A simple way to improve AGN feedback prescription in SPH simulations

TL;DR

This paper compares two AGN feedback models in SPH simulations, demonstrating that virtual particle feedback creates more realistic structures and proposing a simple cone injection method to improve thermal feedback accuracy.

Contribution

The study introduces a simple cone-based energy injection method to enhance thermal feedback models, aligning them more closely with virtual particle feedback results in SPH simulations.

Findings

Virtual particle feedback produces complex SMBH surroundings.

Cone injection improves thermal feedback realism.

Minimal additional computational cost.

Abstract

AGN feedback is an important ingredient in galaxy evolution, however its treatment in numerical simulations is necessarily approximate, requiring subgrid prescriptions due to the dynamical range involved in the calculations. We present a suite of SPH simulations designed to showcase the importance of the choice of a particular subgrid prescription for AGN feedback. We concentrate on two approaches to treating wide-angle AGN outflows: thermal feedback, where thermal and kinetic energy is injected into the gas surrounding the SMBH particle, and virtual particle feedback, where energy is carried by tracer particles radially away from the AGN. We show that the latter model produces a far more complex structure around the SMBH, which we argue is a more physically correct outcome. We suggest a simple improvement to the thermal feedback model - injecting the energy into a cone, rather than spherically symmetrically - and show that this markedly improves the agreement between the two prescriptions, without requiring any noticeable increase in the computational cost of the simulation.