jump to navigation

A last batch of talks before the summaries March 29, 2007

Posted by dorigo in astronomy, news, physics, science.

Today is the last day of the Outstanding Questions in Cosmology conference. A lot of short talks have been scheduled for this morning, while in the early afternoon there will be two conference summaries, one by Albert Stebbins and the other by Richard Lieu.

Most of this morning’s talks are quite technical, and I feel I would either have to be cryptic or write baloney if I were to report about them. But let me give some flavor of the talks in a few lines below anyhow. As with other posts this week, I have to apologize to those of you who find these discussions too hard to follow – I will resume outreach-style blogging after this conference is over!.

A talk I followed with some attention was Damien Hutsemekers’ “Large-Scale Alignments of Quasar Polarization Vectors: Observational Evidence and Possible Implications” (a good example of the inverse proportionality existing in conference talks between the length of the title and the duration of the talk allotted by the organizers – 15′ in talks during today’s morning session; compare it for instance to “CMB and the Dark Ages”, “3-year results from WMAP”, or “Does WMAP require Dark Energy”, all 30′ talks given in the previous days).

Hutsemekers explained that most quasars (a well-established contraption for quasi-stellar radio sources) exhibit only a small linear polarization in their optical emission, of the order of a few percent. The polarization vectors appear to be aligned over scales of about one giga-parsec, at redshifts of about 1: therefore their study could shed some light on the existence of large scale structures in the Universe, something which is all the rage in Cosmology these days. 

A study of 355 polarized quasars with redshifts up to z=2.5 measured the correlation of the polarization vector of each quasar with its neighbors – or with the average polarization in its neighborhood. The result is that there not only seems to be evidence for a correlation at large distance scales, but that the quasars contributing the most to the correlation appear to line along an axis (see graph in the picture above, where quasars are plotted as spheres whose size depends on the amount of correlation) which is not too far from another axis of symmetry that seems to exist between the low-l multipoles of cosmic microwave radiation. Hutsemekers dismissed the possibility of instrumental bias (the polarization measurements are consistent across different experiments) or due to some interstellar polarization effect. I personally found the evidence shown that the effect is not due to an interstellar medium quite unconvincing…

S.Batthcharya talked on “Cosmological Constraints from Galaxy Cluster Velocity Statistics”. He explained that while the Sunyaev-Zeldovich effect (a modification of the energy of the photons due to collisions with electrons) is responsible for most of the shift of the power spectrum of cosmic microwave radiation, an interaction with moving electrons of the microwaves causes deformation of the shift. From the deformations you can obtain information on the velocity of galaxy clusters. The kinematical effects on the temperature spectrum of cosmic microwave radiation are small, but ACT, a new detector which should provide a survey of 150 squared degrees of sky, could detect the 10 micro-kelvin fluctuations with a good enough angular resolution. The combination of ACT data with WMAP3 and other results is expected to thus provide a 5-fold improvement in the current constraints on some cosmological parameters.

Daniel Sudarsky gave an interesting talk titled “Do we really understand the quantum origin of the seeds cosmic structure?“. Another instance of a rethorical question in a talk title. He mentioned that the last decade saw a big success for inflationary cosmology. Primordial inhomogeneity predicted by inflation is in agreement with observations. However, there is something strange: the universe starts homogeneous and isotropic, has a scalar field (the inflaton) which also satisfies a homogeneous and isotropic potential, and still, we get in a situation which is not homogeneous and isotropic (H&I), despite the fact that the dynamics should preserve that symmetry.

Daniel explained that the departure from H&I cannot be ascribed to quantum mechanics alone, since the universe is the prime example of a closed system. So popular arguments include asking you to accept that quantum mechanics does not describe our universe, but a superposition of universes. Others, Daniel argues, will try to convince you that our universe is actually still H&I. Or that it really does not matter. But the question is not purely philosophical: some of these issues can be studied and tested xperimentally – he remarked that even inflation is at times attacked as being a philosophical motivation.

Sudarsky quoted Penrose, who said that quantum mechanics is incomplete: that is the conclusion you arrive at if quantum mechanics is not just a description of our information about reality, but it is tout court a description of reality.

After discussing at length these issues, Sudarsky found out he had only three minutes left for his talk, and there followed a shooting of slides actually full of complicated quantum mechanical formulas which I could not quite grasp. But then he concluded: Something, related to quantum mechanics, is missing in our understanding of the origin of cosmic structure. Something like a self-induced collapse is required to take us from a state which is completely homogeneous and isotropic to another one lacking those symmetries. Present analysis seems to offer a path to alleviate the need of fine tuning.

I must say I did not understand much of the above discussion, but listening to Sudarsky’s talk was still quite interesting to me…


1. Shantanu - March 29, 2007

thanks for the detailed summaries. Do you know if the slides of the talks
are up ?

2. Carl Brannen - March 30, 2007

The correlation of polarizations to me is easily the most fascinating of the talks. I’ve seen the papers, for example the 355 quasar paper is from 2005 and now has references you might want to read: astro-ph/0507274 citactions.

The reason I find these interesting is that, well, I don’t believe in relativity except as an approximation, and so the concept that the polarization is a simple function of distance, with a correction for the motion of the earth relative to a frame of reference, is attractive to me.

I (like all other humans) have a tendency to believe my own BS. Of course I’m afraid fo this, and to help counteract it, I’d love to hear you express your doubts more fully. It would give me an excuse to look more carefully at the data.

3. Carl Brannen - March 30, 2007

Ooops I screwed up the link. astro-ph/0507274 try again here.

4. dorigo - March 30, 2007

Hi Shantanu,

no, but I intend to find out… Anyhow, the conference site is the best place to check for the slides. I think I have a link to it in an earlier post – am answering this from an internet cafe in Tottenham Court Road and I do not have my usual means of handling links. I’ll post it here again tonight from home…


5. dorigo - March 30, 2007

Hi Carl,

yes the talk on the polarization correlations was intriguing. As for my doubts, I have to say most of my understanding in Cosmology so far is based on doubts. Since I do not have a solid background in the theory, I have to rely on reported effects, the source of which I only guess. I need more time before I can have an opinion of my own… Before that, it is not worth discussing them.


6. Tony Smith - March 30, 2007

Tommaso said that Hutsemekers said “… quasars … polarization vectors appear to be aligned over scales of about one giga-parsec, at redshifts of about 1: therefore their study could shed some light on the existence of large scale structures in the Universe … there not only seems to be evidence for a correlation at large distance scales, but that the quasars contributing the most to the correlation appear to line along
an axis … which is not too far from another axis of symmetry that seems to exist between the low-l multipoles of cosmic microwave radiation. …”.

In astro-ph/0302496, Max Tegmark, Angelica de Oliveira-Costa, and Andrew Hamilton said:
“… not only is the quadrupole power low,
but both it and the octopole have almost all their power perpendicular to a common axis in space, as if some process has suppressed large-scale power in the direction of this axis … roughly towards … Virgo …
A generic quadrupole has three orthogonal pairs of extrema (two maxima, two minima and two saddle points).
We see that the actual CMB quadrupole has its strongest pair of lobes, apparently coincidentally, fall near the Galactic plane.
Applying a Galaxy cut therefore removes a substantial fraction of the quadrupole power.
The saddle point is seen to be close to zero.
In other words, there is a preferred axis in space along which the observed quadrupole has almost no power.
the observed quadrupole is the sum of the cosmic quadrupole and the dynamic quadrupole due to our motion relative to the CMB rest frame …
Since this motion is accurately known from the CMB dipole measurement, the dynamic quadrupole can and should be subtracted when studying the cosmic contribution. … This is a small correction …[that]… reduces the cosmic quadrupole slightly.
although the overall octopole power is large, not suppressed like the quadrupole, it too displays the unusual property of a preferred axis along which power is suppressed. Moreover, this axis is seen to be approximately aligned with that for the quadrupole. …
In contrast,
the hexadecapole is seen to exhibit the more generic behavior we expect of an isotropic random field, with no obvious preferred axis. …
What does this all mean?
… for instance … models … involving a flat “small Universe” with a compact topology … and one of the three dimensions being relatively small (of order the Horizon size or smaller) … could have the effect of
suppressing the large-scale power in this particular spatial direction …”.

What if an axis in the direction of the Virgo cluster is an axis along which there are more galaxies, clusters of galaxies, etc ?
Consider a raisin-bread model of our expanding universe,
in which the raisins are bound things that don’t expand (galaxies, clusters of galaxies, etc)
and the bread dough is where the expansion of our universe takes place.

If there were a relative concentration of non-expanding raisins in part of the bread (call it Virgo),
if we were a bug in the bread looking toward the Virgo concentration,
would we see a lesser relative amount of expanding bread dough in that direction, and so less expansion in that direction,
so that our universe would have expanded less in the direction of the Virgo axis,
which would correspond to “… one of the three dimensions …” that is “… relatively small …”?

As to quasar polarization vectors being aligned along an axis “not too far” from the Virgo axis with an excess of matter/raisins,
maybe Tommaso’s feeling that he “… personally found the evidence shown that the effect is not due to an interstellar medium quite unconvincing …” is relevant.

Tony Smith

7. Carl Brannen - March 31, 2007

Dorigo, regarding the photon polarization thing, there is a bit of numerology I should share with you. A paper referencing the PVLAS data is astro-ph/0607294 which supposes an axion explains it. They estimate a copuling constant for photon-axion mixing in equation 9 as 3.8 x 10^-6/GeV.

To get a dimensionless number from this you can square it and divide by the Fermi coupling constant. The numerology is that if you replace the 3.8 with 4.6 you get an exact power of 3^6 = 729 in the dimensionless ratio. That is,

(4.6 x 10^-6)^2/(1.16637 x 10^-5) = 3^12

8. dorigo - March 31, 2007

Hi Carl,

beware of getting too enamoured with factors of three… To me, the “significance” of your observation above is really small – I would not even call it a coincidence.

Regardless of whether the 3.8 has an error bar large enough to cover 4.6 with some probability, it is quite far from it (20%). To be compelling to me, it would have to be at least a factor of several hundred times closer, and with an error bar smaller still. I know you did not claim it was significant, but I would keep it in the drawer for now…


9. dorigo - March 31, 2007

Hi Tony,

yes, doing cosmology these days is a tough job I reckon, since one has only one universe to study, just one particular point of view to study it from, many concurring theories equally unconvincing, and claims that that the observation point is not just random to boot.
Your biscuit model of the universe does draw home a point, and the most annoying thing is that we will probably never know…


10. Carl Brannen - March 31, 2007

Ah Dorigo, I see that I wrote poorly. It is not factors of 3, but factors of 3^6 = 729, which are much more difficult to find randomly. I.e. G_F/G_N = 3^72 = (3^6)^12 quite exactly, assuming you treat the two coupling constants the same with regard to the 4\pi factor.

The 3 comes from a 1/3 probability for ending up in any particular color, i.e. R/(R+G+B). The 6 comes from assuming the elementary particles composite with 6 components. The weakness of G_N comes from the 12, which says that gravity is a complicated force.

By the way, an example of a factor of 3 that is not a 3^6 and which I have not called significant is the Planck mass, which in GeV turns out to be 1.2209×10^19 = 3^40.0039, surely a coincidence that nature created to laugh at me.

11. Carl Brannen - April 5, 2007


Regarding the powers of 3, you might find equation (54) of the following paper interesting: gr-qc/0212096

The paper has to do with the normal modes of a black hole.

Sorry comments are closed for this entry

%d bloggers like this: