2022ApJ...936..171R

We use medium-resolution Keck/Echellette Spectrograph and Imager spectroscopy of bright quasars to study cool gas traced by Ca IIλλ3934, 3969 and Na Iλλ5891, 5897 absorption in the interstellar/circumgalactic media of 21 foreground star-forming galaxies at redshifts 0.03 < z < 0.20 with stellar masses 7.4 ≤ log M_*/M_⊙ ≤ 10.6. The quasar-galaxy pairs were drawn from a unique sample of Sloan Digital Sky Survey quasar spectra with intervening nebular emission, and thus have exceptionally close impact parameters (R_⊥ < 13 kpc). The strength of this line emission implies that the galaxies' star formation rates (SFRs) span a broad range, with several lying well above the star-forming sequence. We use Voigt profile modeling to derive column densities and component velocities for each absorber, finding that column densities N(Ca II) > 10^12.5 cm^–2 (N(Na I) > 10^12.0 cm^–2) occur with an incidence f_C(Ca II) = 0.63^+0.10_–0.11 (f_C(Na I) = 0.57^+0.10_–0.11). We find no evidence for a dependence of f_C or the rest-frame equivalent widths W_r(Ca II K) or W_r(Na I 5891) on R_⊥ or M_*. Instead, W_r(Ca II K) is correlated with local SFR at >3σ significance, suggesting that Ca II traces star formation-driven outflows. While most of the absorbers have velocities within ±50 km s^–1 of the host redshift, their velocity widths (characterized by Δv₉₀) are universally 30-177 km s^–1 larger than that implied by tilted-ring modeling of the velocities of interstellar material. These kinematics must trace galactic fountain flows and demonstrate that they persist at R_⊥ > 5 kpc. Finally, we assess the relationship between dust reddening and W_r(Ca II K) (W_r(Na I 5891)), finding that 33% (24%) of the absorbers are inconsistent with the best-fit Milky Way E(B-V)-W_r relations at >3σ significance.