Faint-end Quasar Luminosity Functions from Cosmological Hydrodynamic Simulations

TL;DR

This study uses cosmological hydrodynamic simulations to predict the faint-end quasar luminosity function and its evolution, showing good agreement with observations at lower redshifts and highlighting the complex black hole activity contributing to faint quasars.

Contribution

First simulation-based analysis of faint-end quasar luminosity functions incorporating detailed black hole growth and feedback processes, aligning well with observed data at z<2.

Findings

Good agreement with observed faint-end optical and X-ray QLFs at z<2

Overestimation of faint-end QLF at higher redshifts

Faint sources at high redshift likely affected by resolution effects

Abstract

We investigate the predictions for the faint-end quasar luminosity function (QLF) and its evolution using fully cosmological hydrodynamic simulations which self-consistently follow star formation, black hole growth and associated feedback processes. We find remarkably good agreement between predicted and observed faint end of the optical and X-ray QLFs (the bright end is not accessible in our simulated volumes) at z < 2. At higher redshifts our simulations tend to overestimate the QLF at the faintest luminosities. We show that although the low (high) luminosity ranges of the faint-end QLF are dominated by low (high) mass black holes, a wide range of black hole masses still contributes to any given luminosity range. This is consistent with the complex lightcurves of black holes resulting from the detailed hydrodynamics followed in the simulations. Consistent with the results on the QLFs, we find good agreement for the evolution of the comoving number density (in optical, soft and hard X-ray bands) of AGN for luminosities above 10^43 erg/s. However, the luminosity density evolution from the simulation appears to imply a peak at higher redshift than constrained from hard X-ray data (but not in optical). Our predicted excess at the faintest fluxes at z >= 2 does not lead to an overestimate to the total X-ray background and its contribution is at most a factor of two larger than the unresolved fraction of the 2-8 keV background. Even though this could be explained by some yet undetected, perhaps heavily obscured faint quasar population, we show that our predictions for the faint sources at high redshifts (which are dominated by the low mass black holes) in the simulations are likely affected by resolution effects.