FUSE Deuterium Observations: A Strong Case For Galactic Infall

Measurements of deuterium in the local interstellar medium have revealed large variations in D/H along different lines of sight. Moreover, recent {\it Far Ultraviolet Spectroscopic Explorer} (FUSE) measurements find D/H to be anticorrelated with several indicators of dust formation and survival, suggesting that interstellar deuterium suffers significant depletion onto dust grains. This in turn implies that the total deuterium abundance in the local Galactic disk could be as high as $\sim 84$ of the primordial D abundance. It was proposed that the infall/accretion of pristine gas is needed to explain such a high deuterium abundance. However, we point out that the infall needed to maintain a high present-day D/H is, within the preferred models, in tension with observations that gas represents only some $\sim 20$ of Galactic baryons, with the balance in stars. We study this tension in the context of a wide class of Galactic evolution models for baryonic processing through stars, which show that deuterium destruction is strongly and cleanly correlated with the drop in the gas fraction. We find that FUSE deuterium observations and Galactic gas fraction estimates can be reconciled in some models; these demand a significant infall rate of pristine material that almost completely balances the rate of star formation. These successful models also require that the average fraction of gas that is returned by dying stars is less than 40% of the initial stellar mass. Cosmological implications of dust depletion of D in high-redshift systems are discussed.