Do you see parallels between climate models and what I do at work?

Question

I work in the area of computer network security and I'm an engineer, a designer of solutions to meet business needs. The problem that my group runs into very often is when the owners of networks find out about a new hot product like a new whiz-bang firewall or some encryption system and then ask us to use that product to design a solution.

We keep telling them that security engineering doesn't work that way. You need to identify a requirement and we can design a secure solution to meet that requirement. After the solution has been designed, THEN we pick the product(s) that are necessary to implement the solution. It may be the new whiz bang firewall or it may not be depending on the security requirements we have designed into the solution.

So given what I have described above about my work, do you see a parallel with the development of climate models where the end "product" is pre-selected and the rest of the work goes into filling in the blanks?

And if the answer is yes, can you see how that may not actually be the best approach?

Answer 1

It kind of reminds me of those medicine advertisements that tell you to ask your doctor about Celicin instead of just asking for your doctors opinion to solve a problem.



I think the laws of evolution go into play and we get an "unnatural selection" process where the studies that fit the agenda survive. Those studies and theories that are unfit and don't fit the agenda die a death of unfunding or scorn. There is abundant evidence for those who look at AGW without rosy colored glasses that the evidence is cherry picked and "selected."



It is too bad that instead of looking at all the facts and then deciding on the best theory to explain that set of facts, too often some will sort through facts and evidence to best argue their pet theory. When money and funding influences the outcome of the study, there is more incentive to sort through and find the convenient evidence IMO.

Answer 2

I have one question about computer climate models. Is there a computer made that could accurately track all the variables in the world wide weather system? Do we even understand all the variables and how they interact with each other? Some of the variables don't even have any direct connection with the weather. Such as volcanoes or sunspot activity

Answer 3

<<>>



That's climateprediction.net, part of the BOINC community which has other famous computing networks like SETI @ home.



Climateprediction is pretty cool in that you can actually see it running in real time, complete with a graphical image of the Earth, where you can see the temperature gradients, precipitation, clouds, and pressure being computed as it goes through half-hour time steps. A lot of it is hindcasting: the ones that I have running now are currently at 1988, 1994, and 1810 (you get one running on each core of your processor).



I have a pretty fast processor, yet it still takes me over 600 hours to complete a single model run. It definitely gives you a sense of how much goes into them.



I would think twice before running it on a laptop though, since laptops tend to overheat and these models use 100% of your processor (although you can limit it if you want). You'll want to keep a close eye on your temperature.

Answer 4

I see paralels between take action sections in civic participation and local news papers which claim to be community connection yet do nothing to encourage students and other community members to submit articles.

Also reminds me of the mailer demon program where people fail to adapt proper security but some cannot afford best security but should for global warming.

I guess what you are saying is they should ask security experts to design constitutions as security programs to get rid of car ads and fail to use own standard freedoms as their security.

also that magnets and production in doing so for removing vehicles from roads or creating vehicle engines idea. more simple to use magnets to remove cars from roads than build news ones.

Answer 5

Sorry, Mike, the answer is "No". And I'm not even a scientist. I started that way, but life intervened, and I wound up in a job where my "other hat" was being a translator between you techie types and the rest of humanity. I must say your question exemplifies the problems I had in doing that other hat job. I see it this way. Engineers focus on the specific job at hand, and only on that. Their goal, very commendable and supposedly exactly what we want, is to analyze a situation, solve the problems, and implement the solutions in the most efficient, effective way.



Owners, managers, ties of most sorts, tend to think of the whole business, not just about the specific problem. They bring in all sorts of "extraneous material", and agonize over irrelevant minor details while missing "the big picture" of the situation at hand. This leads to exactly the situation you discuss in your question. Every techie type I've ever met has said the same thing:"I gave them exactly what they asked for. They wanted something else." If there's an engineering/techie type here who doesn't agree, you'd likely be the first!



The problem is proper communication, and it is on both ends. Business and engineering use different types of thought. This leads to different expectations, even different meanings for words - "theory" being a fine example. What does "theory" mean to you? Further, in general, each side truly believes the other side should obviously see things their way, and so it's the other side's responsibility to change to conform to the principles of our side. Again, does anybody really disagree with this [and why, if you do?]



Your second-to-last paragraph contains the key error in your line of reasoning here. As an engineer, you see everything in terms of 1 specific problem and its solutions. This one problem may have many aspects [How do you get from East coast to West coast in the 1800's? Railroads cross all sorts of terrain and political boundaries...] but you still see it in terms of 1 problem with a desired solution. Engineers have an aim, a goal, in all their work.



Science doesn't work this way. Science is speculative. Someone who was only an engineer would have thrown that first accidental batch of Lucite away, because it was not what they were looking for, and it's inefficient to keep going off on sidetracks. True or false? Science tests many things, and does not always let preconceptions stand in the way. Einstein's explanation of Michaelson-Morley is an example that *should* be known to all, as was his explanation of the photoelectric effect [for which he got the Nobel].



If you look objectively at the history of the investigation into anthropogenic global warming, you will see over and over again that scientists were not expecting the results they got. Most scientists assumed AGW was a potential problem for the far future. I think it's apparent that privately, many were appalled at what they found. I know that when I got into serious recent study of climate changes a few years ago, as a reaction to all the denialism I saw, I became much more worried, and felt I had to do something. This is the common experience of those who are involved in the science. So it's not surprising those of us who follow the science come to feel the same way.



Climate change is obvious, and all around us. You can't make this go away by attempting to cast doubt on the motives or methods of a few scientists.

*********************************

EDIT: Mike, my answer was long enough, I had to simplify where I could. There's a limit to the absolute size of these answers... I've been on both ends of that Business Requirements Analysis, and I can honestly say it does not often work as well as hoped or planned for. This has been true in my career and in my private ventures. If businessmen could project anywhere close to perfectly, each would be a multi-millionaire at least. When the guy tells you, the engineer, he wants the world, with the sun, the moon, and the stars as accessories, as his wish, what he really needs is also the asteroids, comets and Oort cloud, or nothing will work right. Businesses need to be flexible in more ways than people can anticipate. You cannot engineer the future perfectly.



As far as science and speculation, my speculation tends to fall within or near the error bars. The "other side" often tries to expand those error bars, seemingly without limit, certainly without adequate justification.

Answer 6

Baccheus's answer is not satire. It is exactly what happens in real life, exactly what your ilk does. It's noble (or disingenuous depending on your point of view) of you to present yourself as a moderate, but I've seen enough of your answers to know otherwise.



I happen to be a support professional with 18 years experience and I've seen it over and over. Person A asks question, looking for answer or solution. Person B, the acknowledged expert, gives expert opinion. Because Person B is an expert, you will get a correct or nearly correct answer as there are only so many technically feasible answers (or scientifically, there are only so many candidates of theory that fit the data). If you ask another expert you will get a similar answer or set of answers.



Often, Person A doesn't like the answer or solution given because: it costs too much or costs too little (Isn't this stuff supposed to be complicated and expensive? Not necessarily.), is too disruptive to the current paradigm, they just don't like the vendor or the person selling the solution. This dislike often has nothing to do with reality, rationality or the fitness of the answer or proposed solution to the current situation.



And there is absolutely nothing you can do about it. You do everything you can to help the client but in the end just say "This is the best we can do, if you ask around you will get similar answers, I'm sorry we can't help you, we can't put our reputation on the line by installing a solution for you that isn't going to work. Sorry, we can't put our name on this."



Management pinheads only interested in quarterly profit and the latest sale really really hate this and you put your job on the line doing it. The greatest complement ever paid to me was from the best manager my company ever had. In front of his boss, and the client, he said "Bob always does what's in the best interest of the client". That is how we maintain long term relationships, and contracts, and how our business grows by word of mouth. We can be trusted.



But in the case of AGW, because you don't like the answer, you would turn all of this on its head.



You're not fooling anybody who knows better.

Answer 7

What the "scientists" here apparently don't get is that certain assumptions are designed into all simulations and models.



This is why we get headlines like "The Mystery of Global Warming's Missing Heat"



http://www.npr.org/templates/transcript/transcript.php?storyId=88520025



They assumed that there would be warming, programmed it into their models, and then were baffled when their model didn't work. You're exactly right when you suggest they've "pre-selected" the end product.

Answer 8

The parallel would be in a potential client who is currently being routinely hacked. They already have suffered lost hours due to computer breakdowns. You can see that all of their financial records and even their bank accounts are at risk and will certainly be accessed by hackers in the next year.



Your recommendation is to install the newest and best security solution. But they say that is too expensive. So you say OK, lets install something less expensive, it will protect your most sensitive data for maybe another year, but you are still at great risk. They then tell you that you and every other security engineer are evil crooks; in fact everyone who has ever studied security engineering is a liberal, just scheming for for their money. The very act of study causes all of you smart people to become liberal conspirators. So they decide to do nothing other than sit around debating what an evil liberal you are. Only after the company is defunct does somebody realize that education does not make one evil, and that they really should have installed some security.

Answer 9

I really don't understand why you think the "end product" of climate models is pre-selected. The end product of a climate model is a projection based on the inputs (fundamental physics) of how the global climate will change in the future under the chosen scenario.



I think you should take david's suggestion and learn more about climate models.

Answer 10

I'm not sure there is anyone who actually works with climate models (or any mechanistic model for that matter) that would agree that a model is designed to produce a specific outcome.



Not only would it be an enormous waste of time to design a complex set of equations based off of physical principles just to give you the output that you want, but it would be moronic and against every fundamental, philosophical tenet of science.



Here are some sites where you can download and play around with various climate models, I'm sure there are several others that hopefully some of the other users will post.



http://sourceforge.net/projects/climate-model/



http://climateprediction.net/



Several months ago, in another question, I stated that by beginning to work with models I had garnered a new respect for them and their complexities. You were confused by that and thought working with models should have made me more skeptical than anything, that always seemed like a wierd conclusion to come to for me...



Anyways, for shiz and giggles, here is a link to a website devoted to the model I work with.



http://www.bio.mq.edu.au/maestra/



It is a spatially explicit, 3 dimensional, plant canopy transpiration model. At this site you may download the model and all of the source code (which is great if you can read fortran). I guarantee that it was not designed to fit a preconceived output.

Answer 11

It sounds like your work is much more complex than climate models as climate models have only two important variables CO2 and Temperatures which are both directly related everything else is irrelevant and these two are massaged or should I say recalibrated until directly linked!



I imagine that computer security has many more variables and is therefore a much more complex and fulfilling career... plus I'd definitely stick with computer security as I'm sure it has a much longer and prosperous future!

Answer 12

I hope you are better at your own job than you are at understanding climate models.



The "end product" that climate models strive for is matching up with reality using the equations of physics.



EDIT: Does it ever occur to you that you slander hard-working, dedicated people almost every day in here? I don't think that you recommend computer security software produced by a particular company because you get kickbacks from it, but that is essentially what you are suggesting about climate scientists.