Using cosmological simulations and synthetic absorption spectra to assess the accuracy of observationally derived CGM metallicities

We used adaptive mesh refinement hydrodynamic cosmological simulations of a $z=1$ Milky Way-type galaxy and a $z=0$ Dwarf galaxy and generated synthetic quasar absorption-line spectra of their circumgalactic medium (CGM). Our goal is to assess whether standard observational spectroscopic analysis methods accurately reproduce intrinsic column densities, metallicities [Si/H], and hydrogen densities $n_{H}$, in simulated absorption-line systems. Without knowledge of the intrinsic simulated properties (blind study), we analysed synthetic COS and HIRES spectra with fixed $S/N=30$ to determine the column densities, metallicity, and $n_{H}$, using Voigt profile fitting combined with Markov-Chain Monte-Carlo single-phase {\sc Cloudy} modelling techniques. To quantify the intrinsic simulated absorbing gas properties, we objectively determined which gas cells along a line of sight (LOS) contribute to detected absorption in the spectra and adopt the unweighted geometric mean of these gas cell properties. For this pilot study, we performed this experiment for five LOS in the Milky-Way galaxy and five LOS in the Dwarf galaxy. We found an average agreement between the "observed" and intrinsic metallicity overestimated within $0.8σ$ or $0.2$ dex for the "Milky-Way"' CGM and overestimated within $1.4σ$ or $0.2$ dex for the Dwarf galaxy CGM. We found that the spectroscopically-derived $n_{H}$ are underestimated within $0.8σ$ or $0.4$ dex of the intrinsic $n_{H}$ for the "Milky-Way" CGM and overestimated within $0.3σ$ or $0.3$ dex for the Dwarf galaxy CGM. The overall agreement suggests that, for single-phase ionisation modelling of systems where there is substantial spread in gas properties, global metallicity measurements from quasar absorption line studies are capturing the average metallicity and ionisation parameters in a given astrophysical environment.