Friday, 24 April 2009
This issue has been visited several times in the past, and now most recently by Lin, Ostriker, & Miller using clusters detected in SDSS. Their basic method is to create mock clusters by scrambling the galaxies among the observed clusters, and then to compare the luminosities of the brightest galaxies in the mock clusters to the luminosities of the actual BCGs. The purple circles in the top panel of this figure show the observed BCG luminosities as a function of total cluster luminosity. The squares show the BCG luminosities using a running mean, and the green crosses show the running mean computed from the mock clusters. It appears that BCGs are more luminous than expected based on the mock clusters, a difference that becomes more apparent at high luminosities.
This conclusion is also apparent in the bottom panel, which shows the difference between the observed BCG luminosities and the the expected luminosity (filled purple symbols), compared to the difference observed in one realization of the mock cluster sample (open red).
The authors conclude that, in a flux limited sample of LRGs, BCGs will become more dominant at high redshifts. So baryonic accoustic oscillation studies will have to take this into account.
In arXiv:0904.3098 Lin, Ostriker & Miller use a cluster catalog, assemblked from SDSS DR5 to examine the question whether or not BCGs are just the statistical extremes from the luminosity distributions of cluster galaxies, or whether they are really distinct objects. They do a very simple test: they throw together all cluster galaxies, and randomly sample luminosities from this distribution, adding up to the total luminosities of clusters. They then compare the statistic d in this plot. d is the difference in log between the mean luminosity of observed(blue)/sampled(red) BCGs and of 200 MC realizationsof this process. The red line therefore should be centered on zero, and it is. The offset of the blue histogram is significant in the total sample (494 clusters) and in the sample of bright clusters (124 clusters). FOr the total sample they calculate a probablity of less than 0.8% that the distributions come from the same parent distribution, for the bright subsample it is 0.03%. For low luminosity clusters there is no significant deviation between the two (P = 54.5%). They therefore claim that (at least for the bright ones) BCGs must evolve distinct from the rest of the galaxy population, making them brighter than their non-BCG cluster companions.
Friday, 17 April 2009
Interestingly, the authors state that two of the BCGs may be undergoing major mergers, which will make them even more massive.
One worry I have is that this analysis relies on matching the high-redshift cluster sample to an appropriate local sample. The authors don't discuss this matching in detail. If the local sample was selected to have the same total cluster mass (which I think is the case), then all that the authors have shown is that the ratio of BCG stellar mass to total cluster mass doesn't evolve with redshift. But, as the authors note (although in a different context), previous studies have already arrived at this conclusion out to z~0.8.
There are other uncertainties in this analysis. The authors use a crude method to estimate the stellar masses (they seem to use only one or two observed bands). Additionally, it is well-known that the total mass of extended low-redshift galaxies is difficult to measure (the authors don't address this issue).
Thursday, 9 April 2009
The Balloon-borne Large-aperture Submillimeter Telescope (BLAST) released a slew of papers on astro-ph today. This instrument has surveyed 8.7 deg2 in GOODS and ECDFS at 250, 350, and 500 microns. The authors also incorporate IRAC and MIPS data from SIMPLE and FIDEL, and select sources based on 24um flux. The additional wavelength coverage from BLAST allows the authors to better constrain L_IR as well as directly constrain the average dust temperatures of sources as a function of redshift. It turns out that the cosmic IR background can be almost (or entirely) resolved into individual sources, with most of the 70um background coming from z<1 galaxies and the 500um background coming from z>1.
The above plot shows the inferred star formation density evolution from this study (black circles), UV/optical measurements (triangles; dashed error bars have extinction corrections applied). About 70% of their 24um sources had reliable UV/NIR photometric redshifts, and the remaining 30% were estimated through IRAC SED fitting; the grey circles show the SFH with these "IRAC redshifts" excluded. The solid and dashed lines show a luminosity function model with (dashed) and without (solid) taking into account the 24um flux limit of 20 microJansky. As the BLAST measurements fall below the model at thehighest redshifts, the authors conclude that they're missing a population of faint 24um sources.
Thursday, 2 April 2009
Figure 1 from "The M-sigma and M-L Relations in Galactic Bulges and Determinations of their Intrinsic Scatter" by Gueltekin et al. (http://arxiv.org/abs/0903.4897v1):
Caption (Abridged): "The M–sigma relation for galaxies with dynamical measurements. The symbol indicates the method of BH mass measurement: stellar dynamical (pentagrams), gas dynamical (circles), masers (asterisks). Arrows indicate 3sigma_68 upper limits to BH mass. The color of the error ellipse indicates the Hubble type of the host galaxy: elliptical (red), S0 (green), and spiral (blue). The saturation of the colors in the error ellipses or boxes is inversely proportional to the area of the ellipse or box. Squares are galaxies that we do not include in our ﬁt. The line is the best ﬁt relation to the full sample: MBH = 108.12 Msun (sigma/200 km s−1 )4.24. The mass uncertainty for NGC 4258 has been plotted much larger than its actual value so that it will show on this plot."
Interesting Plot showing the latest and most reasonable collection of galaxies with dynamical BH mass estimates, along with a statistically sound investigation of the Msigma relation and its intrinsic scatter. They find a scatter of (log-normal) e_0=0.44±0.06 (0.31±0.06 for ellipticals only), which could be physical or systematic uncertainties in the observations.