paper

The impact of supernovae driven winds on stream-fed protogalaxies

arXiv:1012.2839 · doi:10.1111/j.1365-2966.2011.18668.x

Abstract

SNe driven winds are widely thought to be very influential in the high-redshift Universe, shaping the properties of the circum-galactic medium, enriching the IGM with metals and driving the evolution of low-mass galaxies. However, it is not yet fully understood how SNe driven winds interact with their surroundings in a cosmological context, nor is it clear whether they are able to significantly impact the evolution of low-mass galaxies from which they originate by altering the amount of cold material these accrete from the cosmic web. We implement a standard Taylor-Sedov type solution, widely used in the community to depict the combined action of many SN explosions, in a cosmological resimulation of a low mass galaxy at z =9 from the 'Nut' suite. However, in contrast with previous work, we achieve a resolution high enough to capture individual SN remnants in the Taylor-Sedov phase, for which the solution provides an accurate description of the expansion. We report the development of a high-velocity, far-reaching galactic wind produced by the combined action of SNe in the main galaxy and its satellites, which are located in the same or a neighbouring dark matter halo. Despite this, we find that (i) this wind carries out very little mass (the measured outflow is of the order of a tenth of the inflow/star formation rate) and (ii) the cold gas inflow rate remains essentially unchanged from the run without SNe feedback. Moreover, there are epochs during which star formation is enhanced in the feedback run relative to its radiative cooling only counterpart. We attribute this 'positive' feedback to the metal enrichment that is present only in the former. We conclude that at very high redshift, efficient SNe feedback can drive large-scale galactic winds but does not prevent massive cold gas inflow from fuelling galaxies, resulting in long-lived episodes of intense star formation.(abridged)

Accepted for publication in MNRAS. 21 pages, 14 figures, 1 table. Some clarifications/changes to presentation following referee report, results and conclusions unchanged. This version contains low resolution figures, see journal for high resolution version

References in corpus (11)

Cited by in corpus (60)