UC Santa Barbara

We present the first direct calibration of strong-line metallicity diagnostics at significant cosmological distances using a sample at z ≃ 0.8 drawn from the DEEP2 Galaxy Redshift Survey. Oxygen and neon abundances are derived from measurements of electron temperature and density. We directly compare various commonly used relations between gas-phase metallicity and strong line ratios of O, Ne, and H at z ≃ 0.8 and z = 0. There is no evolution with redshift at high precision (δ log O H = -0.01 ± 0.03, D log Ne O = 0.01 ± 0.01). O, Ne, and H line ratios follow the same locus at z ≃ 0.8 as at z = 0, with ≲0.02 dex evolution and low scatter (≲0.04 dex). This suggests little or no evolution in physical conditions of H II regions at fixed oxygen abundance, in contrast to models which invoke more extreme properties at high redshifts. We speculate that offsets observed in the [N II]/Ha versus [O III]/Hβ diagram at high redshift are therefore due to [N II] emission, likely as a result of relatively high N/O abundance. If this is indeed the case, then nitrogen-based metallicity diagnostics suffer from systematic errors at high redshift. Our findings indicate that locally calibrated abundance diagnostics based on α-capture elements can be reliably applied at z ≃ 1 and possibly at much higher redshifts. This constitutes the first firm basis for the widespread use of empirical calibrations in high redshift metallicity studies.