Introducing the THESAN-ZOOM project: radiation-hydrodynamic simulations of high-redshift galaxies with a multi-phase interstellar medium

TL;DR

THESAN-ZOOM presents a suite of 60 high-resolution radiation-hydrodynamic zoom-in simulations of 14 high-redshift galaxies, resolving a wide range of halo masses ($M_ ext{halo}\sim10^8-10^{13}\,M_\odot$) and the multi-phase ISM. The simulations couple AREPO-RT-based radiative transfer with a non-equilibrium thermochemical network, detailed dust physics, and comprehensive stellar feedback, including an empirically motivated Early Stellar Feedback channel, within patchy reionization fields inherited from the parent THESAN run. They reproduce key observables such as the stellar-to-halo mass relation, the star-forming main sequence, the Kennicutt-Schmidt relation, metal and dust enrichment, and the star-formation rate density from $z=3$ to $14$, while identifying tensions at the low-mass end and the need for empirical dust attenuation at certain epochs. The patchy external radiation field significantly alters the low-density, cold gas content, and the results provide a realistic framework to interpret JWST data and to study the interplay between radiation, dust, and star formation in the early Universe. Future work will incorporate additional physics (e.g., black hole feedback) and extend public data releases to enable broader community use.

Abstract

We introduce the THESAN-ZOOM project, a comprehensive suite of high-resolution zoom-in simulations of $14$ high-redshift ($z>3$) galaxies selected from the THESAN simulation volume. This sample encompasses a diverse range of halo masses, with $M_\mathrm{halo} \approx 10^8 - 10^{13}~\mathrm{M}_\odot$ at $z=3$. At the highest-resolution, the simulations achieve a baryonic mass of $142~\mathrm{M}_\odot$ and a gravitational softening length of $17~\mathrm{cpc}$. We employ a state-of-the-art multi-phase interstellar medium (ISM) model that self-consistently includes stellar feedback, radiation fields, dust physics, and low-temperature cooling through a non-equilibrium thermochemical network. Our unique framework incorporates the impact of patchy reionization by adopting the large-scale radiation field topology from the parent THESAN simulation box rather than assuming a spatially uniform UV background. In total, THESAN-ZOOM comprises $60$ simulations, including both fiducial runs and complementary variations designed to investigate the impact of numerical and physical parameters on galaxy properties. The fiducial simulation set reproduces a wealth of high-redshift observational data such as the stellar-to-halo-mass relation, the star-forming main sequence, the Kennicutt-Schmidt relation, and the mass-metallicity relation. While our simulations slightly overestimate the abundance of low-mass and low-luminosity galaxies they agree well with observed stellar and UV luminosity functions at the higher mass end. Moreover, the star-formation rate density closely matches the observational estimates from $z=3-14$. These results indicate that the simulations effectively reproduce many of the essential characteristics of high-redshift galaxies, providing a realistic framework to interpret the exciting new observations from JWST.

Figures (16)

• Figure 1: Treatment of the boundary conditions for the radiation field at the edge of the high-resolution region (marked by the blue contour) in the $\mathrm{m}10.4\_4\mathrm{x}$ simulation at $z=9.9$ (left two panels) and $z=7.9$ (right two panels). The maps show the H i fraction and the log of the H i photoionization rate ($\Gamma_\text{H\,i}$) as indicated. The white arrows in the H i photoionization rate plot represent the effective net velocity of the radiation field (photon flux divided by photon density). The plot clearly shows that the high-intensity external ionizing source present in the thesan-1 simulation, shown on the left of the zoom-in region, reionizes part of the targeted high-resolution region, outside-in, at early times ($z=9.9$), before the local radiation field takes over and ionizes the majority of the region by $z=7.9$.
• Figure 2: A qualitative illustration of the various physical quantities predicted by the thesan-zoom simulations at $z=3.2$. The central figure shows the large-scale ($2~\mathrm{cMpc}$) gas density distribution around the target galaxy in the m$12.6\_4\mathrm{x}$ simulation, with the white circle denoting its virial radius. The smaller figures showcase a smaller region of $100~\mathrm{ckpc}$ around the target galaxy, and the circles indicate the half-mass radius of stars. Clockwise from bottom-left, we plot the H i fraction, H $_2$ fraction, gas phase metallicity, dust-to-gas-ratio (DGR), dust temperature distribution, gas temperature distribution, the LyC photon density (combining all three LyC bands) and a mock JWST image generated from the F277W, F356W, F444W bands. Except for the last image, all plots show the mass-weighted averages along the line of sight, which extends over the same length as the other two dimensions of the figure.
• Figure 3: The stellar mass (within twice the stellar half mass radius)--halo mass (within $R_{200c}$) relation for the galaxies in the thesan-zoom simulations. The colored symbols show this relation for the target galaxies, while the different marker styles indicate different resolution levels: $4$x, $8$x, and $16$x, indicated by circles, squares, and inverted triangles, respectively. The dark gray points are the centrals, and the lighter gray points show this relation for the satellite population. For comparison, we also show estimates from abundance matching techniques like UNIVERSEMACHINE Behroozi2019 and EMERGE Moster2018 and the empirical model by Tacchella2018. Finally, the solid black line plots the SHMR for the galaxies in the parent thesan-1KannanThesan simulation. The stellar masses of thesan-zoom galaxies align with predictions from abundance matching results within the mass ($M_{200c} \gtrsim 10^{10}~\mathrm{M}_\odot$) and redshift ($z<10$) range they overlap.
• Figure 4: Left Panel: The star-formation rate (SFR) -- stellar mass relation for all the central galaxies in the thesan-zoom simulations at $z=3-15$ in steps of $\Delta z=2$. The solid colored lines show the SFR averaged over $100~\mathrm{Myr} ~ (\mathrm{SFR_{100}})$ and the shaded region is the one sigma distribution around the mean. The simulated star-forming main sequence shows a much stronger evolution compared to the fits to the observed main sequence (colored dashed lines) outlined in Popesso2023. Right Panel: The mean ratio of the SFRs averaged over $10~\mathrm{Myr}$ and $100~\mathrm{Myr}$ as a function of the stellar mass of the galaxy and redshift. The lines are shifted by $(z-3)/2$ for clarity. The horizontal dashed lines mark the zero point of the shifted values.
• Figure 5: The star-formation rate surface density plotted as a function of the total neutral and molecular gas surface density for the target galaxies, at $z=3$ (red points). The observational estimates for the Milky-Way and other nearby galaxies Kennicutt2012 are plotted as black crosses while the blue and orange lines indicate the relation derived from observations of local spirals and star-bursting galaxies Kennicutt2021. The gas depletion times are $\tau_\mathrm{dep}\gtrsim 1~\mathrm{Gyr}$, in low gas surface density environments ($\Sigma_\mathrm{HI+H_2} \lesssim 300~\text{M}_{\astrosun}~\mathrm{pc}^{-2}$), but decreases to about $0.1~\mathrm{Gyr}$ at higher surface densities.