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.
Dr. Ball, thanks for your eloquent comments. I agree that there are cycles.
I found the comment of one solar physicist on this subject particularly interesting. He claimed there are actually an infinite number of cycles of the sun – they transpire over periods infinitely (by our standards) long and infinitesimally short, and thus necessarily overlap each other. Some must cancel each other out, Others, such as the ones Leif has made a career of studying, are quite obvious even to laymen such as myself, and their periodicities can be understood and predicted in a human lifetime. The term “cyclomania”, used to denigrate (even inaccurate) theories of cycles is, in my opinion, a bit disingenuous, since we are all looking for the causes of climate, whether they are cyclical or chaotic. It would be a very uncomfortable irony if the skeptical writers of WUWT overlooked (some of the) actual, historical causes of climate change just because they fall under the heading of cyclical.
A governing mechanism that kicks in and causes thunderstorms when the ocean surface overheats is cyclical, since, by some accounts, it appears to “turn on” during the heat of the day, and gradually turn off, as evening approaches.
Scafetta stuck out his neck in 2011 with predictions. I really would like to see how he’s done 20 months later. If he’s done well, let’s cut him a break. If not, then he needs to explain why not. (The 2012 update to his 2011 post is here:
http://wattsupwiththat.com/2012/03/11/scafetta-prediction-widget-update/)
If it agrees with hadcrut at any point, it cannot be correct!
I’d like to see an ocean cycle or two, or three. For example, where does the warm water from the Mediterranean Sea go after entering the Eastern Atlantic Ocean? What about the water from the Amazon Basin? There are many other on-Earth things that seem to get little attention but might have influence on things that are indexed as of climatic interest, such as the AMO or ENSO.
—–
Regarding the research questioned with this post: Turn the law of the instrument (aka Maslow’s hammer) around; namely
If you have a nail, everything about you begins to look like a hammer. I’ve pounded nails with a tennis racket, a boot, a wine bottle, rocks, a pipe wrench, pliers, and more. Do I need to list 20 to make the point. They all sort-of do the job but when working with nails a modern claw hammer is what you want. Hammer head meets nail – – there is a clear physical picture of what is desired.
Willis, I think, rightly claims that Dr. Scafetta takes the surface temperature of Earth and looks about for anything or many things to hit it, without reason.
——————
Willis is banned at Tallbloke’s – how funny; made my day.
To give Willis a more technical response, that he may not understand because too physical but many readers may understand, the climate system is essentially a complex system regulated by some set of differential equations. To solve differential equations one need the boundary conditions.
In the case od the climate the boundary conditions are provided by the forcings coming from the space plus some other internal forcing such as volcano activity and GHG variation.
The forcings coming from the space contain necessarily a complex harmonic component due to the fact that the moon, the Sun and the planets are moving around. Some of the harmonics are those that I use.
If a system described by set of differential equations has a boundary condition that contain an harmonic complex component this harmonic component will be reflected also in the solution of the set of differential equations, that is in our case in the climate system,
So, my theory, which was in principle already advocated by science giants such as Ptolemy and Kepler and Newton, and constitute the traditional way to understand climate change has nothing strange or wrong, in principle.
Willis is just panicking and my papers are just making him to have some severe nightmares, we need to understand the guy.
In any case, for the interested readers I have made an interesting graph comparing my model with the IPCC models here
http://people.duke.edu/~ns2002/#astronomical_model_1
Anthony, may you please show my new figure here? [Done. -w]
You may also show it in a more interesting post.
I’ve been visiting this site since nearly day one. Havent read many of the comments above but would like to add that I also find the Willis arrogance really distasteful. And at the end of the day you need to convince common people like me to believe in what you are saying.
60 years: kinda like three times the synodic period of Jupiter/Saturn. Why three times? Why not?
Third Harmonic?
A sine-wave component having three times the fundamental frequency of a complex signal.
For the genuine science in the subject of climate cycles, there is the Cambridge University Press book, Climate Cycles Real Or Imaginary? Second Edition c2003 by William James Burroughs.
Bob Shapiro says: July 23, 2013 at 1:14 pm
“Scafetta stuck out his neck in 2011 with predictions. I really would like to see how he’s done 20 months later. If he’s done well, let’s cut him a break. If not, then he needs to explain why not. (The 2012 update to his 2011 post is here:
http://wattsupwiththat.com/2012/03/11/scafetta-prediction-widget-update/)”
Actually my forecast is working very well. See here
http://people.duke.edu/~ns2002/#astronomical_model_1
my updated widget is here
http://people.duke.edu/~ns2002/scafetta-forecast.png
Anthony, please, show the figure of my updated widget.
For the benefit of the readers. So that they may form also a right opinion about Willis post.
It appears to me that climate is cyclic on most if not all time scales. The cycles are not always or even mainly regular, as chaotic elements also influence its changes.
The opening of deep sea channels around Antarctica at the Eocene-Oligocene boundary isolated the continent & started formation of ice sheets there. The Isthmus of Panama altered oceanic circulation, initiating the quasi-regular Pleistocene glacial-interglacial cycles, modulated by orbital mechanics. So clearly, there’s a role for geography, geology & oceanography in controlling climate.
There also appear to be longer-term icehouse/hothouse climatic cycles. You don’t have to ascribe to Svensmark’s cosmoclimatology to recognize the cyclic nature of at least some climatic phenomena. The icehouse portions of the mostly torrid Mesozoic didn’t develop continental ice sheets, as during the Paleozoic (at least twice) & Cenozoic, but their signal is still there.
Climatology is still in its infancy. Good data need to be gathered & testable hypotheses erected thereupon. But what is known IMO shows that CO2 is not the main driver of climate forcings on Earth, or even a very important one.
milodonharlani says: ….
What you say is absolutely correct. One just need some little objectivity.
To the other readers, do not trust Willis post.
Just read my paper instead
http://people.duke.edu/~ns2002/pdf/Scafetta_EE_2013.pdf
The doubts of many readers are actually explained in my paper quite in details.
I cannot past and copy the entire paper content here. You just need to take some time to read the paper that is quite informative.
John West says:
July 23, 2013 at 10:43 am
Thanks for the clarification, John, and clearly many climate phenomena are cyclical, including days and seasons. And the Milankovitch cycles. And I’ve spent a lot of time looking for astronomical correlations with any features of the climate … unfortunately, they are weak, few, and far between. As usual with the climate, the noise seems to be larger than the signal (if one exists).
My point was that the Milankovitch cycles are real astronomical cycles, while Scafetta is just picking numbers. You can’t just choose cycles because they work, that’s curve fitting.
All the best,
w.
G. Karst says:
July 23, 2013 at 12:22 pm
Yeah, you’re likely right, but his kind of pseudoscience pushes my buttons. It’s exacerbated by his refusal to share his data or his calculations or to answer questions about his work. But I likely should have gone easier on the poor man.
w.
If Dr. Scaffeta wants to make his code and data available for replication, I’ll be glad to feature his work in a top post.
To Willis, who says:
“You can’t just choose cycles because they work”
Yet, the scientific method only requires that a proposed model works.
Sorry Willies, your own comments prove that at the end you do not have arguments.
You are just handwaving.
And you are acting out of arrogance and malevolence to mislead the readers of this blog about a scientific research that in truth you do not understand.
Science is not done as you think, Willis. Learn to be humble when you are not able to understand things.
By the way, I am not doing curve fitting. Read my paper well before criticizing them with a list of non-senses.
Nicola,
Thanks for the link to your paper. I’ve read through it a couple of times and am still thinking it through.
I would like to ask you a few question though that may assist me in understanding your paper:
1) From the arguments I read in Section 7 ‘PLANETARY CONTROL ON SOLAR AND CLIMATE CHANGE OSCILLATIONS THROUGHOUT THE HOLOCENE’, I walk away with the impression that the specific physical mechanisms for what you propose are unknown, since you list some possibilities there, is this correct?
2) If this is correct, how did you select the specific astronomical cycles to use? Was it a matter of finding those that matched frequency elements you were looking for, or was there another method?
3) Is the claim that you refuse to share the data, calculations, and code you used true? If so, would you explain why? If not, would you make these available?
Thank you,
Mark Bofill
Anthony Watts says:
July 23, 2013 at 2:39 pm
“If Dr. Scaffeta wants to make his code and data available for replication, I’ll be glad to feature his work in a top post.”
Anthony, you talk as if I run a mysterious GCMs made of thousand equations and mysterious data. The equation of my model is very simple and is Eq. 7 in the paper.
Sum six harmonics listed in eq. 1-6 and add the radiative component (which is a first approximation of GHG, aerosol and volcano component) as explained in the paper which s based on the CMIP5 ensemble mean.
The HadCRUT4 record you can download from their website.
The model is a good first order approximation
I am making an exel file asked by Christopher Monckton
milodonharlani says:
July 23, 2013 at 10:58 am
Volker Doormann says:
July 23, 2013 at 10:40 am
Re: Pseudoscience.
Does a curve-matcher necessarily have to understand the science behind his number-crunching? Or does making falsifiable predictions based upon it suffice? A good match could possibly lead to identifying candidate explanatory physical phenomena. Or not. The exercise may not be pseudoscientific, assuming that appellation is anything more meaningful than a nasty name.
That are questions and sayings, but they miss relevant points.
One point is that science is searching for coherence. If then a structure is discovered, which shows coherence between time functions, it is possible then to make predictions alike eclipses of the Sun by the Moon.
The next point is that the accuracy of the prediction in time an amplitude depends on the knowledge of the accuracy of the real function.
This is fulfilled by showing the solar tide time function of neighbour planets and the inverse square root function of the tide strength as well known from FFT spectra over 10000 years.
Using this functions you can not only simulate the solar tide functions for the last 5000 years, but also for the next 1000 years, simple because the astronomical NASA ephemerides are precise known to that date.
If you do not know what the nature of mass is or what the nature of gravity is, you still can predict eclipses because of knowing the geometrical laws of Kepler. And in Keplers laws are no physical forces involved. Just geometry. No one would call the prediction of eclipses pseudoscience, only because the are no physical forces in his formulas.
I have discovered the solar tide functions three years ago and they are known here to the regulars.
http://www.volker-doormann.org/ghi_solar_s.pdf
and an update from this year
http://www.volker-doormann.org/frequencies_of_climate.doc
Is is a tragedy that in this time there was absolutely no discussion on the subject in WUWT, neither by Anthony, nor by Dr. Scafetta, nor by anybody else.
V.
People attribute this quote (or something like it) to Einstein:
“It can scarcely be denied that the supreme goal of all theory is to make the irreducible basic elements as simple and as few as possible without having to surrender the adequate representation of a single datum of experience.”
1) What Scafetta is doing is not simple.
2) It has way to many elements
3) It is not “elegant” in anyway shape or form. It’s an elephant. It’s more like “a dog’s breakfast”.
4)But most of all, there is no justification for any of the things he has done, because there is no statistically robust prediction.
Scafetta, why don’t you take up Willis challenge (and actually predict something?)
Update to prior post, I see now that the discussion in section 7 only relates to the 11 year cycle. Still, the general form of my first question remains, am I just not understanding the physical mechanisms you propose, or are they not in there?
Questions 2 and 3 can stand as they are I think.
Thanks,
Nicola Scafetta says:
July 23, 2013 at 2:21 pm
I should have qualified my remark about CO2 by saying “in the Phanerozoic Eon”. CO2 around 90,000 ppm in the Pre-Cambrian “Snowball Earth” glaciations may well have been a or even the major driver of deglaciation.
Willis’ challenge:
“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.”
Hmm.. to me looks like another version of Mann’s Hokey Stick. How well does your hindcasting do all the way back to the MWP and LIA? Seems your natural variation parameters are remarkable stable…. like a hockey stick.
Nicola, thank you for showing up to defend your work.
Next, as a number of people have pointed out, I’ve been over the top in some of the things I said. So I apologize for that wherever it has happened.
I also note Anthony’s generous offer that if you make your code and data available for replication, he would be happy to feature your work. I do look forward to that, as it will answer many questions.
Now, I understand your claim, that the various cycles you are using are all astronomical in nature. Let me take one of them, the 9.1 year cycle. According to you,
Now, you are right in your figures and calculations. But I’d have to ask, what is the physical meaning of the equation you are using, which is:
Yes, you are correct, that equation of yours does calculate out to 9.0751 years … but how is that anything but an artificial construct? Why not
Is that second one an astronomical cycle as well?
I ask because that second one calculates to about 10.2, and at various times you have used cycles of 9.1, 9.98, “10-11”, 10.2, 10.9, and 11.86 years for your cycles.
The problem is that when you do that, you have an infinite range of cycles that you can claim as being “based on astronomy” …
Here’s your explanation of the 60 year cycle:
Nicola Scafetta says: March 14, 2012 at 8:13 pm
Um … er … well, then why are you now using a 61 year cycle? In any case, I assume you are referring to three times the synodic cycle of Jupiter, which is 59.5766 years … why not use the real number, why use 60 or 61 years if you claim the synodic cycle is important?
In any case, this will all be clear when you reveal your data and your code. Then we can discuss this in detail.
All the best, and again, my apologies for where I was over the line. The problem is, when you post a graphic which squeezes in four different small graphs, and not one of them has any units of any kind, it’s hard not to be dismissive. I’m working on it, however, and my apology is sincere.
w.
Volker Doormann says:
July 23, 2013 at 3:06 pm
The remarkable fact is that a number of cultures around the world were able to predict eclipses even without Kepler’s work. In March 1504, Columbus used an eclipse to his advantage before Copernicus’ work or Kepler’s birth. He impressed indigenous Jamaicans but would not have done the Maya.
Stonehenge for instance could have been used to keep track of eclipses & hence predict in which month (or moon) lunar eclipses would occur, based solely upon observation, without understanding of celestial mechanics.
“And I’ve spent a lot of time looking for astronomical correlations with any features of the climate … unfortunately, they are weak, few, and far between. As usual with the climate, the noise seems to be larger than the signal (if one exists).”
Me too, but I figured that I need to know what is driving the noise. Like how this March was so cold, but the previous March was baking, and why the heat wave this July etc.