CosmoCoffee

Any comments on this paper please?

Another small and ignorable `glitch' for the standard
cosmological model?

I had not appreciated this paper, until Anze brought it up in another post. Another paper that has brought up the possibility of a huge void (~ 150 Mpc) is astro-ph/0612347. Does anybody know what is the probability of a void with \delta ~ -1, and this size within our horizon. Since this is probably very unlikely, what would be the required f_NL non-Gaussianity parameter that would lead to one of these in our horizon?

I think that Inoue and Silk have only 30% or so underdensity compared to these guys who require -1. I am pretty sure that any such void is completely impossible as a random fluctuation in a standard theory and that if true such (primordial?) voids must also be extremely rare if you want to avoid seeing them in the CMB and other power spectra. It would also be good to see an independent confirmation of this result.

Voids do tend to be very underdense, [tex]\delta\lesssim-0.9[/tex]. The radius of the largest observed void is around 40 Mpc (see astro-ph/0312533), which is a factor 3-4 smaller than the void suggested in the paper. I think it's true that it is unlikely to have so large so empty voids as random fluctuations in the standard picture. However, in the standard picture it is also difficult to get voids which as empty and as large as those which are indeed observed, so one should not be too confident about such statistics. Peebles has even called this a crisis of the CDM model, astro-ph/0101127.

(An easy way to get bigger and emptier voids might simply be to change the power spectrum on the relevant scales. Of course, this would also change the abundance of large overdense structures, which might not be unwanted, see astro-ph/0605393, astro-ph/0609686.)

Syksy Rasanen wrote:Voids do tend to be very underdense, [tex]\delta\lesssim-0.9[/tex]. The radius of the largest observed void is around 40 Mpc (see astro-ph/0312533), which is a factor 3-4 smaller than the void suggested in the paper.
(...)

I think there's a danger here of mixing up [tex]\delta[/tex] for galaxies and [tex]\delta[/tex] for total mass. The observations tend to report [tex]\delta[/tex] in galaxy counts or something similar; to get the same [tex]\delta[/tex] in total mass you have to assume that the bias is the same in voids as it is elsewhere. As far as I know (though I freely admit I've not kept up on this topic), simulations suggest that this isn't true; see for example Ostriker et al. astro-ph/0305203. So you might not even need to change the power spectrum. Maybe better simulations will make the whole "void problem" go away entirely?

I have predicted the 'large' void at z~1 in the direction to the 3-sigma cold spot in astro-ph/0602478 before this paper. The chance of having such a large void should be very small but the 'volume effect' and the 'non-linear gravitational effect' could somehow enhance the chance of having such voids(here a 'void' means a low density region surrounded by a spherically symmetric non-linear wall). For linear fluctuations, a void with 30% underdense region with 200h^{-1}Mpc radius is a 11-13 sigma object at z=0(astro-ph/0612347). A spiky power spectrum could make such a structure without having any 'troubles' with other observations.

I'd like to point out that Laura Mersini-Houghton, Tomo Takahashi and I
predicted the existence of a void at [tex]z\leq 1[/tex] of size roughly 200 Mpc, in a couple of papers (arXiv: hep-th/0611223, hep-th/0612142) which studied the astrophysical signatures of our proposal for a dynamical selection of inflationary initial conditions. This involved considerations of both the backreaction of massive modes of the inflaton, as well as non-local entanglement of our inflating patch and other patches.

The net effect of all this, from the point of large scale
structure, is to add a negative, scale dependent, contribution to
the Newtonian potential [tex]\Phi[/tex], which in turn gives rise to a negative density contrast superimposed upon the positive density perturbations inside the Hubble radius, resulting in voids at scales roughly present Hubble scale [tex]z\leq 1[/tex] and size of ~140-200 Mpc today.

This appears to be what the authors
of astro-ph/07004.0908 observed. (I think there were also observations by WMAP and SDSS previously). These
observations seem to be in agreement with our theoretical prediction of the effects of nonlocal entanglements between inflationary patches. If true, then the cold spot discovery would be a very exciting test of such quantum gravitational effects and may provide the first indirect tests for mechanisms for the selection of initial conditions and open a new window of physics beyond the horizon.

Among other tests, our model can be tested independently by considering
correlations between cosmic shear and temperature anisotropies. It would be interesting to see whether our results are consistent with the work quoted by Inoue.

It seems to me that addition of a negative scale dependent contribution to the Newtonian potential &#934;, leads to a suppression of overall fluctuations on large-angle scales (larger than acoustic horizon scales at LSS). Therefore, it leads to a suppression of generating voids on 140-200Mpc at z<1. Any comments?

Maybe not many people are aware of this, but if such voids exists (radius [tex]200/h[/tex] Mpc and [tex]\delta\approx -0.4[/tex]) and if we happen to live near the center (with a precision of about 10Mpc), this would give a quite good fit of cosmological observations (supernova Hubble diagram and other things) without the need for Dark Energy.

We have tried to fit the supernovas in astro-ph/0606703 (although in that paper there is a slightly different geometry, the result is almost the same for spherical voids). The [tex]chi^2[/tex] is worst than [tex]Lambda[/tex]CDM, but still reasonable, and it can be made slightly better than that playing with the density profile.