Zusammenfassung
We show how the interplay between active galactic nuclei (AGN) and merger
history determines whether a galaxy quenches star formation at high redshift.
We first simulate, in a full cosmological context, a galaxy of total dynamical
mass $10^12\,M_ødot$ at $z=2$. Then we systematically alter the accretion
history of the galaxy by minimally changing the linear overdensity in the
initial conditions. This "genetic modification" approach allows the generation
of three sets of $Łambda$CDM initial conditions leading to maximum merger
ratios of 1:10, 1:5 and 2:3 respectively. The changes leave the final halo
mass, large scale structure and local environment unchanged, providing a
controlled numerical experiment. Interaction between the AGN physics and
mergers in the three cases lead respectively to a star-forming,
temporarily-quenched and permanently-quenched galaxy. However the differences
do not primarily lie in the black hole accretion rates, but in the kinetic
effects of the merger: the galaxy is resilient against AGN feedback unless its
gaseous disk is first disrupted. Typical accretion rates are comparable in the
three cases, falling below $0.1\,M_ødot$ yr$^-1$, equivalent to around
$2\%$ of the Eddington rate or $10^-3$ times the pre-quenching star formation
rate, in agreement with observations. This low level of black hole accretion
can be sustained even when there is insufficient dense cold gas for star
formation. Conversely, supernova feedback is too distributed to generate
outflows in high-mass systems, and cannot maintain quenching over periods
longer than the halo gas cooling time.
Beschreibung
[1607.02507] How to quench a galaxy
Links und Ressourcen
Tags