Why I think Climate Change Denial is little more than pseudoscience. - Part 1

[FuzzyLumpkins]

No.

I have no need to prove to you anything. I waste too much of my time with you as it is. If you missed it, too bad. I will probably repeat the facts again in some future point, and you will probably still not understand.

Ask me if I care anymore.

Intellectual cowardice.

You start off saying they didn't account for changing temperature. I then point out several studies that do just that and then you claim it doesn't account for long term changes. I explain how modeling works and you now claim some mythical facts.

Now you cannot come up with where to move the goalposts this time so you posture with bull like this.

If I have time later on I will quote you post by post just as I did when you postured like this when I called you on your lack of knowledge on capacitors.

It's good for a laugh. Last time i did it you put me on ignore. Run that white flag right up the pole.

[Wild Cobra]

Intellectual cowardice.

You start off saying they didn't account for changing temperature. I then point out several studies that do just that and then you claim it doesn't account for long term changes. I explain how modeling works and you now claim some mythical facts.

See, you miss the meat of what's said. You favor to find any piddly ass reason you can to attack me, or fabricate reasons to attack me. It was clear that when I made the statement that "they didn't consider temperature changes of the ocean for CO2 changes," it was over the long term. All you did was bring all these seasonal studies in a lame attempt to discredit me.

Believe it or not, I understand what you say. The problem is, you are to simple minded to understand what I say. Worse yet, you are a cyber bully.

Why should i do anything beyond typing a few words out to you? Why should I waste my time looking anything up for you, when you are too pathetic?

[MannyIsGod]

You don't understand . When I talk to people who understand, they don't say sea ice is an ice sheet. They don't say that global warming is caused by CO2 creating energy in the atmosphere. They understand Celsius and Kelvin are the same unit on different scales. They don't pull studies out that the authors themselves have admitted to have been wrong.

So no, WC, you don't understand. You can repeat it all you want but it won't make it anymore true.

[DarrinS]

Spatial resolution of climate models isn't high so what does it matter if spatial resolution of data sucks? Weather models don't break down due to lack of data, Darrin. They break up because the variables -

Are not some of these variables DATA?

They break up because the variables - which I've stated are much higher due to spatial and temporal resolution than climate models - multiply with each ongoing hour and the error grows.

Exactly.

Climate models look at how an overall system behaves in the long term and don't need to worry about short term variability. Weather models on the other hand, handle nothing but short term variability.

Short-term variability affects the long-term behavior.

So, I'll state once again. This argument is born out of ignorance. You disagree with availability of data needed for climate models but just why does your opinion carry weight on the subject. I don't even mean that as an insult. If I went to the doctor and told him I disagreed with his diagnosis it sure as would not carry much weight to his patient, would it?

Because I've been modeling complex dynamic systems in the areas of thermo, heat transfer, fluid mechanics, structural dynamics, etc. for over 20 years. I think I my opinion does carry weight on the strengths and limitations of these types of computer models. Just as a statistician might not be an expert in climate science, he can weigh in on the use (or misuse) of a particular type of statistical analysis.

[MannyIsGod]

Your second to last statement completely s up your last. You proved your ignorance right there.

[MannyIsGod]

If you flip a coin and 7 of the first 10 flips are tails does it affect the long term result? No.

[ElNono]

I agree with Darrin when it comes to statistical errors. The scope of climate models are simply huge. Even if you do have access to a lot of the data, you simply don't have the capacity right now to simulate everything fast enough (ie: faster than realtime, which you need to do if you're in the prediction business). You have to statistically approximate, which is fine short term, but the longer you move into the future, the higher the statistical error grows.

But this is also why computing power is critical. The statistical fudge factors in simulations normally come in two flavors: A) too much data (more than what you can process now) or B) not enough data. Increasing computing power is critical to both. In the A case, it allows for more real data processing, and less statistical fudging. In the B case, it allows to create avenues to gather, catalog and research the missing data.

Manny is certainly right that models in the past 20/30 years have improved dramatically, and IMO it's a no brainer one of the biggest factors for that is increased processing power.

Just my 2c

[DarrinS]

If you flip a coin and 7 of the first 10 flips are tails does it affect the long term result? No.

Yeah, good one. We all know chaotic dynamic systems only have 2 possible outcomes.

[ElNono]

If you flip a coin and 7 of the first 10 flips are tails does it affect the long term result? No.

If you were to predict coin flips in the future and you hard-coded 70% probability of tails over heads, the error rate would undeniably grow larger the longer you move into the future, and thus affect the simulation and the result from it.

[FuzzyLumpkins]

See, you miss the meat of what's said. You favor to find any piddly ass reason you can to attack me, or fabricate reasons to attack me. It was clear that when I made the statement that "they didn't consider temperature changes of the ocean for CO2 changes," it was over the long term. All you did was bring all these seasonal studies in a lame attempt to discredit me.

Believe it or not, I understand what you say. The problem is, you are to simple minded to understand what I say. Worse yet, you are a cyber bully.

Why should i do anything beyond typing a few words out to you? Why should I waste my time looking anything up for you, when you are too pathetic?

When all you do is speak in generalities that speak of nothing other than posturing it is not compelling. When I get home I will do a quote by quote review like I said. It's not very hard to show how you were full of .

You in no way demonstrate an understanding. As for specifics: partial derivatives, sampling and periodic functions are beyond your pea-brain. You took them holding a variable for a constant for a calculus operation to mean that the considered said variable constant, you think that sampling endpoints of decades of phenomenon is more telling --because you keep repeating it-- over sampling a period 4 times every year and just considering the variables as they come.

Notice how I gave specifics? Notice how I did not just posture that you didn't know but instead pointed out specifically?

All you have done is move your goal posts from they don't consider it to they do not consider it long term. As if they only run the models for a year. Prima facia it's stupid.

[MannyIsGod]

If you were to predict coin flips in the future and you hard-coded 70% probability of tails over heads, the error rate would undeniably grow larger the longer you move into the future, and thus affect the simulation and the result from it.

No doubt, but Darrin's argument (at least THIS time) wasn't that the model itself was wrong but rather that their inability to resolve short term variability affected the long term. There's nothing correct about that statement.

Ignoring the short term variability is quite easy and not a problem with climate models because you're not trying to forecast the temp in a particular location but rather the amount of energy in the earth. On the simplest level, you can model that in ing excel and come out with reasonably accurate results.

Global Climate Models today are much more complicated beasts and their output their output should be reviewed very skeptically (in the actual sense of the word - not dismissed out of hand but rather taken with a grain or two of salt) but they are trying to make predictions at the local and regional level more than simpler climate models. I've never once made an argument that we can accurately predict the outcome of AGW at any one place on the globe (this is the current aim of GCMs) but when people dismiss climate models and refer to the models that forecast global temperature increase and a lack of data they are typically speaking out of complete ignorance.

[FuzzyLumpkins]

If you were to predict coin flips in the future and you hard-coded 70% probability of tails over heads, the error rate would undeniably grow larger the longer you move into the future, and thus affect the simulation and the result from it.

The error frames they give are based off of the terminations of projections are defined. All you are doing is speaking from generalities. 70% is a made up number and one thing that is clear is that over time those error ranges have shrank.

If the error estimates are wrong then feel free to point to it but all I see is incredulity versus scientists that are giving clear ranges where expected values can fall.

If they are not compounding the probabilities correctly thats kinda the point of peer-review and i am sure the Heritage Foundation has people working to double check it.

[MannyIsGod]

Yeah, good one. We all know chaotic dynamic systems only have 2 possible outcomes.

Its an analogy fool. The point was that short term variability does not affect the long term outcome.

[MannyIsGod]

I agree with Darrin when it comes to statistical errors. The scope of climate models are simply huge. Even if you do have access to a lot of the data, you simply don't have the capacity right now to simulate everything fast enough (ie: faster than realtime, which you need to do if you're in the prediction business). You have to statistically approximate, which is fine short term, but the longer you move into the future, the higher the statistical error grows.

As I stated above, GCMs that try to make predictions on a regional or local level are definitely deserving of this type of criticism. Models that are looking at the global energy budget are really not.

But this is also why computing power is critical. The statistical fudge factors in simulations normally come in two flavors: A) too much data (more than what you can process now) or B) not enough data. Increasing computing power is critical to both. In the A case, it allows for more real data processing, and less statistical fudging. In the B case, it allows to create avenues to gather, catalog and research the missing data.

Manny is certainly right that models in the past 20/30 years have improved dramatically, and IMO it's a no brainer one of the biggest factors for that is increased processing power.

Just my 2c

[FuzzyLumpkins]

Darrin is talking out of his ass. Non-periodic phenomenon can have any number of outputs just as any other arbitrary phenomenon. Him being exclusionary is just posturing.

Force them to talk of specifics and actually apply these principles they claim to espouse. When you do it rings very hollow.

[MannyIsGod]

Darrin has never given specifics on so I don't expect him to start anytime soon unless he finds a youtube.

[ElNono]

No doubt, but Darrin's argument (at least THIS time) wasn't that the model itself was wrong but rather that their inability to resolve short term variability affected the long term. There's nothing correct about that statement.

The error frames they give are based off of the terminations of projections are defined. All you are doing is speaking from generalities. 70% is a made up number and one thing that is clear is that over time those error ranges have shrank.

I was given a general example, thus you get a general breakdown.

This isn't rocket science. If you build a model where one component is "coin flip" and you define "coin flip" = 70% tails (Manny's example, not mine), then it's a model with a statistically crappy component that when you use it to simulate the entire model over time will continue to creep in error feedback into the main model. The longer you simulate, the bigger the error creep, which translates into a less realistic simulation.

If you instead define "coin flip" as a physical calculation including the properties of the coin (mass, size, shape, features), the given force when tossed (vector, impulse) of the coin at a given time, the air pressure, the features of where the coin lands, etc, etc, etc and gather the value from that, then you'll likely have a much better simulation and approximation of that variable, because you know a whole lot more about the conditions of the coin flip. That translates into less statistical error for the individual component and less error creep onto the main model.

Unfortunately, there's simply not enough processing power to go with minutiae on every component. Now, some lend themselves better to approximation, but others unfortunately do not. And if you want to simulate, you're going to have to make compromises into the accuracy in order to get a simulation going. This doesn't apply to just climate model, you can see the same effect on a plethora of other fields. For example, ray tracing on computers have had to deal with the limits of computing for the same exact reason.

Now, there's no doubt that increased processing power opens the door for better processing, better simulation and obviously better results.

[Wild Cobra]

You don't understand . When I talk to people who understand, they don't say sea ice is an ice sheet.

I call it what theuy call it. Do prominent universities call it sea ice or not?



I see "Sea Ice" in the le. Don't you?

They don't say that global warming is caused by CO2 creating energy in the atmosphere.

Where did I say the energy is created?

That's either one of the many incorrectly paraphrased things you guys do because you cannot debate me legitimately, or it was a one time mistake I had since corrected.

Why do you use such unethical tactics?

that will really get you far to be so unethical in future career choices.

They understand Celsius and Kelvin are the same unit on different scales.

And you think I don't?

No wonder you cannot relate. No matter what I say, you do not understand my words.

Where have I ever indicated otherwise? I make the distinction that the scales have to be treated if multiplied.

That's because C = K -273.15.

If I interchange C and K in mathematics using the temperature, will my results be the same?

no. The formula must be tailored for one unit or the other.

You can repeat it all you want but it won't make it anymore true.

Yet you have not shown me to be wrong yet.

Your problem is the same as FuzzNutz. You twist what I say to be something incorrect, just because you fail to understand my intent. Doesn't a wise person ask for clarification rather than accuse and attack?

[MannyIsGod]

that wasn't my example at all, EN.

My initial post was that climate models don't need to resolve the short term in order to get an accurate picture of the long term. Darrin countered with saying that short term variability affects the long term. It does not. Thats why I gave the coin example. I never said anything about programming a model with an incorrect behavior.

A model that is programmed to generate a 0 or 1 at random will generate those numbers equally in the long run regardless of which number is at a higher rate through 30 iterations. THAT was my point. Not that programing it to prefer 0 or 1 would not matter.

[MannyIsGod]

I call it what theuy call it. Do prominent universities call it sea ice or not?



I see "Sea Ice" in the le. Don't you?

Where did I say the energy is created?

That's either one of the many incorrectly paraphrased things you guys do because you cannot debate me legitimately, or it was a one time mistake I had since corrected.

Why do you use such unethical tactics?

that will really get you far to be so unethical in future career choices.

And you think I don't?

No wonder you cannot relate. No matter what I say, you do not understand my words.

Where have I ever indicated otherwise? I make the distinction that the scales have to be treated if multiplied.

That's because C = K -273.15.

If I interchange C and K in mathematics using the temperature, will my results be the same?

no. The formula must be tailored for one unit or the other.

Yet you have not shown me to be wrong yet.

Your problem is the same as FuzzNutz. You twist what I say to be something incorrect, just because you fail to understand my intent. Doesn't a wise person ask for clarification rather than accuse and attack?

Sure, no one has ever shown you to be wrong.

[Wild Cobra]

Manny is certainly right that models in the past 20/30 years have improved dramatically, and IMO it's a no brainer one of the biggest factors for that is increased processing power.

Just my 2c

But you still have to have good and complete data.

garbage in - garbage out.

[MannyIsGod]

Tell me again what you call sea ice, WC?

http://spurstalk.com/forums/showpost...postcount=2003

[FuzzyLumpkins]

I was given a general example, thus you get a general breakdown.

This isn't rocket science. If you build a model where one component is "coin flip" and you define "coin flip" = 70% tails (Manny's example, not mine), then it's a model with a statistically crappy component that when you use it to simulate the entire model over time will continue to creep in error feedback into the main model. The longer you simulate, the bigger the error creep, which translates into a less realistic simulation.

If you instead define "coin flip" as a physical calculation including the properties of the coin (mass, size, shape, features), the given force when tossed (vector, impulse) of the coin at a given time, the air pressure, the features of where the coin lands, etc, etc, etc and gather the value from that, then you'll likely have a much better simulation and approximation of that variable, because you know a whole lot more about the conditions of the coin flip. That translates into less statistical error for the individual component and less error creep onto the main model.

Unfortunately, there's simply not enough processing power to go with minutiae on every component. Now, some lend themselves better to approximation, but others unfortunately do not. And if you want to simulate, you're going to have to make compromises into the accuracy in order to get a simulation going. This doesn't apply to just climate model, you can see the same effect on a plethora of other fields. For example, ray tracing on computers have had to deal with the limits of computing for the same exact reason.

Now, there's no doubt that increased processing power opens the door for better processing, better simulation and obviously better results.

I probably should have edited that post for vehemence.

[MannyIsGod]

LOL tests

http://spurstalk.com/forums/showpost...&postcount=729

[FuzzyLumpkins]

But you still have to have good and complete data.

garbage in - garbage out.

What empirical observations are wrong?

Perhaps you could give a specific example?

You do not even know what the formulas are. You didn't think that they considered temperature. You thought they viewed temperature as a constant until I had to beat your head with studies that showed they didn't. But here you are talking about the input values.

Also, you are acting like aspie. There is an element of degree here. There is completely incorrect and then there is too general. An example of this would be to sample something every 40 years and to sample things every season. Which one paints a more clear picture do you think?

Oh wait, you are the one that goes 100 years and a solubility chart.