We investigate the extent to which the Palomar-Green (PG) Bright Quasar Survey (BQS) is complete and representative of the general quasar population by comparing with imaging and spectroscopy from the Sloan Digital Sky Survey. A comparison of SDSS and PG photometry of both stars and quasars reveals the need to apply a color and magnitude recalibration to the PG data. Using the SDSS photometric catalog, we define the PGs parent sample of objects that are not main-sequence stars and simulate the selection of objects from this parent sample using the PG photometric criteria and errors. This simulation shows that the effective U-B cut in the PG survey is U-B < -0.71 (rather than the intended U-B < -0.44), implying a color-related incompleteness. As the color distribution of bright quasars peaks near U-B=-0.7 and the 2-sigma error in U-B is comparable to the full width of the color distribution of quasars, the color incompleteness of the BQS is approximately 50% and essentially random with respect to U-B color for z<0.5. There is, however, a bias against bright quasars at 0.5 < z < 1, which is induced by the color-redshift relation of quasars (although quasars at z>0.5 are inherently rare in bright surveys in any case). We find no evidence for any other systematic incompleteness when comparing the distributions in color, redshift, and FIRST radio properties of the BQS and a BQS-like subsample of the SDSS quasar sample. However, the application of a bright magnitude limit biases the BQS toward the inclusion of objects which are blue in g-i, in particular compared to the full range of g-i colors found among the i-band limited SDSS quasars, and even at i-band magnitudes comparable to those of the BQS objects.
Download