Marvelous Metals: Surveying the Circumgalactic Medium of Simulated Dwarf Galaxies

TL;DR

This work investigates the CGM of 64 isolated dwarf galaxies at z=0 using two high-resolution cosmological zoom suites, focusing on how mass and metals are partitioned among the disk, CGM, and halo. By computing synthetic HI and metal-absorption column densities and tracking metal production, the study finds that roughly half of halo baryons reside in the CGM (within R_200c), with the CGM hosting a warm phase that dominates the mass budget beyond ~0.5R_200c. Metals produced by dwarfs are distributed with ~15–20% retained in the CGM across the mass range, while disk retention and metal loss vary with stellar mass; the CGM remains a substantial metal reservoir even as the halo evolves. Ionization fractions imply HI, C II, and Si II trace inner, cool gas, whereas CIV and O VI trace warmer and more extended CGM, though O VI is underrepresented compared to some observations, hinting at missing warm/hot gas or stronger feedback. The results underscore the CGM, especially the warm component, as a key component of the baryon cycle in low-mass galaxies and provide forecasts for UV-observable mass to guide future observational programs around dwarfs.

Abstract

Dwarf galaxies are uniquely sensitive to energetic feedback processes and are known to experience substantial mass and metal loss from their disk. Here, we investigate the circumgalactic medium (CGM) of 64 isolated dwarf galaxies ($6.0<$log(M$_*/M_{\odot}$)$<9.5$) at $z=0$ from the Marvel-ous Dwarfs and Marvelous Massive Dwarfs hydrodynamic simulations. Our galaxies produce column densities broadly consistent with current observations. We investigate these column densities in the context of mass and metal retention rates and the physical properties of the CGM. We find $48\pm11\%$ of all baryons within $R_{200c}$ reside in the CGM, with $\sim70\%$ of CGM mass existing in a warm gas phase, $10^{4.5}<T<10^{5.5}$ K that dominates beyond $r/R_{200c}\sim0.5$. Further, the warm and cool ($10^{4.0}<T<10^{4.5}$ K) gas phases each retain $5-10\%$ of metals formed by the dwarf galaxy. The significant fraction of mass and metals residing in the warm CGM phase provides an interpretation for the lack of observed low ion detections beyond $b/R_{200c}\sim0.5$ at $z\sim0$, as the majority of mass in this region exists in higher ions. We find a weak correlation between galaxy mass and total CGM metal retention despite the fraction of metals lost from the halo increasing from $\sim10\%$ to $>40\%$ towards lower masses. Our findings highlight the CGM (primarily its warm component) as a key reservoir of mass and metals for dwarf galaxies across stellar masses and underscore its importance in understanding the baryon cycle in the low-mass regime. Finally, we provide individual galaxy properties of our full sample and quantify the fraction of ultraviolet observable mass to support future observational programs, particularly those aimed at performing a metal budget around dwarf galaxies.

Figures (15)

• Figure 1: Top panel: Summary of the virial mass (blue circle), total gas mass in the halo (green square), and stellar mass (yellow star) for each galaxy in the Marvel simulation (open-faced markers) and the Marvelous Massive Dwarfs simulations (filled markers). The galaxies in the Marvelous Massive Dwarfs have slightly greater virial masses for a given virial radius than the Marvel galaxies due to the difference in cosmology (see Section \ref{['sec:Simulations']}). Bottom panel: Histogram of stellar masses for observed galaxies (Zheng_24: solid black line, Mishra_24 dashed grey line) and the histogram of stellar masses for the M+M sample (orange). All simulated stellar masses are 0.6 of the total sum of star particle masses within a halo, as this is a better match to photometric stellar mass determinations 2013ApJ...766...56M. All galaxies included in Zheng_24 and the $z<0.3$ galaxies included in Mishra_24 are shown in this plot.
• Figure 2: Luminosity - Metallicity relation (Fe/H vs $L_V$, left) and galaxy specific star formation rates (sSFR vs $M_*$, right). Red circles represent galaxies from the Marvel suite, and blue diamond markers represent the Marvelous Massive suite; median values of both suites are shown as solid lines. Left panel: Shown is the logarithm of the mean stellar metallicity relative to solar abundance. Error bars on simulated points show the 16th-84th percentile of the simulated stellar metallicity distribution. Open-face pentagon markers are observational data from McConnachie12, and open-face square markers are observational data from the Local Volume Database Pace2024LVDB. Right panel: sSFR values for the M+M sample are calculated over a timescale of 100 Myr. Open-face markers are observations that derive SFRs as follows: Bordoloi_14 uses detected nebular emission lines, and LiangChen_14 uses rest-frame UV absolute magnitude. Karachentsev2013 and Lee2011 values utilize far UV magnitudes and are corrected for Galactic extinction in Zheng_24 following the methods presented in their Appendix. We find adequate agreement between the M+M sample and observations, as well as similar behavior across suites.
• Figure 3: Projected maps of CGM gas density (upper left), gas temperature (lower left), H i column density (upper center), Si ii column density (upper right), C iv column density (lower center), and O vi column density (lower right) using TRIDENTTrident and ytyt. Maps are made for a typical galaxy in the M+M sample (a log$M_*/M_{\odot}=8.82\,$ galaxy from the Marvelous Massive suite). In each panel, we show the $R_{200c}$ as the dashed white line and $0.15R_{200c}$ as the solid black line. The galaxy disk is oriented in the panel such that the disk is edge-on.
• Figure 4: Adapted Figure 4 from Zheng_24 showing column densities (N) for low ions (H i, C ii, and Si iii) as a function of normalized impact parameter ($b/R_{200c}$). The column density results from the M+M sample are shown as 2-D probability density histograms with the brightest cells showing a greater probability at a given $b/R_{200c}$. The median of all simulated columns as a function of $b/R_{200c}$ is shown as a solid line with two shaded regions representing the 16th-84th ($1\sigma$) and 5th-95th percentiles ($2\sigma$). Observed column densities are shown as points where each marker is attributed to a certain observational paper (legend shown at the bottom of the Figure). Open markers are non-detections and are treated as upper limits, while filled markers are detections; filled markers with upward arrows are saturations and are treated as lower limits. Black bars are intervals provided in Mishra_24 that show the observationally constrained range (upper and lower limits) of column densities for saturated systems. Each column shows all simulated and observational data for the CGM around galaxies within a given stellar mass bin (denoted at the top). From left to right, the three stellar mass bins (and the number of simulated galaxies in each bin) are as follows: $M_* = 10^{6.0-7.5} ~M_{{\odot}}$ (N=19), $M_* = 10^{7.5-8.5} ~M_{{\odot}}$ (N=29), and $M_* = 10^{8.5-9.5} ~M_{{\odot}}$ (N=16). We find the M+M sample reproduces the majority of existing column density observations for the low ions included in this work (see also Table \ref{['tab:N_stats']}).
• Figure 5: Same as Figure \ref{['fig:LowIonColumnDensitiesResults']} (adapted Figure 4 from Zheng_24) now showing column densities for higher ions (Si iv, C iv, and O vi). We find the M+M sample reproduces the majority of existing Si iv column density observations, while the simulations tend to underpredict C iv and O vi column densities (see Section \ref{['subsec:Discussion_OVI']} for further discussion).