Wednesday, March 12, 2014 ... Deutsch/Español/Related posts from blogosphere

Barycentric climatology was never a good science

In a new blog post titled

Death blow to Barycentrism: ‘On the alleged coherence between the global temperature and the sun’s movement’
Anthony Watts talks about the publication of a July 2013 preprint
On the alleged coherence between the global temperature and the sun's movement
by Sverre Holm of Oslo in the April 2014 issue of Journal of Atmospheric and Solar-Terrestrial Physics. Holm concludes that claims by Nicola Scafetta about the "impact of the Jupiter- and Saturn-affected motion of the Solar System's center-of-mass relatively to the Sun on the Earth's climate" are not only unsupported by any plausible physical mechanism.

They are also unsupported by significant evidence – the evidence used as a justification seems to be all about some random flukes that are fully compatible with the word "noise". And Holm's words "...due to a combination of model overfitting and smearing" suggests that he also thinks that the noise was slightly "helped" to emerge.

In June 2011, I would publish an – unedited – interview by V. Kremlík with Dr/Ms Ivanka Charvátová
Interview: Is climate change caused by solar inertial motion?
about these "barycentric theories of the climate". For a few hours, I was seriously thinking about it because it would be really "cool". But after those hours, I came back to my senses. It is really impossible physically.

For Anthony Watts, the barycentric climactic game is over, too. He was challenged by some WUWT readers but I do subscribe to Anthony's words.

Realize that the Earth, the Sun, and every other celestial body (I mean its center-of-mass) is moving along a trajectory that is accelerated, via \(\vec F=m\vec a\), according to the average gravitational acceleration \(\vec g\) calculated in the whole mass of the terrestrial body (the mass provides you with the right weighting for the "average"). Note that the "other side" of Newton's equations determining the force says that \(\vec F = m\vec g\) and \(\vec g = -(GM/R^2)\times \hat R\) in the radial direction etc.

Objects that are attached to the celestial body are effectively "freely falling", like in Einstein's elevator. The fictitious forces exactly "cancel" the gravitational ones. So you don't feel anything. The internal processes such as the climatic ones cannot feel anything, either.

Only the non-uniformities of the gravitational acceleration inside the celestial body have implications for the internal processes inside/at the celestial body. When \(\vec g\) is different on one side of the Earth or the Sun than another, the two end points want to be dragged in "slightly different directions" or by "slightly different accelerations". These are the "tidal forces" that are responsible for the tides on Earth, among other things.

However, you must realize that those are very small forces, especially if they are caused by very distant gravitational sources. Let's sit on the Earth and recall some scaling laws. A celestial body of mass \(M\) that is at distance \(R\) is contributing a negative gravitational potential\[

\Delta \Phi = -\frac{GM}{R}

\] where \(G\) is Newton's gravitational constant. However, the (radial, attractive) force is the \(R\)-derivative of that potential, so the acceleration scales according to the "inverse square" law,\[

|\Delta \vec g| = \frac{GM}{R^2}.

\] The tidal forces are another derivative of that so they go like \(1/R^3\). You may convert them to the differences between accelerations on two sides of the Earth – which are close to Jupiter and far from Jupiter – and these two sides are \(2R_E\) from one another where \(R_E\) is the Earth radius. So we have\[

|\vec g_{\text{one side}}- \vec g_{\text{other side}}| = \frac{4GM R_E}{R^3}

\] The factor of four comes from two in \(2R_E\) and from another \(2\) resulting from the differentiation of \(1/R^2\). The numerical factor is far less important than the scaling with \(1/R^3\). In particular, you may say that the tidal forces are suppressed by \(R_E/R\) relatively to the gravitational acceleration itself. And because the Earth radius is much smaller than the Earth-Jupiter distance, for example, it is a huge suppression, indeed.

In practice, only the Sun and the Moon are close enough and heavy enough to significantly influence tides on Earth.

Similarly, there are tidal forces on the Sun but they are just tiny because everyone is very far from the Sun. So the gravitational acceleration on the Sun and exerted by Jupiter is just \(2.08\times 10^{-7}{\rm m/s}^2\). If you change the distance by two Sun's radii, the acceleration only changes by \(10^{-9}{\rm m/s}^2\) relatively to the acceleration from the previous sentence. This is really tiny, more than \(10^{11}\) times smaller than the Sun-self-induced gravitational acceleration on the Surface of the Sun, about \(274\,{\rm m/s}^2\).

You may compare the tiny tidal forces from Jupiter (and even tinier ones from Saturn) with many other, potentially periodic, sources of acceleration coming from the changing electric and magnetic fields inside the Sun and related forces. If you think rationally, you will end up agreeing that the tidal forces have to be negligible. And as I tried to permanently convince you earlier, a direct impact of the "location" of the barycenter or a direct impact of the potential or the gravitational acceleration is impossible because it contradicts a highly precisely verified (and theoretically essential, in general relativity) principle, the equivalence principle. The principle says that experiments in a freely falling elevator (or inside/at a freely moving celestial body) proceed exactly as experiments outside any gravitational fields (without external sources of gravity). As long as you "freely fall", the gravity is undetectable by your stomach, by your clocks, by your measuring apparatuses, or by your skin (that could feel a temperature etc.). Or by any other local experiment.

In the paper, Sverre Holm would argue that the periodicities near 10-10.5 years, 20-21 years, 30 years, and 60-62 years are just "slight" excesses that are not suggestive of anything. If you Fourier-transform a random function (the nature of the randomness has to be carefully picked: there are "different" kinds of randomness and noise and different distributions and autocorrelations etc.) of time, you inevitably see that the Fourier transform is a function that is higher for some periodicities and lower for some other periodicities. This is statistically inevitable and it doesn't imply that the periodicities really "mean" something or uncover an important, repeatable law of physics. Especially if they're "modestly high", and they are modestly high in this case, they mean nothing and they will go away (or be replaced by other flukes) when more data is collected.

WUWT reader JBird was dissatisfied with Anthony's "game is over" comment:
Game over? Isn’t that a little bit like saying that the “debate is over?” :-)

REPLY: Oh, people will still debate it I’m sure. Tallbloke and his group of cyclists will try to prop it up, but I’d say it pretty much has reached the end of credulity as a workable theory.

Some years ago I thought the theory had some merit, and I dabbled with it a bit, but then just like with CAGW, things didn’t quite add up. Now I’m quite convinced it’s junk. – Anthony
I think that Anthony's reply is perfectly OK. Well, I do know skeptics – otherwise fine people, in some cases – who are ready to embrace anything if it "weakens" the claims by the alarmists. Well, I am not one of them and I am happy to see that Anthony Watts isn't one of them, either. The real reason why we oppose the catastrophic proclamations about the climate change is that they are indefensible by the theory and by the available empirical evidence – and the same reason implies that the barycentric theory of the climate has to be abandoned, too.

We may have seen some imperfect disagreement with Lord Monckton, too.
LM: I just want to make sure that Anthony doesn’t look like a lonely tyrant. I also think that these theories were ludicrous, for physical reasons.

Such an influence would heavily violate the equivalence principle (all bodies accelerate at the same rate in a gravitational field – so on Earth, you can’t locally measure the potential or gravitational acceleration but only the next derivative, the tidal forces), something we experimentally know to be extremely accurate (better than a part in trillion) and that seems to be exactly required for general relativity to be consistent. So it can’t be surprising that all the empirical support for such correlations had to be an accident that has to go away after a closer scrutiny. This was discussed e.g. under an interview with a “barycentrist” here

http://motls... (2011)

Search for “equivalence principle”, to get to my comments.

It is like saying “the debate is over” except that we have strong theoretical *and* empirical reason for stopping being interested in such potential explanations. People may keep on talking about everything but science isn’t an unrestricted babbling. Science eliminates ideas that disagree with known experiments or observations. Some barycentrists may fail to do so, and climate alarmists also fail to do so quite often, but Anthony usually doesn’t and I want to support him in his “blasphemous” decision to disagree with these explanations. Theories can’t be viable just because they are alarmist or just because they are “skeptical”.
Lord Monckton of Brenchley would say:
It is not improper to look for patterns in physical observations, for they may (or may not) reveal a physical law.

It would be interesting to know, for instance, what causes the 30-year periods of warming followed by 30-year periods of cooling that seem to characterize the global mean surface temperature anomaly record since 1850 that are in phase with the great ocean oscillations.

I do not say that these cycles – if they are more than mere coincidences – must be caused by the infinitesimal gravitational influence of the planets on the Sun. However, that the planets are capable of influencing each other gravitationally if the influence is exerted for long enough is suggested by the coplanarity of nearly all the planetary orbits.

And it is climate science that first gave us the notion that has come to be known as mathematical chaos – the observation that in certain objects, the climate arguably among them (Lorenz, 1963, Giorgii, 2005), even the most minuscule perturbation of the initial conditions at any chosen t-zero can exert a disproportionately large influence on the evolution of the object over time.

In short, both theory and observation indicate that it is not impossible for the planets to influence the Sun and, via the Sun, the Earth/Moon system. However, merely because it is not impossible, it ain’t necessarily so.
Another user named "kenmoonman" offered another comment where he or she (completely incorrectly) claimed that the periodic impact of the planets on the Earth's climate is more or less inevitable. Processes in physics are inevitable as long as they don't violate any principles or symmetries etc. (That's Gell-Mann's totalitarian principle: everything that isn't forbidden is mandatory.)

But the effects in which the Sun's or Earth's internal dynamics directly depend on the relative location of the Sun's center and the Solar System's center-of-mass is forbidden by the equivalence principle – or, if you prefer symmetries, by the diffemorphism symmetry of general relativity. Such an influence of other bodies may only be exerted through the non-uniformities of the gravitational fields, i.e. via the tidal forces, and they are demonstrably tiny and beaten by lots of other effects. Moreover, the the Earth and the Sun may easily and instantly adjust to these changes, even if they mattered – the Sun may get prolonged in one direction by a billionth of a percent – and that's it. This can't have a major impact on the Earth's equilibrium temperature balanced with the solar radiation at the accuracy 0.3 °C which is not a lot but it is still 0.1% of the absolute temperature, a vastly higher fraction than the relative increases of the Sun's semiaxis or acceleration resulting from the tidal forces.

So these effects at the significant strength that is assumed by the barycentric climatology are really impossible. Let me just copy-and-paste an answer to Lord Monckton:
LM: Dear Lord Monckton, you wrote that “It is not improper to look for patterns in physical observations, for they may (or may not) reveal a physical law.”

A priori, you are right. A posteriori, you are not. Before 1666 AD, any signal observed on Earth could have revealed a physical law describing the influence of planets on each other. But since 1666 AD, the actual law governing the mutual influence of planets on each other has been known, verified with a huge accuracy, and it unquestionably implies that the relative positions of Jupiter, Saturn, the Sun, or the center-of-mass of the Solar System cannot measurably influence any observation done locally on Earth. The whole Earth just accelerates according to the average vector of gravitational acceleration (caused by the Sun, the Moon, Jupiter, Saturn, and others) in its volume, and only deviations of the local gravitational acceleration from the average one – i.e. the tidal forces – may imply new effects measurable by those who are moving along with the Earth through space because they tend to push pieces of the Earth in different directions.

Similarly, the Sun itself is only affected by the tidal forces caused by the planets etc. Already the acceleration from Jupiter – that still doesn’t matter for the internal processes in the Sun – is about 10 billionths of the gravitational acceleration on Earth only. The tidal forces go like \(1/r^3\) and they imply “relative accelerations” of pieces of the Sun that are smaller by additional two orders of magnitude or so, like \(10^{-9}\,{\rm m/s}^2\). This is vastly smaller than any other source of acceleration that actually operates within the Sun.

The right law describing the forces exerted by one planet or star or another have simply been known (and Einstein’s refinements to Newton’s theory in the general theory of relativity are demonstrably too small to matter at the level that one could “feel”, e.g. through the climate), so it is misguided to look for or to expect completely different, incompatible explanations of the same kind of effects (forces exerted by one planet on another). It doesn’t mean that there can’t be any repeatable 30-year periodicity. But it does mean that if there is such a periodicity, the correct explanation does not include the forces exerted in between different celestial bodies of the Solar System because such influences have been ruled out by the same experiments that have confirmed Newton’s law of gravitation.

Chaos may be relevant for producing various hard-to-predict patterns. But it must be chaos in the “internal processes” in the Earth’s atmosphere (plus oceans) and I am afraid that by definition, “chaos” is the opposite than what you suggest. (Deterministic) chaos is when the past determines the future but the approximate past does *not* determine the approximate future. In effect, it means that it is impossible to determine the future because it depends – pretty much randomly – on *every* (exponentially) tiny variation of the initial conditions. And the negligible variations in the barycentric etc. quantities are not the only tiny variations. There are other, also tiny, but larger effects that matter. In effect, you are saying that one may restore the controllable predictability of the Earth’s climate etc. (following a simple law that may be described in these papers) by studying some details about the initial conditions. But that’s exactly what chaos prohibits if it is present.
It seems that I forgot to mention the "status" of the 60-year (or 30-year) periodicity, a general observation people often make about the 20th century temperature record.

It looks cool that the temperatures warmed in 1900-1940 and then again in 1975-2010 by a similar amount, so you might think that there is some 75-year (that's my counting here) periodicity. But while this explanation may be intriguing, the arguments for such a periodicity (as a general law) are de facto non-existent. If you see some effect (in this case warming) twice, it is not a terribly strong piece of evidence in favor of periodicity and in favor of declaring the distance between them to be the numerical value of the period. It's so weak evidence that I call it "the absence of evidence". The climate variations similar to the 1900-1940 warming that was "seemingly repeated" may be periodic for some reason but it seems more likely to me that they are not periodic (i.e. the "apparent" periodicity may sometimes be 60 years, sometimes 70, sometimes 40, sometimes 100). They are just a form of a noise dominated by a timescale in a ballpark but there's no reason why the periodicity should be exact or predictable or constant.

Some skeptics' claims are perfectly fine and compatible with everything we know (Lindzen's or Spencer's claims about thermostats and clouds; many theories involving ocean cycles in various ways, and so on); some skeptics' claims are completely or almost completely indefensible and the "barycentric climatology" is arguably among them (and so are the general claims that "the greenhouse effect doesn't exist as a matter of principle", as the slayed dragons and their pals say). Some theories are somewhere in between where I count cosmoclimatology (the role of cosmic rays whose effect is magnified by a radiation feedback in the atmosphere which may also allow to amplify the variations in the solar output). Cosmoclimatology is a potential, rather far-reaching and revolutionary explanation that cannot be ruled out by similarly strong arguments as the barycentric climatology and that has a mechanism that is in principle compatible with everything we know about physics but it's still plausible that a detailed argument shows that the strength of the effect is an order of magnitude or several orders of magnitude weaker than what is required.

I think that the climate debate has been going on for a sufficiently long time so that refined enough participants should be a bit more selective. Positive feedbacks strengthening the bare greenhouse effect by a factor of 5 are empirically excluded by now; but some theories including the barycentric one are excluded even more reliably, I think. The barycentric folks don't want the taxpayers to pay trillions of dollars a year to the altar of the climatological political correctness and that's their "practical advantage" over the climate alarmists. But when these political and economic implications are subtracted, and they should be subtracted when we debate "pure science", their science is at least as bad as the alarmists' science.

And that's the memo.

Add to Digg this Add to reddit

snail feedback (0) :

(function(i,s,o,g,r,a,m){i['GoogleAnalyticsObject']=r;i[r]=i[r]||function(){ (i[r].q=i[r].q||[]).push(arguments)},i[r].l=1*new Date();a=s.createElement(o), m=s.getElementsByTagName(o)[0];a.async=1;a.src=g;m.parentNode.insertBefore(a,m) })(window,document,'script','//','ga'); ga('create', 'UA-1828728-1', 'auto'); ga('send', 'pageview');