The evolution of optically selected QSOs – II

We present a new determination of the QSO luminosity function and its evolution based primarily on a large, complete catalogue of 420 faint (B<20.9 mag) ultraviolet excess selected QSOs, for which spectroscopic identification and redshifts have been obtained using the fibreoptic system at the Anglo–Australian Telescope. We confirm that, at faint magnitudes (B>20 mag), the number–magnitude relation for QSOs exhibits a departure from the steep power-law slope seen at brighter (B<19.5 mag) magnitudes, with the surface density of QSOs at B<20.9 mag being equal to 37±3 deg2. Using maximum-likelihood techniques we find that luminosity evolution models of the form |L(z)∝ L₀(1+z)3.2±0.1| adequately describe the evolution of bright (MB<−23) low redshift (z<2.2) QSOs. The QSO luminosity function, Φ(L), is most simply parameterized by two power laws, with |Φ(L)∝ L-3.7±0.1| at high luminosities and |Φ(L)∝ L-1.4±0.2| at low luminosities. There may be some evidence that additional density evolution is required to fit the faintest (MB>−23) QSOs in the survey. However, incompleteness in the original catalogue at these magnitudes may be responsible for this effect.