A new paper in Astronomy and Astrophysics (A&A) finds that cosmogenic nucleides in a section of Epica dome C covering 325-336 kyrs ago doesn’t exhibit Abreu et al’s planetary periods. They conclude that solar variability might have been different then. They don’t consider that Epica ice cores might not be telling them what they think they’re telling them.
Here’s the abstract
No evidence for planetary influence on solar activity 330 000 years ago
A. Cauquoin1⋆, G. M. Raisbeck2, J. Jouzel1, E. Bard3 and ASTER Team3⋆⋆
Context. Abreu et al. (2012, A&A. 548, A88) have recently compared the periodicities in a 14C – 10Be proxy record of solar variability during the Holocene and found a strong similarity with the periodicities predicted on the basis of a model of the time-dependent torque exerted by the planets on the sun’s tachocline. If verified, this effect would represent a dramatic advance not only in the basic understanding of the Sun’s variable activity, but also in the potential influence of this variability on the Earth’s climate. Cameron and Schussler (2013, A&A. 557, A83) have seriously criticized the statistical treatment used by Abreu et al. to test the significance of the coincidences between the periodicities of their model with the Holocene proxy record.
Aims. If the Abreu et al. hypothesis is correct, it should be possible to find the same periodicities in the records of cosmogenic nuclides at earlier times.
Methods. We present here a high-resolution record of 10Be in the EPICA Dome C (EDC) ice core from Antarctica during the Marine Interglacial Stage 9.3 (MIS 9.3), 325–336 kyr ago, and investigate its spectral properties.
Results. We find very limited similarity with the periodicities seen in the proxy record of solar variability during the Holocene, or with that of the model of Abreu et al.
Conclusions. We find no support for the hypothesis of a planetary influence on solar activity, and raise the question of whether the centennial periodicities of solar activity observed during the Holocene are representative of solar activity variability in general.
Tallbloke's Talkshop 
This pretty much shuts down the planetary effect ideas.
They talk about a ” high-resolution record of 10Be in the EPICA Dome C (EDC) ice core from Antarctica during the Marine Interglacial Stage 9.3 (MIS 9.3), 325–336 kyr ago”
However, at that time (325–336 kyr ago) the typical resolution is very low. One point may indicate a smooth of 500 yr and more, so the 100-yr scale patterns disappear. See here samples of EPICA data:
http://www.ncdc.noaa.gov/paleo/icecore/antarctica/domec/domec_epica_data.html
So, one needs to understand first how they get the “high resolution” record. Did they just interpolated the points in some way to get the high resolution? The ice itself works as a smoothing tools because particles diffuse in the ice and the lower and lower frequency patterns are lost as we go back in time.
From the paper: “The resolution of our measurements between 325–336 kyr (20–29 yr) is comparable to that (resampled at 22 yr) used by Abreu et al. (2012)”
It reflects badly on you that you do not dare publish my comments. Tallbloke=CENSORSHIP. Bad. Bad.
[Reply] Leif I’ve let them through, but you gave up your chance to post unhindered here when you failed to apologise for libeling me over Poppenhager’s words.
The authors of this paper are incredibly naive to think that the periodicities of planetary orbits during the Holocene (the last 11,700 years) should closely resemble the periodicities 325-326 kyr ago, or that the latter periodicities can even be reconstructed sufficiently accurately with General Perturbations Theory. General Perturbations Theory uses an analytic extrapolation of current oscillating orbit elements. In this case, investigating the periodicities all the way back to 314 thousand years before the Holocene would require extrapolating oscillating elements from the current (Holocene) period all the way back to 325 kyr before present.
GIGO! Even the far more accurate method of Special Perturbations used to create JPL’s planetary ephemerides doesn’t hold up very well for longer than 12,000 years. Special Perturbations reconstruction of past planetary orbit periods requires backwards-numerical-integration of the equations of motion of all the planetary orbits, and must include all ongoing mutual perturbations of all the planet-moon systems. At some point before 300 thousand years, even SP becomes chaotic.
-Gerry Pease
BTW, I used the familiar word “oscillating” instead of the technically more correct terminology “osculating orbital elements.” If I were writing a technical paper instead of just making a general comment, I would have adhered to protocol.
Right guys. So their ice cores are useless for the task and so is their ephemeris?
Next!
It does make you wonder who’s paying for all these failed Abreu et al rebuttals, and why?
Is it because they don’t want a credible solar prediction on the table for political reasons by any chance?
Perhaps the paper was “nepotistically” reviewed. Inform the publisher. The journal must be terminated ASAP!
Nicola: 🙂 🙂
I have this paper:
Manzi V., R. Gennari, S, Lugli, M. Roveri, N. Scafetta and C. Schreiber, 2012. High-frequency cyclicity in the Mediterranean Messinian evaporites: evidence for solar-lunar climate forcing. Journal of Sedimentary Research 82, 991-1005.
Click to access 991.full.pdf
where I found some of the solar harmonics in a record of several million year ago, but you cannot do this with the porous ice core nucleotides, You need a different type of sediment records that allows a real high resolution record, as that used here
Look Who I found lurking in the sinbin!
Evening Leif! Ready to apologise for libeling me yet?
Nicola that’s an interesting paper. Gerry, do you think those findings could help astronomy fix the degree to which perturbations affect the planetary system organisation on the multi-million year timescale? It seems to me to indicate the system is more stable than a gravity only perturbation theory can account for. It’s like there’s something pushing the planets back into place as well as a perturbatory force pushing them out of place. Maybe this helps with the why phi? question. The tendency to get back to lognormal distribution brought about by ???.
Roger, I think the system is probably inherently reasonably stable for 326 thousand years and, not having read their paper, I don’t know what level of accuracy is expected by the authors of the paper in order to draw conclusions about how well the ice core proxy periodicities should compare to Holocene era periodicities. Chaotic behavior in the interim would destroy the precision of the periods but not the general pattern, unless a large comet or asteroid from the Oort Cloud (or from an even greater distance) entered the solar system and impacted a planet in the 314 thousand year interim time frame. The probability that such a collision actually happened is not zero.
“Results. We find very limited similarity with the periodicities seen in the proxy record of solar variability during the Holocene.” They did find some “limitied similarity,” it seems. The devil is in the details, which I haven’t read.
It is likely that most of the differences result from problems with their proxy method, as you and Nicola have noted.
[snip] Here is a paper on the accuracy of the orbit calculations, they are good several million years back: http://www.leif.org/EOS/laskar-orbit-stability.pdf
[Reply] From now on, I’m going to moderate like McIntyre. Be warned.
This would a vast improvement.
Reply OK leif, you’re out of sinbin and on moderation. I can’t hold a grudge against you forever. Keep it polite or it goes in the bitbucket. 😉
The conclusion section of the paper:
“To test the hypothesis of Abreu et al. (2012) that there is a planetary influence on solar variability, we have investigated the spectral properties of a high-resolution 10Be record in an Antarctic ice core during the period 325–336 kyr BP. We find very limited similarity between the periodicities in this record compared to those found in a proxy record of solar variability during the Holocene, or those predicted by the model of Abreu et al. Since our record is only of a single nuclide from one core, we cannot definitively exclude the hypothesis of Abreu et al. Our results do suggest however, that it is important to test other proxy records of solar variability for times other than the Holocene. If observed periodicities in variability are not constant and stationary over time, this would rule out that any regular forcing factor, such as planetary positions, has a significant influence on solar activity.”
You have to actually read the paper to comment on it.
A question: if the data is good and if the orbit calculations are correct, would you agree with “this would rule out that any regular forcing factor, such as planetary positions, has a significant influence on solar activity”.
Well Leif, Nicola has written to the authors for clarification on how they’ve turned something with a resolution of around 500yrs into a “high-resolution 10Be record in an Antarctic ice core”. The quality of the orbital calcs depends on perturbation theory, which is incomplete.
Beyond that, contrast conclusion:
“we cannot definitively exclude the hypothesis of Abreu et al. Our results do suggest however, that it is important to test other proxy records of solar variability for times other than the Holocene.”
With title:
“No evidence for planetary influence on solar activity 330 000 years ago”
Nicola has already shown that planetary periodicities are present in a proxy which doesn’t suffer the diffusion issues found in ice. Why didn’t they cite him?
It’s paywalled so I have to flag up this small proviso:
If hindcasted, the planetary torque model would presumably be hindcasting beats and phases to within a few years, certainly tens of years. In other words, provided the astrodynamics is correct (thus making the pure mechanics of the torque model correct), Abreu et al can pinpoint an actual date to start a beat cycle and observe the beats and phases going forwards and backwards from that date.
This paper surely can’t nominate a start date as accurately as an astrodynamical model. Epica is dated by means of models (see Wiki quote below). This being the case, it can only look for and observe the periods of the beats themselves without truly knowing their start dates. I’m not even talking about spectral resolution which is probably another difficulty. I’m talking about absolute dating.
So here’s the potential problem. The authors say they have found “very limited similarity” between the core data and the periodicities in the Abreu et al torque model. “Very limited similarity” means some similarity and the ‘limited’ part could just mean they found similar periodicities but phase-shifted. Of course, it would be dishonest not to mention that you saw the same periodicities and not mention the phase shift, especially if that shift was likely due to the accuracy of absolute dating models compared with astrodynamical models.
I doubt there is any sleight of hand here and I’m not going shell out to get behind the paywall and find out. I’m not making accusations, just pointing out the way in which carefully constructed truthful statements can sometimes conceal a multitude of sins. Sins that capitalise on the complexity the authors have plumbed and we haven’t. I’ve seen too much of it not to see the red flags, I’m afraid.
From the Epica Wiki page.
The core time scale is derived from the measured depth scale by a model incorporating surface snow accumulation variations, ice thinning, basal heat fluxes etc., and is empirically “tied” at 4 times by matches to the marine isotopic record.
For all I know, Epica may pinpoint to the year, once it has derived it from its models. That’s useful for communicating with other scientists who know the accuracy isn’t perfect but I doubt very much it could be relied on for pinpointing a planetary opposition date. Epica would therefore be out of phase. If that is all they are really saying, that is a big problem.
I should just say why I have faith in one model and not another. Epica, like climate models, uses all sorts of estimates for snowfall variability and melt which have inherent inaccuracies. Astrodynamics programmes crunch pretty well one well tested equation ad infinitum. They have inherent inaccuracies too (such as the gravitational constant) but those are well- characterised and very small.
Leif says: “A question: if the data is good and if the orbit calculations are correct,…”
Sorry Leif, but only a person with no knowledge in ice core data can think that nucleotide ice core deduced data of 300,000 year ago might have the precision necessary for carefully determine the secular scale harmonics and their timing.
Particles diffuse in the ice core, and when you extract the ice core you cause a lot of fractures. Here it is not a fact of layer of ice that you can count more or less, the problem are the micro-fractures of the ice that let particles to move around. Moreover, for the initial 50 to 100 years the gas almost freely circulates within the snow which is not sufficiently compressed yet.
What ever record you extract from the ice core it is always a multidecadal-secular smooth record.
See for example these data
ftp://ftp.ncdc.noaa.gov/pub/data/paleo/icecore/antarctica/epica_domec/edc-ch4-2008.txt
Methane is a molecule bigger than Beryllium, so it should diffuse less. If you see in the file the data around 300000 year ago have a sampling of between 100 to 600 years.
lsvalgaard said:
January 24, 2014 at 7:41 pm
This pretty much shuts down the planetary effect ideas.
The sad thing is that Leif is completely blind to the possibility that Be10 atoms could slowly diffuse over time from their point of origin in the ice column i.e. on time scales of 10,000’s of years. Until he shows that this likely scenario can be ruled out he has not case for a rebuttal.
Beyond that, contrast conclusion:
“we cannot definitively exclude the hypothesis of Abreu et al. Our results do suggest however, that it is important to test other proxy records of solar variability for times other than the Holocene.”
With title:
“No evidence for planetary influence on solar activity 330 000 years ago”
****************
I smell a “nepotistic” review. Leif, did you review this paper?
Remember, Pal-Review is revealed when one can easily find an error in the paper. Roger is not even a professional scientist but he could easily spot a curious contradiction. The title contradicts the conclusion of the paper and clearly overstates it. The authors are clearly biased on the topic.
In science you can almost never ‘definitely exclude’ or ‘definitely include or prove’ anything, so the caveat is appropriate. [snip] Now, if other ice cores also do not show the same pattern as in the holocene would you then agree that there is no planetary influence. [Snip]
Leif: Snarky sentences removed. What’s left is logically defective. In effect you are saying that if one inadequate proxy can’t find x, then if the other similarly inadequate proxies can’t find x either, then x doesn’t exist.
Do you see the problem?
Speaking with the Abreu team on a similar topic it became evident that ice cores are not good for dating. They admit to using the 14C record as a base dataset to set their timeline. Tree rings are easier to count.
Using an ice core record 330,000 years old is total junk.
Even the 14C dating is under question right now, it might be the topic of some combined research soon.
Geoff,
I found that wikipedia has the 14C record upside down. Courtesy of one William Connolley.
Solanki has it the right way up though.
All these proxies have us looking through a glass darkly, but hopefully, once the planetary theory is better accepted, it will be used to clean up and improve the proxy records.
tallbloke says:
January 27, 2014 at 12:16 pm
Geoff,
I found that wikipedia has the 14C record upside down. Courtesy of one William Connolley.
Solanki has it the right way up though.
I am confused on this, the isotope record of any kind has to be inverted to match solar activity. I must be missing something but the WIKI record and Solanki (INTCAL98) both look inverted to me?
It’s a confusing subject. I found evidence that opposite to expectations, more 14C is found in tree samples at solar max than at solar min. Possibly because the density is greater in slower growing rings so more 14C is present per unit volume.
Review this (confusing) thread if you want to wade through it.
Tim Channon changed his mind and agreed with me in a later post
Dig down into the comments on this later post to see Tim’s 14C recon matched to Salvador’s solar recon
I remember reading through that thread and think there is a conflict of different record sets. There is currently a paper under review dealing with the late onset of the Maunder which may answer some of your questions in relation to the WIKI graph.
The Maunder took some time to get going when looking at the isotope record that is still an area of confusion, but Svalgaard has taken full advantage of this by stating many times that the dynamo during the Maunder Minimum modulated cosmic rays perhaps stronger than we see today. As usual he is not giving all the detail, as this only applies to the beginning of the Maunder Minimum.
tallbloke says:
January 27, 2014 at 1:43 pm
It’s a confusing subject. I found evidence that opposite to expectations, more 14C is found in tree samples at solar max than at solar min. Possibly because the density is greater in slower growing rings so more 14C is present per unit volume.
I am not convinced you have found evidence. The Stuiver, M., and P. D. Quay. 1981 and Kocharov 1995 data is raw and not inverted, plus I think you need to overlay the Stuiver data up to 1950 against the sunspot record which will show it doesn’t fit.
The 10Be record is not affected by growth cycles but still agrees with the 14C record, which I think is pretty conclusive. If you were right the whole Holocene record for both proxies would be wrong which is highly unlikely, the planet positions also support the proxy data.
I am meeting with Ken McCracken on Wednesday and we will be discussing this topic, I will ask him about the resolution of the 10Be data and also show him the Kocharov data, which does show an earlier onset of the Maunder.
Geoff, Excellent. I hope you have a productive meeting. I think if you consider how much the temperature rose 1900-1950 it seems likely a reduction in cloud was partly responsible, probably partly due to increasing solar activity. If Ms Dengel is correct, that would correlate to slower tree growth inside forests, which prefer diffuse sub-cloud lighting and moisture. So that would correlate with falling 14C values, which is what you see if you orientate the curve as I have. Just sayin’.
Geoff Sharp says:
January 27, 2014 at 1:03 pm
“I am confused on this, the isotope record of any kind has to be inverted to match solar activity. I must be missing something but the WIKI record and Solanki (INTCAL98) both look inverted to me?”
A while ago, I think it was on this site, it was pointed out that galactic cosmic rays decrease on high solar activity, while lower energy solar cosmic rays follow solar activity closely. If carbon 14 is produced by galactic cosmic rays, then would it not follow that high 14C would indicate low solar activity?
Pochas, high atmospheric C14 is indicative of low solar activity. The confusion arises because it appears that the measurement of C14 in wood samples goes the other way. This may be linked with the density of wood growth. See the threads I linked for Geoff, but be ready for more confusion…
TB:
I think a basic question here is whether ice core 10Be really measures changes in the level of solar activity. After studying the 10Be record in GISP2, the longest 10Be ice core record we have, Yiou et al concluded that 10Be in GISP2 in fact measures mostly changes in precipitation rate:
“the results confirm that the first-order origin of 10Be concentration variations is changes in precipitation rate associated with different climate regimes.”
ftp://ftp.ncdc.noaa.gov/pub/data/paleo/icecore/greenland/summit/grip/cosmoiso/grip_10be.txt
And if 10Be in Greenland ice cores measures mostly changes in precipitation rate it’s reasonable to suppose that 10Be in Antarctic ice cores might be measuring mostly changes in precipitation rate too.
There’s also the question of whether the GISP2 18O and cation concentrations aren’t also measuring precipitation rate changes. The fact that they move pretty much in lockstep with 10Be strongly suggests a common causation:
Results. “We find very limited similarity with the periodicities seen in the proxy record of solar variability during the Holocene, or with that of the model of Abreu et al.”
So they found some similarity even with a bias against Abreu et al model, interesting!
Their interpretation of finding a similarity and calling it a “very limited similarity” is very telling in itself.
Roger Andrews says:
March 27, 2014 at 11:12 pm
……
Agree.
In addition any reconstruction either of solar activity or climate parameters based on the cosmogenic nuclide production has serious inherent flaw.
Steinhilber et al: “Note that the variation of Φ depends on the geomagnetic field. If another geomagnetic field reconstruction like for example Korte M & Constable CG (2005) were used Φ would show another trend…”
More details here: http://www.vukcevic.talktalk.net/GCR-recon.pdf
Roger A: You missed off the “associated with different climate regimes” bit.
More Sun > more evaporation > faster hydrological cycle > more precipitation
So more 10Be is a proxy for more precip, and more precip is a proxy for more sun (and likely less cloud).
When I was young and the Sun was strong,
the rain fell quick then the clouds were gone…
Vukcevic :
Other possibilities that could be considered:
– Solar magnetic field variability (the cause of TSI variability) modulates cosmogenic production of C14 and 10Be
and ‘oscillates’ at a slower rate than the geomagnetic field, while frequency shift (at the lower end of the spectrum,
from ~ 1/400 year– 1) between two is a mere but extraordinary coincidence.
– Geomagnetic field is the primary modulator of cosmogenic production, while a transfer of energy from higher to the
lower frequencies causes the frequency shift.
Theoretically such process could occur when a collection of nonlinear processes act together on an ‘pulsed’ beam,
in this case on the galactic cosmic rays GCRs, which are generated by supernovae explosions, powerful but short
lasting events.
TB:
“more 10Be is a proxy for more precip, and more precip is a proxy for more sun (and likely less cloud).”
Agree with the first bit, dubious about the second. (Surely less cloud would cause less precip, not more?)
But here’s the point. The fact that GISP2 10Be correlates so closely with other variables that aren’t cosmogenic (d18O, cations, also methane) strongly suggests that while the 10Be in GISP2 may indeed be cosmogenic the fluctuations it measures aren’t. The same probably goes for Taylor Dome 10Be in the Antarctic – take a look at it.
And if EPICA 10Be also isn’t measuring cosmogenic variations then Cauquion et al’s results are explained.
Roger A: (Surely less cloud would cause less precip, not more?)
Ain’t necessarily so.