Composite Spectral Energy Distributions and Infrared–Optical Colors of Type 1 and Type 2 Quasars

Abstract

We present observed mid-infrared and optical colors and composite spectral energy distributions (SEDs) of type 1 (broad-line) and 2 (narrow-line) quasars selected from Sloan Digital Sky Survey (SDSS) spectroscopy. A significant fraction of powerful quasars are obscured by dust and are difficult to detect in optical photometric or spectroscopic surveys. However, these may be more easily identified on the basis of mid-infrared (MIR) colors and SEDs. Using samples of SDSS type 1 and 2 matched in redshift and [O III] luminosity, we produce composite restframe 0.2-15 μm SEDs based on SDSS, UKIDSS, and Wide-field Infrared Survey Explorer photometry and perform model fits using simple galaxy and quasar SED templates. The SEDs of type 1 and 2 quasars are remarkably similar, with the differences explained primarily by the extinction of the quasar component in the type 2 systems. For both types of quasar, the flux of the active galactic nucleus (AGN) relative to the host galaxy increases with AGN luminosity (L O III [ ] ) and redder observed MIR color, but we find only weak dependencies of the composite SEDs on mechanical jet power as determined through radio luminosity. We conclude that luminous quasars can be effectively selected using simple MIR color criteria similar to those identified previously (W1 W2 0.7; > -Vega), although these criteria miss many heavily obscured objects. Obscured quasars can be further identified based on optical-IR colors (for example, u W3 AB). These results illustrate the power of large statistical studies of obscured quasars selected on the basis of MIR and optical photometry.

Physical Models for the Clustering of Obscured and Unobscured Quasars

“…Hickox et al (2017) showed that SED models are able to predict optical/mid-IR colors for obscured and unobscured quasars that are consistent with observations. The color cut used in DiPompeo et al (2017a) corresponds to the output of Hickox et al (2017) SED model that assumed A V = 20. Based on equation (3) in Draine (2003), this gives N H ∼ 3.7 × 10 22 cm −2 , which is consistent with our adopted definition of quasar obscuration.…”

“…Here, the black hole's change in mass determines the quasar's evolution from obscuration 1 We note that DiPompeo et al (2017a) defined quasar obscuration using an optical/mid-IR color cut of r − W 2 = 6 (Vega) (e.g., Hickox et al 2007Hickox et al , 2017 This cut takes advantage of the fact that obscured and unobscured quasars occupy different parts of SDSS/WISE color space (e.g., Hickox et al 2007). Hickox et al (2017) showed that SED models are able to predict optical/mid-IR colors for obscured and unobscured quasars that are consistent with observations. The color cut used in DiPompeo et al (2017a) corresponds to the output of Hickox et al (2017) SED model that assumed A V = 20.…”

The XXL Survey

“…In other words, the optically red source selection does not correspond to the same level of obscuring column densities as those seen for X-ray wavelengths. Hickox et al (2017) proposed an additional criterion for the selection of obscured AGNs based on the combination of the u and WISE W3 band. Since there is not a unique definition of how obscured an AGN must be to be classified as absorbed, the various selection criteria target different ranges of obscuration.…”

WISE mid-infrared properties of compact active galactic nuclei selected from the high radio frequency AT20G survey

“…Other authors (Jarrett et al 2011;Mateos et al 2012) have defined wedges in the WISE colour-colour plot (such as Figure 7) to idenfity AGNs. Hickox et al (2017) conclude that luminous quasars can be effectively selected using a simple criteria such as those identified by Stern et al (2012), although they warn that such criteria may miss heavily obscured quasars. In Figure 8, we plot the histogram of W1-W2 colours of all sources with redshift in our sample.…”