The metallicity of galaxy disks: Infall versus outflow

Both gas accretion (infall) and winds (outflow) change a galaxy's metallicity and gas fraction, lowering the effective yield. Low effective yields in galaxies with rotation speeds < 120 km s(-1) have been widely interpreted as due to the onset of supernova-driven winds below a characteristic galaxy mass, but gas accretion is also a viable explanation. However, calculations presented here prove that (1) metal-enriched outflows are the only mechanism that can significantly reduce the effective yield, but only for gas-rich systems; (2) it is nearly impossible to reduce the effective yield of a gas-poor system, no matter how much gas is lost or accreted; and (3) any subsequent star formation drives the effective yield back to the closed-box value. Thus, only gas-rich systems with low star formation rates (such as dwarf irregulars) can produce and maintain low effective yields. The drop in effective yield seen in low-mass galaxies is therefore due to these galaxies' gas richnesses and low star formation rates, which result from their surface densities falling entirely below the Kennicutt SF threshold. Additional calculations confirm that the fraction of baryonic mass lost through winds varies only weakly with galaxy mass, shows no sharp upturn at any mass scale, and does not require that > 15% of the baryons have been lost by galaxies of any mass. Supernova feedback is therefore unlikely to be effective for removing large amounts of gas from low-mass disk galaxies. In addition, the dependence between metal loss and galaxy mass is sufficiently weak that massive galaxies dominate metal enrichment of the IGM. The calculations in this paper provide limiting cases for any arbitrary chemical evolution history, as is proven in an Appendix.