Cosmological Galaxy Evolution with Superbubble Feedback II: The Limits of Supernovae

TL;DR

This study investigates the limitations of supernova feedback in regulating star formation in galaxies, showing that supernovae alone are insufficient for galaxies above a certain mass, necessitating other feedback mechanisms like AGN feedback.

Contribution

It introduces simulations including evaporation and conduction physics to assess supernova feedback limits across galaxy masses, highlighting the need for additional feedback in massive galaxies.

Findings

Supernovae cannot prevent excessive star formation in galaxies with mass >10^{12} M_\

,

,

Abstract

We explore when supernovae can (and cannot) regulate the star formation and bulge growth in galaxies based on a sample of 18 simulated galaxies. The simulations include key physics such as evaporation and conduction, neglected in prior work, and required to correctly model superbubbles resulting from stellar feedback. We show that for galaxies with virial masses $>10^{12}\;M_\odot$, supernovae alone cannot prevent excessive star formation. This failure occurs due to a shutdown of galactic winds, with wind mass loadings falling from $\eta\sim10$ to $\eta<1$. In more massive systems, this transfer of baryons to the circumgalactic medium falters earlier on and the galaxies diverge significantly from observed galaxy scaling relations and morphologies. The decreasing efficiency is simply due to a deepening potential well preventing gas escape. This implies that non-supernova feedback mechanisms must become dominant for galaxies with stellar masses greater than $\sim4\times10^{10}\;M_\odot$. The runaway growth of the central stellar bulge, strongly linked to black hole growth, suggests that feedback from active galactic nuclei is the probable mechanism. Below this mass, supernovae alone are able to produce a realistic stellar mass fraction, star formation history and disc morphology.