Ultra-strong MgII absorbers trace both inflowing and outflowing gas: insights from dual down-the-barrel and quasar sightlines

TL;DR

To study baryon cycling around galaxies, the paper uses seven isolated galaxy–quasar pairs with ultra-strong MgII absorbers ($W_{r,2796} \\ge 3$ Å) at $0.4 \\le z \\le 0.6$, combining down-the-barrel spectroscopy with background-quasar absorption to trace inflows and outflows. The results reveal inflows in five galaxies with central velocities $\\Delta v$ in the range $61$–$361$ km s$^{-1}$ and outflows in three with $\\Delta v$ from $-311$ to $-119$ km s$^{-1}$, with quasar sightlines showing counterparts out to impact parameters $D = 15$–$31$ kpc. The inferred mass flow rates imply mass loading factors $\\eta = \\dot{M}/\\mathrm{SFR} < 1$, suggesting the detected cool MgII gas alone cannot sustain current SFRs, though other gas phases may contribute. The unexpectedly high incidence of inflows suggests ultra-strong MgII absorbers offer a powerful strategy to map inflow and outflow cycles across cosmic time.

Abstract

We present Keck/LRIS spectroscopy of seven isolated galaxy-quasar pairs at $0.4 \leq z \leq 0.6$, each exhibiting ultra-strong MgII absorption ($W_{r,2796} \geq 3$ Å), probing both down-the-barrel and transverse gas flows. Down-the-barrel galaxy spectra reveal outflows in three galaxies ($v = 19$ to $311$ km s$^{-1}$) and inflows in five ($v = 61$ to $361$ km s$^{-1}$), including one system showing inflows and outflows simultaneously. All galaxies with detected inflows are below the star-forming main sequence, suggesting that they might be actively replenishing their gas reservoirs. Outflows have a mean covering fraction of $C_{f, \rm out}=0.5$, whereas inflows show a lower average of $C_{f, \rm in}=0.3$. Mass flow rates span $\dot{M}_{\rm in} = 0.01-1.18$ $M_{\odot} \mathrm{yr}^{-1}$ for inflows and $\dot{M}_{\rm out} = 0.23-1.03$ $M_{\odot}\mathrm{yr}^{-1}$ for outflows, yielding mass loading factors below unity and implying these galaxies cannot sustain their current level of star-formation rates. These results are based on the T $\sim 10^4$ K photoionised gas phase traced by MgII; additional accreting/outflowing material in other gas phases may also be present, but remains undetected in this study. Quasar sightlines consistently show redshifted inflow components and blueshifted outflow components, demonstrating that ultra-strong MgII absorbers trace baryon cycling out to impact parameters of $D = 15$-31 kpc. Moreover, the unexpectedly high prevalence of inflows suggests that ultra-strong MgII absorbers offer a powerful strategy for future surveys to systematically map inflow and outflow cycles across cosmic time.

Figures (9)

• Figure 1: equivalent width as a function of impact parameter of our sample (large pink circles) composed of ultra-strong absorbers. For comparison, we also include measurements from other studies Nielsen_2013bKacprzak2013Dutta2020Huang2021. Detections are displayed as circles, while $3\sigma$ upper limits are shown as downward arrows. The solid black curve is the log-linear best fit to the MAGIICAT data presented in Nielsen2013a.
• Figure 2: Top left panels: DECaLS $grz$ image of each quasar field with their respective foreground galaxies. The location of the quasar and galaxy is indicated with the blue and red arrows, respectively. The white lines represent the location of the LRIS slit. Top right panels: $\lambda\lambda 2796,2803$ absorption doublet found in the spectra of the background quasars. Individual absorption components are shown in purple, orange and dark green when present. Bottom left panels: emission line of each galaxy. In the case of J155003, the emission line is not covered in our wavelength range, so we present instead. Bottom right panels: $\lambda\lambda 2796,2803$ absorption doublet of each galaxy. In all spectral panels, the zero-velocity point is defined as the central velocity of the (or ) emission line. The galaxy spectra are shown in black and the 1$\sigma$ errors are presented in light grey. Individual absorption components are colour-coded: green traces the ISM, blue represents outflows, and brown indicates inflows. Vertical lines highlight the central velocity of the ISM, inflows and outflows, as labelled. The total fitted model is shown in dark grey.
• Figure 3: (continued)
• Figure 4: (continued)
• Figure 5: Location of our targets relative to the star-forming main sequence (SFMS). Grey contours represent the density distribution of star-forming galaxies from Barro2011. The black solid line shows the SFMS at $z=0.5$ from Whitaker2012, while the black dashed lines represent the $1\sigma$ scatter around this relation. Our targets are shown as blue and orange squares, depending on whether we detect outflows or inflows in them, respectively. The sizes of these squares are proportional to the equivalent width of the flow in the down-the-barrel observations. In the case of J000413 and J024008, where two flow components are present, the square size is proportional to the component with the largest equivalent width. The blue and orange diamonds and the green circles correspond to the Rubin_2014 sample, colour coded in the same way as our squares, while green circles represent targets where no flows are detected. There are no such cases in the present work. These diamonds are also scaled proportionally to the equivalent width of the flow. Their sample contains galaxies at a similar redshift to ours.