Page 1 of 1

[1009.3884] Too big, too early? Multiple High-Redshift Gala

Posted: September 22 2010
by Syksy Rasanen
The authors consider 15 massive clusters observed at redshifts >1. In the vanilla [tex]\Lambda[/tex]CDM model, you don't expect a lot of massive clusters early on. The authors report a probability of [tex]3\times10^{-4}[/tex] for having so many so massive clusters in the observed region.

The authors focus on the influence of non-Gaussianity: the probability of massive clusters rises with [tex]f_{NL}[/tex], and the authors find that for [tex]f_{NL}=550[/tex] the probability rises to one. Personally, I am not that much into in large non-Gaussianity, but the observation of the discrepancy is interesting. I would imagine that for example slower expansion at high redshifts would increase the number of clusters; one could check whether the required change is consistent with the supernova data.

As a caveat, the authors note that the observations would be consistent with the standard picture with no non-Gaussianity if the cluster masses are over-estimated by 1.5[tex]\sigma[/tex], which doesn't seem like a lot to me. I am not too familiar with this, but I recall there has been controversy over the cluster abundance and the X-ray mass determinations before ( ; see also

[1009.3884] Too big, too early? Multiple High-Redshift Gala

Posted: September 23 2010
by Florian Pacaud
Hi Syksy,

although the idea behind this article has received a growing interest in the past months/year, I would be careful on the interpretation of these results. Cluster mass measurements at z>1 are extremely difficult and, as you say, a 1.5[tex]\sigma[/tex] deviation is not at all unexpected.

In addition, the mass definition itself needs some carefull handling for these kind of studies, and I am not sure they really did it properly. More exactly, they dont give us the necessary information to be confident that it was done correctly.

If you look at the caption of table 1, they mention the "mass (converted to [tex]M_{200}[/tex])" without ever saying how this conversion was performed. This conversion is nothing but straightforward since you need to assume a cluster mass profile and be very careful that you properly understand what convention people have used in the past articles.

All these cluster masses were obtained from spherical mass overdendity considerations, while the Jenkins mass function that they use in their paper uses Friend-of-Friend halo masses. Although many people don't care much about that, I think the translation from one to the other is also not straightforward (and the reason why the galaxy cluster community likes to use the recent mass functions from Tinker et al. 2008).

A frequent assumption is that FOF halo finder give you a mass that is roughly M200. Even neglecting the possible inaccuracy of this assumption, one has to be careful that the overdensity of 200 here is with respect to the mean matter density, while cluster papers often give results with respect to the critical density.

All in all, I'm not saying that the paper is not good and the results may well be valid and rigorous. But I am just desapointed that many important caveats of the analysis are not even mentionned.