MUSEQuBES: The Column Density, Covering Fraction, Mass, and Environmental Dependence of Cool HI Gas Around Low-Redshift Galaxies

TL;DR

This study maps cool neutral hydrogen around 256 low-redshift galaxies using HI Lyman-series absorption seen in COS/HST data and MUSEQuBES galaxy spectroscopy. It characterizes the N(HI) profile, covering fraction, and HI mass in the outer CGM, finding a steep power-law decline with normalized impact parameter (slope ≈ −3) for isolated star-forming galaxies and a mass-dependent extension of HI-rich gas up to ~1.5 Rvir. The analysis reveals that high-mass and non-isolated environments suppress HI within the virial radius but enhance it at larger radii, with the CGM mass budget in the outer regions comparable to the galaxy’s own HI reservoir once ionization corrections are applied. The results highlight the role of environment in shaping the distribution, kinematics, and phase structure (including BLAs tracing warm-hot gas) of the CGM, with implications for gas accretion and feedback processes across a broad stellar-mass range.

Abstract

We investigate cool HI gas traced by Lyman series absorption around 256 galaxies at z ~ 0.48 (median stellar mass, log10(M*/Msun) = 8.7) using 15 background quasars (median impact parameter, D = 140 pkpc), as part of the MUSE Quasar-fields Blind Emitters Survey (MUSEQuBES). We find that the HI column density (N(HI)) profile around isolated star-forming galaxies spanning ~3 dex in stellar mass is well described by a power law with slope ~ -3 when expressed as a function of normalized impact parameter D/Rvir. The HI covering fraction (k) within the virial radius for log10(N(HI)/cm^{-2}) = 14 is significantly lower in high-mass passive galaxies than in isolated star-forming galaxies. The k-profile of isolated star-forming galaxies suggests a characteristic size of the HI-rich CGM of ~ 1.5 Rvir across the stellar mass range. The mean HI mass in the outer CGM (0.3-1 Rvir ) increases with stellar mass, ranging from ~ 10^5 to 10^6.6 Msun. The b-parameters of the strongest HI components correlate and anti-correlate with specific star-formation rate (sSFR) and mass, respectively, with >2 sigma significance. Broad Lya absorbers (BLAs) with b > 60 km/s are predominantly associated with high-mass galaxies, likely tracing the warm-hot phase of the CGM. The velocity centroids of H i components indicate that absorbers at D < Rvir are largely consistent with being gravitationally bound to their galaxies, independent of stellar mass. Finally, leveraging ~ 3000 galaxies from the wide-field Magellan follow-up of six MUSEQuBES fields, we find that non-isolated galaxies exhibit an HI-rich environment extending roughly three times farther than in isolated counterparts.

Figures (18)

• Figure 1: The various galaxy properties for the MUSEQuBES sample are plotted against each other with solid blue circles. The red downward arrows in the last row indicate the SFR upper limits. The panels along the diagonal show the probability density of the corresponding galaxy property with black solid lines. The upper limits are considered as measurements for the SFR probability density.
• Figure 2: Left: The H i column densities of the 227 components with $Q>1$ used in this work plotted against the $b$-parameters. The red line shows the limiting $N(\hbox{H,{\sc i}})$ as a function of $b-$parameter for a fiducial $S/N$ of 10 per pixel. The histograms on the top and right show the distribution of $b$ parameters and component column densities, respectively, with blue dashed lines indicating the corresponding median values. Right: The total column densities ($N(\hbox{H,{\sc i}})$) of absorption systems, defined by grouping components within $\Delta v = 300$$\rm km~s^{-1}$, are plotted as a function of galaxy redshift using filled circles. The $3\sigma$ upper limits in cases of non-detections are shown by the open triangles. The observed gap between the detections and upper limits arises partly because the upper limits correspond to single-component systems, whereas the detected systems generally consist of multiple H i components. In addition, these detections are located in the vicinity of galaxies and therefore tend to have higher column densities than the general population of H i absorbers. The top and side panels show the distributions of galaxy redshift and $N(\hbox{H,{\sc i}})$, respectively, with open black histograms representing galaxies with H i detections and filled grey histograms indicating non-detections.
• Figure 3: (A): Total $N(\hbox{H,{\sc i}})$ for the 256 MUSEQuBES galaxies plotted against impact parameters $D$. The solid and hollow points represent the detection and 3$\sigma$ upper limits on $N(\hbox{H,{\sc i}})$, respectively. (B) Total $N(\hbox{H,{\sc i}})$ for the subsample constructed from the isolated and group samples, where the group galaxy with the smallest $D$ is considered as the host. (C) Same as panel (B), but considering the galaxy with the smallest $D/R_{\rm vir}$ as the host. Panels (D), (E), and (F) are similar to panels (A), (B), and (C), respectively, but plotted against $D/R_{\rm vir}$. The results of generalized Kendall-$\tau$ tests for each case are indicated at the top of the respective panels. The strongest anticorrelation is observed between $N(\hbox{H,{\sc i}})$ and $D/R_{\rm vir}$ in panel (F).
• Figure 4: Top: Total $N(\hbox{H,{\sc i}})$ for low-mass (median $\log_{10}~(M_{\star}/\rm M_{\odot})$$=7.9$; left panel) and high-mass (median $\log_{10}~(M_{\star}/\rm M_{\odot})$$=9.4$; right panel) galaxies plotted against $D/R_{\rm vir}$. The solid and open circles indicate detection and 3$\sigma$ upper limits on $N(\hbox{H,{\sc i}})$, respectively. The circles inside black squares indicated the minimum $D/R_{\rm vir}$ galaxies for the groups, and the circles without black squares indicate isolated galaxies. Bottom: Similar to the top panels, but for star-forming (SF; left) and passive (E; right) galaxies.
• Figure 5: The $N(\hbox{H,{\sc i}})-$profile for the isolated, star-forming MUSEQuBES sample. The solid and hollow circles represent the detection and $3\sigma$ upper limits, respectively. The data points are color-coded by the stellar mass of the associated galaxy. The red solid line and shaded region represent the best-fit power law and the corresponding 68% confidence interval. The shaded magenta region represents the best-fit intrinsic scatter.