Observational Nonstationarity of AGN Variability: The Only Way to Go Is Down!

Abstract

To gain insights into long-term variability of Active Galactic Nuclei (AGN), we analyze an AGN sample from the Sloan Digital Sky Survey (SDSS) and compare their photometry with observations from the Hyper Suprime-Cam survey (HSC) observed 14.85 years after SDSS. On average, the AGN are fainter in HSC than SDSS. We demonstrate that the difference is not due to subtle differences in the SDSS versus HSC filters or photometry. The decrease in mean brightness is redshift dependent, consistent with expectations for a change that is a function of the rest-frame time separation between observations. At a given redshift, the mean decrease in brightness is stronger for more luminous AGN and for objects with longer time separation between measurements. We demonstrate that the dependence on redshift and luminosity of measured mean brightness decrease is consistent with simple models of Eddington ratio variability in AGN on long (Myr, Gyr) timescales. We show how our results can be used to constrain the variability and demographic properties of AGN populations.

A Sample Bias in Quasar Variability Studies

“…Both predictions based on simulated data are similar to observed trends (Voevodkin 2011;Caplar et al 2020;Tachibana et al 2020). The selection of a flux-limited quasar sample at certain epoch inevitably incorporates an asymmetry in variability, since there are more temporarily brightened quasars above the flux limit than those falling below.…”

“…While the asymmetry in the numbers of dimmed and brightened quasars is prominent, the mean magnitude change of the whole sample is mild, since most of the time pairs have small differences in magnitude. The mean magnitude differences after 10 years in the observed frame for our simulated sample of i t0 < 19 is ∼ 0.1 mag fainter, roughly matching the observed mean magnitude change in Caplar et al (2020). This mean magnitude bias is stronger at lower redshifts (at fixed flux limit), since the RMS variability amplitude is higher (Fig.…”

“…The quasar sample defined earlier on is typically flux limited, and contains more fainter (in terms of the mean luminosity, since quasars are variable) quasars scattered above the flux limit than those scattered below, which would eventually scatter down given sufficient time 1 . Therefore, statistically the sample will have more dimmed quasars than brightened ones when observed at later epochs, as confirmed in observations (e.g., Rumbaugh et al 2018;Caplar et al 2020). This asymmetry in quasar brightness changes at later times was interpreted as a selection bias and modeled with simulated light curves (e.g., Rumbaugh et al 2018;Luo et al 2020).…”

“…While our results do not necessarily imply that the DRW model is correct, they provide a cautionary note on recent claims that there is a time asymmetry in quasar light curves based on bright, (approximately) flux-limited samples. The manifestation of this bias, in terms of the mean magnitude changes and the asymmetry in the SF, is generally small and only detectable with large statistical quasar samples as those studied in, e.g., Caplar et al (2020) and Tachibana et al (2020).…”

“…, where SF ∞ is the asymptotic structure function on very long timescales. Given a typical bright-end slope of the LF γ = 0.8 and σ lc = 0.2, we have A ∼ −0.1 mag, which is a small effect to measure with a large sample of quasars (Caplar et al 2020), and an overabundance of dimmed quasars of N dimmed /N brightened ≈ 1.8, 3.5 for |∆m| = 0, 0.25. The comparisons above are between the epoch of sample selection and a different epoch.…”