Australia Telescope National Facility

QSO-galaxy correlations: lensing or dust?
Scott M. Croom, PASA, 18 (2), in press.

Next Section: References
Title/Abstract Page: QSO-galaxy correlations: lensing or
Previous Section: Theoretical expectation
Contents Page: Volume 18, Number 2

QSO-galaxy cross-correlations

The cross-correlation function between QSOs with z>1 and galaxies with B<23 is shown in Fig. 2a (the samples were chosen for minimal redshift overlap). The dotted line is the amplitude of the integral constraint correction due to our normalization of the density by the number of galaxies in the field. This is derived assuming

$\omega(\theta)=A\theta^{-0.8}$ , from which we find the integral constraint to be $(6.5\pm0.1)A$ . When fitting the data we therefore fit

$\omega(\theta)=A(\theta^{-0.8}-6.5)$ . The best fit model has

$A=0.0028\pm0.0006$ . Using Eq. 1 this then implies that

$(2.5\alpha-1)/b_{\rm g}=-7.8\pm0.7$ (EdS) or $-18.3\pm1.6$ ( $\Lambda$ ). For realistic values of $\alpha$ ( $\sim0.2$ for the QSO samples used) this implies a bias $b_{\rm g}<0.1$ , an order of magnitude or more smaller than expected in standard models.

If we consider a dust interpretation, then we see that $A_{\rm B}$ must be a function of $\theta$ , but on scales $\sim1-5'$ we find

$\omega_{\rm qg}(\theta)\simeq-0.06$ . This would imply

$A_{\rm B}\simeq0.13$ ( $\alpha=0.2$ ), and therefore

$E(B-V)\simeq0.02$ . This reddening is within the upper limit of E(B-V)<0.06 (90%) found by Ferguson (1993) in clusters and groups. There appears to be a significant anti-correlation out to $\simeq8'$ , which corresponds to a physical scale of

$\sim1.2-1.3~h^{-1}~{\rm Mpc}$ (depending on the cosmological model used) at the median redshift of the B<23 galaxies,

$z_{\rm med}\simeq0.25$ . This is a scale typical of galaxy clusters, although currently there are no direct detections of intra-cluster dust distributed on these scales. However, there is tentative evidence of dust in the central parts of the Coma cluster (within

$\sim0.1~h^{-1}~{\rm Mpc}$ ) from ISO observations, with a reddening of

$A_{\rm V}\sim0.01-0.26$ mag inferred (Stickel et al. 1998).

At fainter flux limits (B<26), we see that the anti-correlation disappears (see Fig. 2b). First, this is good evidence that the B<23 anti-correlation is not caused by any systematic errors in the galaxy catalogue. It appears that the whatever the source of the anti-correlation with the brighter galaxies, it is compensated for by those fainter than B=23. A randomly distributed population at 23<B<26 is sufficient to remove the anti-correlation, due to the larger number of faint galaxies. As the redshift distribution of the B<26 galaxies and the QSOs is likely to be similar (Fig. 1), both the lensing and dust effects will be reduced. We will present detailed models describing this effect in Croom & Shanks 2001.

A much clearer conclusion will be available soon with large, homogeneous QSO surveys such as the 2dF QSO Redshift Survey (Croom et al. 2001) and the Sloan Digital Sky Survey (York et al. 2000) which will be analysed as a function of luminosity, as the break in the QSO number count is a good probe of the physics behind these correlations. These new large data sets will have sufficient numbers of QSOs to show whether the correlation changes sign as it goes past the break, as expected in lensing, or only changes in amplitude, as in a simple dust model.

Next Section: References
Title/Abstract Page: QSO-galaxy correlations: lensing or
Previous Section: Theoretical expectation
Contents Page: Volume 18, Number 2

ASKAP

Public