Winds versus jets: a comparison between black hole feedback modes in simulations of idealized galaxy groups and clusters

TL;DR

This study compares the effectiveness of thermal isotropic and kinetic jet AGN feedback modes in galaxy simulations, revealing that jets are more efficient at quenching star formation and cooling cores, with implications for modeling feedback in cosmology.

Contribution

It introduces a physically motivated model of black hole spin evolution and compares thermal and kinetic AGN feedback modes in idealized galaxy simulations, highlighting their differences and sensitivities.

Findings

Kinetic jet feedback is more effective at quenching star formation.

Jet feedback results in cooler cores in galaxy clusters.

Isotropic feedback effects are insensitive to parameter variations.

Abstract

Using the SWIFT simulation code we study different forms of active galactic nuclei (AGN) feedback in idealized galaxy groups and clusters. We first present a physically motivated model of black hole (BH) spin evolution and a numerical implementation of thermal isotropic feedback (representing the effects of energy-driven winds) and collimated kinetic jets that they launch at different accretion rates. We find that kinetic jet feedback is more efficient at quenching star formation in the brightest cluster galaxies (BCGs) than thermal isotropic feedback, while simultaneously yielding cooler cores in the intracluster medium (ICM). A hybrid model with both types of AGN feedback yields moderate star formation rates, while having the coolest cores. We then consider a simplified implementation of AGN feedback by fixing the feedback efficiencies and the jet direction, finding that the same general conclusions hold. We vary the feedback energetics (the kick velocity and the heating temperature), the fixed efficiencies and the type of energy (kinetic versus thermal) in both the isotropic and the jet case. The isotropic case is largely insensitive to these variations. In particular, we highlight that kinetic isotropic feedback (used e.g. in IllustrisTNG) is similar in its effects to its thermal counterpart (used e.g. in EAGLE). On the other hand, jet feedback must be kinetic in order to be efficient at quenching. We also find that it is much more sensitive to the choice of energy per feedback event (the jet velocity), as well as the efficiency. The former indicates that jet velocities need to be carefully chosen in cosmological simulations, while the latter motivates the use of BH spin evolution models.