Guest Post by Willis Eschenbach
In one of my late-night somnambulistic ramblings through the climate literature, I came across a 2021 study entitled “Evidence of solar 11-year cycle from Sea Surface Temperature (SST)” by Daniele Mazza and Enrico Canuto, hereinafter MC21. They claim that they can use Fourier analysis to show that there is a solar cycle signal in the 1948-2021 sea surface temperature in the tropical Pacific Nino 4 area (5°N to 5°S, 160°E – 150°W). They also make the same claim for a wider band in the same Nino 4 area of the tropics (10°N to 10°S, 160°E – 150°W).
In this analysis, I’ll show two very different reasons why their conclusions are not justified.
Let me start with the question, what is Fourier analysis when it’s at home? Well, back in 1807, a French genius named Joseph Fourier published his brilliant insight. He had realized that any signal, like say a time series of temperature observations, could be decomposed into a number of perfect sine waves. These sine waves, each with their own amplitude, period, and phase, all add together to exactly reconstruct the original signal.
Fourier analysis is a very powerful form of decomposing a signal. It has been immensely successful in analyzing, synthesizing, and understanding signals of all types, and it is used in a huge variety of analyses. However, it’s not the only game in town.
CEEMD stands for Complete Ensemble Empirical Mode Decomposition. Like Fourier analysis, it decomposes a signal into a number of simpler signals. However, Fourier decomposition breaks a signal into only pure, unvaryingly regular sine waves. CEEMD, on the other hand, breaks the signal into
“empirically determined” underlying signals. This means that the range of cycle lengths of the signals in each group is determined by the data itself. I discuss CEEMD in detail in my post entitled Noise Assisted Data Analysis.
What does this mean in practice? Well, let me apply CEEMD to the data that they have used in their analysis. First, here is the monthly sunspot data from 1948 to mid-2021.
Figure 1. Monthly sunspot counts, January 1948 — June 2021.
And here is the complete ensemble empirical mode decomposition (CEEMD) of the same sunspot signal.
Figure 2. CEEMD Decomposition of the monthly sunspot counts.
So what are we looking at here? The top panel shows the raw data. Panels C1 to C9 show the signal in each of the nine empirical modes. These, together with the residual trend in the bottom panel, will add together to perfectly reconstruct the original signal. Clearly, the majority of the sunspot signal is in empirical mode C6.
Below is another way to look at the CEEMD sunspot decomposition. This is to look at the periodograms of each of the empirical mode signals. Periodograms show the periods (cycle lengths) of the signals that make up that empirical mode. Figure 3 shows that result.
Figure 3. Periodograms of the CEEMD Decomposition of the monthly sunspot counts.
Here, we see that empirical mode C6 contains a strong sunspot cycle peaking at just under 11 years, with a bit more sunspot-related strength in empirical mode C7. Other than that, there’s little to be seen.
Now, we can recreate the long-period variations in the sunspot signal simply by adding together the empirical modes of ~ 11 years and longer shown in Figure 2. These are the modes C6 to C9 plus the residual trend. Figure 4 shows that result, superimposed on the underlying sunspot data.
Figure 4. Sunspot counts as in Figure 1, overlaid with the sum of the empirical modes of 11 years and longer.
So far, so good. You can see how well the CEEMD shows the variations in the sunspot data. Next, I’ll show the same kind of analysis for the wide NINO4 area used in the MC21 analysis. To start with, here’s the raw 1948-2021 data.
Figure 5. Monthly NINO4 10°N-10°S temperatures, January 1948 — June 2021.
Next, the complete ensemble empirical mode decomposition (CEEMD) of the same temperature signal.
Figure 6. CEEMD Decomposition of the monthly NINO4 temperatures.
Then we have the periodograms of each of the empirical mode signals.
Figure 7. Periodograms of the CEEMD Decomposition of the monthly NINO4 temperature.
Here, we can see that there is a strong cycle peaking at 12 years … and this is the reason that the authors of MC 21 claim a “solar signal” in the ocean temperatures. However, there is no 12-year cycle in the sunspot data. Look at Figure 3. It’s a few months shorter than an 11-year cycle.
Finally, as in Figure 4, we can reconstruct the underlying temperature cycle in the NINO4 10°N-10°S data by adding the 11-year and longer empirical modes plus the residual. Figure 8 shows that result.
Figure 8. Temperatures as in Figure 1, overlaid with the sum of the empirical modes of 11 years and longer.
As you can see, there’s a signal in there, and the CEEMD analysis gives a very good fit … but it’s very unlike the signal in the sunspots. To highlight the differences, let me show the sums of the eleven-year-plus CEEMD modes for the sunspots and the temperatures.
Figure 9. Comparison of underlying cycles in sunspots and NINO4 10°N-10°S temperatures. To allow direct comparison, the CEEMD residual trends have not been included in either result.
Here, the errors in their analysis become quite evident. They’ve claimed that a 12-year cycle in the temperature data is due to solar variations. But as you can see, although there is passable agreement up to 1980, even in that section the peaks and troughs of the temperature signal sometimes lead the solar signal by up to a year and a half. This would imply an impossibility, that the NINO4 temperature is causing the sunspot cycle … and at other times, the spots lead the temps by up to three and a half years.
Worse yet, post-1980 the NINO4 temperatures start shifting more and more to the right. This is a reflection of the difference between the 10.75-year cycle of the sunspot data over the Jan 1948 – Jun 2021 period, and the 12-year cycle of the NINO4 temperatures over the same period.
In short, while the cycles are close, they do not show any connection between the 11-year sunspot cycle and the 12-year temperature cycle over that period.
So why is there a similarity? They reveal the reason in their study, viz:
After having downloaded and analysed hundreds of temperature records of the earth surface, eventually, we found clear evidence for the sun’s 11-years cycle signature in some few cases, while for the vast majority of the others this wasn’t detectable, buried under other oscillations (seasonal or El-Nino related) or noise.
(In passing, I love their totally unsupported claim that the solar effect is everywhere but it just isn’t “detectable” because it’s “buried” under reasons … but I digress.)
The problem is that if you look in enough places you’ll eventually find a similar signal … but that’s probably not statistically significant. Here’s an example.
Suppose you have a random number generator that generates a new random number from one to one hundred each time it’s used. A man says “I can guess the range of the next number. It will be between one and five”. And sure enough, the next number is three.
Since the odds of him guessing it right by chance are only one in twenty (0.05), that result is said to be statistically significant at a “p-value” of 0.05, and perhaps the man is right that he can guess the number. Of course, with a p-value of 0.05, there’s still a 5% chance it was just dumb luck.
But suppose, on the other hand, that the next random number is thirty-two. The man is wrong. So he says “Let me try again” … and again he fails. So he tries again, and again, and of course, eventually he gets it right.
Is that result statistically significant? Does he get to claim success?
Well … no. As my dad used to say when I was a kid, “Even a blind hog will find an acorn once in a while”. (In my youth, I always misheard it as him saying “a blind hawk will find an acorn”, and so I spent years wondering what a hawk, blind or not, would do with an acorn anyhow … but again, I digress.)
No, it’s not significant and he can’t claim success, because if you make enough attempts, or in the current sunspot case if you look in enough places, you’ll eventually get a positive result.
To adjust for this, we use what is called the “Bonferroni Correction”. This was an extension of work by the Italian mathematician Carlo Emilio Bonferroni (1892-1960). The correction itself was developed by a woman named Olive Jean Dunn and published in 1961. She only mentioned Bonferroni once in her analysis, but she was a woman during the 1960’s so Bonferroni got the glory … go figure.
In any case, the Bonferroni/Dunn Correction says that if you are looking for statistical significance at some specified p-value of “α” (say 0.05 as in the example above, a value commonly used in climate science) and you look for it in “n” places, you need to adjust your p-value downwards as follows:
By their own description, the authors looked for the solar signal in “hundreds of temperature records” … so to find something statistically significant, it needs to be a very, very good match, with a Bonferroni-corrected p-value of
α of 0.05 / n of 100 = corrected p-value of 0.0005
This is a level of correspondence rarely seen in climate science … and in their analysis, they don’t even mention statistical significance.
So that’s my analysis of the MC21 study. And heck, if they’d just used plain old Fourier analysis and instead of just doing it on the NINO4 temperatures they’d compared it to the Fourier analysis of sunspots for the same time period, they’d have seen the problem right away:
Figure 10. Fourier periodograms of sunspots and NINO4 10°N-10°S temperatures.
As you can see, the periods are far from the same, which means that they will go into and out of phase with each other. This in turn means that, as shown above in Fig. 9, at times the changes in the NINO4 temperatures will lead the changes in the sunspots … and that means the sunspots cannot possibly be the cause of the NINO4 temperatures.
And another solar study bites the dust.
For reference, I started investigating this question of a sunspot-weather connection a couple of decades ago, and I was a true believer in the sunspot-weather connection. I thought it would be easy to find evidence that sunspots affected surface weather in some form.
But despite looking at a bunch of temperatures, rainfall, river levels, lake levels, ocean levels, and other phenomena which were claimed to contain a sunspot signal, I’ve never found one claim that stood up to close examination. See here for links to 24 of my sunspot analyses, all of which showed … nothing. Doesn’t mean there isn’t a connection between sunspots and surface phenomena—it just means that if it exists, I’ve been unable to find it.
Here, it’s a glorious spring day. I’m going outside. These are the redwood tree and the flowers in our front yard, basking in the abundant solar radiation and enjoying the warmth. Life is good.
Best to all,
w.
The Usual: When you comment, please quote the exact words you are discussing. It avoids heaps of misunderstandings.
A Sidenote For Those Interested: As has happened to me a couple of times with other discoveries, I independently derived the Bonferroni/Dunn correction from basic principles long before I ever heard about Bonferroni. I saw the problem and calculated the proper response.
However, the form I derived gives an exact answer, and the usual Bonferroni/Dunn correction is a greatly simplified approximation of that exact form.
If alpha is the desired p-value, and N is the number of tries, the exactly accurate form that I had independently derived is:
Corrected p-value = 1 – exp( log(1 – alpha) / N )
Now, the usual Bonferroni/Dunn correction is:
Corrected p-value = alpha / N
When I found out about the Bonferroni/Dunn correction a decade or so ago, I got to wondering how good an approximation it is. So I calculated the errors (actual minus approximation). Here are the errors for alpha = 0.05 and N from 2 to 10
N Error
2 0.00032
3 0.00029
4 0.00024
5 0.00021
6 0.00018
7 0.00016
8 0.00014
9 0.00013
10 0.00012These errors are all definitely within tolerance. So I gave up using my own method and went for the approximation, much easier to remember and use.
Addendum: In addition to the Bonferroni Correction, I also independently derived the Koutsoyiannis method for determining effective N, and the Date-Compensated Discrete Fourier Transform, or DCDFT (Ferraz-Mello, S. 1981, Astron. J., 86, 619). A couple of people have asked me if it bothers me to find out that someone preceded me in deriving those methods, meaning that I was not the first one over the line.
Quite the opposite. I take it as evidence that I actually do understand the effective N statistics, the Bonferroni Correction, and the Fourier transform. I understand them well enough to derive them independently. And since I’m totally self-taught in these matters, never took even one statistics or signal analysis class, that’s an important validation for me.
That photo looks a bit rude. Sorry, but someone had to say it.
“That photo looks a bit rude.”
Which one … & why ???
A little bit phallic, you think? I thought it was ok.
no, not phallic.
Phallic? At least you’re not thinking from the same end as the other guys…
Someone does not understand the word phallic.
Just from appearance, not to mention the general trend of top illustrations on this blog, I would guess it is not a photo but an AI created object d’art, except that object d’art is not , I think, normally applied to drawings, so it is probably an AI created drawing.
Try an image search with this “solar sunspot photos”
Do you mean the nude woman trying to hide behind the bush?
Internal ocean cycles. Sometimes they offset and sometimes compound any solar influences.
Solar cycles vary in length from eight to nine years to 14 years, so 12 is perfectly reasonable. The average is a bit less than 11 years.
It’s also to be expected that solar cycles would be reflected in tropical Pacific SST, as well as in air pressure, hence both components of ENSO, a coupled oceanic-atmospheric phenomenon.
Please see below:
https://agupubs.onlinelibrary.wiley.com/doi/10.1029/2020GL091369
The Asian monsoons have long been associated with the solar cycle:
https://link.springer.com/article/10.1007/s13351-016-6046-6
Amplification of the solar signal in the summer monsoon rainband in China by synergistic actions of different dynamical responses
https://www.nature.com/articles/srep02753
Solar forcing of the Indian summer monsoon variability during the Ållerød period
Your answer was the best. Willis’ expectations of perfection here are the problem. He should have recognized the variation in cycle length as a source for error in analysis.
Furthermore the ocean lags are important as Stephen said, and lags are not on a clock with perfect timing as Willis’ expectations seem to indicate they should be.
Bob Weber February 20, 2023 5:03 pm
Bob, the reason I prefer the CEEMD is that it shows very graphically the variations in cycle lenghth. I dealt with that question in detail in the head post. Your claim that I don’t recognize it merely means you didn’t read or didn’t follow the head post.
Again you are either ignoring or unaware of the data I showed. The difference in the two cycles varies from the ocean lagging the sunspots by up to +3.5 years, to the sunspots lagging the ocean by up to 1.5 years.
More importantly, since it shows the ocean leading sunspots by up to 1.5 years … are you going to argue the ocean is causing the sunspots?
Look, I understand you want to hold on to your beliefs. But at some point, you need to realize that there is very, very, very little evidence that the sunspot cycle is related to earth weather. If any kind of clear evidence were there, this discussion would have ended fifty or a hundred years ago.
Look, I started out believing that Herschel was right 220 years ago when he said that sunspots affected wheat prices … but I actually looked at the evidence with a critical eye and found nothing. Nor am I the only one. See, e.g., “On the insignificance of Herschel’s sunspot correlation“.
w.
PS—John Tillman and others, don’t bother linking to yet another innumerate study claiming that e.g. in some small corner of the planet there’s some random dataset with a small correlation with sunspots. DO YOUR OWN DAMN HOMEWORK and look very critically at the studies. I’m tired of doing it for others.
Unless the study specifically and clearly deals with a) autocorrelation, b) Bonferroni, and c) direct comparisons of sunspot cycles with putative weather cycles as in Figure 9 above, I’m not interested. I’ve wasted far too much time pointing out the egregious errors in such studies. Read the 24 analyses I linked to above to get an idea of just how lousy most solar studies are.
Y’all truly don’t seem to realize the limitations of any Fourier analysis. For example, here’s a Fourier periodogram of a weather phenomenon I’ll call “FGN”.
Note that there is a strong 11.75-year cycle in the data. Does this show that FGN is being affected by sunspots?
Well … no. Why?
Because in this case, “FGN” is simply fractional Gaussian noise, which is just autocorrelated random data. The dataset looks like this:
Looks like temperature observations, no? But it is …
Just.
Random.
Numbers.
Like I said, ALL signals can be decomposed into simpler signals … but that does NOT mean the simpler signals have meaning.
Did I have to dig far to find this? Nope. This was the third run of FGN I tested.
So please, stop with the links to simplistic Fourier analyses.
w.
You are implying that temperature measurements and solar influences are “random noise” in which phantom frequency components can be found.
That’s not a very good reason for disbelief in the study.
The influence of the solar cycle is found in every corner of the Earth, but especially in the tropics and subtropics.
You just can’t be bothered to look at the overwhelming evidence which repeatedly falsifies your pet but evidence free ideology..
John Tillman February 20, 2023 6:10 pm
“Can’t be bothered to look”? I linked above to 24 separate analyses of the question, looking in 24 separate “corners of the earth”, that I’ve done. Plus this one, of course.
As a part of that, I asked people to send me what they consider to be the one strongest study of the question they knew of, and then I analyzed the study. In every case, as in this one, I found fatal weaknesses of all types.
Come back when YOU have done 25 separate analyses of the sunspot question and we’ll talk about it.
Until then, shut your lying mouth.
w.
I’ve done a lot more than that,
There is no bigger liar on this blog than you. Please link to each of the 25 papers you claim to have “analyzed”.
You studiously ignore all papers which you don’t want to attempt to challenge. This is the exact opposite of the scientific method, in which those are exactly the studies which you shoul most wish to tackle.
What do you find wrong with the papers to which I’ve linked today?
Which leaves aside the fact that your “analyses” are worse than worthless, as youi’re a statistical and scientific ignoramus pretending to know what you’re talking about.
When you’ve actually studied statistics and any relevant scientific discipline for at least eight years, get back to me.
True citizen scientists actually study their disciplines, rather than imagining they’ve made discoveries, out of their profound ignorance of the literature. There’s a reason why real scientific papers begin with surveys of prior work, which you have openly denied doing, out of false pride.
Science requires extinction of ego, at least temporarily, whereas with you, it’s all about ego and not search for explanations of reality.
“There is no bigger liar on this blog than you …
You studiously ignore all papers which you don’t want to attempt to challenge. … Which leaves aside the fact that your “analyses” are worse than worthless, as you’re a statistical and scientific ignoramus … Science requires extinction of ego, at least temporarily, whereas with you, it’s all about ego and not search for explanations of reality.”
The traditional approach to debate is to quote at least one sentence in the article and then try to refute it. What you have provided is a meaningless, generic, leftist-style character attack. Willie E. does a good job with his articles, in my opinion, and you are an angry fool. That’s two correct opinions in one sentence.
Me: Come back when YOU have done 25 separate analyses of the sunspot question and we’ll talk about it.
John TIllman: I’ve done a lot more than that,
I’ve provided links to all 28 of my analyses of claimed sunspot effects.
John, please provide the links to your 30+ analyses of claimed sunspot effects, and we’ll talk.
w.
PS to John TIllman:
In addition to this post and the 24 posts I linked to above, I looked and found three other recent posts about sunspots here, here, and here. That makes 28 analyses of sunspots I’ve done.
I’ve looked at the claimed effect of sunspots on El Nino, clouds, lake levels, river levels, rainfall, sea levels, tree rings, the Little Ice Age, beryllium deposition rates, pan evaporation rates, sea surface temperatures, volcanic eruptions, Norwegian child mortality rates, and cosmic rays.
And despite all of that, you falsely claim that I “can’t be bothered to look at the evidence”?
Seriously?
I’ve looked harder at the evidence than anyone I know of, definitely including you.
Like I said, come back when you’ve posted up 28 sunspot-related analyses for public inspection and critical review, and we can talk about your findings.
w.
It’s not just about sunspots and magnetism, but about the solar spectrum. Had you ever studied solar physics, you’d know that.
But, being a scientific ignoramus, for you it’s all about numbers the significance of which you’re incapable of understanding. Same as your cartoonish, willfull misunderstanding of ENSO, based solely upon dubious numbers rather than an attempt to grasp the physics behind the phenomenon. Indeed, lack of interest in so doing.
You find what you want to find and disregard the rest. The Michael Mann school of “climate science”.
You blather specious sophistry, then try to claim it is science?
Then you stretch that blather into ad hominems.
Stop the specious nonsense and start listing facts and definitive analyses, not generic feelings.
Projection.
Ad hom.
Throwing insults is one of the quickest ways to convince me you have NO logical of scientific argument to make at all.
Here’s a good place to start your education in the effects of the sun on earth’s climate, again from Reviews of Geophysics. As I said, get back to me in eight years, after you’ve first studied physics, chemistry, astronomy, geology, meteorology and climatology, before undertaking the grad level work you so presumptuously deem beneath your polymathic genius:
https://agupubs.onlinelibrary.wiley.com/doi/pdfdirect/10.1029/2009RG000282
Oh no , here we go again.
I was just about to post about what an excellently argued article this was.
Even if you feel John Tillman’s comments are unfounded spitting flames does NOT help your case.
Ah, but that was why you started blogging, not so? If we knew how to do the homework, we would not need the likes of you.
That’s why I thank all the varied gods that people like you DO exist, and I thank you for doing my homework for me. I promise, I learn it by heart before I put it on Teacher’s table…
As far I am concerned Wills, you haven’t done the kind of work necessary to rule out solar forcing.
“…it merely means you didn’t read or didn’t follow the head post.”
It’s not me, it’s you Willis. I read your post before; several times now.
“…very graphically the variations in cycle length. I dealt with that question in detail in the head post.”
“This means that the range of cycle lengths of the signals in each group is determined by the data itself.” – your post
You really didn’t talk much about the variation in solar cycle length, you used “11-year” in every instance, neglecting the actual variation in length.
“More importantly, since it shows the ocean leading sunspots by up to 1.5 years … are you going to argue the ocean is causing the sunspots?”
Don’t be ridiculous with your gaslighting and your thinly disguised anti-solar activism. Your myopic analysis leaves out every other area of the ocean that the sun can influence first before affecting the Nino4 region.
“Look, I understand you want to hold on to your beliefs. But at some point, you need to realize that there is very, very, very little evidence that the sunspot cycle is related to earth weather. If any kind of clear evidence were there, this discussion would have ended fifty or a hundred years ago.”
OK it’s my turn. You are projecting, as you want to hold on to your beliefs that you’ve made so very public in your many disingenious solar analyses.
_________________________________________________
Solar Cycle Induced Tropical Warming/Cooling Pattern Found*
Asymmetrical Tropical ~1°C step-up/down was found by differencing:
•Step-up: subtract previous solar min years SST from solar maxs SST
•Step-down: subtract previous solar max years SST from solar mins SST
Typically the ENSO region warms (cools) first, leading to general global ocean warming (cooling), following solar activity increases (decreases) above (below) the solar threshold, demonstrated below with ERSSTv5.
The odds are 1.9×10^11:1 against this pattern recurring 9 times in a row.
The current solar cycle is following the same pattern so far, as predicted, which will make it ten cycles in a row following this general pattern.
The pattern wouldn’t exist without solar forcing.
Further, the Nino4 region is currently responding well to solar irradiance (TSIS TSI) above my decadal warming threshold (white line in next plot), with new tropical warming eroding the La Nina, building the next El Nino making Kelvin wave(s), predictably according to my solar activity work.
*This work of mine was presented at the 2022 Sun-Climate Symposium and the 2022 AGU Frontiers in Hydrology meeting.
_________________________________________________
I remember how you, Willis Eschenbach, blasted me in the comments section in 2018, telling me you didn’t want to hear about my first Sun-Climate Symposium poster. You are now 5 years behind the curve Willis.
“ But at some point, you need to realize that there is very, very, very little evidence that the sunspot cycle is related to earth weather.”
What a completely ignorant statement Willis; you’re projecting.
No Willis, it’s you who needs to do the “realizing”. I don’t want you to be ignorant anymore Willis, there’s no need for ignorance now.
This is what it all boils down to, based on my 2022 presentations:
The difference between (1) droughts and (2) deluges is the time spent while the sun’s activity is either (1) below or (2) above my decadal sun-ocean warming threshold
Recent evidence for this: (1) droughts brought on by the La Nina due to low solar minimum activity, droughts which are now breaking after (2) heavy rains & snows have ensued after high solar activity.
The atmospheric rivers that dumped on California this year were powered by Nino4 evaporation driven by recently high TSI, partly alleviating the La Nina drought driven by low TSI during the solar minimum years. This is a prime example of the sun’s TSI effects being layered and time sensitive.
The past year’s solar activity has been very strongly above the threshold for precipitation, driving the weather with more rain/snow.
The values in the plot were calculated by averaging sunspot number between zero-crossings of the cumulative departure from average function for each of the climate indices listed.
Bob, you have made a host of vile accusations that I am “disingenuous”, “ridiculous”, engaged in “gaslighting”, “ignorant”, “projecting”, and other such ugly attacks.
So I’ll let you rave on in peace. I won’t debate with a man who accuses me of acting in bad faith. There’s no cheese at the end of that maze. You’ve now proven that you’re subject to the First Rule Of Pig Wrestling, viz:
“Never wrestle with a pig. The pig enjoys it, and you just get dirty”.
Have a good life,
w.
You have 24 negative results, but you are up against true believers who claim thousands of positive results of solar activity ruling the climate, the weather, the stock market, etc.
Their usual excuse is that a lot of other ’causes’ “compound the solar signal”. You can’t win against that.
Thanks, Leif, always good to hear from you. You are right. As the famous philosopher Homer observed:
Stay well, my friend,
w.
You guys still have not seen or grasped that sun spots are not the alpha and omega. It is really all about the solar polar field strengths and the interaction of the sun’s radiation coming through with the atmosphere. Less magnetic field strengths means that more of the most energetic particles are able to escape from the sun. That means that more ozone, NOx and HxOx are formed TOA. In its turn, more of these components in the air means that less UV will be able to get into the oceans. It is simple, really. Just plot the solar polar field strengths against the ozone data, e.g. Arosa.
I am sorry. I lost the work I did on that.
Sorry. I lost the plot here (discussion with John Tillman). I think maths won’t solve the problems here. I am with John Tillman. Who defined what is a (sun) spot, anyway. The best we can do now is look at the solar polar magnetic field strengths and compare it with [ozone] and Tmax (global)
You mean like these and the area under the curve?
http://climate4you.com/images/DIFF12-SpecificHumidity300mb_DIFF12-CO2-MaunaLoa_NOAA%20CPC%20OceanicNinoIndexMonthly1979%20With37monthRunningAverage_FloatingBars.gif
No idea what you think that means, RG.
w.
The ONI chart occurrences of super El Nino events are compressed in the decades of higher solar cycle intensity. This is followed by increased incidence of La Nina events in the more recent, weak solar cycle years. The problem for cycle comparisons is a) the solar cycle periodicity is stable but changing intensity across multiple cycles and b) the ONI displays shorter and more variable cycles with super El Nino cycles within the ONI cycles. That’s too much for an underspecified model and leads to oversimplified test conclusions.
I can see the face of a dog. Sunspots are the cause of dogs.
He has 24 negative results because he is looking at sunspots. He is simply looking at the wrong aspect of the Sun.
The cause of the sunspots is the pull of the planets and that has a dominant component at 19.8 years when Jupiter and Saturn are in conjunction and a half period of that when they are in opposition. The tangential acceleration and braking creates the solar activity.
The Sun’s signal in the temperature record around 11.8 years is due to the Sun’s orbit, which is essentially in opposition to Jupiter but oscillates due to the other large planets. The Sun’s orbit is not circular so the influence of the distance varies from one cycle to the next but with the underlying 11.8 year period..
The movement of the planets move the Sun and create the sunspots. Same driver but it is the distance between Sun and Earth that modulates the ToA EMR over annual and 11.8 year periods that influences surface temperature.
It seems to me he is looking at a cycles display by the sun, as were the authors of the paper. While the cycles are revealed by a count of variations in sunspot number, isn’t that what the paper he is critiquing did? Does he, or the paper, make any other claim about sunspots, such as energy delivered to Earth or appearance of interstellar visitors?
This is what Willis stated. I am showing the connection. Both sunspots and the Sun orbit around the barycentre are driven by the planetary forces.
Sunspots have a basic period driven by the conjunction of Jupiter and Saturn at 19.8 years – it varies because they are not the only planets influencing the tangential acceleration and deceleration of the sun that is generating the sunspots.
Solar EMR has an 11.8 year beat associated with the orbit of the Sun; also a function of the planets but predominantly in opposition to Jupiter, which has an 11.8 year cycle.
The beat of the Sun’s orbit on Earths ToA EMR is most noticeable in individual months from year to year as that reduces the large influence of Earth’s annual orbit in the signal. The monthly variation over the hundred or so years of the Sun’s repeating cycle is only a couple of W/m^2 but it is significant enough to show up in good temperature records.
So I have shown him why they are related. Both caused by the influence of the planets on the sun. Earth’s temperature is affected by the changing distance Sun to Earth due to the Sun’s orbit.
It is worth noting that the influence on ToA EMR is most noticeable in January when Earth is closest to the sun but not the same distance ever year. January has the most distinctive 11.8 year peak in the Fourier analysis. The variation over the last 50 years is only 0.5W/m^2 but that is clearly enough to show up in temperature records.
Over the past 50 years, March ToA EMR has risen steadily; up 0.5W/m^2 on average but a peak range of over 1W/m^2.
Rick, I spent time discussing this a day or two ago. You ended up by posting your supposed “correlation”.
When I pointed out that two things that go in and out of phase due to similar but different periods is proof that they are NOT correlated you went quiet. I hoped that you had realised it was not working.
This is exactly what Willis is pointing out in his discussion of figure 9.
I was directed to the JPL ephemeris web application that simplified the inertial analysis on my part. It is fast and I assume more accurate than my effort.
Attached is the result for 2000 to 2030 showing absolute tangential acceleration in MKS units.
You can decide whether it has significance as such matters are not permitted to be discussed on WUWT.
I’m glad you found my JPL suggestion useful. The first thing is hindcast. Do the same thing from 1850 onwards, for example, and see whether seems to tie into anything else.
Rick, I know you’re dense, but are you truly too dense to understand the following simple paragraph of the Site Policy? I’ve highlighted the important part so you don’t miss it:
Take it elsewhere, please. This is not the site for specious claims of “acceleration” and “braking”.
w.
[SNIPPED]. As I just said (emphasis mine), the site policy says:
w.
In addition to JPL you can actually get measured ToA TSI from the SORCE database without the distance correction. Although it is limited time frame, you will see the months from year-to-year vary.
https://lasp.colorado.edu/data/tsis/tsi_data/tsis_tsi_L3_c24h_latest.txt
Column 10 for unadjusted.
The data is not long enough to get the beat of the Sun’s orbit of 11.8 years so you have to generate the distance and work from that. But it does show variation from year-to-year. And look at months from year-to-year rather than annual cycles. The annual swing swamps the Sun’s signature in the data.
I will take it up with Charles. But referencing one paper as the death blow when there is an obvious connection is truly sad.
The paper does not even consider the tangential acceleration.
It is nonsense to rule out planetary forces based on a single paper that does not example the rapid changes in tangential force.
The planets move the sun by more than its own radius it is ridiculous to consider the forces trivial.
[SNIPPED] As I just quoted from the site policy (emphasis mine):
w.
I comments\ed about unknow objects in the atmosphere being on the list
A big no no here
Comment disappeared
More moderator bait.
Scientists say “The Big Cheese Moderator” deletes every 25th comment just to show us who is the boss. And scientists are never wrong. … This thread might have been more interesting if every third comment had been deleted !
[SNIPPED] Same reason. Either you are foolish or dense. It is AGAINST SITE POLICY to rave on about barycentrism. There are lots of sites where such discussion is not only permitted, it is encouraged. Please take it there.
w.
w.
So you accept that the planets move the sun because JPL produce the orbital data – there you go.
Gravitational effects of orbiting planets have no significant effect on the sun. The “force-change” involved is negligible there — it could not be measured.
It is a stretch tp claim a force of 8E23N negligible.
The 26th Willis E. article is going to contradict his prior 25 articles.
Based on the Rule of 26
Which I just invented.
“You have 24 negative results,…”
OK, I’ll bite (realizing you’ll probably not respond).
Wills has 24 negative results that were often made using CYCLOMANIA as their basis.
Willis’ CEEMD analyses are by very definition cyclomania, dependent on perfect timing for everything, the ocean, the sunspot cycles, cycles that are not always 10 years, 9 months long.
Oh, the irony…
Bob, I very specifically dealt with the variations in cycle lengths. See Fig. 9, which shows the differences cycle by cycle. Fun fact—the variations in the solar cycle are NOT MATCHED by the variations in the temperature cycles. I’ll leave you to ponder why that is, and why sometimes the temperature changes BEFORE the sunspots change.
w.
Interesting… no arguments with the analyses present.
I am curious now about the 12 year periodicity in the NINO 4 data. This is enough cycles analyzed that it doesn’t appear to be random. Would love to dive into this deeper at some point.
Thanks, Jeff. Actually, it’s not enough cycles to see what’s going on. In natural datasets, we often have what I term “pseudocycles”, which fade into and out of existence. One of the advantages of CEEMD is that we can see that happening directly.
Here’s the Fourier periodogram of the full Nino 4 dataset. There are no cycles that are greater than ~ 12% of the range of the data.
And here are the CEEMD underlying cycles and their periodograms.
Note how the cycles come and go in empirical modes C6 to C8.
Note how, just as in the Fourier periodogram above, there’s no main underlying signal.
What is crucial to remember is that EVERY signal will decompose into sine waves (Fourier) or empirical modes (CEEMD). The fact that we find these underlying signals does NOT make them suddenly significant.
w.
It simply relates to the Sun’s orbit around the barycentre that alters the distance to Earth in conjunction with Earth’s orbit around the barycentre. The Sun’s orbital period oscillates around 11.8 years because the sun orbits mostly in opposition to Jupiter but with oscillations caused by the bigger planets. The Sun’s orbit is almost chaotic compared to the planetary orbits but it moves up to 0.01AU from the barycentre in the extremes of its orbit. That is enough to alter the ToA solar EMR reaching Earth.
I pointed out yesterday that the Earth-moon system orbits the SUN and thus mimics the latter’s three-leaf clover, 57y cycle with respect to the SSB. That does NOT change E-M to sun distance by anything and thus cannot alter the ToA anything.
This kind of beligerent stupidity and refusal to recognise FACTS and enter into a logical science based argument is why this kind of thing gets banned. A policy which I was about to complain about because nothing should be excluded from consideration.
Figures Nine and Ten pretty much make the argument. The cycles were coinciding part of the time, but were clearly different.
Like maybe charge and discharge issues for heat
“However, Fourier decomposition breaks a signal into only pure, unvaryingly regular sine waves. CEEMD, on the other hand, breaks the signal into “empirically determined” underlying signals. This means that the range of cycle lengths of the signals in each group is determined by the data itself.”
A nit-pick. Cycle length *is* frequency, i.e. the frequency of the “regular sine wave”. Not all components found in a Fourier analysis have to have the same cycle length/frequency with just offset phases.
By the same token, in your Fig 10 the difference between a 12 year cycle and an 11 year 9mo signal is pretty small compared to the cycle length (about 2.5%?). There are lots of physical explanations as to why this could be so. It could even just be measurement uncertainty cropping up. It’s the same around the peaks just shorter than 5 years for temp and just longer than 5 years for sunspots.
I’m not saying your analysis doesn’t raise questions about the study. But neither am I completely persuaded that the study doesn’t bring forth some issues that need further study.
My bad, the annotation should have said “10 years 9 months”. I’ve fixed it in the head post.
The Nino 4 cycle is 10.6% longer than the sunspot cycle. And claiming “there are a lot of physical explanations why this could be” is just handwaving unless you specify them.
Finally, take a look at my comment where I show the analysis for the full length of the Nino 4 dataset … there, the cycle is 13 years, not 12, and the amplitude of that cycle is down in the noise.
w.
Tim, I had similar thoughts. We argue all the time about the uncertainty of measurement averages, with sampling bias being a potential contributor to such issues. The temperature data are treated as though they are exact values, when — like all measurements — there are associated uncertainties which probably vary with time. Whatever the method of decomposition, the ability to reconstruct the original time-series says nothing about the reliability of the original data. So, varying uncertainties in averages can account for an appearance of shifts in phase of the component cycles.
While we can probably characterize sunspots better than ocean temperatures, we might still miss some small ones on the far side of the sun, or as Henry Pool suggested above, sun spots may be an imperfectly correlated proxy for another solar influence(s) that actually is/are responsible for temperature modulation on Earth. Thus, the claim that there isn’t a correlation between sun spot numbers and climate is a straw man argument because the actual solar driver is imperfectly correlated with sun spots.
Willis, I believe at Figure 10 you have an error in the label for the sunspot period. “11 yrs 9 mo” should be “10 yrs 9 mo”. That also agrees with your CEEMD results at Figure 3.
Otherwise, a very understandable and succinct refutation.
True, thanks, fixed.
w.
Willis,
I agree with your conclusion, I think, and the use of the Bonferroni correction. But I’m puzzled about where CEEMD comes in. You express the signal in terms of empirical modes, and then basically do a Fourier analysis of the modes – via the periodogram. But what happens if you just do the periodogram of the original signal?
Thanks, Nick. I did do a Fourier periodogram of the original signal, it’s Fig. 10.
The advantage of CEEMD is that often in natural climate datasets, a certain cycle will be strong for a while and then fade out entirely. Fourier analysis will show that cycle is there, whereas CEEMD shows it’s just what I call a “pseudocycle”, a cycle that grows and decays that is not a constant feature of the data.
w.
I think that a question that needs to be asked and answered is why pseudocycles exist in natural phenomena. A couple of reasons that come to mind is that the pseudocycles are the result of random constructive interference and have no physical meaning, or alternatively, there is a base cycle that gets obscured by noise or other pseudocycles. The fact that cycles around 11 +/-2 years are being found suggests to me they aren’t artifacts of random constructive interference. Rather, the base cycle is being corrupted by other signals. I think that what is needed is a way to rigorously deal with how uncertainty in time-series measurements can influence the decomposition of a signal.
There is a simple answer. Climate is a nonlinear dynamic system.
By nonlinear is meant feedbacks. By dynamic is meant time lagged feedback. So by definition climate is mathematically chaotic (even IPCC AR3 said so). In any chaotic system there are pseudocyclic events that arise and then fade. In Gleick’s very layman accessible book ‘Chaos’ the archetypical example is a dripping faucet. AKA periodic bifurcations.
The Sun is not static in the solar system. It orbits the barycentre mostly in opposition to Jupiter but at varying distance and angular velocity. The orbit appears almost chaotic.
The period close to 11.8 years in temperature records is in response to the changing distance between Sun and Earth due to the Sun’s movement rather than the sunspots.
The sunspots are primarily caused by the big changes in pull as Jupiter moves in and out of conjunction and opposition at a full period, conjunction to conjunction, of 19.8 years. Sunspots and the Sun’s orbit are caused by the planetary forces but at different periods.
NOTHING “orbits” the SSB because the SSB does not attract anything: it has zero MASS. A “centre of mass” has no mass !!
We can plot it’s movement relative to the SSB and that movement is primarily determined by the major planets as you point out.
during the small duration of time analyzed. That doesn’t say it isn’t a significant feature of some longer period, say the Holocene. Unfortunately the data isn’t available to be looked at.
Great post, WE. I find it amazing that so many climate papers simply do not hold up to basic scrutiny. I did not independently derive those corrections as you did, I was simply taught them in graduate level probability and statistics courses at my university. But how can so many authors and reviewers not know them, or forget to apply them, in obvious circumstances as here? The shoddy climate science is unsettling.
Tru’ ‘dat, my friend …
Best to you and yours,
w.
Likewise to you and your beautiful ex-fiancée.
“Graduate level” .. “university” .. how reactionary. Please understand that “university” has been redefined to mean an educational institution formerly known as “kindergarten”. For more details, google “safe spaces”.
otherwise the moms complain
Didn’t the National Academy of Sciences address the issue a few years back?:
most of the publicity seeking “climate scientists” are ignorant of statistics beyond knowing how to select functions from Excel, Matlab, and a variety of other packages.
With tongue firmly in cheek. There is one other object with a periodicity that is closer to 12 years, actually 11.862 years, and that is the orbital period of the planet Jupiter!
Graph @ February 20, 2023 11:00am looks likes showing one of the peaks at 13 years, so if you add another 13 months (when J & E are in opposition) to 11.862 which (I think) is ~12.95 years.
Hi Vuk.
could you explain in a slightly rigorous way how you get to “add” these two periods to get 12.95 ? This circa 13y periodicity occurs in many ocean basins and I’ve never found what it could be caused by or attributed to.
I wrote an article ten years ago which got copied to Judith Curry’s site.
https://climategrog.wordpress.com/2013/03/01/61/
I’ve looked at all sorts of combinations of planetary periods which seems to be an obvious place to look, but I could never get 13. How is this calculated. Thanks.
Tha Sun orbits mostly in opposition to Jupiter and is is the ORBIT of the sun that influences the distance to Earth and hence ToA EMR on Earth.
The orbit is not circular and does not have anywhere near constant angular velocity. The cycle takes hundreds of years to actually repeat but there is a dominant component at Jupiter’s period.
Willis is simply looking at the wrong aspect of the Sun.
The sunspots are also caused by planetary pulls but primarily driven by the conjunction of Jupiter and Saturn when the pull reaches a maximum at a period of 19.8 years. The pull is at a minimum when Jupiter and Saturn are in opposition and the associated deceleration also generates sunspot activity. The tangential acceleration and deceleration of the Sun is best correlated to sunspots rather than the radial pull, which is always positive.
Little known fact that Charles Keeling was using similar technique to explore lunisolar tidal effects on temperature and CO2. He wasn’t much into what climatology is doing with his data. He was trying to understand the multicausal factors of variability in his observations at Mauna Loa.
https://www.pnas.org/doi/full/10.1073/pnas.94.16.8321
Willis- Just curious, but what software are you using to do the CEEMD analysis? Also, MATLAB refers to it as Complementary Ensemble Empirical Mode Decomposition. Is this the same thing? Have you ever used it to look at historical price data such as the stock market or futures markets?
Good questions, Tom. I do all of my work in the computer language R. The CEEMD analysis is included in the hht package (Hilbert-Huang Transform).
For anyone who uses Excel or another computer language, give R a try. It’s free, cross-platform, has a killer free user interface “RStudio”, and has heaps of free packages that cover just about everything under the sun … including CEEMD.
I have no idea why Matlab calls CEEMD that. It’s called “Complete Ensemble Empirical Mode Decomposition” to distinguish it from the simpler form, EEMD. It’s called “Complete” because, unlike its predecessor “EEMD”, adding up all the signals in the decomposition reconstructs the original signal exactly.
I’ve never used it to look at the stock markets, might be interesting … so many topics, so little time.
w.
For non R users, here is a link to the CEEMD function.
https://rdrr.io/cran/hht/man/CEEMD.html
R is the way to go for data analysis
Ah, to be able to predict the stock market! What a dream!
Of course, if you are going to look for ‘natural cycles’ in a system purposely manipulated by fraud, monopolism and warmongers, I wish you all the luck of the Irish.
“If you had the luck of the Irish,
you’d wish you was English instead” Yoko Ono
P.S. I am sure you know that the “stock market” is a market, a shop, with an owner. It is not some kind of impartial machine that facilitates trade, it is a privately owned business. Just like a bottle store or abortion clinic. With profit incentives for the owner and his workers, right?
There is no Fourier analysis to be done, friend! But maybe you can derive the programming parameters of their computer, which is sitting there, betting against itself (and your pension fund) to artificially increase the prices on selected stocks…
Might as well analyse football scores. Or cow patties per square meter. (Oh, shucks, someone’s gone done that!)
There is a reason why this was published in a bottom feeder journal like Academia Letters and Willis has shown clearly what it is. There is no point in wasting anyone’s time reading anything published there.
A 2021 study published by the AGU’s prestigious Geophysical Research Letters found the same result:
The Footprint of the 11-Year Solar Cycle in Northeastern Pacific SSTs and Its Influence on the Central Pacific El Niño
https://agupubs.onlinelibrary.wiley.com/doi/10.1029/2020GL091369
AbstractApplying statistical analyses to reanalysis products during the period 1900–2018, this study finds the 11-year solar cycle to have a significant correlation with sea surface temperature (SST) variations in the Northeastern Pacific. The solar influence is first manifested and amplified in the lower stratosphere, which then alters the strength of Hadley circulation in the troposphere. Lastly, the changes in the sinking branch of the Hadley circulation modulate surface heat fluxes to give rise to the SST footprint. The footprint has a structure similar to that of the Pacific meridional mode (PMM) that is known to be an important trigger of the central Pacific (CP) type of the El Niño-Southern Oscillation (ENSO). The 11-year solar cycle is thus shown to contribute to the slow modulation of the CP ENSO and, in particular, to be associated with more CP El Niño (La Niña) events during the active (inactive) phase of the cycle.
Key Points
So they allegedly find a “footprint” in one subsection of one ocean basin. Then some hand-waving hypothetical causational link.
This this “footprint”, “modulates” ENSO ( but only during the “inactive” solar phase ).
So the sun, allegedly, affects one specific corner of climate but only when it is INACTIVE.
That is snake oil salesman BS of the highest order. That pseudo-scientific garbage like that gets published in PR journals is just another indication of the end of the age of reason. We are truly witnessing the end of our civilisation.
Spring? We won’t get flowers like that for a another four weeks (?)
1:45 they took down the tribute to Jay Lehr. Wondered if you might know why.
I live on the Pacific face of the first range in from the ocean, an hour and a half north of San Francisco, at an elevation of about 700 feet. The area on this face of this range from about 600 to 800 feet is locally called the “banana belt”, because it almost never freezes. It’s frozen here maybe three times in forty years.
A quarter mile from my house, just over the ridge on the other face, it freezes half a dozen times each year. In the valley just below me, a dozen times a year.
But here, we grow avocados, passion fruit, and all things tropical. And spring comes early.
Microclimates. Gotta love them.
w.
Re: Microclimates. I’ve been to the West coast once – Santa Barbara / Carpenteria – to visit an uncle. We bought a miniature kite with about 30 feet of thread on a spool the size of a thimble. One of us held the kite overhead and the offshore wind took it and carried it out to its full length. We tied it to an umbrella pole and it hovered about 45 degrees off the horizon for hours while my daughter played in the sand and I tried (very briefly) to swim.
This kite doesn’t work in Colorado.
I live just east of the Cascade Crest in Washington State.
My microclimate is expected to be at 6°F this week, Friday am.
🧑🎄 😒
Chilly swimming all year round.
I found it cold just to wade in the surf.
A pack of hardy youngsters swam out 30 yards from shore and treaded water and waiting for the right wave to body surf to shore, then back out again. They spent at least an hour out there bobbing around.
Willis thanks for taking the time to analyze this. I am just curious if the pre-2015 SILSO data shows a similar result. Modified temperature and sunspot data could be problematic to statistical analysis if the data tampering is unscientific.
The changes in the SILSO data are as far from “unscientific” as is possible.The changes were made for very valid reasons.
Temperature data, on the other hand …
w.
This is proof of the solar signal in good temperature measurements like the Nino34 region. It is related to the orbit of the sun that alters the distance to Earth in addition to the orbit of Earth around the barycentre. The Sun’s orbit is mostly in opposition to Jup[iter at 11.8 years but it oscillates around that primarily due to Saturn. So all you have done is confirmed the influence of the Sun’s orbit.
The conjunction of Jupiter and Saturn occurs every 19.8 years. That is a full sunspot cycle but the opposition occurs at approximately half that interval. So sunspots occur close to half the frequency of Jupiter – Saturn conjunction with peak pull at conjunction and minimum pull at opposition. The best correlation between sunspots and the motion comes from the tangential acceleration and deceleration as the alignment approaches..
You are confusing the Sun’s orbit at approximately 11.8 years and sunspots that have a period between 10 and 11 years.
The Sun’s orbit is not circular. The influence on ToA EMR on Earth of the Sun’s orbit from one cycle to the next changes but it does create a signal varying around 11.8 years in good temperature records.
Am I correct to note that sunspot signal is instantaneous while SSTs lag – which could help explain fig 10, but not the early data in fig 9?
If SSN is somehow related to a variation in flux (W/m^2), temperature would be some kind of cumulative integral of these changes. The periodicity would still be there. The lag would vary as much as the period varies but would NOT change sign. A causal lag will always be a causal lag.
Willis,
Just curious, and lazy by not looking for myself, have you found a ~12 year cycle in any other of your investigations?
Willis,
Nostalgia. About 1976, my colleagues were interested in the developing theme of Geostatistics led by the French like Matheron and Fritz Agterberg, who we brought out to Australia. I was attracted to Geostatistics because of its logic and simplicity. Then, sunspots cropped up somewhere in my reading, so I went looking for why so many researchers wanted to correlate them with earthly events. I went through my numerous copies of Scientific American and manually digitised as many multi-year graphs of anything as I could find. Market prices of tomatoes in California, copper prices on NYSE, wheat yields in the western world .. these are all I can remember. I did geostats cross correlations of each against sunspots and did not find enough to be of interest. It was before the age of personal computers so the calcs were manual & I tired of them quickly.
Sometimes your mentions of your past work resonate with my recollections.
I still cannot understand why so many people have tried to relate sunspot counts to earthly matters. Maybe it is because they are there. Maybe the cult of Astrology dies hard. Dunno.
Geoff S
Because as that rarely cited Stott (Philip) refers to it, it’s that great big fiery ball in the sky during the day.
One of the querries of the GPTChat computer a few days ago was: “What are the best scientific arguments skeptical of the man-made CO2 climate change hypothesis?” It cited four as bullet points: Natural Climate Variability, Temperature Data Uncertainty, Role of the Sun and CO2 Lagging Temperature. Nowhere does the computer acknowledge that the sun’s effect on the oceans, which are a mystery as far as I know, and deep ocean upwelling are high in the skeptic’s list of unknowns.
No problem with the physics of Earth/Sun/Planetary relationships and to mate. My puzzle was, why study sunspot numbers? Geoff S
Studying sunspots is not a problem. Making spurious correlations is another.
It seems to me many people try to relate sunspot numbers to energy delivered to earth from the sun and that to weather properties. It seems very likely that such relationships exist on some planets in some planetary systems, such as those that consist of multiple stars or have variable stars, so why not this planet?
There is little doubt that the sun’s radiation has a very large influence on the surface of the earth. Investigating whether or not changes on the sun cause corresponding changes on the earth seems like a very natural question. One has to look in order to find out.
What’s the sense of applying high end analysis techniques to data that is known/understood to be faulty? The early part of the record, pre-satellite, has been documented to be questionable at best and cannot be validly be compared to satellite SST(skin).
The data is incommensurable within the posited data set.
If so, and not everyone will agree, why analyze it at all?
If people do a ‘low-end’ analysis, then other people would say they’re not thorough.
Willis does stuff for fun and people love his style. Hence his appearance here.
Fellow author Hansen waltzes in to challenge the accuracy of the data and whether it is useful for the conclusion. That’s old school science Hansen, but I think you are right. I realize Willie E. does not agree, and I respect him too, so I was waiting for someone I trust to say that, so I could agree. Which you did after all the “solar energy war” comments ended.
In college, while getting my BS degree,
I learned maybe four things:
1) How ro spot BS
(2) Scientists say lots of things that prove to be wrong
(2) That there’s no such thing as a high quality conclusion based on poor quality data, and
(3) Real science requires at least three decimal places.
That’s why the claimed +1.5 degree C, tipping point does not bother me, If they had said +1.486 degrees C., or 1.528 degrees C., I’d be very worried, and head for the hills, or build an ark. (Note: (3) does not work in a science lab experiment unless the professor is very gullible)
Richard Greene February 21, 2023 8:07 am
Richard and Kip, in order to challenge someone’s way of arriving at their scientific claims, it’s necessary to use their data to compare apples to apples.
Yes, I could just stand on the sideline and say “Bad scientists, your temperature data is faulty.” But that’s not the real problem with their study.
The real problem with their study is in the methods and the math used to reach their conclusions, regardless of whether their data is valid or it is garbage.
And that’s why I didn’t challenge their data. I wanted to show that whether their data is good or bad, their methods are statistically incorrect and mathematically improper.
w.
“See here for links to 24 of my sunspot analyses, all of which showed … nothing.”
I can’t imagine many scientists in the world, after 24 articles finding no climate related effect of solar cycles, would have written article #25.
Of course it is possible they were 25 false conclusions, Willis E. — your 26th article could find that sunspots control the climate, cause cancer, and warts too. I look forward to article #26. I always knew sunspots were important — that must be why people count them?
This is a serious comment, not satire.
Counting sunspots is probably 1% of studying the Sun. If a statistics geek can’t find a correlation between sunspots and weather, then I’m on the side of the geek.
Back to chaos theory: There surely is an attractor, a strange one, at that, why are we so hell-bent on trying to find cause between effects? Should there not be an attempt, a competition, if you will, to find the largest, most determinate signal? Sun spots are probably just solar weather. With remarkable periodicity….
Of course! There is no CO2 to perturb the solar weather. 🙂
Willis –
I think the following relationship holds between global monthly temperatures and monthly sunspots.
Take the absolute first difference of both series. So, S(T,i) = ABS(T(i+1)-T(i)) etc. These series are a local measure of the volatility of, respectively, sunspots and temperature.
Take an annual average of these series, and plot the scatter diagram of the resulting series against the other.
I find a wide scatter, but a large deficit of datapoints where sunspot volatility is low and temperature volatility is high.
So the relationship is that low sunspot volatility is correlated with low temperature volatility, but explicitly not that there is a linear relationship across all levels of volatility. It even makes some sense physically, because sunspots create volatility in TSI.
This is based on staring at charts: I haven’t done statistics worth a damn. But you might be able to. You know what they say in business, it’s best if you eat your own lunch, rather than have someone else do it. If someone is going to falsify your assertion about sunspots, maybe it should be you!
All the best,
R.
I just gave your method a try, using the HadCRUT monthly global temperature and the SILSO sunspot data. No significant relationship (p-value=0.18).
w.
From the article: “A couple of people have asked me if it bothers me to find out that someone preceded me in deriving those methods, meaning that I was not the first one over the line.”
Just being able to derive it independently would be enough for me. 🙂
Very good post, Willis.
deriving that formula is impressive. I was instinctively suspicious of simply dividing by n because , in my experience, things are never that simple. Having done it rigorously and shown 1/n is pretty close is much more acceptable to my gut neurons.
Thanks, CG. I was mondo bummed when I first learned about the simple version of the Bonferroni/Dunn correction, but that turned to happiness when I found that it’s just a lovely approximation.
w.
There sure are a lot of “sun worshippers” here.
The sun is responsible for our climate.
But not necessarily for our climate changes.
That you can find an article that doesn’t demonstrate what it says, and you can’t find an effect of the solar cycle on ocean temperatures, it says nothing about the existence of such an effect. The article’s weakness says nothing about the subjacent hypothesis’ strength, which does not rest on that article.
The most important article regarding this matter is probably:
White, W.B., Dettinger, M.D. and Cayan, D.R., 2003. Sources of global warming of the upper ocean on decadal period scales. Journal of Geophysical Research: Oceans, 108(C8).
Another important article is:
White, W.B. and Liu, Z., 2008. Non‐linear alignment of El Nino to the 11‐yr solar cycle. Geophysical Research Letters, 35(19).
You are embracing a wrong paradigm by insisting on using the wrong tool to analyze the effect of the Sun on the ocean’s temperature.
The detected quasi-decadal changes in subsurface ocean heat content are phase-locked to the solar cycle, whether you like it or not. There’s ample bibliography on that which you ignore. Even the models that are usually clueless about solar things detect it:
Misios, S. and Schmidt, H., 2012. Mechanisms involved in the amplification of the 11-yr solar cycle signal in the tropical Pacific Ocean. Journal of Climate, 25(14), pp.5102-5118.
You are just cherry-picking bad studies as if they could demonstrate anything but the existence of shoddy science.
How can an 11 year average solar cycle overcome the ocean’s thermal inertia?
And how would you see a sunspot signal with all other possible simultaneous causes of ocean temperature changes?
In the ice core record. ocean temperature changes took hundreds of years to cause atmospheric CO2 level changes.
It doesn’t overcome the ocean’s thermal inertia. It’s that thermal inertia that causes lag in the cycle. The sunspot cycle can have a phase shift from the temperature due to thermal inertia. The amount of heat transferred is a rate. Q = k * time where k is BTU/hour and time is in hour (or whatever units you want). It’s like heating one end of a metal rod. The other end doesn’t get hot for a while. So the cycle frequency at the far end of the rod doesn’t have to be the same as at the near end of the rod.
Solar variability has a strong effect on climate. Paleoclimatology is very clear about that. It is not surprising that the solar cycle has some effect on surface temperatures, and the effect is found to be about 0.1ºC. It must come from somewhere on the surface, and the ocean is 71% of the surface.
Clearly not doing the types of analysis Willis does.
The surface layer of a water body can change temperature significantly within a matter of hours. Generally, we are talking about sea surface temperatures, not the bulk average.
I remember as a teenager swimming in Lake Shasta (Calif.) in the early afternoon in July. It was probably about 75-80 deg F at the surface. I dove under the water and at about 6′ passed through a thermocline where the water was probably closer to 50 deg.
Javier Vinós February 21, 2023 7:09 am
I read that article. It is based entirely on computer model simulations of the upper ocean. I suppose there are people out there who believe that modelworld = real world. I’m not one of them.
For example, it says:
None of those variables are actual observations. They are all the output of a computer model. The problem with computer models is that they are quite mindless—if you put something in the input it is bound to show up in the output. This means that when they include solar variables in the input, they’ll show up in the output.
You, and Professor White, seem to think this means something about the real world.
Me, I think it simply means something about computer models.
True, but ONLY in the models. We simply don’t have the data to determine if that’s true in the real world.
That paper is hilarious. It starts out by saying:
You’re gonna believe that the average of the output of a bunch of untested computer models which even you admit are “clueless” are going to tell us something about reality?
Seriously? That’s your authority? They don’t even try to use the actual data, they just grab a bunch of computer models and proudly announce their results … and you believe them?
Really?
What is it with charming folks like you who want to accuse me of bad faith? I tell the truth as best I know it. I don’t “cherry pick” anything.
For years I’ve invited people to give me the best studies that they know of, and I’ve analyzed them. As with your studies listed above, they’ve turned out to be nonsense, full of bad math, bad computer models, statistical errors, and incorrect logic.
I see that you don’t like that, but that’s no reason to attack me personally.
w.
That means you read it but did not understand it. Reanalysis is fed real data on sea temperature and all sorts of variables. It is not simulated data.
That you don’t believe the data is there, does not mean the data is not there. That’s your problem, not science’s problem.
No, but I believe a solar effect was not introduced by the programming, so if it is coming out it means the solar forcing prescribed is having an effect. Models are not the truth we are told, but they are a tool for learning what we know and what we don’t know. I don’t think you can learn anything from models with your attitude. But then you don’t learn much from scientific articles either, so no surprise.
Perhaps because of your strong anti-solar bias. I am not charming nor do I pretend to be. Charmness has no value in science. Trying to be unbiased has a lot of value in science.
Javier Vinós February 21, 2023 11:52 am
No, it means you don’t understand reanalysis climate models. Yes, a small amount of real data is fed into a reanalysis climate model. But then, the reanalysis climate model just guesses the values between the real data points. And models are lousy at such guessing, because they don’t do edges, they do gradations. If you have a value of say 10 at one point and 4 at another, the computer will estimate a value of about 7 … but nature does edges, not gradations. The average between the middle of a cloud and the clear air surrounding the cloud will either be cloud or clear, but not the average of cloud and clear. The poet puts it best about how nature does edges:
Model that, sucka!
And given that they are talking about a bunch of underwater ocean variables, many of which have only been measured very sporadically in both time and space, that means that MOST of what you and they laughably call “data” is just modeled guesses, and not very good guesses at that.
Dang, you are credulous. You call the models “clueless” and then go on to believe them implicity … say what?
First, the authors are using the output of a reanalysis computer model as input to another computer model, which is most definitely a very sketchy procedure and is not recommended by sane modelers. It ups the errors exponentially.
Next, as I pointed out and have demonstrated in posts such as “Life is Like a Black Box of Chocolates“, the output of a computer climate model is just a lagged, resized version of what you put into it. This is NOT true of the real earth. So when they add solar input to their stack of models, they’re pretty much guaranteed to find it in the output. They don’t have to “introduce it by the programming”. They introduce it as input, with GIGO being the analysis method.
You seem to think this means something about the real world. It doesn’t. It only means something about modelworld.
Some models will do that. We have very good models for a variety of phenomena. Climate models, on the other hand, have proven beyond doubt that they are mechanistic garbage that can teach us very little. The climate is far and away the most complex system we’ve ever tried to model. It has six independent subsystems—atmosphere, hydrosphere, cryosphere, biosphere, lithosphere, and electrosphere. All of these have internals resonances and cycles. Each of them is constantly interchanging energy with all of the others. It is all driven by solar energy which is constantly changing at all temporal and spatial scales.
And there are important phenomena operating at timescales from nanoseconds to millennia, and spatial scales from molecular to planet-wide.
So it’s no surprise we’re failing miserably in modeling the climate. But it gets worse. The models haven’t been subjected to any kind of V&V, which means your average model running a bank of elevators has been tested much more rigorously than the climate models. Heck, we haven’t even shown that the Navier-Stokes equations used in the models actually converge.
To add to that, they are far from “physics-based”. As I showed in “Meandering Through A Climate Muddle“, without a host of kludges, guardrails, tuning, and inbuilt limits, they’d run off the rails in a minute.
Your faith in these failed models is touching, but misplaced.
Back to personal accusations, are you? Charming, as I said. When people start throwing mud like you’re doing, it’s clear evidence that they’re out of scientific ammunition.
I have no “anti-solar bias”, that’s just another bit of evidence-free personal mud-throwing. I’ve looked heaps of places for a solar signal, and I haven’t found it. So sue me.
w.
Your opinion about models only matters to you, really. And it uncovers another one of your multiple biases.
Thanks, Willis. And a beautiful coastal redwood. There’s a very few examples here in the east US — they suffer from fungal diseases.
So do climatologists.
CEEMD does seem compellingly; isn’t another advantage it’s ability to overcome end truncation artefacts by extrapolating beyond the artificially cut ends?
Another question:
However, Fourier decomposition breaks a signal into only pure, unvaryingly regular sine waves. CEEMD, on the other hand, breaks the signal into
“empirically determined” underlying signals.
Is there a limit to the complexity of the underlying signal that CEEMD uses? For instance what is to stop an “underlying signal” being just a copy of the original signal being studied?
Argumentam ad absurdam of course but there’s a point in there somewhere.
Regarding “the periods are far from the same, which means that they will go into and out of phase with each other”: I also saw “The difference in the two cycles varies from the ocean lagging the sunspots by up to +3.5 years, to the sunspots lagging the ocean by up to 1.5 years”. I see this as explainable by the lag averaging 1 year with noise causing the lag to have +/- of 2.5 years.
As for what this noise is? The Figure 6 CEEMD of SST in Nino 4 longitude range and in latitude range of 10S to 10N has strong C5 and C6 components, and the Figure 7 periodogram shows periods of C5 and C6 mostly typical of ENSO.
As for the Figure 10 Fourier spectral analysis of Nino 4 (10S to 10N) temperature having a double peak and the stronger peak having period more than a year longer than that of sunspot count: There is the matter that CEEMD splits an apparent ENSO signal into two components.
Something else I saw: The MC21 study claims for both 5S-5N and 10S-10N latitude ranges in the Nino 4 longitude range, and the effort to refute MC21 was only done on one of these latitude ranges. I ask for an explanation for lack of attempt (as I saw this) to refute the claim for the latitude range of 5S-5N.
As for further study that I suggest that can reinforce or counter the claims of the MC21 study:
1: Considering a longer time period that starts before 1948. If the lag of Nino 4 temperature from sunspot number stays mostly close to 1 year +/- 2.5 years, and agreement in Fourier spectral analysis between Nino 4 temperature does not decrease much, then this would counter an anticipatable claim that start time of 1948 was done for cherrypicking.
2: Repeat Fourier spectral analysis and CEEMD for each hemisphere of the sun’s sunspot number, and repeat again for both options for alternating from one hemisphere to the other at solar minimum to consider only the solar hemisphere that has one of the two main solar hemisphere magnetic polarities. I admit this is “P hacking” cherrypicking, but if an extreme P value (around .001?) for correlation with Nino 4 temperature (or some other weather record, preferably for a large area at least as large as Nino 4) can be found, then I think it’s worth exploring if P value much below .05 is repeatable and looking mostly reliable over multiple different time periods.