical chip), this was what got me looking into GCM’s, it is a behavior model, and a big issue is that behavior models are developed to work the way the modeler thinks they work, not necessarily how they really work.

notfubar
June 4, 2014 5:33 am

Villiage Idiot:
“In general, we look for a new law by the following process: First we guess it; then we compute the consequences of the guess to see what would be implied if this law that we guessed is right; then we compare the result of the computation to nature, with experiment or experience, compare it directly with observation, to see if it works. If it disagrees with experiment, it is wrong. In that simple statement is the key to science. It does not make any difference how beautiful your guess is, it does not make any difference how smart you are, who made the guess, or what his name is — if it disagrees with experiment, it is wrong.” – Richard Feynman

captainfish
June 4, 2014 6:10 am

@notfubar says:
June 4, 2014 at 5:33 am
I hate it when Nature doesn’t agree with our experiments and we have to throw nature out. I can’t tell you how aggravating that is: to have nature not agree with models or experiments forcing nature to start over.
If only nature were as good as our models we wouldn’t be in this mess.

Mike Bromley the Kurd
June 4, 2014 6:31 am

Village Idiot says:
June 3, 2014 at 12:07 pm
Yours is the achilles heel of the warmist movement. The basic falsehood that we should act ANYWAY. Act to do what? What will the outcome be? Do we even have an inkling? No? Yes? I think it’s you who is the bitter one here. Bitter because no matter how hard you ignore the cognitive dissonance that you foist upon yourself, you fail utterly at foisting it upon others. And that’s just not fair, is it. Your self-deprecating handle is redundant.

Winston
June 4, 2014 7:45 am

richardscourtney says:
June 3, 2014 at 3:14 pm
And all of the above is why the indications of a model cannot be assumed to indicate anything about reality unless those indications have been demonstrated to match reality.
——
Isn’t it actually even worse that that? A model of a phenomenally complex system like Earth’s climate could by chance match reality for an interval and then, suddenly, wildly diverge from it due to some unknown factor or factors not included in the model.

Winston
June 4, 2014 7:51 am

Ceri says:
June 3, 2014 at 12:58 pm
I used to work in the electricity industry in the UK and one of my roles was forecasting electricity prices…
Moral of the story: Models with hundreds of parameters can tell any story you want, you just have to be able to sell it.
——–
Ceri, many thanks for that wonderful example of the too often misplaced trust in computer models.

bwanajohn
June 4, 2014 8:01 am

@Village Idiot
You visit your stock broker. He shows a stock with flat performance (reality) but superimposes the computer projection of value showing year-on-year increase (models) and recommends you buy the stock, at a premium of course. Here’s the kicker, he doesn’t have an explanation as to why the projections don’t match the performance. Let’s say the company’s main product (CO2) isn’t producing the expected results. DO YOU BUY THE STOCK? Not me, not in a million years.

Keith Sketchley
June 4, 2014 3:30 pm

Rational artists recognize what their work is and is not. For example, http://www.cordair.com. The Romantic Manifesto by Ayn Rand is a good reference. Both references know what reality is.

June 5, 2014 3:14 am

Winston:
You raise an important point in your post at June 4, 2014 at 7:45 am which asks me

richardscourtney says:
June 3, 2014 at 3:14 pm

And all of the above is why the indications of a model cannot be assumed to indicate anything about reality unless those indications have been demonstrated to match reality.
——

Isn’t it actually even worse that that? A model of a phenomenally complex system like Earth’s climate could by chance match reality for an interval and then, suddenly, wildly diverge from it due to some unknown factor or factors not included in the model.

Yes. Please note that I wrote “the indications of a model cannot be assumed to indicate anything about reality unless those indications have been demonstrated to match reality” (emphasis added, RSC).
Extrapolations and/or predictions cannot be demonstrated to match a reality which has yet to occur.
There is an infinite number of ways to accurately model the past
and
there is an infinite number of possible futures
but
there is only one future that will occur.
Richard

June 7, 2014 7:51 pm

Those who reify mathematics live in fantasy worlds.

Verified by MonsterInsights