Comparing Simple Quasar Demographics Models

TL;DR

This study compares various simple models linking quasars, galaxies, and dark matter halos across redshifts 1 to 6, focusing on black hole growth and variability to fit observed quasar luminosity functions.

Contribution

It introduces a set of simple, self-consistent quasar models with different assumptions, fitting observed data and analyzing degeneracies among models.

Findings

All models successfully fit the observed quasar luminosity function.

Model shape is influenced by intrinsic luminosity at faint end and variability at bright end.

Degeneracies among models highlight the need for additional observational constraints.

Abstract

This paper explores several simple model variations for the connections among quasars, galaxies, and dark matter halos for redshifts 1 < z < 6. A key component of these models is that we enforce a self-consistent black hole (BH) history by tracking both BH mass and BH growth rate at all redshifts. We connect objects across redshift with a simple constant-number-density procedure, and choose a fiducial model with a relationship between BH and galaxy growth rates that is linear and evolves in a simple way with redshift. Within this fiducial model, we find the quasar luminosity function (QLF) by calculating an "intrinsic" luminosity based on either the BH mass or BH growth rate, and then choosing a model of quasar variability with either a lognormal or truncated power-law distribution of instantaneous luminosities. This gives four model variations, which we fit to the observed QLF at each redshift. With the best-fit models in hand, we undertake a detailed comparison of the four fiducial models, and explore changes to our fiducial model of the BH-galaxy relationship. Each model variation can successfully fit the observed QLF, the shape of which is generally set by the "intrinsic" luminosity at the faint end and by the scatter due to variability at the bright end. We focus on accounting for the reasons that physically different models can make such similar predictions, and on identifying what observational data or physical arguments are most essential in breaking the degeneracies among models.