The standard model of blazars presumes that they are active galaxies with a jet closely aligned to the observer’s line of sight. Observationally, blazars can be distinguished from other active galaxies with a variety of diagnostics, including extreme luminosities and non-thermal spectra dominated by relativistic beaming.
The population of blazars displays significant observational diversity, however. The primary historical method of classification has been through the optical spectrum of the blazar, dividing them into two groups: (1) BL Lacs, characterized by weak or non-existent optical emission lines, and (2) flat-spectrum radio quasars (FSRQs), which have strong, broad emission lines in their optical spectrum. An enduring question for many years has been whether BL Lacs and FSRQs comprise physically distinct populations of galaxies, or whether a sequence exists between the two groups.
Since highly-beamed emission from the relativistic jet often dominates the light observed from the blazar, studies of the host galaxy itself are challenging. As an alternative approach, we study the clustering properties of blazars in an attempt to determine their parent populations. Clustering studies benefit from the fact that they are assumed to be independent of the blazar’s orientation to our line-of-sight. Differences in the clustering properties on scales of hundreds of kpc are already known to exist, for example, in the density-morphology relation \citep{dre80} and in the populations of powerful radio galaxies \citep{pre88}.
Previous clustering studies of blazars have been limited by sample sizes of only a few tens. \citet{wur97} carried out a deep, largely subarcsec imaging survey of BL Lac objects conducted at the CFHT. \citet{wur93} described the results pertaining to the host galaxies of 50 BL Lac objects at \(z<0.65\); \citet{wur97} report on the clustering environment of 45 of these 50 BL Lacs. The remaining five objects either have unknown or very uncertain redshifts or we were unable to obtain deep, photometrically calibrated images of them, which prohibited successful clustering analysis.
With this substantially larger (45 vs. 5) sample, \citet{wur97} confirmed the early result of \citet{pre88} that BL Lac objects largely avoid rich clusters at low redshift and have distributions in \(B\) richness measurements much more consistent with those of FR II radio galaxies than with FR I’s. The typical environment of a BL Lac object is a sub-Abell richness class 0 cluster with a CFHT sample mean of \(\langle\)\(B_{gB}\)\(\rangle=209\) Mpc\(^{1.77}\). Further, these results apply to all types of BL Lac objects regardless of selection method (e.g., radio or X-ray selection) or detailed property (e.g., high or low optical polarization percentage, presence/absence of emission lines etc.) because no BL Lac subtype has statistically distinct \(B_{gB}\) values. The only exceptions to this statement are correlations between redshift and \(B_{gB}\) and between host galaxy luminosity and \(B_{gB}\), and a possible anticorrelation between \(B_{gB}\) and radio core dominance.
\citet{smi95} computed \(B_{gB}\) for 16 BL Lac galaxies and six FR I galaxies; their respective correlation amplitudes were statistically indistinguishable for the small sample sizes, consistent with an Abell richness class of 0.
Mention blazar envelope/sequence work, and possibility of how environment studies play into the larger scientific goal.
With the release of the Sloan Digital Sky Survey (SDSS), we have for the first time a large and deep catalog with nearly uniform photometry and coverage out to redshifts greater than 1. We combine the SDSS data with the latest identifications of blazars, for which counts are now in the thousands. The larger sample size allows for the first time robust classification of blazar clustering properties, which we compare to other data to explore the possible parent population(s).