The Definitive Abundance of Interstellar Oxygen

Abstract

Using the Goddard High Resolution Spectrograph (GHRS) onboard HST, we have obtained high S/N echelle observations of the weak interstellar O I 1356 A absorption toward the stars Gamma Cas, Epsilon Per, Delta Ori, Epsilon Ori, 15 Mon, Tau CMa, and Gamma Ara. In combination with previous GHRS measurements in six other sightlines (Zeta Per, Xi Per, Lambda Ori, Iota Ori, Kappa Ori, and Zeta Oph), these new observations yield a mean interstellar gas-phase oxygen abundance (per 10$^6$ H atoms) of 10$^6$ O/H = 319 +/- 14. The largest deviation from the mean is less than 18%, and there are no statistically significant variations in the measured O abundances from sightline to sightline and no evidence of density-dependent oxygen depletion from the gas phase. Assuming various mixtures of silicates and oxides, the abundance of interstellar oxygen tied up in dust grains is unlikely to surpass 10$^6$ O/H $\approx$ 180. Consequently, the GHRS observations imply that the total abundance of interstellar oxygen (gas plus grains) is homogeneous in the vicinity of the Sun and about 2/3 of the solar value of 10$^6$ O/H = 741 +/- 130. This oxygen deficit is consistent with that observed in nearby B stars and similar to that recently found for interstellar krypton with GHRS. Possible explanations for this deficit include: (1) early solar system enrichment by a local supernova, (2) a recent infall of metal-poor gas in the local Milky Way, or (3) an outward diffusion of the Sun from a smaller galactocentric distance.