5. HIGH-REDSHIFT DWARF GALAXIES

The first galaxies are generally defined as halos that can undergo atomic line cooling, are metal-enriched, and can host sustained star formation [17]. Here I present some of the highlights of our latest numerical work on the formation of the first galaxies [72]. These radiation hydrodynamics AMR simulations tracked the formation and feedback of over 300 Population III stars and the buildup of 38 low-mass galaxies in a 1 comoving Mpc³ volume until z = 7. The cosmic Population III star formation rate (SFR) is nearly constant at 3 × 10^-5 M yr^-1 Mpc^-3 from z = 15 to z = 7. The largest galaxy has a final total and stellar mass of 1.0 × 10⁹ M and 2.1 × 10⁶ M, respectively. Galaxies above 10⁸ M generally have a mass-to-light ratio between 5 and 30, whereas the very low-mass galaxies have mass-to-light ratios between 100 and 3000 because of their inability to efficiently form stars.

The evolution of the density, temperature, and metallicity of the entire volume is shown in Figure 3. At z = 7, 76% of the volume is ionized, and 6.5% (1.9%) of the mass (volume) is enriched above 10^-3 Z. We focused on the buildup of the largest galaxy and an isolated dwarf galaxy with a total mass of 10⁸ M. Figure 4 shows the metallicity of the star formation history and metallicity distribution functions in both halos. The mass resolution of this simulation captures the formation of all star-forming minihalos with M > 10⁵ M.

Figure 3. Evolution of the entire simulation volume (Lbox = 1 Mpc) at redshifts 15, 12, 10, 8, and 7 (left to right) that follows the formation of 38 dwarf galaxies and over 300 Population III stars. Pictured here are the density-weighted projections of density (top), temperature (middle), and metallicity (bottom). Note how the stellar radiative feedback from low-mass galaxies reionize the majority of the volume. The metallicity projections are a composite image of metals originating from Pop II (red) and III (blue) stars with magenta indicating a mixture of the two. From [72].

The smaller galaxy experiences rapid mass accretion until z ~ 12 and afterward it evolves in relative isolation. It begins forming metal-enriched stars after a nearby pair-instability SN enriches a nearby halo to ~ 10³ Z. This may be a peculiar case at high redshift, where a halo is enriched from a neighboring halo and does not form any Population III stars itself. It begins to form stars in a bursts at a rate of 5 × 10^-4 M yr^-1 Mpc^-3, peaking at 2 × 10^-3 M yr^-1 Mpc^-3 at z = 10. The galaxy is self-enriched by these stars, gradually increasing from 10^-3 Z to 10^-2 Z by z = 10. Afterward there is an equilibrium between metal-rich outflows and metal-poor accretion from the filaments, illustrated by the plateau in stellar metallicities in Figure 4.

Figure 4. The scatter plots show the metal-enriched (Pop II) star formation history of a 109 M (left) and a 108 M (right) halos as a function of total metallicity, i.e. the sum of metal ejecta from both Pop II and Pop III SNe, at z = 7. Each circle represents a star cluster, whose area is proportional to its mass. The open circles in the upper right represent 103 and 104 M star clusters. The upper histogram shows the SFR. The right histogram depicts the stellar metallicity distribution. The larger halo shows a large spread in metallicity at z > 10 because these stars formed in progenitor halos that were enriched by different SN explosions. At z < 10, the majority of stellar metallicities increase as the halo is self-enriched. The spikes in metallicity at t = 620, 650, and 700 Myr show induced star formation with enhanced metallicities in SN remnant shells. The dashed lines in the left panel guide the eye to two stellar populations that were formed in two satellite halos, merging at z = 7.5. The smaller halo evolves in relative isolation and steadily increases its metallicity to [Z/H] ~ -2 until there is an equilibrium between in-situ star formation and metal-poor inflows from filaments. From [72].

The larger galaxy forms in a biased region of 50 comoving kpc on a side with ~ 25 halos with M ~ 10⁶ M at z = 10. About half of these halos form Population III stars with a third producing pair-instability SNe, enriching the region to 10^-3 Z, the metallicity floor that has been extensively studied in previous works. However, the metal-rich ejecta does not fully escape from the biased region, and most of it falls back into the galaxies or surrounding IGM, leaving the voids pristine. After z = 10, these ~ 25 halos hierarchically merge to form a 10⁹ M halo at z = 7 with two major mergers at z = 10 and z = 7.9. At late times, this galaxy grows mainly through mergers with halos above the filtering mass [26, 25, 70], i.e. gas-rich halos that are not photo-evaporated, and the gas fraction increases from 0.08 to 0.15 over the last 200 Myr of the simulation. The left panel of Figure 4 shows a large scatter in metallicity at early times, which is caused by inhomogeneous metal enrichment of its progenitors. Once it hosts sustained star formation after z = 10, the metallicity trends upwards as the stars enriches its host galaxy. In contrast with the smaller halo, the larger galaxy undergoes a few mergers with halos with an established stellar population. This creates a superposition of age-metallicity tracks in the star formation history.

This simulation of the early stages of galaxy formation only covered a handful of galaxies and did not explore the differing galaxy populations. However, it has given us a clear picture of the inner workings of these galaxies and the important physical processes involved in shaping the first galaxies and their connections to the first stars. We hope to improve on this work to survey a larger galaxy population and focus on larger galaxies that the Hubble Space Telescope has already observed and the James Webb Space Telescope will observe at z > 6.