Table of Contents
Fetching ...

Chemodynamical properties of gas-rich galaxies: a comparison of observations and simulations

Anna Velichko, Yves Revaz, Annalisa De Cia, Cédric Ledoux, Jens-Kristian Krogager, Céline Péroux, Benedetta Casavecchia

Abstract

We perform a comprehensive analysis of the chemical and dynamical properties of quasar-damped Lyman-α (DLA) galaxies and compare these to the GEAR chemodynamical simulations. Specifically, we aim to constrain the behavior of α-element enhancements with metallicity, the dependence of [α/Fe] on the specific star formation rate (sSFR), and the absorption-line velocity widths (Δv90) vs. stellar mass, Δv90 vs. metallicity, and mass-metallicity relations. For the comparison, we select five galaxies simulated with the chemodynamical Tree-SPH code GEAR with stellar masses in the range of log(M*/Msol) between 6.1 and 10.8, and at six different redshifts between 0.33 and 4.12. We find that the abundance ratios [α/Fe] and [M/H] observed in the interstellar medium (ISM) of DLA galaxies overlap with the abundance trends in gas of the simulated galaxies. Our findings corroborate a picture in which DLAs with Δv90 below and above 100 km/s trace galaxies with masses in the ranges of log M* 6 - 8 and 8 - 11 solar masses, respectively. We suggest that observations should be used with caution when constraining the theoretical [α/Fe] vs. sSFR relations because of systematics (if abundances are obtained from emission lines) or differences in the gas properties as probed by a DLA and its counterpart. So far, only the observations in absorption of inner gas of the LMC and SMC are in agreement with the simulated data. We confirm that DLAs detected at large impact parameters most likely probe the gas of satellite or other halo galaxies which are adjacent to the central galaxy. We further find that the velocity widths vs. stellar masses and mass-metallicity relations agree well with observations, while GEAR should be calibrated more carefully to reproduce the Δv90 vs. metallicity relation.

Chemodynamical properties of gas-rich galaxies: a comparison of observations and simulations

Abstract

We perform a comprehensive analysis of the chemical and dynamical properties of quasar-damped Lyman-α (DLA) galaxies and compare these to the GEAR chemodynamical simulations. Specifically, we aim to constrain the behavior of α-element enhancements with metallicity, the dependence of [α/Fe] on the specific star formation rate (sSFR), and the absorption-line velocity widths (Δv90) vs. stellar mass, Δv90 vs. metallicity, and mass-metallicity relations. For the comparison, we select five galaxies simulated with the chemodynamical Tree-SPH code GEAR with stellar masses in the range of log(M*/Msol) between 6.1 and 10.8, and at six different redshifts between 0.33 and 4.12. We find that the abundance ratios [α/Fe] and [M/H] observed in the interstellar medium (ISM) of DLA galaxies overlap with the abundance trends in gas of the simulated galaxies. Our findings corroborate a picture in which DLAs with Δv90 below and above 100 km/s trace galaxies with masses in the ranges of log M* 6 - 8 and 8 - 11 solar masses, respectively. We suggest that observations should be used with caution when constraining the theoretical [α/Fe] vs. sSFR relations because of systematics (if abundances are obtained from emission lines) or differences in the gas properties as probed by a DLA and its counterpart. So far, only the observations in absorption of inner gas of the LMC and SMC are in agreement with the simulated data. We confirm that DLAs detected at large impact parameters most likely probe the gas of satellite or other halo galaxies which are adjacent to the central galaxy. We further find that the velocity widths vs. stellar masses and mass-metallicity relations agree well with observations, while GEAR should be calibrated more carefully to reproduce the Δv90 vs. metallicity relation.
Paper Structure (26 sections, 15 figures, 3 tables)

This paper contains 26 sections, 15 figures, 3 tables.

Figures (15)

  • Figure 1: The surface density of all gas for the five GEAR galaxies Revaz2018Roca-Fabrega2021 at redshifts 4.12 (upper panel) and 0.33 (lower panel). For h074, the lower rightmost panel is empty, as all the gas has been removed by photoevaporation.
  • Figure 2: [$\alpha$/Fe] vs. metallicity. Shaded area: [Mg/Fe] versus [Fe/H] in gas of five simulated galaxies h000, h026, h076, h050 and h074 at $z$ from 4.12 to 0.33 (see the numbers). In the case of h074, there are no gas particles left at $z<2.01$. Orange open diamonds and blue open triangles are the values [$\alpha$/Fe]$_{\rm nucl}$ obtained by Velichko2024 from the observations of the ISM in QSO-DLAs in the high- and low-$\Delta v_{90}$ subsamples, respectively. Filled symbols mark DLAs masses of which have been determined by Moller2020 or Christensen2014 from observations in emission of the DLA counterparts (see Table \ref{['tab:DLA_measured']}). The blue and orange curves are three-piecewise fits to the data for the low- and high-$\Delta v_{90}$ subsamples obtained in Velichko2024. The gray dashed curve shows the $\alpha$-element enhancements obtained for the MW by McWilliam1997.
  • Figure 3: [$\alpha$/Fe] or [O/Fe] vs. specific star formation rate in gas. Small filled diamonds represent [O/Fe] from GEAR Revaz2018Roca-Fabrega2021 which we fitted with the blue solid line and a 3$\sigma$ probability distribution function (blue shaded area). For comparison, the brown and green dashed curves, with corresponding contours, show [O/Fe] vs. sSFR from the TNG100 Chruslinska2024 and EAGLE Matthee2018Chruslinska2024 cosmological simulations, respectively. The measurements of [$\alpha$/Fe] from absorption spectra of DLAs are taken from Velichko2024 (circles) and from Krogager2013Fynbo2011Fynbo2013, corrected for dust depletion by us (stars). The measurements of sSFRs have been obtained by Moller2020Christensen2014Rahmati2014Krogager2013Fynbo2013Fynbo2011. The data for nearby galaxies (LMC, SMC, IZw 18, Sextans A, NGC 3109) are taken from DeCia2024 and/or Chruslinska2024. All the abundances are converted to the Asplund2021 solar scale.
  • Figure 4: H I column densities vs. impact parameter at $z>1.9$. Solid curves and 1$\sigma$ confidence area correspond to the GEAR simulations Revaz2018Roca-Fabrega2021 at $z=2.55$ averaged over 18 viewing angles. Blue diamonds are high-$z$ observations from Christensen2014Moller2020 and compilations of observed data by Rahmati2014Krogager2020. The highlighted symbols show the objects belonging to the golden sample from Velichko2024. Dashed curves are the data from hydrodynamical simulations performed with the use of the GADGET-3 code by Rahmati2014 in the stellar mass ranges of $7.0<$ log$(M_\star/M_\odot)<8.5$ (orange) and $10.0<$ log$(M_\star/M_\odot)<11.5$ (green). The horizontal colored stripes indicate intervals of the H I column density corresponding to DLA, sub-DLA and Ly-limit systems (see text for details).
  • Figure 5: H I column densities vs. impact parameter at $z\sim0$. Gray curve: the GEAR galaxy h000 Roca-Fabrega2021 taken at $z = 0.33$. Blue curve: the median 2D radial profile of log N(H I) for an AURIGA galaxy at $z=0-0.3$ with stellar mass of log$M_\star = 10.7 M_\odot$ resimulated by vandeVoort2019. Green curve: cumulative covering fractions of H I from the TNG50 simulations of a galaxy with log$M_\star = 11.8 M_\odot$Nelson2020. The low-$z$ observations taken from Christensen2014Moller2020Kulkarni2022Weng2023Berg2023 are shown by blue diamonds.
  • ...and 10 more figures