First Measurement of the Clustering Evolution of Photometrically Classified Quasars

Abstract

We present new measurements of the quasar angular autocorrelation function from a sample of ~80,000 photometrically classified quasarstaken from the First Data Release of the Sloan Digital Sky Survey. Wefind a best-fit model ofomega(theta)=(0.066 +0.026 -0.024)theta -(0.98+/-0.15) for the angular correlation function, consistent with estimates of theslope from spectroscopic quasar surveys. We show that only models withlittle or no evolution in the clustering of quasars in comovingcoordinates since a median redshift of z~1.4 can recover a scale lengthconsistent with local galaxies and active galactic nuclei (AGNs). Amodel with little evolution of quasar clustering in comoving coordinatesis best explained in the current cosmological paradigm by rapidevolution in quasar bias. We show that quasar biasing must have changedfrom b Q ~3 at a (photometric) redshift ofz� phot =2.2 to b Q ~1.2-1.3 byz� phot =0.75. Such a rapid increase with redshift inbiasing implies that quasars at z~2 cannot be the progenitors of modernL* objects; rather they must now reside in dense environments, such asclusters. Similarly, the duration of the UVX (ultraviolet-excess) quasarphase must be short enough to explain why local UVX quasars reside inessentially unbiased structures. Our estimates of b Q are ingood agreement with recent spectroscopic results (Croom et al. 2005),which demonstrate that the implied evolution in b Q isconsistent with quasars inhabiting halos of similar mass at everyredshift. Treating quasar clustering as a bivariate function of bothredshift and luminosity, we find no evidence for luminosity dependencein quasar clustering, and that redshift evolution thus affects quasarclustering more than changes in quasars' luminosity. Our results arerobust against a range of systematic uncertainties. We provide a newmethod for quantifying stellar contamination in photometricallyclassified quasar catalogs via the correlation function.