Аннотация
Using high resolution cosmological zoom-in simulations of galaxy formation,
we investigate the star formation-feedback cycle at high redshifts ($z>1$),
focusing on progenitors of Milky Way-sized galaxies. Our star formation model
is based on the local density of molecular hydrogen (H$_2$) forming on dust
grains, as this may be an important ingredient for regulating star formation in
the high redshift, metal-poor regime of galaxy formation. Our stellar feedback
model accounts for energy and momentum from supernovae, stellar winds and
radiation pressure. We use a suite of simulations with different parameters and
assumptions about star formation and prescription recipes. We find that in
order to reproduce global properties of the Milky Way progenitors, such as star
formation history and stellar mass-halo mass relation, simulations should
include 1) a combination of local early ($t4$ Myr) momentum feedback
via radiation pressure and stellar winds and subsequent efficient supernovae
feedback, and 2) the global star formation efficiency on kiloparsec scales
should be feedback regulated. In particular, we find that in models with
efficient feedback, the local efficiency of star formation per free fall time
can be substantially larger than the global star formation efficiency inferred
from the Kennicutt-Schmidt relation. We find that simulations that adopt
inefficient star formation inferred from such relation fail to produce vigorous
outflows and eject sufficient amounts of enriched gas in order to regulate the
galactic baryon content. This illustrates the importance of understanding the
complex interplay between star formation and feedback and the detailed
processes that contribute to the feedback-regulated formation of galaxies.
(Abridged for arXiv)
Пользователи данного ресурса
Пожалуйста,
войдите в систему, чтобы принять участие в дискуссии (добавить собственные рецензию, или комментарий)