Highly ionized gas in the Galactic halo: A FUSE survey of O VI absorption toward 22 halo stars

Far Ultraviolet Spectroscopic Explorer (FUSE) spectra of 22 Galactic halo stars are studied to determine the amount of O VI in the Galactic halo between ∼0.3 and ∼10 kpc from the Galactic midplane. Strong O VI λ1031.93 absorption was detected toward 21 stars, and a reliable 3 σ upper limit was obtained toward HD 97991. The weaker member of the O VI doublet at 1037.62 Å could be studied toward only six stars because of stellar and interstellar blending problems. The measured logarithmic total column densities vary from 13.65 to 14.57 with 〈logN〉 = 14.17 ± 0.28 (1 σ). The observed columns are reasonably consistent with a patchy exponential O VI distribution with a midplane density of 1.7 × 10^-8 cm^-3 and scale height between 2.3 and 4 kpc. We do not see clear signs of strong high-velocity components in O VI absorption along the Galactic sight lines, which indicates the general absence of high-velocity O VI within 2-5 kpc of the Galactic midplane. This result is in marked contrast to the findings of Sembach et al., who reported high-velocity O VI absorption toward ∼60% of the complete halo sight lines observed by FUSE. The line centroid velocities of the O VI absorption do not reflect Galactic rotation well. The O VI velocity dispersions range from 33 to 78 km s^-1, with an average of (b) = 45 ± 11 km s^-1 (1 σ). These values are much higher than the value of ∼18 km s^-1 expected from thermal broadening for gas at T ∼ 3 × 10⁵ K, the temperature at which O VI is expected to reach its peak abundance in collisional ionization equilibrium. Turbulence, inflow, and outflow must have an effect on the shape of the O VI profiles. Kinematical comparisons of O VI with Ar I reveal that eight of 21 sight lines are closely aligned in LSR velocity (|ΔV_LSR| ≤ 5 km s^-1), while nine of 21 exhibit significant velocity differences (|ΔV_LSR| ≥ 15 km s^-1). This dual behavior may indicate the presence of two different types of O VI-bearing environments toward the Galactic sight lines. The correlation between the H I and O VI intermediate-velocity absorption is poor. We could identify the known H I intermediate-velocity components in the Ar I absorption but not in the O VI absorption in most cases. Comparison of O VI with other highly ionized species suggests that the high ions are produced primarily by cooling hot gas in the Galactic fountain flow and that turbulent mixing also has a significant contribution. The role of turbulent mixing varies from negligible to dominant. It is most important toward sight lines that sample supernova remnants like Loops I and IV. The average N(C IV)/N(O VI) ratios for the nearby halo (this work) and complete halo (Savage et al.) are similar (∼0.6), but the dispersion is larger in the sample of nearby halo sight lines. We are able to show that the O VI enhancement toward the Galactic center region that was observed in the FUSE survey of complete halo sight lines (Savage et al.) is likely associated with processes occurring near the Galactic center by comparing the observations toward the nearby HD 177566 sight line to those toward extragalactic targets.