2013ApJ...773...12S

Type II supernovae (SNe) can be used as a star formation tracer to probe the metallicity distribution of global low-redshift star formation. We present oxygen and iron abundance distributions of Type II SN progenitor regions that avoid many previous sources of bias. Because iron abundance, rather than oxygen abundance, is of key importance for the late stage evolution of the massive stars that are the progenitors of core-collapse supernovae, and because iron enrichment lags oxygen enrichment, we find a general conversion from oxygen abundance to iron abundance. The distributions we present here are the best yet observational standard of comparison for evaluating how different classes of supernovae depend on progenitor metallicity. We spectroscopically measure the gas-phase oxygen abundance near a representative subsample of the hosts of Type II SNe from the first-year Palomar Transient Factory (PTF) SN search, using a combination of Sloan Digital Sky Survey (SDSS) spectra near the SN location (9 hosts) and new longslit spectroscopy (25 hosts). The median metallicity of these 34 hosts at or near the SN location is 12+log(O/H) = 8.65, with a median error of 0.09. The median host galaxy stellar mass from fits to SDSS photometry is 10^9.9 M_☉. They do not show a systematic offset in metallicity or mass from a redshift-matched sample of the MPA/JHU value-added catalog. In contrast to previous SN host metallicity studies, this sample is drawn from a single survey. It is also drawn from an areal rather than a targeted survey, so SNe in the lowest-mass galaxies are not systematically excluded. Indeed, the PTF SN search has a slight bias toward following up transients in low mass galaxies. The progenitor region metallicity distribution we find is statistically indistinguishable from the metallicity distribution of Type II SN hosts found by targeted surveys and by samples from multiple surveys with different selection functions. Using the relationship between iron and oxygen abundances found for Milky Way disk, bulge, and halo stars, we translate our distribution of Type II SN environments as a function of oxygen abundance into an estimate of the iron abundance, since iron varies more steeply than oxygen. We find that though this sample spans only 0.65 dex in oxygen abundance, the gap between the iron and oxygen abundance is 50% wider at the low-metallicity end of our sample than at the high-metallicity end.