The Zharkova et al paper that incorrectly purported to link solar-climate effects to movements of the Sun around the barycenter has been retracted.
This paper generated an enormous thread on @PubPeer where the authors continued to defend the indefensible and even added in new errors (such as a claim that the Earth’s seasonal cycles are due to variations in the Earth-Sun distance). Additionally, it seeded multiple nonsense newspaper articles in the UK and elsewhere (some of which were quietly deleted or corrected).
But the interesting thing is that this cycle of very public solar claim/counter-claim/claim/retraction was totally predictable.
Why is this? What is it about solar-climate links in particular that brings out the confirmation bias and the defend-at-all-costs responses? Why are the UK tabloids so excited about mini-ice age stories?
First off, it has to be clearly said that there is an enormous amount of good work done on this question. People like Judith Lean, Greg Kopp, Jo Haigh, Lesley Gray, Leif Svalgaard have been building better and better records of historical solar activity, improving the calibrations and observations of current measurements, and really drilling down into the mechanisms of possible climate impacts. Second, I have published multiple papers on the topic from a modeling perspective.
But, there has been a long history of people assuming that they *know* that solar cycles have an effect and then just looking every more deeply for the mechanism. Indeed, solar-climate links might be the ur-topic of the current p-hacking scandal that is troubling a lot of science these days.
There must be a pony in there somewhere
This goes back a very long way. Indeed, the first modern “sunspot-climate” claim (published by William Herschel in 1801), was in fact insignificant (Love, 2013), though in Herschel’s defense, statistical significance wasn’t really understood in the late 18th Century.
Slightly more recently, a classic of the genre was published in Science (Friis-Christensen and Lassen, 1991) which not only misrepresented the analysis they did to “prove” a link between climate and “solar-cycle length”, but in correcting it made even more arithmetic errors (Laut, 2003). That this massively cited paper (> 1300 cites) is still unretracted is continuing mystery.
Needless to say, very few (if any) of these solar-climate links are predictive. That is, once new data comes in, the purported correlations evaporate as fast as the credibility of the authors. And yet, the next paper that ‘fixes’ the correlation still gets published. We have, of course, discussed this before.
It’s (not) the sun
Some of this is related to a desire to find something other than human activities as the cause of the climate changes since the late 19th Century. Folks who really, really, really, don’t want climate change to impact societal choices [newsflash, it already has] often grab on to speculative solar effects as a last ditch throw of the uncertainty dice. But obviously, solar-cycle mania predates any of those concerns. For instance, what was Oscar Wilde responding to?
Why does not science, instead of troubling itself about sunspots, which nobody ever saw, or, if they did, ought not to speak about; why does not science busy itself with drainage and sanitary engineering?
Oscar Wilde (1882)
These days, the intransigence of climate change contrarians comes as no surprise. So the warm welcome afforded to solar-climate proponents by the GWPF et al is to be expected. But for anyone serious, making whoopee with such strange bedfellows is probably unwise.
It’s a minefield
To anyone who is working on this topic (including me), the conclusion that you must tread carefully is inescapable. The need for self-criticism in the design and publication of results and the importance of real peer review cannot be overstated. The normal human tendencies to rush, or be excited by a new finding, have to be tempered by the knowledge that this has led many authors to make mistakes and be premature (and wrong) in their conclusions. Zharkova et al are merely the latest in a long line of people who have fallen into this trap.
They won’t be the last.
References
- V.V. Zharkova, S.J. Shepherd, S.I. Zharkov, and E. Popova, "Retraction Note: Oscillations of the baseline of solar magnetic field and solar irradiance on a millennial timescale", Scientific Reports, vol. 10, 2020. http://dx.doi.org/10.1038/s41598-020-61020-3
- J.J. Love, "On the insignificance of Herschel's sunspot correlation", Geophysical Research Letters, vol. 40, pp. 4171-4176, 2013. http://dx.doi.org/10.1002/grl.50846
- E. Friis-Christensen, and K. Lassen, "Length of the Solar Cycle: An Indicator of Solar Activity Closely Associated with Climate", Science, vol. 254, pp. 698-700, 1991. http://dx.doi.org/10.1126/science.254.5032.698
- P. Laut, "Solar activity and terrestrial climate: an analysis of some purported correlations", Journal of Atmospheric and Solar-Terrestrial Physics, vol. 65, pp. 801-812, 2003. http://dx.doi.org/10.1016/S1364-6826(03)00041-5
#49 Paul Pukite
Paul, in #30 I talked about communication and it really does seem to be an issue… if you want to explain something to a non-specialist like me you have to listen to their questions.
I understand the underlying physical principles. I just can’t see how you are putting them together, cause-and-effect-wise, possibly because you keep switching terminology. How about providing these plots with clear labels, on one page:
1. The “spring annual impulse”.
2. The “varying tidal forcing”.
3. The “annual cycle”.
4. The variation in ENSO with which you are trying to correlate any or all of the above.
#45, AB–
Thanks! Not sure my wife would agree, though.
zebra said:
This seems to be a bit of a “raising the bar” argument, but here is an appropriate single-page set of charts which includes everything that you asked for:
https://imagizer.imageshack.com/img922/568/0uN5cm.png
The top curve is the model of the long period tidal forcing over an 18 year interval, which is largely comprised of the fortnightly tropical tidal cycle of 13.66 days (well known to anyone that studies tidal charts). This is calibrated against the precise measurement of the earth’s change in length-of-day (LOD), which is also well known to follow the tidal forcing. That is, the same lunar forcing that is able to change the earth’s rotation rate can also cause the ocean to slosh.
The middle curve right hand side is the top curve modulated by the strict annual impulse and integrated as an inertially lagged response of the fluid, as water does not respond the same as a solid. Note that this no longer falls along an annual time step since the fortnightly tidal signal is not always strong at the annual impulse time — in fact this is largely equivalent to a spring tide cycle and how it varies in strength from year to year, with the period equal to the 3.8 year octon of the metonic eclipse cycle — can look this up at the NASA JPL eclipse page.
The left middle curve is the result of applying the right middle curve as a combined forcing to Laplace’s Tidal Equations and comparing to an ENSO time series. The solution to Laplace’s Tidal Equations is essentially a GCM computation, but analytically solved as a closed form expression.
The bottom panel is the frequency spectrum of the right middle curve showing the simplicity of the tidal forcing — note the strong spike at 1/3.8 /yr and 1-1/3.8 /yr and then the continued harmonics of this forcing. As I said, this is mainly a result of the fortnightly tropical tide modulated by the annual impulse. Can see clearly the sideband spitting in the forcing.
This was all presented at the 2016, 2017, and 2018 Fall Meetings of the AGU and in a monograph that was published by AGU/Wiley in 2019. Information available by clicking on my profile.
#53 Paul Pukite,
Your effort is greatly appreciated. I apologize if you posted it previously and I missed it, but the middle-right plot is what I was looking for. Also, I couldn’t access your publication.
I will assume others have checked your application of it that produced the middle-left correlation with the ENSO data. I’ll take some time to think about the “sloshing” in three dimensions and see if I can create a narrative and visualization that I am comfortable with.
But I do want to thank you for helping me demonstrate what an actual scientific or engineering discussion might be like. That’s what I attempt to contribute here… even if we ignore the childish ranting that goes on, there is still too much rhetorical talking past each other and dodging questions. So it was gratifying to have a focused dialogue that produced a clear result.
#53, PP–
Thanks for the graphs. They do help in grasping what you are trying to say here.
A couple of questions, though:
1) It appears in the upper graphs that there are three traces, but only two are labeled. (dLOD, which must be delta in length of day, and model forcing, which must relate to the tidal forces. The latter is clear, as it’s in red, but what is the other blue trace? And which blue trace is which?)
2) Can you name/explain the units on the y-axes? This may well be a ‘newbie question’, but in this I’m a newbie, so…
Kevin asks:
There is only one blue trace. I think what you are seeing is an optical illusion, kind of related to a moire fringing pattern. The beats from the fortnightly and monthly tidal cycles are alternating as a pair, so the high value looks like one curve and the low value looks like another. It’s a high res image so you can blow it up. This pattern is common in tidal gauge readings at all scales — daily on up (at the daily scale it’s a beat of the diurnal and semidiurnal cycles).
The units on the y-axes are always an amplitude or intensity in terms of a standard scale of “arbitrary units”. The absolute value matters in terms of reconciling the forcing required but doesn’t matter for fitting a model to a cyclic pattern. This is well known from fitting a tidal time-series, as the absolute value of the tidal gauge readings are irrelevant to matching the pattern and then extending it. The utility of the A.U. label is one of the first things one learns when working in a research lab and is further explained on this wikipedia page – https://en.wikipedia.org/wiki/Arbitrary_unit.
There’s much research on studying the impact of climate change on wildlife populations in the northern latitudes. One issue is that some of the populations fluctuate wildly and so it is difficult to isolate the impact of a warming trend. This link describes research from last year claiming a lunar+ENSO link to erratic cyclic arctic lemming populations:
https://geoenergymath.com/2020/03/29/lemming-fox-dynamics-not-lotka-volterra/