The ALMA REBELS Survey: The Cosmic HI Gas Mass Density in Galaxies at $z\approx 7$

TL;DR

This study estimates the HI gas content in galaxies at redshift 7 using [CII] emission, revealing a high HI fraction in galaxies and a lower cosmic HI density than previous DLA-based estimates, aligning with recent simulations.

Contribution

First to estimate cosmic HI density at z~7 using [CII] emission, providing new insights into HI distribution and its evolution in early galaxies.

Findings

HI gas mass fraction in galaxies at z~7 is about 90%

Cosmic HI density at z~7 is roughly 10 times lower than DLA extrapolations

Most HI at high redshift resides in galactic halos, not star-forming ISM

Abstract

The neutral atomic gas content of individual galaxies at large cosmological distances has until recently been difficult to measure due to the weakness of the hyperfine HI 21-cm transition. Here we estimate the HI gas mass of a sample of main-sequence star-forming galaxies at $z\sim 6.5 - 7.8$ surveyed for [CII]$-158\mu$m emission as part of the Reionization Era Bright Emission Line Survey (REBELS), using a recent calibration of the [CII]-to-HI conversion factor. We find that the HI gas mass excess in galaxies increases as a function of redshift, with an average of $M_{\rm HI} / M_\star \approx 10$, corresponding to HI gas mass fractions of $f_{\rm HI} = M_{\rm HI} / (M_\star + M_{\rm HI}) = 90\%$, at $z\approx 7$. Based on the [CII]-$158\mu$m luminosity function (LF) derived from the same sample of galaxies, we further place constraints on the cosmic HI gas mass density in galaxies ($\rho_{\rm HI}$) at this redshift, which we measure to be $\rho_{\rm HI} = 7.1^{+6.4}_{-3.0} \times 10^{6}\,M_{\odot}\,{\rm Mpc^{-3}}$. This estimate is substantially lower by a factor of $\approx 10$ than that inferred from an extrapolation of damped Lyman-$\alpha$ absorber (DLA) measurements, and largely depend on the exact [CII] LF adopted. However, we find this decrease in $\rho_{\rm HI}$ to be consistent with recent simulations and argue that this apparent discrepancy is likely a consequence of the DLA sightlines predominantly probing the substantial fraction of HI gas in high-$z$ galactic halos, whereas [CII] traces the HI in the ISM associated with star formation. We make predictions for this build-up of neutral gas in galaxies as a function of redshift, showing that at $z\gtrsim 5$ only $\approx 10\%$ of the cosmic HI gas content is confined in galaxies and associated with the star-forming ISM.