The Multi-Phase Circumgalactic Medium of DESI Emission-Line Galaxies at z~1.5

TL;DR

This study probes the multiphase circumgalactic medium around DESI emission-line galaxies at z~1.5 using MgII and CIV absorption seen in ~7,741 ELG–quasar sightlines. It finds that MgII and CIV distributions and covering fractions correlate with stellar mass, but their radial profiles largely align when scaled by the halo virial radius, indicating gravity-driven halo extents. CIV gas is more extended than MgII, and the two tracers show only a loose coupling, with inner regions exhibiting elevated FeII/MgII and suppressed CIV/MgII, implying a non-co-spatial multiphase CGM. Velocity dispersions rise from ~100 km/s inside halos to ~200 km/s beyond, and the absorption properties reveal a complex, three-phase CGM with distinct densities and locations. The work also documents redshift evolution in CGM properties and estimates substantial neutral gas, metal, and dust masses in the CGM, roughly comparable to ISM budgets, highlighting the CGM’s important role in baryon cycling and galaxy evolution at high redshift.

Abstract

We study the multi-phase circumgalactic medium (CGM) of emission line galaxies (ELGs) at $z\sim1.5$, traced by MgII$\lambda2796$, $\lambda2803$ and CIV$\lambda1548$, $\lambda1550$ absorption lines, using approximately 7,000 ELG-quasar pairs from the Dark Energy Spectroscopic Instrument. Our results show that both the mean rest equivalent width ($W_{0}$) profiles and covering fractions of MgII and CIV increase with ELG stellar mass at similar impact parameters, but show similar distributions when normalized by the virial radius. Moreover, warm CIV gas has a more extended distribution than cool MgII gas. The dispersion of MgII and CIV gas velocity offsets relative to the galaxy redshifts rises from $\sim100 \, \rm km \, s^{-1}$ within halos to $\sim 200 \, \rm km \, s^{-1}$ beyond. We explore the relationships between MgII and CIV $W_{0}$ and show that the two are not tightly coupled: at a fixed absorption strength of one species, the other varies by several-fold, indicating distinct kinematics between the gas phases traced by each. We measure the line ratios, FeII/MgII and CIV/MgII, of strong MgII absorbers and find that at $<0.2$ virial radius, the FeII/MgII ratio is elevated, while the CIV/MgII ratio is suppressed compared with the measurements on larger scales, both with $\sim4-5\, σ$ significance. We argue that multiphase gas that is not co-spatial is required to explain the observational results. Finally, by combining with measurements from the literature, we investigate the redshift evolution of CGM properties and estimate the neutral hydrogen, metal, and dust masses in the CGM of DESI ELGs -- found to be comparable to those in the ISM.

Figures (15)

• Figure 1: Rest equivalent width distributions around ELGs. The upper panels show the rest equivalent width distribution in physical space and the lower panels show the rest equivalent width distribution in halo space $r_{p}/r_{vir}$. The left and right panels show the results for MgII and CIV respectively. The colors indicate measurements around DESI ELGs with different stellar mass as listed in the figure. The large data points show the mean rest equivalent widths with uncertainty based on bootstrapping the sample 500 times. The blue dashed lines and the red dashed lines show the best-fit functions (Equation 1 and Equation 2) for MgII and CIV respectively. In the upper left panel, the grey dashed and dotted lines show the best-fit functions of MgII profiles at $z\sim0.3-0.4$ from Nielsen13 and Huang21 respectively. In the upper right panel, the grey dashed and data points show the CIV measurements at $z<0.1$ from Bordoloi14 and Liang14 respectively. The purple and green data points in the two panels are measurements at $z>2$ from QPQ14 and Steidel10. CII absorption strength is used to estimate the expected MgII absorption strength.
• Figure 2: Covering fraction of MgII and CIV absorption lines as a function of stellar mass. The upper and lower panels are the measurements of MgII and CIV absorbers respectively. The left, middle and right panels show the results for weak absorbers with $\rm 0.4<W_{0}<1 \, \AA$, strong absorbers with $\rm W_{0}>1 \, \AA$, and all systems with $\rm W_{0}>0.4 \, \AA$ respectively. The stellar mass of galaxies is indicated by the colors. The dashed lines are the best-fit curves with functional forms described in the text. The uncertainties are estimated based on binomial statistics.
• Figure 3: Maximum impact parameters, $r_{p}$, with MgII and CIV covering fractions being 0.2 and 0.4. Left: Maximum $r_{p}$ with 0.2 covering fractions of weak absorbers ($0.4<W_{0}<1 \rm \, \AA$). Middle: Maximum $r_{p}$ with 0.2 covering fractions of strong absorbers ($W_{0}>1 \rm \, \AA$). Right: Maximum $r_{p}$ with 0.4 covering fractions of all systems ($W_{0}>0.4 \rm \, \AA$). MgII and CIV measurements are indicated by blue and red data points respectively. The dashed lines show the best-fit power laws. The uncertainties of the data points are obtained by bootstrapping the samples 500 times.
• Figure 4: Covering fraction of MgII and CIV absorption lines as a function of $r_{p}/r_{vir}$. The upper and lower panels are the measurements of MgII and CIV absorbers respectively. The left, middle and right panels show the results for weak absorbers with $\rm 0.4<W_{0}<1 \, \AA$, strong absorbers with $\rm W_{0}>1 \, \AA$, and all systems with $\rm W_{0}>0.4 \, \AA$ respectively. The stellar mass of galaxies is indicated by the colors. The dashed lines are the best-fit curves with functional forms described in the text. The uncertainties are estimated based on binomial statistics.
• Figure 5: Covering fraction difference (covering fraction of high-SFR galaxies - covering fraction of low-SFR galaxies). The right, middle and left panels show the covering fraction differences for galaxies with $10^{9.3}<M_{*}<10^{10} \, M_{\odot}$, $10^{10}<M_{*}<10^{10.4} \, M_{\odot}$ and $M_{*}>10^{10.4}\, M_{\odot}$ respectively. The upper and lower panels show the results for MgII absorbers and CIV absorbers with $W_{0}>0.4 \rm \, \AA$ respectively. The color bands show the best-fit values from inner $r_{p}<50 \, \rm kpc$ (green), intermediate $r_{p}<100 \, \rm kpc$ (orange), to outer regions $r_{p}<300 \, \rm kpc$ (blue).