Selection Effects in the SDSS Quasar Sample: The Filter Gap Footprint

TL;DR

The paper investigates how the SDSS quasar selection process is affected by filter gaps and emission lines, creating a periodic bias in redshift distribution that impacts cosmological analyses.

Contribution

It demonstrates that filter gap effects induce a periodicity in the quasar redshift distribution, highlighting a selection bias in SDSS data.

Findings

A periodicity of Δz~0.6 is imprinted on the redshift distribution.

Filter gaps cause systematic undercounting of quasars at specific redshifts.

The effect complicates using SDSS quasars to test redshift periodicity claims.

Abstract

In the Sloan Digital Sky Survey (SDSS) quasars are targeted using colors and anything that can cause the identifying characteristics of the colors to disappear can create problems in the source selection process. Quasar spectra contain strong emission lines that can seriously affect the colors in photometric systems in which the transmission characteristics vary abruptly and significantly with redshift. When a strong line crosses a gap between two filter passbands the color effects induced by the line change abruptly, and there is also a dimming in apparent brightness compared to those redshifts where the strong line is inside a filter passband where the transmission is high. The strong emission lines in quasars, combined with the varying detectability introduced by the transmission pattern of the five filters, will result in a filter-gap footprint being imprinted on the N(z) distribution, with more quasars being missed when a strong line falls in a filter gap. It is shown here that a periodicity of Delta(z)~0.6 is imprinted on the redshift-number distribution by this selection effect. Because this effect cannot be rigorously corrected for, astronomers need to be aware of it in any investigation that uses the SDSS N(z) distribution. Its presence also means that the SDSS quasar data cannot be used either to confirm or to rule out the Delta(z)~0.6 redshift period reported previously in other, unrelated quasar data.