Lensing Probabilities for Spectroscopically Selected Galaxy-Galaxy Strong Lenses

TL;DR

This paper models the probabilities of detecting spectroscopic galaxy-galaxy strong lenses, accounting for selection effects like fiber size, and estimates the lensing rate, revealing higher probabilities than previous surveys due to different selection criteria.

Contribution

It introduces a detailed probabilistic framework for spectroscopic lens searches that incorporates unique selection effects and provides updated lensing rate estimates.

Findings

Total rogue line detection probability: 0.9-3.0%

Lensing rate estimate: 0.5-1.3%

Discrepancy with SLACS due to selection criteria

Abstract

Spectroscopic galaxy-galaxy lens searches are presently the most prolific method of identifying strong lens systems in large data sets. We study the probabilities associated with these lens searches, namely the probability of identifying a candidate with rogue [OII] emission lines in a galaxy's spectrum, and the probability that the candidate will show features of strong lensing in follow-up photometric observations. We include selection effects unique to spectroscopic data, and apply them to the Sloan Lens ACS (SLACS) survey (Bolton et al. 2006). The most significant selection effect is the finite size of the spectroscopic fiber which selects against large separation lenses and results in a non-monotonic dependence of the rogue line probability on velocity dispersion. For example, with the 3 arcsec diameter SDSS fiber and 2 arcsec FWHM seeing, we find that the probability that a given LRG has a rogue [OII] line in its spectrum decreases with velocity dispersion from 150 km/s to 300 km/s and then increases up to 400 km/s for a given source size. The total probability for observing a rogue line in a single survey spectrum is ~0.9-3.0%, and the total lensing rate is ~0.5-1.3%. The range is due to uncertainties in the physical size of [OII] emission regions, and in the evolution of the [OII] luminosity function. Our estimates are a factor of ~5 higher than the results of the SLACS survey, a discrepancy which we attribute to the SLACS requirement that multiple rogue lines be observed simultaneously.