Feedback effect on the observable properties of $z>6$ AGN

TL;DR

This study explores how different kinetic feedback models influence the observable characteristics of $z>6$ AGN and their host galaxies, emphasizing the importance of feedback geometry and seeding prescriptions.

Contribution

It demonstrates the impact of kinetic feedback on ISM properties and highlights the significance of outflow geometry and seeding methods in cosmological simulations of high-redshift AGN.

Findings

Kinetic feedback reduces ISM column density by up to a factor of 10.

It is necessary for matching observed gas reservoir sizes in $z>6$ AGN hosts.

Biconical outflow geometry better reproduces observed AGN properties but overestimates multiple AGN systems.

Abstract

Active galactic nuclei (AGN) feedback has a major impact onto the supermassive black-hole (SMBH) growth, the properties of the host galaxies, and their cosmic evolution. We investigate the effects of different kinetic feedback prescriptions on the observable properties of AGN and their host galaxies at $z>6$ in a suite of zoom-in cosmological simulations. We find that kinetic feedback decreases the column density of the interstellar medium (ISM) in the host galaxy by up to a factor of $\approx10$, especially when the SMBHs reach high accretion rates ($\approx10-30\,\mathrm{M_\odot\,yr^{-1}}$). In particular, kinetic feedback is required to extend the ISM size to $>1$ kpc and match the observed sizes of the gas reservoirs in $z>6$ AGN host galaxies. Moreover, it produces unobscured lines of sight along which the AGN can be detected in the rest-frame UV band with magnitudes consistent with observed values of $z>6$ AGN. The assumed geometry of the outflow plays an important role in shaping the observed properties of high-redshift AGN. We find that a biconical geometry is favored over a spherical one to reproduce the observed properties, but it overestimates the number of multiple AGN systems detectable in X-ray observations. This result suggests that simplistic BH seeding recipes widely employed in cosmological simulations produce too many X-ray detectable multiple AGN at $z=6-7$, thus soliciting the adoption of more physically motivated seeding prescriptions.