The Column Densities of Molecular Gas across Cosmic Time: Bridging Observations and Simulations

TL;DR

This paper compares observations and simulations to study the evolution of molecular hydrogen column densities across cosmic time, revealing how denser regions and higher redshifts influence molecular gas distribution in galaxies.

Contribution

It provides a comprehensive analysis of $f(N_{H_2})$ evolution using both observational data and state-of-the-art simulations, including new post-processing models and analytical approaches.

Findings

Higher molecular gas densities are reached at z=3 compared to z=0.

Simulations broadly reproduce observed $f(N_{H_2})$ shapes with some differences.

H$_2$ dominates over HI at high column densities above log(N)~21.8-22.

Abstract

Observations of the cosmic evolution of different gas phases across time indicate a marked increase in the molecular gas mass density towards $z\sim 2-3$. Such a transformation implies an accompanied change in the global distribution of molecular hydrogen column densities ($N_{\rm{H_2}}$). Using observations by PHANGS-ALMA/SDSS and simulations by GRIFFIN/IllustrisTNG we explore the evolution of this H$_2$ column density distribution function [$f(N_{\rm{H}_2})$]. The H$_2$ (and HI) column density maps for TNG50 and TNG100 are derived in post-processing and are made available through the IllustrisTNG online API. The shape and normalization of $f(N_{\rm{H}_2})$ of individual main-sequence star-forming galaxies are correlated with the star formation rate (SFR), stellar mass (${M_*}$), and H$_2$ mass ($M_{\rm{H}_2}$) in both observations and simulations. TNG100, combined with H$_2$ post-processing models, broadly reproduces observations, albeit with differences in slope and normalization. Also, an analytically modelled $f(N)$, based on exponential gas disks, matches well with the simulations. The GRIFFIN simulation gives first indications that the slope of $f(N_{\rm{H}_2})$ might not majorly differ when including non-equilibrium chemistry in simulations. The $f(N_{\rm{H}_2})$ by TNG100 implies that higher molecular gas column densities are reached at $z=3$ than at $z=0$. Further, denser regions contribute more to the molecular mass density at $z=3$. Finally, H$_2$ starts dominating compared to HI only at column densities above log($N_{\rm{H}_2} / \rm{cm}^{-2}) \sim 21.8-22$ at both redshifts. These results imply that neutral atomic gas is an important contributor to the overall cold gas mass found in the ISM of galaxies including at densities typical for molecular clouds at $z=0$ and $z=3$.