paper

The Birth of a Galaxy. II. The Role of Radiation Pressure

arXiv:1206.1043 · doi:10.1111/j.1365-2966.2012.21809.x

Abstract

Massive stars provide feedback that shapes the interstellar medium of galaxies at all redshifts and their resulting stellar populations. Here we present three adaptive mesh refinement radiation hydrodynamics simulations that illustrate the impact of momentum transfer from ionising radiation to the absorbing gas on star formation in high-redshift dwarf galaxies. Momentum transfer is calculated by solving the radiative transfer equation with a ray tracing algorithm that is adaptive in spatial and angular coordinates. We find that momentum input partially affects star formation by increasing the turbulent support to a three-dimensional rms velocity equal to the circular velocity of early haloes. Compared to a calculation that neglects radiation pressure, the star formation rate is decreased by a factor of five to 1.8 x 10^{-2} Msun/yr in a dwarf galaxy with a dark matter and stellar mass of 2.0 x 10^8 and 4.5 x 10^5 solar masses, respectively, when radiation pressure is included. Its mean metallicity of 10^{-2.1} Z_sun is consistent with the observed dwarf galaxy luminosity-metallicity relation. However, what one may naively expect from the calculation without radiation pressure, the central region of the galaxy overcools and produces a compact, metal-rich stellar population with an average metallicity of 0.3 Z_sun, indicative of an incorrect physical recipe. In addition to photo-heating in HII regions, radiation pressure further drives dense gas from star forming regions, so supernovae feedback occurs in a warmer and more diffuse medium, launching metal-rich outflows. Capturing this aspect and a temporal separation between the start of radiative and supernova feedback are numerically important in the modeling of galaxies to avoid the "overcooling problem". We estimate that dust in early low-mass galaxies is unlikely to aid in momentum transfer from radiation to the gas.

arXiv comment18 pages, 11 figures, replaced with accepted version, MNRAS. Minor changes with the conclusions unaffected

References in corpus (100)

Cited by in corpus (146)