Spatial Variability in the Ratio of Interstellar Atomic Deuterium to Hydrogen. II. Observations toward γ²Velorum and ζ Puppis by the Interstellar Medium Absorption Profile Spectrograph

High-resolution far-ultraviolet spectra of the early-type stars γ2 Vel and ζ Pup were obtained to measure the interstellar deuterium abundances in these directions. The observations were made with the Interstellar Medium Absorption Profile Spectrograph (IMAPS) during the ORFEUS-SPAS II mission in 1996. IMAPS spectra cover the wavelength range 930-1150 Å with λ/Δλ ~ 80,000. The interstellar D I features are resolved and cleanly separated from interstellar H I in the Lyδ and Lyϵ profiles of both sight lines and also in the Lyγ profile of ζ Pup. The D I profiles were modeled using a velocity template derived from several N I lines in the IMAPS spectra recorded at higher signal-to-noise ratio. To find the best D I column density, we minimized χ2 for model D I profiles that included not only the N(D I) as a free parameter, but also the effects of several potential sources of systematic error, which were allowed to vary as free parameters. H I column densities were measured by analyzing Lyα absorption profiles in a large number of IUE high-dispersion spectra for each of these stars and applying this same χ2-minimization technique. Ultimately we found that D/H = 2.18 × 10-5 for γ2 Vel and 1.42 × 10-5 for ζ Pup, values that contrast markedly with D/H derived in Paper I for δ Ori A (the stated errors are 90% confidence limits). Evidently, the atomic D/H ratio in the ISM, averaged over path lengths of 250-500 pc, exhibits significant spatial variability. Furthermore, the observed spatial variations in D/H do not appear to be anticorrelated with N/H, one measure of heavy-element abundances. We briefly discuss some hypotheses to explain the D/H spatial variability. Within the framework of standard big bang nucleosynthesis, the large value of D/H found toward γ2 Vel is equivalent to a cosmic baryon density of ΩBh2 = 0.023 ± 0.002, which we regard as an upper limit since there is no correction for the destruction of deuterium in stars.