Do the skeptics here simply not understand how science (and engineering) have changed, due to computers?

Question

Cover story in Wired this month:

"The quest for knowledge used to begin with grand theories. Now it begins with massives amounts of data. Welcome to the Petabyte age."

The Saturn rockets of 40 years ago were designed the old fashioned way, with engineering formulas and calculations.

Anyone trying to design a modern jet fighter that way will never build a good one. They also have no chance to get the contract. The Air Force doesn't even LOOK at real stuff until very far along in the competition, they just examine the computer simulations. Generally they then pick only two competitors to actually build airplanes, for the final decision.

The use of computer modeling is one reason Japanese car manufacturers have thrashed their American counterparts.

Bottom line; General philosophical arguments about the usefulness of computer models mark someone as locked in the past.

Answer 1

"Do the skeptics here simply not understand how science (and engineering) have changed, due to computers?"

No, I understand. Computer models can be very useful. Can they be used as evidence for a theory? No. Are they a substitute for observations and physical evidence? No.

"One reason the vast majority of scientists agree that man made greenhouse gases are the main cause of the present warming is that the models are, in statistical terms, "robust". That is, different models using different assumptions all get the same answer."

The assumptions are not particularly different. See, for instance:

http://wmbriggs.com/blog/2008/04/08/why-...

Notice that Dr. Gavin Schmidt joins in, but does not disagree with the main conclusion.

-----------

Edit:

""Would these be the same 'models' that showed that the US hurricane seasons would just become worse and worse, yet they have in fact became almost non existent?"

Weather. Climate. Two different things."

Wrong Bob. The trend in hurricane frequency is climate, not weather.

Answer 2

Apples and Oranges Bob.  A pilot can fly in a simulator and when he gets into the real thing, it's like he already flew the 2,000 hours needed.  Computer simulations are great when the science is understood and variables can be calculated.

However with "global warming" the science isn't very well understood, too many assumptions are made, and there isn't a very good understanding on the interactions of the variables.  This is why "global warming" has to be believed, why skepticism needs to be squelched, and no one can predict the climate 6 months, 1 year and 2 years from now.  Believers will only say what the climate may be after everyone is long dead.

Maybe someday, Bob, however we're not there yet.

Answer 3

I am an engineer, I have read most engineers are skeptical of global warming.  I understand the science, and how computers have effected the world in which we live.  We landed a shuttle on the moon with computers about as advanced as your calculator.  We missed mars with super computers.  Why?  Computers are only as good the data input.  I also understand from my schooling I can force any data to fit my model from experiments using computers.  If the link to the PDF report came through you should read.

Computer models for the automobile industry are not that accurate, that is why they always crash test.  Why are the models so important?  They are fairly good at taking existing measured items like an engine and analysing a change to the engine. If computer models for automobile industry were that good we would not need an O2 sensor.  

I don't think that you understand that science has not changed with computers.  Last time I checked the First law of thermo was the same before computers.  All of the calculations done by computers were done by hand at one time.   Computers just changed the productivity of the person doing it.  

By the way below is my temperature model.  It is all scientific.

I am closer than the one from the IPCC, does that make it correct that I used a computer.  I peak about .3 degrees warmer than we are now.  

1962-0.177203622

1963-0.199743193

1964-0.180320327

1965-0.13276472

1966-0.098461896

1967-0.082013022

1968-0.074599828

1969-0.072771556

1970-0.079298659

1971-0.091968703

1972-0.099568413

1973-0.094071567

1974-0.082783331

1975-0.061854856

1976-0.030405823

19770.005718672

19780.036916587

19790.060924227

19800.073089741

19810.076194062

19820.073387489

19830.0711853

19840.075997615

19850.097759369

19860.133577781

19870.168910778

19880.200626427

19890.197167538

19900.166588758

19910.164872526

19920.179135649

19930.173515017

19940.193253518

19950.275895666

19960.342295148

19970.346905255

19980.346882247

19990.375485022

20000.364961684

20010.315066333

Again, I do not think you understand what I am telling you.  Computers allow for advancment through increased productivity of the scientist.  Advancement is not science.

Answer 4

That's a very good point.  The 'skeptics' are always dismissing computer models as no better than random guesses and saying things like 'nobody can predict the future'.  But we rely on computer models in many different fields, including the ones mentioned by WIRED.  In one of my graduate school research jobs, I was simulating entire galaxies and galaxy clusters in a computer model.

A good recent example is with ocean heat content.  There was some noise made when it appeared the data showed the oceans warming less than the models predicted.  As has often been the case with data-AGW model discrepancies, the problem was with the data collection method varying over time.  The data has now been corrected, and fits the models quite well.

http://www.realclimate.org/index.php/arc...

I definitely think one reason the 'skeptics' doubt the AGW theory is that they don't understand the value and accuracy of computer models.  To be blunt, it's one of many aspects of the science which they don't understand.  And I don't think it's that they're not capable of understanding the basic science - I think in many cases they simply don't want to, because they *need* AGW to be wrong.

Answer 5

This is an obvious one- proof again that a little knowledge is a dangerous thing. The principle debate among climatologists isn't about the value of predictive modeling, but rather, on the very 'robustness' that you describe. The models are only robust when compared to models of 5 years ago. The fact is, we can't predict an el nino/la nina event until it's underway, and this is a semi-annual occurance. The current models are still missing factors in the additive input algorithms.

As the smarter scientists have said, wait a little. Imagine how the models will have improved in 5 years.

Answer 6

Why is this under Global Warming again???

Answer 7

Locked in the past?  Ha.  Locked in period.

edit:

I'm reading the posts now.  Engineers are notoriously hard headed.  INTJ personality type.  

Know everything, know what they don't know, won't change a l**k until you show them the proof.  Well guess what people, you don't know everything.  Never can and never will.  At some point, the engineering mind matures and is able to see the big picture.  Is able to become a team leader and be just a little bit flexible - enough to allow other points of view, other ways of doing things.  Admit that maybe, just maybe, somebody else is a better expert.  

When it comes to global warming, it's you against the entire scientific establishment of the world.  I have enough humility to accept that I don't know everything.  Why don't you?  

Because as the theory becomes set in stone you are looking less like a scientist / engineer and more like the crank who believes in free energy, oil is created from geochemical processes in the mantle, gigatons of carbon put into the atmosphere but not man's fault, gigatons of anthropogenic carbon in the environment but the earth will just somehow absorb it all with no effect.  

I'm not even going to waste time arguing the incredible impact of digital simulation.

I mean really, give me a break.

Answer 8

Computer simulations require complete understanding of the system, its components and how each component interacts with all the others.

For example Las Vegas oddsmakers use computers to simulate football games - - what would happen if the Pro Bowl left tackle on one team is injured early in the game and is replaced by a rookie, that sort of thing.     These work fairly well because the tendencies and abilities of the other players and the coaches' blocking schemes are finite and well understood.

We simply do not have that level of knowledge of the various factors that go into the climate or that level of understanding of how they all interrelate.

Answer 9

I appreciate your efforts Bob, but a lot of these folks calling themselves skeptics simply will not be swayed until, well I'm not exactly sure what it will take - even God himself could descend from the heavens and proclaim in a voice everyone in the world can easily hear:  "It is true!" and some folks would still somehow turn it into a politicized debate.

Fortunately we can spot these clones pretty darn easily.  They use labels like "Left" and "Liberals" - they're pretty much in step with their talk show idol indoctrination.

At it's very root it's an ignorance and a true lack and appreciation of what exactly SCIENCE is.  These poor souls simply regurgitate and glom onto the latest think-tank BS or repeat what is spewed on their AM radio dial. - We know it's bunk - we know it's politically motivated - and we know that science in it's purest form is above all this pettiness.

I guess the real debate is who started using science innappropriately - and who is really perpetuating it... but a guy like you Bob - I can easily tell you're taking every effort to be objective.

Which is WAAAAAYYY more than you can say for some of these Answer trolls.

Guess my answer makes me a Communist-Leftist-Hippie-Radical-Enviro.

Oh well, I always thought I was a Bike-Riding Cowboy, but who can argure with a bunch of "skeptics"???

(And is it really worth it?)

Answer 10

Twenty years ago I asked a old engineer how engineering had changed in his lifetime.  He said that when he was younger and needed information for a difficult problem, he would go to a more knowledgable person (an engineer's engineer) and discuss the problem with them, but with computer information so readily available, he only needs to research things on line to find expert information.  Who you knew mattered a lot more in engineering many years ago.

The problem is that with all the good data on line, there is a lot of misleading information also.  Engineering skill requires the ability to analyse the data, come to conclusions, and produce a result.  Using global warming as an example, an analysis of the data indicate the world climate is currently becoming warmer and more variable in most locations.  The common conception is that the cause of this is modern technology, but the information available online indicates that 0.2% of the trees in the world are being burned by people each year.  By calculating the amount of heat and pollution produced by this, you get a bigger number than the pollution and heat caused by cars, power plants, and factories.  This additionally compounds global warming in a way that other sources do not - trees are a heat sink for incident solar energy.  They transform the energy of the sun into their life mass.  So the really cause of global warming is poor people who burn wood as opposed to industrialized societies.

The other effect of computers is that everyone can now get lots of data, so there are a lot more people who have the option of disagreeing with any decision that any engineer makes.  This has had a serious adverse effect on engineering endeavours.  We haven't been able to put a man on the moon in 35 years and it has been almost 50 years since anyone has been to the bottom of the Mariana's Trench.  Computers are indirectly responsible for this lack of technological progress.  We can't even duplicate the efforts of a previous generation.

All that said, I use computers all day, every day to get the information I need.  I couldn't live without them.

Answer 11

Yes. So, you are saying that if you owned a copy of "Microsoft Spy Plane Design Simulator 9.0," you could design better aircraft than Clarence "Kelly" Johnson!

I'll assume your knowledge of aviation history is on par with your knowledge of quantum mechanics & everything else......  

http://www.wvi.com/~sr71webmaster/kelly1...

I'm sorry, coffee just squirted out my nose.

Edit:  False!  Modern design software and simulators assist the engineer, they are not a substitute for expert knowlege.

Answer 12

I agree wholeheartedly.  Our top answerer has been known to refer to the use of computer models as "mysticism" and claim that no one has any basis for global warming predictions because the calculations aren't done with pencil and paper.

I know through my own educational experience, flow diagrams for very well understood chemical processes are usually done using computer models.  And those are not as complicated as this:

http://www.globalchange.umich.edu/global...

Answer 13

You mean that scientists haven't tested climate science on an actual planet, with atmosphere, in a lab somewhere?  I'm shocked.

Well I guess we don't need to worry about the problem then, do we?  I mean if you can't PROVE that someone just shot an arrow at my head, then I can't get hurt if I fail to move when you yell "Duck!", right?

Let's all wait for "proof" and see what happens, shall we?

Here's some interesting history on modelling:

1955-65: Establishment of Atmospheric General Circulation Modeling

http://www.aip.org/history/sloan/gcm/195...

"Atmospheric GCMs simulate the entire global circulation of the atmosphere. But they are not the only mathematical models useful in understanding the relationship between atmospheric composition and radiative transfer (the basis of the greenhouse effect). Other types of models used to study the greenhouse effect include energy-balance models, which compute global average surface temperature, and radiative-convective models, which calculate the vertical structure of the atmosphere. Each of these may have one or two dimensions.  In addition, models of the ocean circulation (also called GCMs) play an increasingly crucial role in climate modeling."

Answer 14

There is an old saying in information tech:  Garbage In Garbage Out.

Your SNR for CLIMATE may be high, but no matter the "signal strength" you can't compensate for absence of data from the past.

Okay, tree rings and ice-cores, but how do know what are good parameters for correlating CO2,CH4, H2O and ice properties?

And the whole thing is funded by the same people who have an interest in the outcome.

And you can't stop it anyway.  Even if you nuked every human on the planet, you can't be sure it would stop.  Incidentally, thats madness.

Answer 15

Right...the F-16 was a TERRIBLE jet.

In order for a computer model to be useful, it has to account for all relevant variables.  We don't know all of the variables.  We don't know how all of the variables interact.  We actually know very little...but many are CERTAIN the result must equal "man's to blame".  When you build a model with a foregone conclusion, the results rarely surprise.

Answer 16

If what you said is true, why is the strongest evidence pointing in the other direction?  Why is the left so fearful of scientific debate?  

Likely because even they have come to realize that they're dead wrong with their data.

Answer 17

okay, thanks for that. computer models and simulations only work to the extent of the data input.  I don't think anyone is denying there usefulness, but when not all the data is collected you have an incomplete model and therefore an inaccurate one

Answer 18

Computers have their use.  They currently are useful for predicting weather about 5 days in advance but even then they need to be checked by humans.  There is much that we don't know.  You cannot expect a computer to accurate predict weather and climate unless the particular variables and assumptions are better understood.  Clearly, they are getting better.  Clearly, they still have a ways to go.

Answer 19

I am not sure why this question is in Global Warming but...

Perhaps it is because of the fact that the vast majority of reputable scientists do NOT agree with the findings the IPCC report.  So many of them disagree with that report in fact that they have had to sue to have their names removed from the report.  The only reason they were on the report to begin with was due to the fact that is was added when they had agreed to attend as a guest.

Answer 20

One reason the vast majority of scientists agree that man made greenhouse gases are the main cause of the present warming is that the models are, in statistical terms, "robust". That is, different models using different assumptions all get the same answer. Even when the parameters in the model are varied within reasonable limits. Also in statistical terms the "signal" in the data from man made greenhouse gases is very strong.

Would these be the same 'models'  that showed that the US hurricane seasons would just become worse and worse, yet they have in fact became almost non existent?

That might have a bit to do with the skepticism don't you think?

Answer 21

Even today, with current computing capability, models are nowhere close to being able to reliably predict climate change 50 to 100 years into the future. If they were, the temperature decline from 1998 to 2008 would have been predicted by the models -- it wasn't.

Answer 22

Sometimes (chaos theory) a small assumption (or misjudgement) at the beginning can lead to big errors, computer models *or* beleif in them.

Answer 23

Ren Hoek nailed this question. Let me answer with a few questions. Do you think climatologist have even a 90% understanding of all the factors that effect climate? Will the computer models be better than they are now in 10, 20, 50 or 100 years? Are they as good as they can be?

You guys are always discounting solar influence by suggesting that the only change is .3w/m² TSI. There are many other influences than TSI. There are many studies that show cloud cover as well as major atmospheric and ocean circulation patterns that change with solar activity. The models don't even consider such changes. Our tides are effected by the moon and the suns mass and magnetic characteristics are effected by celestial bodies that orbit it. The models are over simplified.

Answer 24

Where is this green house gas that idiot said was according to his calculations. He doesn't know that plants take in the CO2 and give us back O2 ,but also the plant keeps the C which becomes the Corbin in fossil fuels.

Then there is Methane , Where is it there has been billions of cubic ft. calculated but where is it. I don't believe if it is there we can measure it. If the green house gas is not there neither is Global Warming by Gore.

Answer 25

Wow, I'll guess I'll put in my two cents worth.

I avoided the modeling aspect of global warming at first,which means I was skirting around a lot of valuable information, and learning how much of the information out there was generated.  

In efforts undertaken thus far to learn about global climate change models, I've learned that they have improved tremendously, by necessity and by design.  Which leads me to want to take to task part of the answer from GW Yahoo Chieftan Jello:

"However with "global warming" the science isn't very well understood, too many assumptions are made, and there isn't a very good understanding on the interactions of the variables."

The science is understood perfectly well by the scientists, thank you very much, the ones that can call themselves Dr. because they really are. Scientists can learn a lot from even an imperfect model.  God forbid they should all get everything exactly right the first time around.  Jeez.

To be fair, I don't think a lot of people give much thought to how valuable models have become in modern life.  When people do take the time to learn about the what why and how of modeling global climate change, they'll emerge with -- in some cases -- a newfound respect for this very wonderful tool.

Answer 26

Hmmm...ever had a problem, with your OS? We all know it's a tool and a good one @ that...but. Would that mean you have no comprehension of how to use a computer? Me thinks, you open a can of worms with this question.

All pretense aside, sure cybernetics has came along way. It's still depends on human discern. The guy's in the white lab coats, still have to come up with the big picture. Even with technology, there's still the cross roads of form, functionally, practicality, implementation, feasibility, interpretation, diagnostics, and last but not least. Is this going to be a standardized assessment or is any type of human discretion going to play a part?