I’ve been looking at the Nikolov and Zeller paper again. Among other things, they claim to be able to calculate the surface temperature Ts of eight different planets and moons from knowing nothing more than the solar irradiation So and the surface pressure Ps for each heavenly body. Dr. Zeller refers to this as their MIRACLE equation. He says:
Why aren’t you all trying to disprove our MIRACLE equation rather than banging your heads against walls trying to prove or disprove who knows what and exclaiming you have problems with this or that? The question is how can we possibly have done it – there is no question that our equations work – if you haven’t verified that it works, why haven’t you? […] Why aren’t you thinking: “hmmmm, N&Z have given us an equation that lo-and-behold when we plug in the measured pressures and calculate Tgb as they suggest, gives us a calculated Ts that also matches measured values! You can’t disprove the equation? So maybe we are cooking the data books somehow, but how?
This is supposed to be evidence that their theory is correct, and people keep telling me ‘but they’ve got real evidence, they can make predictions of planetary temperatures, check it out”. Plus it’s hard to ignore an invitation like Dr. Zellers, so I checked it out.
Figure 1. These are not the equations you are looking for.
They first postulate something called the “Near-surface Atmospheric Thermal Enhancement” or “ATE” effect that makes the earth warmer than it would be without an atmosphere.
The “ATE effect” is measured by something called Nte(Ps), which is defined and estimated in their paper as follows.
where Nte(Ps) is a measure of the “Near-surface Atmospheric Thermal Enhancement” effect.
Nte(Ps) is defined as the actual average surface air temperature of the planet Ts divided by the theoretical “graybody” temperature of the planet Tgb calculated from the total solar insolation So of the planet. Nte(Ps) is estimated using a fitted function of the surface pressure of the planet Ps.
Let me simplify things a bit. Symbolically, the right part of equation (7) can be written as
Nte(Ps) = e^(t1 * Ps ^ t2 + t3 * Ps ^ t4) (7Sym)
where “e” is the base of natural logs and Ps is the surface pressure on the planet or moon. There are four tunable parameters (t1 through t4) that are “fitted” or tuned to the data. In other words, those values are repeatedly adjusted and tuned until the desired fit is obtained. This fitting can be easily done in Excel using the “Solve…” menu item. As you’d expect with four parameters and only eight datapoints, the fit is quite good, and their estimate is quite close to the actual value of Nte(Ps).
Amusingly, the result of equation (7) is then used in another fitted (tuned) equation, number (8). This is:
where So is total solar irradiation.
This is their piece de resistance, their MIRACLE equation, wherein they are saying the surface temperature of eight different planets and moons can be calculated from just two variables— Pr, the surface pressure, and So, the total Solar irradiation. This is what amazes the folks in the crowd so much that they write and tell me there is “evidence” that N&Z are right.
Obviously, there is another tuned parameter in equation (8), so we can rewrite this one symbolically as:
Ts = t5 * (Solar + adjustment ) ^ 1/4 * Nte(Ps). (8Sym)
Let me pause a minute and point something out about equation (8). The total solar irradiation Solar ranges from over 9,000 W/m2 for Mercury down to 1.51 W/m2 for Triton. Look at equation 8. How will adding the adjustment = 0.0001325 to any of those values before taking the fourth root make the slightest bit of difference in the result? That’s just bizarre, that is. They say they put it in so that the formula will be accurate when there is no solar, so it will give the background radiation of 3 Kelvins. Who cares? Truly, it changes Ts by a maximum of a thousandth of a degree for Triton. So for the moment let me remove it, as it makes no practical difference and it’s just confusing things.
Back to the tale. Removing the adjustment and substituting equation 7 into equation 8 we get:
Ts = t5 * Solar^0.25 * e^(t1 * Ps ^ t2 + t3 * Ps ^ t4) (eqn 9)
This is amazing. These guys are seriously claiming that with only eight datapoints and no less than five tunable parameters , they can calculate the surface temperature of the eight planets knowing only their surface pressure and solar irradiation. And with that many knobs to turn, I am sure they can do that. I did it on my own spreadsheet using their figures. I get about the same values for t1 through t5. But that proves nothing at all.
I mean … I can only stand in awe at the sheer effrontery of that claim. They are using only eight datapoints and five tunable parameters with a specially-designed ad-hoc equation with no physical basis. And they don’t think that’s odd in the slightest.
I will return to this question of the number of parameters in a bit, because even though it’s gobsmacking what they’ve done there, it’s not the best part of the story. Here’s the sting in the tale. We can also substitute equation (7) into equation (8) in a slightly different way, using the middle term in equation 7. This yields:
Ts = t5 * Solar^0.25 * Ts / Tgb (eqn 10)
This means that if we start out by knowing the surface temperature Ts on the right side of the equation, we can then calculate Ts on the left side … shocking, I know, who would have guessed. Let’s check the rest of the math in equation (10) to see why that works out.
Upon inspection it can be seen that the first part of the right side of equation (10),
t5 * Solar^0.25
is an alternate form of the familiar Stefan-Boltzmann equation relating temperature and radiation. The S-B equation can be written as
T = (Solar / c1) ^ 0.25.
where T is temperature and c1 is a constant equal to the S-B constant times the emissivity. We can rewrite this as
T = 1/(c1^0.25) * Solar^0.25
Setting another constant c2 equal to 1 / (c1^0.25) gives me the Stefan-Boltzmann equation as:
T = c2 * Solar^0.25
But this is exactly the form of the first part of the right side of equation 10. More to the point, it is an approximation of the graybody temperature of the planet Tgb.
We can check this by observing that if emissivity is .9 then constant c1 is 5.103E-8, and c2 is therefore about 66. However, that value will be reduced by the rotation of the planet. Per the N&Z formula in their latest post, that gives a value of about 27.
Their fitted value is 25, not far from the actual value. So curiously, what it turns out they’ve done is to estimate the Stefan-Boltzmann constant by a bizarre curve fitting method. And they did a decent job of that. Actually, pretty impressive considering the number of steps and parameters involved.
But since t5 * Solar^0.25 is an estimation of the graybody temperature of the planet Tgb, that means that Equation 10 reduces from
Ts = t5 * Solar^0.25 * Ts / Tgb (eqn 10)
to
Ts = Tgb * Ts / Tgb.
and finally to
Ts = Ts
TA-DA!
CONCLUSION
Let me recap the underlying effect of what they have done. They are looking at eight planets and moons.
1. They have used an equation
e^(t1 * Ps ^ t2 + t3 * Ps ^ t4)
with four free parameters to yield an estimate of Ts/Tgb based on surface pressure. As one would expect given the fact that there are half as many free parameters as there are data points, and that they are given free choice to pick any form for their equation without limit, this presents no problem at all, and can be done with virtually any dataset.
2. They have used an equation
t5 * Solar^0.25
with one free parameter in order to put together an estimate of Tgb based on total planetary insolation. Since Tgb does depend inter alia on planetary insolation, again this presents no problem.
3. They have multiplied the two estimates together. Since the result is an estimate of Tgb times an estimate of Ts/Tgb, of course this has the effect of cancelling out Tgb.
4. They note that what remains is Ts, and they declare a MIRACLE.
Look, guys … predicting Ts when you start out with Ts? Not all that hard, and with five free parameters and a choice of any equation no matter how non-physically based, that is no MIRACLE of any kind, just another case of rampant curve fitting …
Finally, there is a famous story in science about this kind of pseudo-scientific use of parameters and equations, told by Freeman Dyson:
We began by calculating meson–proton scattering, using a theory of the strong forces known as pseudoscalar meson theory. By the spring of 1953, after heroic efforts, we had plotted theoretical graphs of meson–proton scattering. We joyfully observed that our calculated numbers agreed pretty well with Fermi’s measured numbers. So I made an appointment to meet with Fermi and show him our results. Proudly, I rode the Greyhound bus from Ithaca to Chicago with a package of our theoretical graphs to show to Fermi.
When I arrived in Fermi’s office, I handed the graphs to Fermi, but he hardly glanced at them. He invited me to sit down, and asked me in a friendly way about the health of my wife and our newborn baby son, now fifty years old. Then he delivered his verdict in a quiet, even voice. “There are two ways of doing calculations in theoretical physics”, he said. “One way, and this is the way I prefer, is to have a clear physical picture of the process that you are calculating. The other way is to have a precise and self-consistent mathematical formalism. You have neither.
I was slightly stunned, but ventured to ask him why he did not consider the pseudoscalar meson theory to be a selfconsistent mathematical formalism. He replied, “Quantum electrodynamics is a good theory because the forces are weak, and when the formalism is ambiguous we have a clear physical picture to guide us. With the pseudoscalar meson theory there is no physical picture, and the forces are so strong that nothing converges. To reach your calculated results, you had to introduce arbitrary cut-off procedures that are not based either on solid physics or on solid mathematics.”
In desperation I asked Fermi whether he was not impressed by the agreement between our calculated numbers and his measured numbers. He replied, “How many arbitrary parameters did you use for your calculations?”
I thought for a moment about our cut-off procedures and said, “Four.”
He said, “I remember my friend Johnny von Neumann used to say, with four parameters I can fit an elephant, and with five I can make him wiggle his trunk.” With that, the conversation was over. I thanked Fermi for his time and trouble, and sadly took the next bus back to Ithaca to tell the bad news to the students.
The Nikolov and Zeller equation contains five parameters and only eight data points. I rest my case that it is not a MIRACLE that they can make the elephant wiggle his trunk, but an expected and trivial result of their faulty procedures.
My regards to everyone,
w.
PS—There is, of course, a technical term for what they have done, as there are no new mistakes under the sun. It is called “overfitting”. As Wikipedia says, “Overfitting generally occurs when a model is excessively complex, such as having too many parameters relative to the number of observations.” Five parameters is far, far too many relative to eight observations, that is a guaranteed overfit.
PPS—One problem with N&Z’s MIRACLE equation is that they have not statistically tested it in any way.
One way to see if their fit is even remotely valid is to leave out some of the datapoints and fit it again. Of course with only eight datapoints to start with, this is problematic … but in any case if the fitted parameters come out radically different when you do that, this casts a lot of doubt on your fit. I encourage N&Z to do this and report back on their results. I’d do it, but they don’t believe me, so what’s the point?
Aother way to check their fit is to divide the dataset in half, do the fit on one half, and then check the results on the other half. This is because fitted equations like they are using are known to perform very poorly “out of sample”, that is to say on data not used to fit the parameters. Given only eight data points and four parameters for equation 7, of course this is again problematic, since if you divide the set in half you end up with as many parameters as data points … you’d think that might be a clue that the procedure is sketchy but what do I know, I was born yesterday. In any case I encourage N&Z to perform that test as well. My results from that test say that their fit is meaningless, but perhaps their test results will be different.
[UPDATE] One of the commenters below said:
Willis – go ahead – fit an elephant. Please!
Seriously N&Z are only demonstrating in algebra what has been observed in experiments, that heating a gas in a sealed container increases both pressure and temperature.
OK, here’s my shot at emulating the surface temperature using nothing but the data in the N&Z chart of planetary body properties:
Figure 1. Willis’s emulation of the surface temperature of the planetary bodies.
My equation contains one more variable and two less parameters than the N&Z equation. Remember their equation was:
Ts = 25.3966 * Solar^0.25 * e^(0.233001 * Pressure ^ 0.0651203 + 0.0015393 * Pressure ^ 0.385232)
My equation, on the other hand, is:
Ts = 0.8 * Tgb + 6.9 * Density + 0.2 * Gravity)
Note that I am absolutely not making any claim that temperature is determined by density and gravity. I am merely showing that fitting a few points with a few variables and a few parameters is not all that difficult. It also shows that one can get the answer without using surface pressure at all. Finally, it shows that neither my emulation nor N&Z’s emulation of the planetary temperatures are worth a bucket of warm spit …
[UPDATE 2] I figured that since I was doing miracles with the N&Z miracle equation, I shouldn’t stop there. I should see if I could beat them at their own game, and make a simpler miracle. Once again, their equation:
Ts = 25.3966 * Solar^0.25 * e^(0.233001 * Pressure ^ 0.0651203 + 0.0015393 * Pressure ^ 0.385232)
My simplified version of their equation looks like this:
Ts = 25.394 * Solar^0.25 * e^(0.092 * Pressure ^ 0.17)
Curiously, my simplified version actually has a slightly lower RMS error than the N&Z version, so I did indeed beat them at their own game. My equation is not only simpler, it is more accurate. They’re free to use my simplified miracle equation, no royalties necessary. Here are the fits:
Figure 2. A simpler version of the N&Z equation 8
Again, I make no claim that this improves things. The mere fact that I can do it with two less tuned parameters (three instead of five) than N&Z used does not suddenly mean that it is not overfitted.
Both the simplified and the complex version of the N&Z equations are nothing but curve fitting. This is proven by the fact that we already have three simple and very different equations that hindcast the planetary temperatures. That’s the beauty of a fitted equation, if you are clever you can fit a lot using only a little … but THAT DOESN’T MEAN THAT PRESSURE DETERMINES TEMPERATURE.
For example, I can do the same thing without using pressure at all, but using density instead. Here’s that equation:
Ts = 25.491 * Solar^0.25 * e^(0.603 * Density ^ 0.201)
And here’s the results:
Figure 3. An emulation of the planetary temperatures, using density instead of pressure.
Does this now mean that the planetary temperature is really controlled by density? Of course not, this whole thing is an exercise in curve fitting.
w.
Discover more from Watts Up With That?
Subscribe to get the latest posts sent to your email.
Konrad Hartmann ran a nice little physical experiment at Tallbloke’s showing the relationship between pressure and temperature in the presence of a radiative source. This should help the understanding of the tire pumpers, like George Smith.
Regarding the curve fitting issue, Wilis seems to be of the opinion that some number of parameters, X, automatically becomes an overfit. I agree with John Day. It depends on what is being modeled and how convoluted the thing that one is fitting to is. In the case of an almost random walk looking data set, like historical temperature, pushing the number of parameters to force the fit is a problem. I don’t see that N&Z are doing that. They are not fitting an elephant. Still, there is a problem with the number of parameters. It could be that their model is wrong and that their results are only due to overfit. But it could also be true that their model is right, in which case the number of parameters do not make it wrong, as Willis seems to think.
From my perspective, N&Zs main point is that higher pressure means more gas density which in turn means more captured radiative energy which then means higher temperature. I can’t see how Willis’ quibbling about curve fitting helps with proving or disproving that idea. Konrad Hartmann’s experiment, however, is a real contribution.
LazyTeenager says: “To summarize: the derivation is all about individual molecules, their kinetic energies and how that determines the impulse on the walls of the container, and hence the pressure. There is no need to consider collisions between molecules at all. It is not relevant.”
Nice try, Lazy. How you do squirm! If it is not relevant, then why did you bring them into the discussion? January 23, 2012 at 9:53 pm:
“The ideal gas theory considers molecules to be too small to collide with each other and ignores that factor as being…” –Lazy Teenager.
William M. Connolley says: “…if you take the GCMs code, and the correct initial conditions, they will predict tomorrow’s weather, or next week’s weather, for you.”
Beyond that, they’re worthless.
Surprising that someone would “over fit” to that extent and not recognize doing it. However, I caught the “non-mathematician scientists” at work doing that sort of thing routinely. So, maybe not so surprising.
Doesn’t mean they’re conclusions are wrong. Simply means their approach gives, at best, extremely weak confirmation of cause and effect. Basically worthless as it stands.
Willis says: “Thanks, tallbloke. Here on this thread I’m focused on what they did in equations 7 and 8. There are many aspects to their work. I oppose the idea that it can all be settled in one thread.”
(1) It’s your thread, Willis; (2) covering all of N&Z in a single thread is proven bogus by what we’ve seen here. Even restricting discussion to two lousy equations has resulted in chaos. We even drew in Der Übertroll, to no particular gain. (3) It’s your thread.
Naive question: Isn’t this pretty close to the ‘elevator speech’ that Willis was asking for a few posts back? Sounds pretty common-sensical. One might ask how the gas captures radiative energy. In Willis’s challenge, the gas was transparent to IR. So it gets energized by conduction from the IR-absorbing surface. No GHGs necessary. And gravity makes it denser. All you need is an external source of radiation.
/Mr Lynn
Good response… That was my first reaction also, and was wondering if I was “out to lunch”. I still may be, but it’s nice to have company. GK
“”””” Edim says:
January 23, 2012 at 11:13 pm
George E. Smith; says:
“So what part of THIS SYTEM IS “NOT” IN EQUILIBRIUM is it that YOU do not understand. It is not even a closed system, since you yourself said it is being illuminated by a star which is constantly feeding energy into the system.
OOoops !! the ideal gas law applies ONLY to closed systems in thermal equilibrium. Gee!! that requires the whole system to be at a single uniform Temperature.”
This is not correct. Ideal gas law applies to any point (x,y,z) of gas when ideal gas approximation can be used with reasonable accuracy (real gas that behaves sufficiently like an ideal gas). When not, there are more detailed equations of state gor real gases. Equilibrium or not is irrelevant. “””””
Well Edim, now we have a real conundrum: “””””
This is not correct. Ideal gas law applies to any point (x,y,z) of gas when ideal gas approximation can be used with reasonable accuracy “””””
How about that: I went to your point (x,y,z) to look for the ideal gas, and I found nothing at all there. Now it is possible that something might at some time have been near there; but Werner Heisenberg tells me it doesn’t stay there any length of time, and I have no idea where the hell it went. Whatever it was, that maybe at one time was at x,y,z, it sure as hell wasn’t a gas, let alone an ideal gas; maybe an atom or molecule came by sometime; but NO, you cannot have a gas at x,y,z .
Whether the gas is real or ideal is somewhat inconsequential Yes we know that real gases consist of a very large number of molecules, or else they don’t have a definable Temperature, and without a definable Temperature, any equation of state, whether ideal or Van der Waals, or other is useless.
If the N&Z miracle equation can foretell the Temperature of the atmosphere on any planet, evidently without any consideration of what energy surces (if any) are present, then clearly I need to look for some alternative work from what I do now.
Now if you have five variables and eight observed data points, then I suppose you can write a formula that recovers those eight data points quite well, well with eight variables you should get all eight with complete accuracy. I have absolutely no qualms about such an equation.
One can even do it in words: eg “Bad boys rape our young girls, but violet gives willingly.”
Hows that for a miracle equation ? Every Radio Shack hobbyist knows that equation derives all the colors of the standard resistor color code; a “cool” way to remember a finite number of simple facts. Unfortunately, it gives a completely wrong set of answers for the colors of the rainbow. Dang ! I thought I was on to something there.
Dr Roy Spencer’s “third order polynomial” which he warns us is for entertainment only crashes and burns outside of the sphere of influence for which Roy designed it. That’s why he told us it was just for fun.
N&Z miracle is for fun only, it replicates the finite set of numbers that were used to create it.
scf (Jan. 24, 2012 at 3:45 pm):
Thanks for taking the time to respond and for the kind words! Your response make me wonder whether my message got across to you as I had intended. Though independent statistical events and datapoints are related, they are different concepts. Though IPCC assessment report 4 references datapoints, it does not reference the complete set of independent statistical events, the so-called “statistical population” of the IPCC’s study yet observed events from this population (if any) provide the sole basis for testing this study’s model.
“”””” LazyTeenager says:
January 23, 2012 at 9:53 pm
George E Smith says
Note that the temperature T of a system in equilibrium can be computed from the just kinetic energy of the moving gas particles and their mutual collisions (density, implying pressure). We don’t need to know the radiative aspects of the system to compute the temperature! What part of the Ideal Gas Law do you not understand here? “””””
——-
Somewhat ironically George claims that temperature can be calculated from —mutual—- collisions of gas molecules. This is not true.
The ideal gas theory considers molecules to be too small to collide with each other and ignores that factor as being
Seems George needs to spend some time actually studying the ideal as law “””””
Now Lazy; in my book, there is no lower form of animal life, than someone, who doesn’t even have the guts or confidence in their own position, to actually put their name to such unmitigated balderdash, as you have alleged here; to whit.
“”””” George E Smith says
Note that the temperature T of a system in equilibrium can be computed from the just kinetic energy of the moving gas particles and their mutual collisions (density, implying pressure). We don’t need to know the radiative aspects of the system to compute the temperature! What part of the Ideal Gas Law do you not understand here? “””””
——-
Somewhat ironically George claims that temperature can be calculated from —mutual—- collisions of gas molecules. This is not true. “””””
So now you truly lazy bladderwort, Why don’t you just go and find exactly WHERE it was that ” I SAID THAT” Just try and find where I said anything even vaguely resembling that.
For a start, my command of the English language, is sufficiently adequate, that I would never write anything quite that garbled.
So do us all a favor, and go back and find WHO IT WAS WHO ACTUALLY DID SAY THAT
Then why don’t you come back and give me an apology.
Do not put words into my mouth. I choose my words carefully; words have meaning; different words have different meaning; I choose to not use different words.
So if you are going to quote me; do so IN MY WORDS They mean exactly what I intended them to mean.
And do it quickly before I decide to post what I REALLY think of people like you.
William M. Connolley says:
January 24, 2012 at 2:44 pm
Ooooh, William doesn’t think that climate models are tuned, excuuuuse me! If this was Wikipedia, he’d just erase my statements …
William, you are an unpleasant man who has made a life out of erasing any opposition to your views from the Wikipedia pages on climate science. You did it in such a slimy way that Wikipedia finally threw you out on your keester.
Let me be clear. I do not like seeing you on my threads. You are a wrecker and a destroyer, not a builder or a teacher. Dissension and disagreement follow you like a bad smell.
I advise everyone, this man is a danger. Google his name, he is quite infamous. He is quite happy to do just about any underhanded thing to advance AGW alarmism. There has rarely been a clearer case of Noble Cause Corruption, as throughout his whole skeevy list of actions that finally cost him his title at Wikipedia, while ruthlessly squelching opposition views, he has always seen himself as being the white knight crusader for saving the planet, and devil take the hindmost.
William, you are so far in the wrong that it is doubtful that you could ever get back in the right. You are a dishonest, untrustworthy, and nasty person that has burnt all possible bridges. You can get back, but you’ll need to build the bridges.
Now, I have no desire to ban you from here, that’s not my style. I would greatly prefer it if you talked about nothing but the science. It would be nice if you put a curb on your tongue and a muzzle on your insinuations, but if you want to be a jerk I wont stop you. I just wanted to let you know how I feel about your presence. It makes me want to put Vaporub on my upper lip.
w.
… aw, man, now I’ve gone and blown my shot at reforming, I’ve fallen back down to the bottom of the twelve steps. Nothin’ else to do but stand up and start climbing all over again …
Ira Glickstein, PhD says:
January 24, 2012 at 6:01 pm
Thanks, Ira. Practical experience is greatly under-rated in the sciences. Being out in the real world as a commercial fisherman and a mechanic and a diver and a builder and the like give a man a pretty good BS detector.
I’d never heard of N&Z until they were posted here. I didn’t read their post until I read yours. My comment on their kind of gravito-thermal theories was:
The other thing to my advantage is that I worked as a bureaucrat for some years. I’ve both written and read dozens and dozens of project proposals and grant reports and replies to overseas donors and the like. Bureaucrats love the kind of bafflegab that N&Z are putting out above, the type that seems to say something but says nothing at all. So for example, sorting through dozens of applications which I could fund, partially fund, or not fund has definitely helped my ability to see through empty rhetoric.
That’s why I pushed people to give an elevator speech, to show that they’re not just all hat and no cattle, that they do understand the theory that they are defending.
All the best, keep up the good fight,
w.
Willis says: “…now I’ve gone and blown my shot at reforming…”
Der Übertroll strikes again!
Willis, you are too kind in your rebuke of Joel Shore. C’mon Joel, which models are out there for skeptics to tune, test, and evaluate first hand? Willis and Tilo really say all that needs to be said about the models. Nostradamus predictions brilliantly hindcast as well..
..
and John Marshall says: “… since there are no GHG’s in the Jovian atmosphere there must be another mechanism to create that extra heat and that must be gravity.”
Well that solves it then! Good grief, Must avoid using “must” on all but laws and first principals lest the word itself becomes your argument.
Seems even more apropos now, so I will repost what I said ten days ago on “A matter of some gravity” thread. Willis kind of echoed the second one. I think eventually someone will also realize what I did about Jelbring’s greenhouse effect definition being actually his own invention.
“Jim D says:
January 14, 2012 at 8:11 pm
Elevator speech on Jelbring:
The atmosphere is warmer as you go down in it because of the adiabatic lapse rate (g/cp) therefore greenhouse gases have nothing to do with the adiabatic lapse rate (true). The adiabatic lapse rate causes the greenhouse effect (false).
Elevator speech on Nikolov and Zeller:
You can fit a four-parameter curve to surface temperature over blackbody temperature ratio of seven solar system atmospheres just as a function of mass (not albedo or composition or clouds!). Therefore these other things don’t matter. It doesn’t matter that you can’t derive this curve except by a mathematical fit to the data. Who needs to explain why it fits?”
Actually, it was just a wee time-out, some misdirection. His friends/pawns continued in his brief absence, and he’s now back on the Editorial Board, making up for lost time.
jorgekafkazar says:
January 24, 2012 at 10:05 pm (Edit)
Surrender, puny humans! No one can resist Der Übertroll!
Very funny, jorge,
w.
WE, you do sound threatened. But thanks for showing up your real side, you’d managed to hide it for quite a while. As for your link; neither you nor Watts understands wiki’s NPOV policy. This may help a bit.
> you are so far in the wrong
Oh dear. Because, as I clearly stated further up, I agree with you on N+Z. Does that mean you’re wrong, too?
“Note that I am absolutely not making any claim that temperature is determined by density and gravity. I am merely showing that fitting a few points with a few variables and a few parameters is not all that difficult. It also shows that one can get the answer without using surface pressure at all.”
1) What does Willis think the definition of the density of a compressible gas would look like without using pressure in the equation?
2) Can Willis or anyone else produce an equation which which will lie along a smooth curve which gives fit to the planetary surface temperatures using the radiative properties of gases rather than surface pressure which is as good as N&Z’s, or Willis’ density solution?
I really think Willis comments need moderation. You can be funny (which Willis often is) or even a bit smug because you know something on this particular niche field, but you should never be demeaning. People make mistakes because they dont know something but not because they are stupid. What is happening here is that they are chased away to places where science does not run high. Take the Conolley comment. Even though he erased many of my contributions he’s not beyond repair as noboddy is. I was actually happyly surprised he popped up here. If he participates here within the rules he should be welcome. He might find out there are other ways to argument then to suppress opinion. I dont think he got an opportunity to learn that here today.
PS: I agree with your scientific opinion on this issue and greatly value your contributions. Just the style sucks big time lately. I guess Lucy was trying to say something similar but was not clear enough. ;-).
I have added an additional update, showing that I can beat their emulation of the planetary temperatures (lower RMS error) using only three parameters instead of five … take a look. I’ve done it with density instead of pressure as well.
w.
One of the mysteries of this paper is where the observed T_s values (which they claim to predict) came from. For most, no information on sources is given at all. Europa? Triton? What did they use and how did they get a global figure?
One of the hallmarks of WUWT has been a policy of eschewing ad hominem attacks on other commenters. From Anthony’s Policy page:
/Mr Lynn
So many errors everywhere, best to just have a guess and then argue with everyone over their guess. Then call it science at work.
Would it not be better if Nikolov, Zeller and Eschenbach and all the other great minds and interested parties here work together to Iron out the kinks and come up with something solid?
As an innocent bystander (lol) the condescending tone and the “I’m right and you’re wrong” attitudes are becoming old and worn out and sooooo boring!!
Ha! I did a quick search and found an article on Contracting (IRYW) disease, some folks may need to be tested.
Beware Contracting “I’m Right, You’re Wrong” Disease?
http://artpetty.com/2010/03/03/
Willis Eschenbach says:
Your usual can-do “do-it-myself” attitude seems to have disappeared here. However, I’ll note that you have already essentially shown the claim you are making here to be wrong. As I recall, you basically showed that for all of its complexity, the GISS Model E global temperature result hindcast / prediction closely follows the radiative forcings. Hence, if you don’t include the fact that radiative forcings have gone up due to the anthropogenic component over the half century, you aren’t going to be able to do a good job reproducing the global temperature record.
The parameters that you speak of are not parameters that are tuned to reproduce the global temperature record. They are parameters that, to the extent they are adjusted, are adjusted to reproduce things like correct cloud fraction or correct sea ice fraction. (It is also worth noting that the climateprediction.net experiment has done lots of runs where they have allowed such parameters to take on a whole spectrum of values.)
I have already addressed this in my previous comment:
So, yes, you are correct that the fact that the models can hindcast does not turn out to constrain the climate sensitivity that much, because there is too much uncertainty in the anthropogenic aerosol forcing and so different estimates of that and different climate sensitivities can pretty much offset each other.
However, there is no evidence that one can reproduce the temperature rise over the last ~40 years without including a steadily increasing radiative forcing and the only known such forcing is the forcing due to added greenhouse gases.