Dr. Roger Pielke Sr. writes about a new paper from Nicola Scafetta.:
A new paper has just appeared
Nicola Scafetta 2011: A shared frequency set between the historical mid-latitude aurora records and the global surface temperature. Journal of Atmospheric and Solar-Terrestrial Physics In Press doi:10.1016/j.jastp.2011.10.013
This paper is certainly going to enlarge the debate on the role of natural climate variability and long term change.
The abstract reads [highlight added]
Herein we show that the historical records of mid-latitude auroras from 1700 to 1966 present oscillations with periods of about 9, 10–11, 20–21, 30 and 60 years. The same frequencies are found in proxy and instrumental global surface temperature records since 1650 and 1850, respectively, and in several planetary and solar records. We argue that the aurora records reveal a physical link between climate change and astronomical oscillations. Likely in addition to a Soli-Lunar tidal effect, there exists a planetary modulation of the heliosphere, of the cosmic ray flux reaching the Earth and/or of the electric properties of the ionosphere. The latter, in turn, has the potentiality of modulating the global cloud cover that ultimately drives the climate oscillations through albedo oscillations. In particular, a quasi-60-year large cycle is quite evident since 1650 in all climate and astronomical records herein studied, which also include a historical record of meteorite fall in China from 619 to 1943. These findings support the thesis that climate oscillations have an astronomical origin. We show that a harmonic constituent model based on the major astronomical frequencies revealed in the aurora records and deduced from the natural gravitational oscillations of the solar system is able to forecast with a reasonable accuracy the decadal and multidecadal temperature oscillations from 1950 to 2010 using the temperature data before 1950, and vice versa. The existence of a natural 60-year cyclical modulation of the global surface temperature induced by astronomical mechanisms, by alone, would imply that at least 60–70% of the warming observed since 1970 has been naturally induced. Moreover, the climate may stay approximately stable during the next decades because the 60-year cycle has entered in its cooling phase.
The highlights listed in the announcement of the paper read
► The paper highlights that global climate and aurora records present a common set of frequencies. ► These frequencies can be used to reconstruct climate oscillations within the time scale of 9–100 years. ► An empirical model based on these cycles can reconstruct and forecast climate oscillations. ► Cyclical astronomical physical phenomena regulate climate change through the electrification of the upper atmosphere. ► Climate cycles have an astronomical origin and are regulated by cloud cover oscillations.
========================================================
Dr. Scafetta writes in and attaches the full paper in email to me (Anthony) this week saying:
I can forecast climate with a good proximity. See figure 11. In this new paper the physical link between astronomical oscillations and climate is further confirmed.
What the paper does is to show that the mid-latitude aurora records present the same oscillations of the climate system and of well-identified astronomical cycles. Thus, the origin of the climatic oscillations is astronomical what ever the mechanisms might be.
In the paper I argue that the record of this kind of aurora can be considered a proxy for the electric properties of the atmosphere which then influence the cloud cover and the albedo and, consequently, causes similar cycles in the surface temperature.
Note that aurora may form at middle latitude or if the magnetosphere is weak, so it is not able to efficiently deviate the solar wind, or if the solar explosions (solar flare etc) are particularly energetic, so they break in by force.
During the solar cycle maxima the magnetosphere gets stronger so the aurora should be pushed toward the poles. However, during the solar maxima a lot of solar flares and highly energetic solar explosions occurs. As a consequence you see an increased number of mid-latitude auroras despite the fact that the magnetosphere is stronger and should push them toward the poles.
On the contrary, when the magnetosphere gets weaker on a multidecadal scale, the mid-latitude aurora forms more likely, and you may see some mid-latitude auroras even during the solar minima as Figure 2 shows.
In the paper I argue that what changes the climate is not the auroras per se but the strength of the magnetosphere that regulates the cosmic ray incoming flux which regulate the clouds.
The strength of the magnetosphere is regulated by the sun (whose activity changes in synchrony with the planets), but perhaps the strength of the Earth’s magnetosphere is also regulated directly by the gravitational/magnetic forces of Jupiter and Saturn and the other planets whose gravitational/magnetic tides may stretch or compress the Earth’s magnetosphere in some way making it easier or more difficult for the Earth’s magnetosphere to deviate the cosmic ray.
So, when Jupiter and Saturn get closer to the Sun, they may do the following things: 1) may make the sun more active; 2) the more active sun makes the magnetosphere stronger; 3) Jupiter and Saturn contribute with their magnetic fiend to make stronger the magnetic field of the inner part of the solar system; 4) the Earth’ magnetosphere is made stronger and larger by both the increased solar activity and the gravitational and magnetic stretching of it caused by the Jupiter and Saturn. Consequently less cosmic ray arrive on the Earth and less cloud form and there is an heating of the climate.
However, explaining in details the above mechanisms is not the topic of the paper which is limited to prove that such kind of mechanisms exist because revealed by the auroras’s behavior.
The good news is that even if we do not know the physical nature of these mechanisms, climate may be in part forecast in the same way as the tides are currently forecast by using geometrical astronomical considerations as I show in Figure 11.
The above point is very important. When trying to predict the tides people were arguing that there was the need to solve the Newtonian Equation of the tides and the other physical equations of fluid-dynamics etc. Of course, nobody was able to do that because of the enormous numerical and theoretical difficulty. Today nobody dreams to use GCMs to predict accurately the tides. To overcome the issue Lord Kelvin argued that it is useless to use the Newtonian mechanics or whatever other physical law to solve the problem. What was important was only to know that a link in some way existed, even if not understood in details. On the basis of this, Lord Kelvin proposed an harmonic constituent model for tidal prediction based on astronomical cycles. And Kelvin method is currently the only method that works for predicting the tides. Look here:
http://en.wikipedia.org/wiki/Tide-predicting_machine
Figure 11 is important because it shows for the first time that climate can be forecast based on astronomical harmonics with a good accuracy. I use a methodology similar to Kelvin’s one and calibrate the model from 1850 to 1950 and I show that the model predicts the climate oscillations from 1950 to 2010, and I show also that the vice-versa is possible.
Of course the proposed harmonic model may be greatly improved with additional harmonics. In comparison the ocean tides are predicted with 35-40 harmonics.
But this does not change the results of the paper that is: 1) a clearer evidence that a physical link between the oscillations of the solar system and the climate exists, as revealed by the auroras’ behavior; 2) this finding justifies the harmonic modeling and forecast of the climate based on astronomical cycles associated to the Sun, the Moon and the Planets.
So, it is also important to understand Kelvin’s argument to fully understand my paper.
…
This work is the natural continuation of my previous work on the topic.
Nicola Scafetta. Empirical evidence for a celestial origin of the climate
oscillations and its implications. Journal of Atmospheric and Solar-Terrestrial Physics Volume 72, Issue 13, August 2010, Pages 951-970
http://www.sciencedirect.com/science/article/pii/S1364682610001495
Abstract
We investigate whether or not the decadal and multi-decadal climate
oscillations have an astronomical origin. Several global surface temperature
records since 1850 and records deduced from the orbits of the planets
present very similar power spectra. Eleven frequencies with period between 5
and 100 years closely correspond in the two records. Among them, large
climate oscillations with peak-to-trough amplitude of about 0.1 and 0.25°C,
and periods of about 20 and 60 years, respectively, are synchronized to the
orbital periods of Jupiter and Saturn. Schwabe and Hale solar cycles are
also visible in the temperature records. A 9.1-year cycle is synchronized to
the Moon’s orbital cycles. A phenomenological model based on these
astronomical cycles can be used to well reconstruct the temperature
oscillations since 1850 and to make partial forecasts for the 21st century.
It is found that at least 60% of the global warming observed since 1970 has
been induced by the combined effect of the above natural climate
oscillations. The partial forecast indicates that climate may stabilize or
cool until 2030–2040. Possible physical mechanisms are qualitatively
discussed with an emphasis on the phenomenon of collective synchronization
of coupled oscillators.
=======================================================
The claims here are pretty bold, and I’ll be frank and say I can’t tell the difference between this and some of the cycl0-mania calculation papers that have been sent to me over the last few years. OTOH, Basil Copeland and I looked at some of the effects of luni-solar on global temperature previously here at WUWT.
While the hindcast seems impressive, a real test would be a series of repeated and proven short-term future forecasts. Time will tell.
Geoff Sharp says:
November 15, 2011 at 3:06 am
The mid-latitude aurora is what Nicola is representing which you conveniently omitted, which has been the pattern of your hand waving exercise. It is far from a red herring, more the elephant in the room. This is not a restriction, more the meat of the paper.
The mid-latitude aurora is well represented by the observations from Denmark, which fully qualifies as mid-latitude [Scafetta even includes the Faroe Islands at 62N]. And what is there to say about this, except that they show what even Nicola claims he now understands, namely that the mid-latitude aurorae follow the sunspot cycle closely.
You could start with the data behind Nicola’s paper and if needed I am sure Nicola could provide other papers as referenced in his paper.
The data is of poor quality compared to the geomagnetic record of the electric currents of which the auroral sightings are but unreliable proxies. I know that auroral data very well, the records being stored in my very office at the Danish Meteorological Institute. We stopped recording the sightings in the late sixties because it was realized how useless an exercise it was now that we had the much better magnetic records. Here are the sightings record for several stations in the North Atlantic: http://www.leif.org/research/Aurorae-DMI-Nord.png . Note how dissimilar they are although looking at the same aurorae. Now, Scafetta claims that the auroral data after 1900 don’t matter for his paper, so they can hardly be any elephant in his room.
Geoff Sharp says:
November 15, 2011 at 3:06 am
The mid-latitude aurora is what Nicola is representing
And he uses the Faroe Islands data as representative of that. The data from Germany collected here shows the very close relation between auroral sightings and geomagnetic activity:
http://www.geofisica.unam.mx/divulgacion/geofinternacional/iframes/anteriores/2011/04/6_schroder.pdf
Returning to the Faroe Islands, note that the number of auroral nights at the height of the second largest solar cycle in 1947 is on par with those reported for some of the deepest solar minima, showing how unreliable the data is. Stations north and south of the Faroes show almost no aurorae at these minima. Silverman [his Figure 3] has compiled whatever data there is from mid-latitudes. His conclusion is plain: the decades 1900-1920 have very few aurorae in agreement with the geomagnetic record, so, effectively demolishing any claims to the contrary.
But, I think you have lost track of the issue, which is whether auroral sightings at mid-latitudes follow the solar cycle [as I and everybody else claim] or not [as Nicola claims].
Geoff Sharp says:
November 15, 2011 at 3:06 am
I am sure Nicola could provide other papers as referenced in his paper.
In fact, he does cite Silverman. Here is the Silverman data [Fig.3] since 1800. http://www.leif.org/research/Silverman-Fig-3.png . Combined with what we know to be large auroral activity after 1950, it is clear that if there is any ‘cycle’ it is a 100-year cycle. So, no 60-yr cycle in mid-latitude aurora.
Geoff Sharp says:
November 15, 2011 at 3:06 am
I am sure Nicola could provide other papers as referenced in his paper.
He does, in fact, cite Silverman, who has compiled the most complete list of all, some 45,000 observations. Here is the past 200 years of Silverman’s data about mid-latitude aurorae: http://www.leif.org/research/Silverman-Fig-3.png . Combined with our knowledge that the cycles after 1950 [19, 21,22] were large and thus had lots of aurorae, you can see there is not a 60-yr cycle, but, if any, a 100+ year cycle. So, that takes take of that elephant.
Leif Svalgaard says:
November 15, 2011 at 6:56 am
………..
Here is the first of three (two to follow) articles:
http://hal.archives-ouvertes.fr/docs/00/64/12/35/PDF/NorthAtlanticOscillations-I.pdf
Currently only in a ‘document technical validation’; dissect it for the ‘not so good, bad and ugly’ bits, so I can move it forward .
If Anthony is about, please have a go too, if you are so inclined.
I better say thank you before you respond, but I can take it, do your ‘worst’, the only way to find out if something is any good (I am still learning how to attach the ‘sad face’). Tnx.
“I think you are getting confused about the brief period when J/S have their greatest effect. Yes there is a point of greatest modulation but it takes 30 years to reach that point before turning back to a baseline position 30 years later.”
Nonsense, what about the other two J/S synods in this supposed 60yr cycle ? Draw a 60yr sign wave and plot the other two synods and see how they fall on the negative portion of the wave. Now there is no logical reason why those two would be negative and the third one producing the whole positive portion of the cycle. And what of the 3 J/S oppositions within the 60yrs, is only the middle one more negative somehow ?
So how do we define “when Jupiter and Saturn are close to the Sun” ? IMO giving it as much as a 3yr window around the synod is stretching it.
“Solar wind speed IMO does not have enough variation over the cycle as I have shown you previously.”
Now that is a problem as solar wind speed it has a direct impact on auroras.
Ulric Lyons says:
November 15, 2011 at 8:35 am
Now that is a problem as solar wind speed it has a direct impact on auroras.
Mid-latitude aurorae only occurs during geomagnetic storms and those depend very little on solar wind speed. The real factors that determine the strength of a storm are the compression of the solar wind magnetic field and density combined with the occurrence of the field pointing southwards.
M.A.Vukcevic says:
November 15, 2011 at 8:25 am
Currently only in a ‘document technical validation’; dissect it for the ‘not so good, bad and ugly’ bits, so I can move it forward .
In order to get you on track early on, you should not calculate R^2 from moving averages. That is not statistically sound. If you have 100 data points, you can compute a moving average over say 5 points, which will give you 96 ‘data points’, but they are not independent, so R^2 [and various other tests for significance] is meaningless [much too high]. You can compute R^2 [and significance] from the 20 averages you would get from grouping the original data points into 20 consecutive intervals of 5 points. Any referee would reject the paper right here because of this, so don’t do this.
Leif Svalgaard says:
November 15, 2011 at 9:16 am
In order to get you on track early on, you should not calculate R^2 from moving averages.
Which is what you do as far as I can see, but I could be wrong. So, specify clearly if the R^2s are calculated from the single data points or the moving averages. You should also bear in mind that too many [and too detailed] Figures tend to obscure your message [counter-intuitive, but a fact] and that Journals often have limitations on the number of Figures for that reason. Finally, we should not hijack this thread to discuss your missive.
Leif Svalgaard says:
…….You are unusually restrained, but thanks for now (smiley face here).
-It is not going into paper print, just electronic version (.pdf file), so number and size of Figures is no problem.
– Yes it is R^2 with moving averages, but that is necessitated by the nature of the data, one refers to atmospheric pressure (very volatile) already heavily smoothed, the other is oceans surface temperature, which again is made of thousands averaged elements.
Nearest you can get to a reasonable data is shown in Fig.11
http://hal.archives-ouvertes.fr/docs/00/64/12/35/PDF/NorthAtlanticOscillations-I.pdf
for which only trends are calculated (unexpectedly exact opposite).
The message :
– the IPCC is hugely deficient in the AMO-NAO relationship stakes, none of the stuff in the numbered graphs is known to climate scientists, including Dr. Mann, the expert authority on the AMO.
– degree of predictability based on the historical records correlation.
Ulric Lyons says:
November 15, 2011 at 8:35 am
Now that is a problem as solar wind speed it has a direct impact on auroras.
Mid-latitude aurorae only occurs during geomagnetic storms and those depend very little on solar wind speed. The real factors that determine the strength of a storm are the compression of the solar wind magnetic field and density combined with the occurrence of the field pointing southwards.
And the number of storms [which determines the number of nights with mid-latitude aurorae] does not depend on the solar wind speed but on the number of CMEs [which depends directly on the number of sunspots]. So, solar wind speed is ‘not a problem’ in this.
M.A.Vukcevic says:
November 15, 2011 at 9:57 am
– Yes it is R^2 with moving averages, but that is necessitated by the nature of the data, one refers to atmospheric pressure (very volatile) already heavily smoothed, the other is oceans surface temperature, which again is made of thousands averaged elements.
The underlying data is not smoothed as moving averages in time, so their smoothing does not matter. Repeat: any referee would reject the paper right here. If you need to smooth, calculate averages over adjacent [but non-overlapping] intervals. Do not correlate moving averages, that is a sure cause for rejection. I’m just saving you grief to come. But, as I said, no more discussion on the thread. Post it as a separate article.
will do!
Leif Svalgaard says:
November 15, 2011 at 9:58 am
“Mid-latitude aurorae only occurs during geomagnetic storms and those depend very little on solar wind speed.”
Well maybe Scafetta should have looked at polar auroras and land temperatures alone instead, it is clear that the “abrupt discontinuities” here: http://www.leif.org/EOS/92RG01571-Aurorae.pdf
correspond regularly to episodes of very low land temperatures.
Ulric Lyons says:
November 15, 2011 at 10:46 am
it is clear that the “abrupt discontinuities” here:
Most of those are observational problems [‘civilization’ bias, reporting issues, rise of newspapers, etc] not real changes in solar input or actual auroral occurrence.
Leif Svalgaard says:
November 14, 2011 at 8:33 pm
Again: please state here for the record that you now fully agree that mid-latitude aurorae follow the sunspot cycle and geomagnetic activity records in concert with the generally accepted view of auroral physics and all the modern data our marvelous space-based monitors have provided us with.
~
Will perhaps mostly due?
I still take issue with the penetration of interstellar neutrals crossing the orbital paths of the planets within the heliosphere, in particular whilst entering and exiting the frontal nose direction of the indented heliosphere bubble at equinox.
Image A in this IBEX (Interstellar Boundary Explorer) image set.
http://www.sciencemag.org/content/326/5955/969/F1.large.jpg
In the image you see the Oxygen trailing the helium in our orbit around the sun.
And speaking of oxygen and those slower alot slower moving electrons visible in the Ozarks in the next image.
Oct 25, 2011
http://www.spaceweather.com/aurora/images2011/24oct11/Brian-Emfinger1.jpg
Lot of stuff hitting the upper atmosphere over a rather extended area of the northern hemisphere hey there.
Density changes ..
Leif Svalgaard says:
November 14, 2011 at 5:55 pm
The Faroe data in Figure 2 are not homogeneous [for whatever reason] and cannot be used as support as you claim. Figure 3 of Silverman 1992 [ http://www.leif.org/EOS/92RG01571-Aurorae.pdf ] show this clearly.
There are power peaks shown in this paper of S.M. Silverman and he has identified the frequencies and the and the time periods from 5.6 years to ~83 years in Fig 4 and Fig 5.
The frequencies [1/years] of the aurora can be assigned to harmonic frequencies or synodic frequencies in the solar system of the planets Jupiter, Saturn, Chiron, Uranus, Neptune and Quaoarm, but also to the main frequency of the sun spot cycle of 1/11.196 y:
Period[y] f[1/y]
84 0.011903 = 0.011903 = 84y
UR
55.6 0.01797 = 0.011903 + 0.00606832 = 0.01797 = 55.6y
UR + NE
33.3 0.03003 = 0.033947 – 0.00403089 = 0.029916 = 33.426y
SA - PL
24.45 0.0408998 = 0.033947 + (2*0.0034749)= 0.0408968 = 24.452y
SA + (2*QU)
18.2 0.054945 = 0.033947 + (6*0.0034749)= 0.054793 = 18.2y
SA + (6*QU)
14.81 0.067522 = 2*0.033947 = 0.067894 = 14.72y
(2*SA)
12.904 0.07775 = 0.084317 – (2*0.0034749)= 0.077367 = 12.92y
JU - (2*QU)
11.39 0.087796 = 0.084317 + 0.0034749 = 0.087792 = 11.39y
JU + QU
10.17 0.098328 = 0.084317 +(4*0.0034749) = 0.098216 = 10.18<y
JU + (4*QU)
9.41 0.106269 = 0.084317 + (2*0.0119032)= 0.108123 = 9.249y
JU + (2*UR)
8.60 0.11625 = 0.084317 + 0.033947788 = 0.118265 = 8.455y
JU + SA
8.0808 0.12375 = 0.084317 + (2*0.019700) = 0.123718 = 8.083y
JU + (2*CH)
5.56 0.1798 = (2*0.08931) = 0.17863 = 5.598y
(2*SS)
It is remarkable that the power of the slow moving synods is high, while the power of the synodic frequencies with the faster Jupiter are small. Also the half frequency of the object Chiron fits in this scheme.
This must not mean that there must be a mechanism were these couples of planets are involved, but it is remarkable that there are these relations, it links again the aurora periods to the planets in the solar system.
V.
Ulric Lyons says:
November 15, 2011 at 8:35 am
Nonsense, what about the other two J/S synods in this supposed 60yr cycle ? Draw a 60yr sign wave and plot the other two synods and see how they fall on the negative portion of the wave.
I think you will find it is not all about the J/S synod, you will notice a rough 10-11 year frequency in the peaks of Nicola’s figure 7A. When considering tidal forces Saturn is about 5% of Jupiter’s tidal force, thus the shape of Jupiter’s orbit can be more important than the J/S conjunction. Nicola and I differ on the orbital mechanics involved at this point but the data is still the same (JPL). I am of the view that both the Earth and Jupiter have the Sun as their orbit axis point and the Jupiter perihelion is controlled solely by planet perturbations with Saturn making up the majority of influence. My Jupiter/Sun distance graph shows that perihelion can occur at J/S conjunction and J/S opposition, which I think proves my point, but you may need to get clarification from Nicola.
http://tinyurl.com/2dg9u22images/j-sun1.jpg
Leif Svalgaard says:
November 15, 2011 at 7:24 am
The Schroder paper clearly shows that mid latitude aurora occur during times of low solar output. Of 171 aurorae 23% were observed with a Kp index lower than 5. You may prefer to rely on magnetic readings but it is clear there can be a disconnect, which is Nicola’s point.
Your claim that there is NOT a 60 year period in the mid-latitude aurora is still yet to be shown. I can see you will not analyze the data, so your claims will continue to be hand waving.
Geoff Sharp says:
November 15, 2011 at 3:05 pm
“I think you will find it is not all about the J/S synod..”
I think I will find it`s got nothing to do with Jupiter`s tidal effects on Earth`s magnetosphere either.
“My Jupiter/Sun distance graph shows that perihelion can occur at J/S conjunction and J/S opposition..”
They precess over 40 synods: http://www.hps.cam.ac.uk/starry/keplerastrolmed.jpg
The last times the syzygies were closest to Jupiter perihelion was in 1702 and 1762. You may well find it was warmer around the 1730`s and 1790`s in that century.
The Silverman paper is interesting and backs up the Scafetta paper. Importantly it is necessary to only look at the data covering mid-latitudes. Two graphs in particular covering from 1700-1948 which cover the New England data (20,000 records taken over several hundred locations 41-45N) and the Fritz data that appears to be >55N show a clear 60 year period (the Dalton minimum has to be allowed for). The graphs agree with fig 2B in Scafetta’s paper.
Clearly the mid-latitude aurora record mentioned does NOT follow the sunspot or aa record of the era (with the exception of the Dalton).
http://tinyurl.com/2dg9u22images/fritz.png
http://tinyurl.com/2dg9u22/images/newengland.png
Ulric Lyons says:
November 15, 2011 at 5:14 pm
You are missing all the points Ulric and making no sense. I tried but invariably failed, I will leave you to it.
Geoff Sharp says:
November 15, 2011 at 6:01 pm
“You are missing all the points Ulric and making no sense.”
Quite the opposite, I showed you the points you had missed, and why the idea that a 60yr cycle could arise from these configurations makes no sense.
Geoff Sharp says:
November 15, 2011 at 5:54 pm
Clearly the mid-latitude aurora record mentioned does NOT follow the sunspot or aa record of the era (with the exception of the Dalton).
Since today we know the physics and we have excellent data and therefore know that the aurorae follow very well the sunspot record, you can conclude that if an auroral record does not agree with this, then that record is faulty and unreliable. This is the point you and Nicola are missing.
I presume this is the data file
aurorae.dat.rev
which is located here
ftp://ftp.ngdc.noaa.gov/STP/SOLAR_DATA/AURORAE
Using R to histogram the years between 1700-1900 and then plotting the probability density, eye balling it, I get 3 gaussians 60 +- 10 years wide at half height.
$breaks
[1] 1700 1720 1740 1760 1780 1800 1820 1840 1860 1880 1900
$counts
[1] 136 497 243 530 1017 135 628 1325 757 109
The peak values are located at roughly 1720, 1780, and 1850 years.
The ratios of the peak values are roughly 1:2:3 (from right to left which is odd.)
The histogram is similar to figure B.
If you stand on your head and squint, you can see the 3 gaussians in figure B.