Table of Contents
Fetching ...

Galaxy Metallicity Gradients in the Reionization Epoch from the FIRE-2 Simulations

Xunda Sun, Xin Wang, Fangzhou Jiang, Houjun Mo, Luis C. Ho, Qianqiao Zhou, Xiangcheng Ma, Hu Zhan, Andrew Wetzel, Russell L. Graf, Philip F. Hopkins, Dusan Keres, Jonathan Stern

TL;DR

This study uses the high-resolution FIRE-2 simulations to quantify gas-phase metallicity gradients in $22$ reionization-era galaxies from $z\sim10$ to $z\sim5$, measuring gradients within $0.25-1\,R_{90}$ and relating them to stellar mass, star formation, and gas kinematics. Gradients are generally negative at high redshift and progressively flatten with time, with a typical slope near $-0.1$ dex per kiloparsec by $z\sim6$, and significant scatter driven by bursty central enrichment and limited mixing. A strong link emerges between gradient steepness and gas-flow strength, as captured by $\Delta v/2\sigma$, while global velocity dispersion and rotational support show weak direct correlations; localized central star formation also fosters steep gradients. The results support an evolutionary picture in which early, bursty star formation yields strong, inhomogeneous chemical enrichment that, aided by feedback-driven large-scale flows, gradually smooths into flatter gradients by cosmic noon, aligning with JWST observations and informing models of early galaxy evolution.

Abstract

We employ the high-redshift suite of FIRE-2 cosmological hydrodynamic zoom-in simulations to investigate the evolution of gas-phase metallicity radial gradients in galaxies in the epoch of reionization (EoR). Our sample consists of 22 galaxies spanning the redshift range $z \sim 10-5$. We find that galaxies at $z\sim10$ exhibit a median metallicity gradient of $-0.15\,\mathrm{dex\cdot kpc^{-1}}$ with substantial scatter, which gradually flatten to $-0.1\,\mathrm{dex\cdot kpc^{-1}}$ at $z\sim6$, accompanied by a reduction in scatter. In the EoR, metallicity gradients correlate positively with stellar mass: more massive galaxies display flatter gradients with smaller scatter, broadly consistent with recent JWST observations. At fixed stellar mass, galaxies with higher star formation rates (SFRs) exhibit steeper negative gradients, while sSFR shows a strong anti-correlation with gradient slope. Because EoR galaxies in FIRE-2 generally lack significant rotational support, we adopt the ratio of peak-to-peak velocity shear to twice the velocity dispersion ($Δv/2σ$) as a proxy for the strength of gas flows. We find a strong positive correlation between metallicity gradients and $Δv/2σ$: galaxies with lower $Δv/2σ$ (i.e., weaker gas flows) tend to exhibit steeper negative gradients. Furthermore, galaxies with steeper gradients display higher central SFR surface densities, suggesting localized star formation with inefficient interstellar medium mixing that drives inside-out chemical enrichment in galaxy evolution in the early Universe.

Galaxy Metallicity Gradients in the Reionization Epoch from the FIRE-2 Simulations

TL;DR

This study uses the high-resolution FIRE-2 simulations to quantify gas-phase metallicity gradients in reionization-era galaxies from to , measuring gradients within and relating them to stellar mass, star formation, and gas kinematics. Gradients are generally negative at high redshift and progressively flatten with time, with a typical slope near dex per kiloparsec by , and significant scatter driven by bursty central enrichment and limited mixing. A strong link emerges between gradient steepness and gas-flow strength, as captured by , while global velocity dispersion and rotational support show weak direct correlations; localized central star formation also fosters steep gradients. The results support an evolutionary picture in which early, bursty star formation yields strong, inhomogeneous chemical enrichment that, aided by feedback-driven large-scale flows, gradually smooths into flatter gradients by cosmic noon, aligning with JWST observations and informing models of early galaxy evolution.

Abstract

We employ the high-redshift suite of FIRE-2 cosmological hydrodynamic zoom-in simulations to investigate the evolution of gas-phase metallicity radial gradients in galaxies in the epoch of reionization (EoR). Our sample consists of 22 galaxies spanning the redshift range . We find that galaxies at exhibit a median metallicity gradient of with substantial scatter, which gradually flatten to at , accompanied by a reduction in scatter. In the EoR, metallicity gradients correlate positively with stellar mass: more massive galaxies display flatter gradients with smaller scatter, broadly consistent with recent JWST observations. At fixed stellar mass, galaxies with higher star formation rates (SFRs) exhibit steeper negative gradients, while sSFR shows a strong anti-correlation with gradient slope. Because EoR galaxies in FIRE-2 generally lack significant rotational support, we adopt the ratio of peak-to-peak velocity shear to twice the velocity dispersion () as a proxy for the strength of gas flows. We find a strong positive correlation between metallicity gradients and : galaxies with lower (i.e., weaker gas flows) tend to exhibit steeper negative gradients. Furthermore, galaxies with steeper gradients display higher central SFR surface densities, suggesting localized star formation with inefficient interstellar medium mixing that drives inside-out chemical enrichment in galaxy evolution in the early Universe.

Paper Structure

This paper contains 14 sections, 3 equations, 11 figures, 1 table.

Figures (11)

  • Figure 1: Example galaxy (z5m12d) from our sample. Left: For each sample, the left panel shows the face-on gas density map, the middle panel displays the stellar density map, and the right presents the gas-phase metallicity map. Here the white dashed circle indicates the radius $R_{90}$, within which we measure the metallicity gradient, $5.8\,\mathrm{kpc}$ at $z=6.2$, $5.1\,\mathrm{kpc}$ at $z=6$, $4.6\,\mathrm{kpc}$ at $z=5.9$. Right: Metallicity gradients of the same sample on the left. The mean values in each bin are shown as black circles, with black error bars representing the corresponding $1-\sigma$ uncertainties. The galaxy exhibits different gradients at different redshift: positive gradient at $z=6.2$, steep negative gradient at $z=6$, flat gradient at $z=5.9$. The evolution from $z=6.2$ to $z=6.0$ shows a rapid rise in central metallicity (with $z=6.1$ representing an intermediate stage of this enrichment not shown here). This localized central enhancement, driven by localized star formation, produces a very steep negative gradient. By $z=5.9$, feedback becomes effective, driving inside-out enrichment, leading to metal enrichment in the outer regions, which results in a flatter overall gradient.
  • Figure 2: The cosmic evolution of metallicity gradients. In general, the FIRE-2 simulations predict that galaxy metallicity gradients flatten from $\sim-0.15~\mathrm{dex\cdot kpc^{-1}}$ in the reionization epoch (i.e. $z>5$) to $\sim$0 (flat radial gradient) during the cosmic noon (i.e. $z\sim2$), broadly consistent with other simulation results Garcia_EAGLE_TNG. The red line and shaded region indicate the median and 1-$\sigma$ spread of our measurements. We also overlay observational results currently available from JWST data. Here the purple points are from Vallini2024_z7 and Arribas2024_z7 at $z\sim7$, orange ones from Venturi2024_z6-8 at $z\sim6-8$, brown ones from Li2025 and cyan ones from Tripodi2024_z4-10. The dark yellow line are result of Acharyya2024_FOGGIE. The color and symbol scheme is used consistently throughout the following figures. Both simulations and observations suggest that as galaxies evolve, their metallicity gradients tend to flatten progressively over time.
  • Figure 3: Metallicity gradient versus stellar mass. The red line indicates the best-fit linear relation. We find a positive correlation between gas-phase metallicity gradient and stellar mass, consistent with recent observational results Vallini2024_z7Venturi2024_z6-8Li2025.
  • Figure 4: Metallicity gradient versus redshift at different stellar masses. Galaxies are divided into three stellar mass bins: $M_\star < 10^6\,\mathrm{M}_\odot$, $10^6\,\mathrm{M}_\odot < M_\star < 10^8\,\mathrm{M}_\odot$, and $M\star > 10^8\,\mathrm{M}_\odot$. This binning allows us to examine how the redshift evolution of gas-phase metallicity gradients depends on galaxy stellar mass. Massive galaxies tend to show a more gradual evolution in their metallicity gradients with redshift, in contrast to low-mass galaxies, which display more pronounced and rapid changes over time.
  • Figure 5: Left: Metallicity gradient versus SFR. SFR are measured as the young stars over the past $50\,\mathrm{Myr}$. The color of each point shows the stellar mass of galaxy. The error-bar shows the means and $1-\sigma$ of all samples. We also divide these galaxies into different stellar mass bins as linear fits in different color $M_\star <10^6\,\mathrm{M_\odot}$, $10^6\,\mathrm{M_\odot} <M_\star <10^8\,\mathrm{M_\odot}$ and $M_\star >10^8\,\mathrm{M_\odot}$. Across different mass bins, the metallicity gradient shows a similar trend, becoming steeper as galaxies become more actively star-forming. Right: Metallicity gradient versus sSFR. Here the red line show the linear fit. The metallicity gradients exhibit a strong negative correlation with sSFR, which is completely opposite to the trend during the cosmic noon epoch Sun2024_z04_3.
  • ...and 6 more figures