2013ApJS..204....6F

We present the results of a search for red QSOs using a selection based on optical imaging from the Sloan Digital Sky Survey (SDSS) and near-infrared imaging from UKIDSS. Our main goal with the selection is to search for QSOs reddened by foreground dusty absorber galaxies. For a sample of 58 candidates (including 20 objects fulfilling our selection criteria that already have spectra in the SDSS), 46 (79%) are confirmed to be QSOs. The QSOs are predominantly dust-reddened except for a handful at redshifts z ≳ 3.5. However, the dust is most likely located in the QSO host galaxies (and for two, the reddening is primarily caused by Galactic dust) rather than in the intervening absorbers. More than half of the QSOs show evidence of associated absorption (BAL absorption). Four (7%) of the candidates turned out to be late-type stars, and another four (7%) are compact galaxies. We could not identify the remaining four objects. In terms of their optical spectra, these QSOs are similar to the QSOs selected in the FIRST-2MASS Red Quasar Survey except they are on average fainter, more distant, and only two are detected in the FIRST survey. As per the usual procedure, we estimate the amount of extinction using the SDSS QSO template reddened by Small-Magellanic-Cloud-(SMC) like dust. It is possible to get a good match to the observed (rest-frame ultraviolet) spectra, but it is not possible to match the observed near-IR photometry from UKIDSS for nearly all the reddened QSOs. The most likely reasons are that the SDSS QSO template is too red at optical wavelengths due to contaminating host galaxy light and because the assumed SMC extinction curve is too shallow. Three of the compact galaxies display old stellar populations with ages of several Gyr and masses of about 10¹⁰ M_☉(based on spectral energy distribution modeling). The inferred stellar densities in these galaxies exceed 10¹⁰ M_☉/kpc2, which is among the highest measured for early-type galaxies. Our survey has demonstrated that selection of QSOs based on near-IR photometry is an efficient way to select QSOs, including reddened QSOs, with only small contamination from late-type stars and compact galaxies. This will be useful with ongoing and future wide-field near-IR surveys such as the VISTA and EUCLID surveys.