Guest Post by Willis Eschenbach
Well, I wasn’t going to mention this paper, but it seems to be getting some play in the blogosphere. Our friend Nicola Scafetta is back again, this time with a paper called “Solar and planetary oscillation control on climate change: hind-cast, forecast and a comparison with the CMIP5 GCMs”. He’s posted it up over at Tallbloke’s Talkshop. Since I’m banned over at Tallbloke’s, I thought I’d discuss it here. The paper itself is here, take your Dramamine before jumping on board. Dr. Scafetta has posted here on WUWT several times before, each time with his latest, greatest, new improved model. Here’s how well Scafetta’s even more latester, greatester new model hindcasts, as well as what it predicts, compared with HadCRUT4:
Figure 1. Figure 16A from Scafetta 2013. This shows his harmonic model alone (black), plus his model added to the average of the CMIP5 models following three different future “Representative Concentration Pathways”, or RCPs. The RCPs give various specified future concentrations of greenhouse gases. HadCRUT4 global surface temperature (GST) is in gray.
So far, in each of his previous three posts on WUWT, Dr. Scafetta has said that the Earth’s surface temperature is ruled by a different combination of cycles depending on the post:
First Post: 20 and 60 year cycles. These were supposed to be related to some astronomical cycles which were never made clear, albeit there was much mumbling about Jupiter and Saturn.
Second Post: 9.1, 10-11, 20 and 60 year cycles. Here are the claims made for these cycles:
9.1 years : this was justified as being sort of near to a calculation of (2X+Y)/4, where X and Y are lunar precession cycles,
“10-11″ years: he never said where he got this one, or why it’s so vague.
20 years: supposedly close to an average of the sun’s barycentric velocity period.
60 years: kinda like three times the synodic period of Jupiter/Saturn. Why three times? Why not?
Third Post: 9.98, 10.9, and 11.86 year cycles. These are claimed to be
9.98 years: slightly different from a long-term average of the spring tidal period of Jupiter and Saturn.
10.9 years: may be related to a quasi 11-year solar cycle … or not.
11.86 years: Jupiter’s sidereal period.
The latest post, however, is simply unbeatable. It has no less than six different cycles, with periods of 9.1, 10.2, 21, 61, 115, and 983 years. I haven’t dared inquire too closely as to the antecedents of those choices, although I do love the “3” in the 983 year cycle. Plus there’s a mystery ingredient, of course.
Seriously, he’s adding together six different cycles. Órale, that’s a lot! Now, each of those cycles has three different parameters that totally define the cycle. These are the period (wavelength), the amplitude (size), and the phase (starting point in time) of the cycle.
This means that not only is Scafetta exercising free choice in the number of cycles that he includes (in this case six). He also has free choice over the three parameters for each cycle (period, amplitude, and phase). That gives him no less than 18 separate tunable parameters.
Just roll that around in your mouth and taste it, “eighteen tunable parameters”. Is there anything that you couldn’t hindcast given 18 different tunable parameters?
Anyhow, if I were handing out awards, I’d certainly give him the first award for having eighteen arbitrary parameters. But then, I’d have to give him another award for his mystery ingredient.
Because of all things, the mystery ingredient in Scafetta’s equation is the average hindcast (and forecast) modeled temperature of the CMIP5 climate models. Plus the mystery ingredient comes with its own amplitude parameter (0.45), along with a hidden parameter for the zero point of the average model temperatures before being multiplied by the amplitude parameter. So that makes twenty different adjustable parameters.
Now, I don’t even know what to say about this method. I’m dumbfounded. He’s starting with the average of the CMIP5 climate models, adjusted by an amplitude parameter and a zeroing parameter. Then he’s figuring the deviations from that adjusted average model result based on his separate 6-cycle, 18-parameter model. The sum of the two is his prediction. I truly lack words to describe that, it’s such an awesome logical jump I can only shake my head in awe at the daring trapeze leaps of faith …
I suppose at this point I need to quote the story again of Freeman Dyson, Enrico Fermi, “Johnny” Von Neumann, and the elephant. Here is Freeman Dyson, with the tale of tragedy:
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 self-consistent 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.
Given that lesson from Dyson, and bearing in mind that Scafetta is using a total of 20 arbitrary parameters … are we supposed to be surprised that Nicola can make an elephant wiggle his trunk? Heck, with that many parameters, he should be able to make that sucker tap dance and spit pickle juice …
Now, you can expect that if Nicola Scafetta shows up, he will argue that somehow the 20 different parameters are not arbitrary, oh, no, they are fixed by the celestial processes. They will likely put forward the same kind of half-ast-ronomical explanation they’ve used before—that this one represents (2X+Y)/4, where X and Y are lunar precession cycles, or that another one’s 60 year cycle is kind of near three times the synodic period of Jupiter and Saturn (59.5766 years) and close is good enough, that kind of thing. Or perhaps they’ll make the argument that Fourier analysis shows peaks that are sort of near to their chosen numbers, and that’s all that’s needed.
The reality is, if you give me a period in years, I can soon come up with several astronomical cycles that can be added, subtracted, and divided to give you something very near the period you’ve given me … which proves nothing.
Scafetta has free choice of how many cycles to include, and free choice as to the length, amplitude, and phase of each those cycles. And even if he can show that the length of one of his cycles is EXACTLY equal to some astronomical constant, not just kind of near it, he still has totally free choice of phase and amplitude for that cycle. So to date, he’s the leading contender for the 2013 Johnny Von Neumann award, which is given for the most tunable parameters in any scientific study.
The other award I’d give this paper would be for Scafetta’s magical Figure 11, which I reproduce below in all its original glory.
Figure 2. Scafetta’s Figure 11 (click to enlarge) ORIGINAL CAPTION: (Left) Schematic representation of the rise and fall of several civilizations since Neolithic times that well correlates with the 14C radio- nucleotide records used for estimating solar activity (adapted from Eddy’s figures in Refs. [90, 91]). Correlated solar-climate multisecular and millennial patterns are recently confirmed [43, 44, 47]. (Right) Kepler’s Trigon diagram of the great Jupiter and Saturn conjunctions between 1583 to 1763 [89], highlighting 20 year and 60 year astronomical cycles, and a slow millennial rotation.
First off, does that graphic, Figure 11 in Scafetta’s opus, make you feel better or worse about Dr. Scafetta’s claims? Does it give you that warm fuzzy feeling about his science? And why are Kepler’s features smooched out sideways and his fingers so long? At least let me give the poor fellow back his original physiognomy.
There, that’s better. Next, you need to consider the stepwise changes he shows in “carbon 14”, and the square-wave nature of the advance and retreat of alpine glaciers at the lower left. That in itself was good, I hadn’t realized that the glaciers advanced and retreated in that regular a fashion, or that carbon 14 was unchanged for years before and after each shift in concentration. And I did appreciate that there were no units for any of the four separate graphs on the page, that counted heavily in his favor. But what I awarded him full style points for was the seamless segue from alpine glaciers to the “winter severity index” in the year 1000 … that was a breathtaking leap.
And as you might expect from a man citing Kepler, Scafetta treats scientific information like fine wine—he doesn’t want anything of recent vintage. Apparently on his planet you have to let science mellow for some decades before you bring it out to breathe … and in that regard, I direct your attention to the citation in the bottom center of his Figure 11, “Source: Geophysical Data, J. Biddy J. B. Eddy (USA) 1978″. (Thanks to Nicola for the correction, the print was too small to read.)
Where he stepped up to the big leagues, though, is in the top line in the chart. Click on the chart to enlarge it if you haven’t done so yet, so you can see all the amazing details. The “Sumeric Maximum”, the collapse of Machu Pichu, the “Greek Minimum”, the end of the Maya civilization, the “Pyramid Maximum” … talk about being “Homeric in scope”, he’s even got the “Homeric Minimum”.
Finally, he highlights the “20 year and 60 year astronomical cycles” in Kepler’s chart at the right. In fact, what he calls the “20 year” cycles shown in Kepler’s dates at the right vary from 10 to 30 years according to Kepler’s own figures shown inside the circle, and what he calls the “60 year astronomical cycles” include cycles from 50 to 70 years …
In any case, I’m posting all of this because I just thought folks might like to know of Nicola Scafetta’s latest stunning success. Using a mere six cycles and only twenty tunable parameters plus the average of a bunch of climate models, he has emulated the historical record with pretty darn good accuracy.
…
And now that he has explained just exactly how to predict the climate into the future, I guess the only mystery left is what he’ll do for an encore performance. Because this most recent paper of his, this one will be very hard to top.
In all seriousness, however, let me make my position clear.
Are there cycles in the climate? Yes, there are cycles. However, they are not regular, clockwork cycles like those of Jupiter and Saturn. Instead, one cycle will appear, and will be around for a while, and then disappear to be replaced by some longer or shorter cycle. It is maddening, frustrating, but that’s the chaotic nature of the beast. The Pacific Decadal Oscillation doesn’t beat like a clock, nor does the El Nino or the Madden-Julian oscillation or any other climate phenomena.
What is the longest cycle that can be detected in a hundred year dataset? My rule of thumb is that even if I have two full cycles, my results are too uncertain to lean on. I want three cycles so I can at least get a sense about the variation. So for a hundred year dataset, any cycle over fifty years in length is a non-starter, and thirty-three years and shorter is what I will start to trust.
Can you successfully hindcast temperatures using other cycles than the ones Scafetta uses? Certainly. He has demonstrated that himself, as this is the fourth combination of arbitrarily chosen cycles that he has used. Note that in each case he has claimed the model was successful. This by no means exhausts the possible cycle combinations that can successfully emulate the historical temperature.
Does Scafetta’s accomplishment mean anything? Sure. It means that with six cycles and no less than twenty tunable parameters, you can do just about anything. Other than that, no. It is meaningless.
Could he actually test his findings? Sure, and I’ve suggested it to him. What you need to do is run the analysis again, but this time using the data from say 1910 to 1959 only. Derive your 20 fitted variables using this data alone.
Then test your 20 fitted variables against the data from 1960 to 2009, and see how the variables pan out.
Then do it the other way around. Train the model on the later data, and see how well it does on the early data. It’s not hard to do. He knows how to do it. But if he has ever done it, I have not seen anywhere that he has reported the results.
How do I know all this? Folks, I can’t tell you how many late nights I’ve spent trying to fit any number and combination of cycles to the historical climate data. I’ve used Fourier analysis and periodicity analysis and machine-learning algorithms and wavelets and stuff I’ve invented myself. Whenever I’ve thought I have something, as soon as it leaves the training data and starts on the out-of-sample data, it starts to diverge from reality. And of course, the divergence increases over time.
But that’s simply the same truth we all know about computer weather forecasting programs—out-of-sample, they don’t do all that well, and quickly become little better than a coin flip.
Finally, even if the cycles fit the data and we ignore the ridiculous number of arbitrary parameters, where is the physical mechanism connecting some (2*X+T)/4 combination of two astronomical cycles, and the climate? As Enrico Fermi pointed out, you need to have either “a clear physical picture of the process that you are calculating” or “a precise and self-consistent mathematical formalism”.
w.
PS—Please don’t write in to say that although Nicola is wrong, you have the proper combination of cycles, based on your special calculations. Also, please don’t try to explain how a cycle of 21 years is really, really similar to the Jupiter-Saturn synodic cycle of 19+ years. I’m not buying cycles of any kind, motorcycles, epicycles, solar cycles, bicycles, circadian cycles, nothing. Sorry. Save them for some other post, they won’t go bad, but please don’t post them here.
One difficulty I see in decomposing climate into cycles (focusing on temperature as the measure) is that there are natural cycles and un-natural cycles.
The natural cycles come from nature. And there are plenty of these.
The un-natural are those that have been introduced through “homogenization” efforts. And there seem to be plenty of these–the perpetrators aren’t saying.
It’s like trying to pick the dirt from a slice of watermelon after dropping it at a picnic.
It is interesting that orbital position ‘theory’ has been used to predict an imminent minimum of solar magnetic cycle activity in 1987. We are of course all aware there as been an abrupt change to the solar magnetic cycle. It appears at this point in time that we are going to observe how solar magnetic cycle changes cause either a Bond cycle or a Heinrich event.
One must be careful not to throw the baby out with the bath water. Anomalies (correlation is an anomaly) are often either ignored (because there is no physical explanation for the anomaly) or covered up with an incorrect theory. An example is the creation of dark matter to explain the spiral galaxy rotational anomaly. The true physical explanation as to what causes the spiral galaxy rotational anomaly is connected with this problem.
There is a physical explanation as to why the orbital position of the planets affects the solar magnetic cycle and why the solar magnetic cycle affects the magnetic fields of the planets.
http://link.springer.com/article/10.1007%2FBF00148211
Solar Physics : 1987, Volume 110, Issue 1, pp 191-210
We employ the JPL long ephemeris DE-102 to study the inertial motion of the Sun for the period A.D. 760–2100. Defining solar orbits with reference to the Sun’s successive close approaches to the solar system barycenter, occurring at mean intervals of 19.86 yr, we find simple relationships linking the inertial orientation of the solar orbit and the amplitude of the precessional rotation of the orbit with the occurrence of the principal prolonged solar activity minima of the current millenium (the Wolf, Spörer, and Maunder minima). The progression of the inertial orientation parameter is controlled by the 900-yr ‘great inequality’ of the motion of Jupiter and Saturn, while the precessional rotation parameter is linked with the 179-yr cycle of the solar inertial motion previously identified by Jose (1965). A new prolonged minimum of solar activity may be imminent.
The solar magnetic cycle changes affect the planet magnetic fields (there is an imprint so to speak on the planets due to solar magnetic cycle changes) which explains the Bond climate cycle (850, 1350, 1850 years between events) and the Heinrich abrupt climate change events (super strong Bond events 8000 to 10000 years between events).
One method of identifying and solving a scientific problem where there are one or more fundamental theory errors is to look for anomalies and then to play with hypotheses to try to make all of the anomalies go away, in that there is a physical explanation for what is observed.
One of the best anomalies to find and solve this problem (solar magnetic cycle causing cyclic planetary magnetic field anomalies) is the anomalous magnetic field orientation of Uranus and Neptune. Very interesting.
Willis,
your great maritime knowledge and experience as seaman is largely missed. The heat for this planet comes from the sun. Any variation in sunray will find some reflection in air temperatures. However, by far the most heat from the sun is taken up by the oceans; some is quickly released to the atmosphere, other is stored temporarily, or for a long time. Once the heat is in the sea, the sea “determine” the release, based on its extreme complex structure, dynamic, and overall mean temperature, which is merely four degree Celsius. Any sun cycle can be traced in air temperature changes. But that will vary according ocean condition, and presumably only by a marginal figure. The sun ray for the earth has been extremely constant over millions of years. The sun has never caused an ice age. The ocean can generate the next ice age, or (any significant) climatic change within a couple of months and years. However, your critic to Dr. Nicola Scafetta’s air temperature cycle assessment should have been as harsh as done.
For the record, I was aware of Dr. Scaffetta’s latest paper several days ago. I chose not to run it because like Willis, I saw the seeds of cyclomania in it. Dr. Scafetta’s motives are noble, but his methods are the issue.
It is tempting to find cycles in data, the human mind looks for such patterns naturally, just as we see shapes in clouds. Certainly there’s nothing wrong with either, but as with predicting what the next cloud will look like based on previous cloud patterns (bunny, dog, bird, flower, angel) so it is equally difficult to predict the next climate pattern based on previous ones.
RealClimate’s show us your code has a bit more weight now. I still think it was a funny joke on his part to reply back that you need to take a class in wavelets.
There is one distinctive sea surface temperature cycle in North Atlantic:
http://blog.sme.sk/blog/560/310249/Natlantic.jpg
It is so profound, it actually drags the whole “global average” with it.
It goes linearly up and down, each period lasting 25-30 years.
Until we do not know, what causes these regular ups and downs, we are nowhere to talk about “climate sensitivity” or whatever. But we can see, that is switched to cold phase in 2006, that it will last probably another 20 years and that the whole AGW BS is only based on warm part of the cycle.
(North Pacific is quite similar).
Planetary “cycles” are just another form of astrology.
bryguyh says:
July 23, 2013 at 6:45 am
I like furrier series.
Would apply especially to the famous lynx-sunspot cycle.
Sorry for this off-topic post.
[snip – sorry doesn’t cut it. We have Tips and Notes for this – Anthony]
Clive E. Birkland says:
July 23, 2013 at 6:53 am
Dribble is not drivel.
Thanks for another excellent post.
Its a timely reminder how easy it is to fit a model to data if you allow to many degrees of freedom, not only IPCC is guilty of that.
Also thanks for the full Version of the Dyson / Fermi/ Neuman story. I only knew Von Neumans claims from before. All budding scientists should make a big mental note not only of von Neumans but also of Fermis words and keep it next to Feynmans advice about scientific honesty.
@GabrielHBay, July 23, 2013 at 6:50 am
There you go again: “My humble apologies to the Willis E. fan club who is clearly out in force”. You may think that’s witty. Perhaps you really are trying to be funny and I’m having a sense of humour failure. Please forgive me then for being so po-faced, but I do take offence at the idea that I’m somehow a Willis groupie. If there’s one thing I suspect would make Willis vomit is the suggestion that he might have a ‘fan club’ here at WUWT.
Why, oh why, is it such a crime in some people’s minds to show genuine respect for another man’s mind? I could after all say the same things about the inimitable Robert Brown, The Good Lord Monckton, Anthony himself, and many others here. I love their minds – big deal, so what? That doesn’t make me some kind of brown-tongue. I guess I just recognize greatness tempered by humility when I see it and want to sit at its feet and learn. That’s a facet of wisdom, not a fault to be ridiculed.
That said, I really do hope that I have got you all wrong and that the hard edge of cynicism I detect in your posts is all down to my own oversensitivity…
…The latest post, however, is simply unbeatable. It has no less than six different cycles, with periods of 9.1, 10.2, 21, 61, 115, and 983 years. …
Two words.
Ptolemy
Epicycles
That is all…
Monckton of Brenchley says:
July 23, 2013 at 3:49 am
For these reasons I should hesitate to dismiss the astronomical approach as astrological.
=========
Indeed. We calculate the ocean tides to great precision using an astrological approach. In contrast, tidal models that use first principles as is done with climate models are hopeless. They cannot deal with the complexity.
Early humans learned to predict the annual cycle of climate long before they understood the process. Again this was done using astrological techniques, building machines (temples) that measured the position of the sun in the heavens.
Have we really progressed so little?
Chaos limits what we can hope to predict from first principles using existing technology. Everything around us moves in cycles. These cycles persist, which suggests they are self-reinforcing and self-organizing. This holds the key to prediction in the face of chaos, as has been demonstrated time and again throughout history.
GabrielHBay says:
“ then interesting lines of though[t] like astronomical cycles, etc, being treated with disdain leaves me, shall one say, edgy.”
Nothing contradicts the alarmist accusation of us being “fake” skeptics as evidenced by our willingness to believe anything that counters CAGW like ardent disagreement among ourselves.
And personally I like sarcasm and enjoyed reading Willis’ take.
Juraj V. says:
July 23, 2013 at 7:07 am
There is one distinctive sea surface temperature cycle in North Atlantic:
http://blog.sme.sk/blog/560/310249/Natlantic.jpg
It is so profound, it actually drags the whole “global average” with it.
You are nearly there:
I would say there two distinctive variable patterns in North Atlantic: sea surface temperature and tectonic movements
http://www.vukcevic.talktalk.net/SST-NAP.htm
It is hard to prove that they are somehow related, on the other hand it would be unwise to dismiss the correlation.
The energy is required to alter thermo-haline circulation of an ocean; we are told that effects of CO2 are uncertain, solar variability as TSI insufficient, Scafetta’s ‘planetarisam’ is suppose to work via solar input which is already considered inadequate, etc etc..
Tectonics by all accounts, if and when concentrated on the narrow Arctic-North Atlantic passages e.g. the Fram and Denmark straits certainly could do the required work.
Here are two quotes from WHOI, the world best oceanographer institute:
1. Crucial to this WARM-TO COLD oceanographic choreography is the DENMARK STRAIT Overflow Water (DSOW), , the largest of the deep, overflow plumes that feed the lower limb of the conveyor belt and return the dense water south through gaps in the Greenland-Scotland Ridge.
2. Fram Strait represents the unique deep water connection between the Arctic Ocean and the rest of the world oceans. Its bathymetry controls the exchange of water masses between the Arctic basin and the North Atlantic. The significant heat flux through water mass exchange and sea ice transport, i.e. transport of fresh water and sea ice southwards and transport of warm saline waters northwards, influences the thermohaline circulation at a global scale.
epicycles provided accurate predictions many centuries before the discovery of the elliptical method. both of these methods were discovered before Newton’s law of gravity provided a mechanism.
Without understanding celestial mechanics, the Maya made eerily accurate eclipse forecasts, even for parts of the world of which they had no knowledge. So of course too did the geocentric Ptolemaic system, not shown false until Galileo’s observations of the phases of Venus.
As soon as I hear Jupiter and Saturn as an explanation for use of particular cycles, I immediately think Scientology.
Willis,
Good post.
I kinda’ thought Nicola would get the hint after his first attempt at curve fitting.
Your commentary does seem a bit harsh, though I see little reason to give credit or kudos to someone who adjusts formula parameters until their hindcast matches the data.
On the other hand you’ve made reference to an early occurrence in Dyson’s career, predating his doctorate, I believe.
It is interesting to note the correlation between that example and this, as it seems to me that quite a bit of Dr. Dyson’s work (extensive, diverse and impressive as it is) originates from playing with formulas and just generally screwing around with the math until something jumps out. By spending time doing something fairly similar to what Scafetta is doing, Dyson discovered several important relationships in math and physics that actually work and are important advancements.
Now I don’t know who Nicola Scafetta is and I haven’t his paper. Further, your point on curve fitting is well taken. Your attempt at playing the part of Enrico Fermi to Dyson’s curve fitting adventure is justified. But the statement that ‘you aren’t buying cycles of any kind’ simply flies in the face of objective facts; we begin with the daily cycle, followed closely by the annual cycle. I am not about to suggest a different set of cycles or an alternate formula, but these are the two basic cycles that impact daily temperature and weather. If climate is a ‘long term average’ of daily weather, this is where to begin your understanding and it certainly relates to cyclical physical phenomena.
We also know that there are many more cycles affecting our climate. There is a massive chaotic system interpreting the interaction of these many complex physical processes, many of which have a normal (if not exact) periodic cycle. We don’t have a solid understanding of some of the physical systems which serve as ‘inputs’ to our climate. (I don’t see any proven models for sun-spot cycles, though we have been working for some time to perfect one, and we have good reason to believe they have an effect on our climate.
With the lack of any rationalization of his formula to known mathematical or physical properties, yes, it is clear that Scafetta is grasping at straws. But that does not mean his efforts will not eventually produce meaningful results. My guess is that if he does it enough times, he will begin to see one term or another fits with a known physical cycle, possibly in a way not previously considered. If you do not spend time playing with the math this recognition will never occur. As Dyson said; “just by trial and error, we found out how to do it, and the error was essential.”
In the end, modeling chaos is akin to reading tea-leaves. What Nicola is doing cannot be said to be worse or better than what mainstream academia climate scientists are doing with their GCMs. To the extent that he does not manipulate his data he is behaving better as a scientist. Eventually he just may be the guy who figures out how to reconcile current GCMs to Milankovich cycles and whatever other, currently unknown or poorly understood, physical mechanisms that drive our climate. By this, perhaps our understanding of “the cable, or patch-cord that links the Sun to the Climate System.” will be improved.
The harshest criticism should be reserved for those who throw out such tripe and then hide behind ‘proprietary data’ and undisclosed methodology while casting aspersions on any who dare question their knowledge and proclamations.
It seems to me you are treating Scafetta like Hansen or Mann.
One of these is as ‘the Great and Powerful Oz’, hiding behind a curtain and demanding reverence and awe while reading tea leaves for fame and fortune.
The other is a ‘quack’ scientist, quietly but diligently working through error after error, occasionally asking others to ‘take a look at what he’s found’. Usually he will be shot down. One time he may succeed. One time is enough.
One of these two may eventually improve our understanding of the world in which we live, possibly even by accident. I’d put my money on the quack over the ‘Great and Powerful Oz’, any day.
I leave you with another quote from Jack Eddy; “But I also think that many of the most significant discoveries in science will be found not in but between the rigid boundaries of the disciplines: the terra incognita where much remains to be learned.”
Still, a good post and a thorough dismantling of ‘curve fitting’ as meaningful advancement of our understanding. Hope Scafetta isn’t discouraged by it. Maybe he’ll now take a look at the physics he is trying to model and see something interesting.
Just my $0.02.
By the way; what did you get kicked off Talkshop for? Just curious.
When I read this post, I was shaking my head. First, there is the silly feud between you and Tallbloke. Second, the venom from that feud seems to have fueled your display of ridicule and sarcasm for attacking Dr. Scafetta. Now, constructive criticism is more than appreciated. Usually, you are spot on. However, it is hard to give your critique a fair regard given the distraction.
Willis, this is beneath you. I hope you will see your way to mend fences, extend your hand, and soldier on. We need robust discussion without the silly nonsense.
the insistence of modern science on a mechanism is largely a nonsense. every revolution in science uncovers a new mechanism underlying the previous explanation. the ultimate underlying mechanism can only be truly known at infinity, which is beyond our grasp.
Is the sky blue because of the nature of the nitrogen molecule? Or is it blue because it had to be some color, and it just accidentally turned out to be blue? Or is it blue because the Creator made it so? Can this question ever be resolved? What difference does it make?
the power and proof of science lies in the accuracy of its prediction, independent of cause.
Nice work willis.
I look at the governing equations he creates. They are all dimensionally incorrect. That means regardless of how well they fit the data, they do not express the physics.
Anyway.
Thought folks might enjoy this
I don’t care what anyone says but this big lump of rock (some of it molten) that goes round the bigger lump(?) of thermonuclear hydrogen that has other big lumps of rock (and gas) going round it at intervals of 88 days, 243 days, 365 days (us), 687 days, 11.86 years, 29.5 years….and I forget how long it takes Uranus and Neptune to go around…Anyway, all this cyclical motion must have an effect on the grand scheme of things. And you can’t deny it. Just how much of an effect it has I have no idea.
Cycles are caused by things going around other things.
I don’t kno why anything would go around another thing though.
Anyway. That’s the layman’s explanation for y’all…
It’s a shame to read such pseudo science nonsense critique from Willis Eschenbach in this blog. None of his arguments are valid scientific arguments. All fallacies or junk or still wrong.
Any scientist can see in the graph Dr. Roy Spencer has hold up with his hands in the video hearing, twice an oscillation of about 900 years in the temperature reconstruction of the last 2000 years. Because such periods do not come out of nothing it is a common thinking, that (i) this period must have a geometric structure and (ii) it must have a physical process which controls the heat. But until the physical process is unknown, this makes the geometric structure not untrue. We dammed have to search for the mechanism, but not to discredit a persons work because there is no mechanism explained.
The argument given related to Fermi is an argument for theoretical physics and doesn’t holds for empirical research of climate frequencies of doubtful amplitudes and doubtful time scale calibration in the time range of 10000 years to month’s. It is not possible to put all functions and effects in one formula. There is no single formula to calculate the tide height correct with a precision of 10 cm; it takes a lot of syntetic Suns and Moons because of functons of the cycles,
To reject comments because of the authority is not a method of science, it is ignorance in a case, were the author has no idea what geometries are behind the temperature frequencies and functions. This blog is not a scientific blog, it is a blog were you can give personal comments about climate and climate change. If Dr. Scafetta has analysed some 20 parameter, that does not mean that this has no scientific basis; each temperature frequency can be analysed in power and frequency or frequencies, if the function is not simple a sine function.
This method is not simple ‘curve fitting’ to geometric function in a mathematic formula of many terms, as Anthony claims. It is an empiric way to come near to the climate frequencies reconstructed in the records. I have done such a parameter fitting in a magneto-optical material, based on the knowledge of the geometry of optical oscillators: http://doormann.tripod.com/jap6871990.
It is no secret here that analysed real geometric solar tide functions can be used to simulate the global climate.
http://www.volker-doormann.org/images/solar_tides_800.gif
You can compare this function with the Fig. 3 A in the paper of Dr. Scafetta.
Because the power strengths of the temperature frequencies are following an inverse square root function of the tide frequency, this indicates that there is Kepler’s third law involved, which is not invalid at far distances like Sedna, which is a times factor 1000 away from the Sun as the Earth.
Slow down.
V.