Introducing the Phoebos simulation: galaxy properties at the dawn of galaxy formation

Understanding the formation and evolution of galaxies from the Big Bang to the present day is one of the most important questions in modern astronomy. The tremendous amount of observational data accumulated in the past decade that probe various properties of galaxies across cosmic time demand a more detailed theoretical understanding of galaxy formation and evolution. In this thesis, I will investigate several open question in this field using state-of-the-art cosmological hydrodynamic zoom-in simulations of galaxy formation from the Feedback in Realistic Environments (FIRE) suite. These high-resolution simulations (10-10 4 M ⊙ , 0.1-10pc) include realistic models of the multi-phase ISM, star formation, and stellar feedback and explicitly capture gas cooling down to 10 K, star formation in dense clumps in giant molecular clouds, and feedback coupling on the smallest resolved scales. These simulations are powerful tools for studying the key physics governing galaxy formation and evolution and understanding the detailed observations of galaxy properties. The first half of this thesis presents three studies on galactic chemical evolution. Chapter 2 focuses on the origin and evolution of the galaxy mass-metallicity relation (MZR), one of the fundamental properties of galaxies. I will show that the FIRE simulations broadly agree with the observed galaxy MZR from z = 0-3. The slope of the MZR is mainly driven by the metal retention fraction in low-mass galaxies, while the amount of redshift evolution of the MZR is mostly determined by the star formation histories of galaxies. Chapter 3 attempts to understanding the diversity of gas-phase metallicity gradients found in intermediate-redshift ( z ~ 0.6-3) galaxies. I will show that the metallicity gradient in a galaxy varies on small timescales driven by bursty star formation and feedback cycle at early times, naturally resulting in the observed diversity of metallicity gradients in z ~ 2 galaxies. The metallicity gradient only reflects the instantaneous dynamics of a galaxy. Chapter 4 will study the structure, stellar age and metallicity gradients, and formation history of Milky Way (MW)-like disk galaxies. At high redshift, star formation happens in a chaotic, bursty mode, which eventually forms a nearly spherical structure by z = 0. Since z ≾ 1, a stable gas disk emerged and stars formed in that disk thereafter. The thickness of the gas disk decreases with time due to lowering gas fraction. Stars formed earlier in this disk are kinematically heated to a thicker, flaring disk. Such a formation history leads to the age and stellar metallicity gradients consistent with what observed in the MW disk. The second half of this thesis focuses on galaxy formation in the first billion years of the Universe, known as the reionization era. Chapters 5 and 6 study the escape fraction of ionizing photons from galaxies at z ≥ 5, which is an important, yet poorly constrained parameter for understanding the reionization history. Most ionizing photons are emitted by the youngest stellar populations in the galaxy, which are usually embedded in their 'birth clouds'. Stellar feedback is required to clear these clouds in a few Myr before ionizing photons are allowed escape. In the meanwhile, the ionizing photon budget decreases rapidly as the most massive stars start to die. The competition of timescales between feedback and stellar evolution is thus the most important physics determines f esc . I will show that canonical single-star stellar population models such as STARBURST99 generally yield a f esc far below what is required for cosmic reionization. Binary models, in contrast, produce more ionizing photons at late times than single-star models and thus lead to a much higher f esc . Chapter 7 presents a new suite of high-resolution cosmological zoom-in simulations of z ≥ 5 galaxies that contains thousands of halos at any time in all zoom-in regions. I will present the stellar mass-halo mass relation, SFR-M halo relation, stellar mass-magnitude relation, stellar mass functions, and multi-band luminosity functions at z = 5-12. These prediction agree well with current observational constraints and can be further tested by future observations with the James Webb Space Telescope . Using these new simulations, Chapter 8 studies the morphology and size evolution of galaxies at z ≥ 5. I will show that the rest-frame UV light from z ≥ 5 galaxies is usually dominated by one or several star-forming clumps that are intrinsically bright and small. Current observations with moderate surface brightness limits tend to only pick up the intrinsically small galaxies or individual clumps but miss the diffuse light in the galaxies. Such a selection effect is likely to result in the extremely small sizes claimed for the faint galaxies in the Hubble Frontier Fields.

In the local Universe, star formation is typically inefficient both globally and when considered as the fraction of gas converted into stars per local free-fall time. An important exception to this inefficiency is regions of high gravitational accelerations g, or equivalently surface densities $\Sigma = g/(\pi \, G)$, where stellar feedback is insufficient to overcome the self-gravity of dense gas clouds. In this paper, I explore whether dark matter can play an analogous role in providing the requisite accelerations on the scale of entire galaxies in the early cosmos. The key insight is that characteristic accelerations in dark matter haloes scale as $(1+z)^2$ at fixed halo mass. I show this is sufficient to make dark matter the source of intense accelerations that might induce efficient star formation on galactic scales at cosmic dawn in sufficiently massive haloes. The mass characterizing this regime scales as $(1+z)^{-6}$ and corresponds to a relatively constant comoving number density of $n(>\!M_{\rm {vir}}) \approx 10^{-4}\, {\rm Mpc}^{-3}$ at $z \gtrsim 8$. For somewhat rarer haloes, this model predicts stellar masses of $M_{\star }\sim 10^{9}\, {\rm M}_{\odot }$ can form in regions that end up with sizes $\mathcal {O}(100\, {\rm pc})$ over $40\, {\rm Myr}$ time-scales at $z\approx 12-14$; these numbers compare well to measurements for some of the brightest galaxies at that epoch from JWST observations. Dark matter and standard cosmological evolution may therefore be crucial for explaining the surprisingly high levels of star formation in the early Universe revealed by JWST.

Dwarf galaxies with stellar masses around 109M⊙ can be explored at high and low redshifts and they give a glimpse of the different conditions of galaxy formation at different epochs. Using a large sample of about 300 zoom-in cosmological hydrodynamical simulations of galaxy formation I will briefly describe the formation of dwarfs at this mass scale at 3 different epochs: cosmic dawn (Ceverino, Klessen, Glover 2018), cosmic noon (Ceverino, Primack, Dekel 2015), and today (Ceverino et al. 2017). I will describe the FirstLight simulations of first galaxies at redshifts 5-15. These first dwarfs have extremely high star formation efficiencies due to high gas fractions and high gas accretion rates. These simulations will make predictions that will be tested for the first time with the James Webb Space Telescope (JWST). At cosmic noon, z = 2, galaxy formation is still a very violent and dynamic process. The VELA simulations have generated a set of dispersion-dominated dwarfs that show an elongated morphology due to their prolate dark-matter halos. Between z = 1 and 0, the AGORA simulation shows the formation of a low-mass disc due to slow gas accretion. The disc agrees with many local scaling relations, such as the stellar-mass-halo-mass and the baryonic Tully-Fisher relation.

The James Webb Space Telescope (JWST) will revolutionize our understanding of early galaxy formation, and could potentially set stringent constraints on the nature of dark matter. We use a semi-empirical model of galaxy formation to investigate the extent to which uncertainties in the implementation of baryonic physics may be degenerate with the predictions of two different models of dark matter – cold dark matter (CDM) and a 7 keV sterile neutrino, which behaves as warm dark matter (WDM). Our models are calibrated to the observed UV luminosity function at z = 4 using two separate dust attenuation prescriptions, which manifest as high and low star formation efficiency in low-mass haloes. These efficiencies capture the net effect of processes that regulate star formation. We find that while at fixed star formation efficiency, ε, there are marked differences in the abundance of faint galaxies in the two dark matter models at high-z; these differences are mimicked easily by varying ε in the same dark matter model. We find that a high ε WDM and a low ε CDM model – which provide equally good fits to the z = 4 UV luminosity function – exhibit nearly identical evolution in the cosmic stellar mass and star formation rate densities. We show that differences in the star formation rate at fixed stellar mass are larger for variations in ε in a given dark matter model than they are between dark matter models; however, the scatter in star formation rates is larger between the two models than they are when varying ε. Our results suggest that JWST will likely be more informative in constraining baryonic processes operating in high-z galaxies than it will be in constraining the nature of dark matter.

Aims The galaxy population shows a characteristic bimodal distribution based on the star formation activity and is sorted into star-forming or quiescent. These two subpopulations have a tendency to be located in different mass halos. The circumgalactic medium (CGM), as the gas repository for star formation, might contain the answer to the mystery of the formation of such bimodality. Here we consider the bimodality of the galaxy population and study the difference between the properties of the hot CGM around star-forming and quiescent galaxies. Methods. We used the X-ray data from the first four SRG/eROSITA all-sky surveys (eRASS:4). We selected central star-forming and quiescent galaxies from the Sloan Digital Sky Survey DR7 with stellar mass 10.0 < log(M*/M⊙) < 11.5 or halo mass 11.5 < log(M200 m/M⊙) < 14.0 within spectroscopic redshift zspec < 0.2, and we built approximately volume-limited galaxy samples. We stacked the X-ray emission around star-forming and quiescent galaxies, respectively. We masked detected point sources and carefully modeled the X-ray emission from unresolved active galaxy nuclei (AGN) and X-ray binaries (XRB) to detect the X-ray emission from the hot CGM. We measured the X-ray surface brightness (SX, CGM) profiles and integrated the X-ray emission from hot CGM within R500c (LX, CGM) to provide the scaling relations between LX, CGM and galaxies’ stellar or halo mass. Results. We detect extended X-ray emission from the hot CGM around star-forming galaxies with log(M*/M⊙) > 11.0 and quiescent galaxies with log(M*/M⊙) > 10.5, extending out to R500c. The SX, CGM profile of quiescent galaxies follows a β model with β ≈ 0.4, where β quantifies the slope of the profile. Star-forming galaxies with median stellar masses log(M*, med/M⊙) = 10.7, 11.1, 11.3 have LX, CGM ≈ 0.8, 2.3, 4.0 × 1040 erg/s, while for quiescent galaxies with log(M*, med/M⊙) = 10.8, 11.1, 11.4, LX, CGM ≈ 1.1, 6.2, 30 × 1040 erg/s. Notably, quiescent galaxies with log(M*, med/M⊙) > 11.0 exhibit brighter hot CGM than their star-forming counterparts. In halo mass bins, we detect similar X-ray emission around star-forming and quiescent galaxies with log(M200 m/M⊙) > 12.5, suggesting that galaxies in the same mass dark matter halos host equally bright hot CGM. We emphasize that the observed LX, CGM − M500c relations of star-forming and quiescent galaxies are sensitive to the stellar-to-halo mass relation (SHMR). A comparison with cosmological hydrodynamical simulations (EAGLE, TNG100, and SIMBA) reveals varying degrees of agreement, contingent on the simulation and the specific stellar or halo mass ranges considered. Conclusions. Either selected in stellar mass or halo mass, the star-forming galaxies do not host brighter stacked X-ray emission from the hot CGM than their quiescent counterparts at the same mass range. The result provides useful constraints on the extent of feedback’s impacts as a mechanism for quenching star formation as implemented in current cosmological simulations.

We present a sample of four emission-line galaxies at z = 6.11–6.35 that were serendipitously discovered using the commissioning data for the James Webb Space Telescope (JWST)/NIRCam wide-field slitless spectroscopy mode. One of them (at z = 6.11) has been reported previously, while the others are new discoveries. These sources are selected by the secure detections of both [O iii] λ5007 and Hα lines with other fainter lines, which were tentatively detected in some cases (e.g., [O ii] λ3727, [O iii] λ4959). In the [O iii]/Hβ–[N ii]/Hα Baldwin–Phillips–Terlevich diagram, these galaxies occupy the same parameter space as that of z ∼ 2 star-forming galaxies, indicating that they have been enriched rapidly to subsolar metallicities (∼0.4 Z ⊙), similar to galaxies with comparable stellar masses at much lower redshifts. The detection of strong Hα lines suggests a higher ionizing photon production efficiency within galaxies in the early universe. We find brightening of the [O iii] λ5007 line-luminosity function (LF) from z = 3 to 6, and weak or no redshift evolution of the Hα line LF from z = 2 to 6. Both LFs are underpredicted at z ∼ 6 by a factor of ∼10 in certain cosmological simulations. This further indicates a global Lyα photon escape fraction of 7%–10% at z ∼ 6, which is slightly lower than previous estimates through the comparison of the UV-derived star formation rate density and Lyα luminosity density. Our sample recovers % of z = 6.0–6.6 galaxies in the survey volume with stellar masses greater than 5 × 108 M ⊙, suggesting the ubiquity of strong Hα and [O iii] line emitters in the Epoch of Reionization, which will be further uncovered in the era of JWST.

Luminous infrared starbursts in the early Universe are thought to be the progenitors of massive quiescent galaxies identified at redshifts 2–4. Using the Mid-IRfrared Instrument (MIRI) on board the James Webb Space Telescope (JWST), we present mid-infrared sub-arcsec imaging and spectroscopy of such a starburst: the slightly lensed hyper-luminous infrared system SPT0311-58 at z = 6.9. The MIRI IMager (MIRIM) and Medium Resolution Spectrometer (MRS) observations target the stellar (rest-frame 1.26 μm emission) structure and ionised (Paα and Hα) medium on kpc scales in the system. The MIRI observations are compared with existing ALMA far-infrared continuum and [C II]158μm imaging at a similar angular resolution. Even though the ALMA observations imply very high star formation rates (SFRs) in the eastern (E) and western (W) galaxies of the system, the Hα line is, strikingly, not detected in our MRS observations. This fact, together with the detection of the ionised gas phase in Paα, implies very high internal nebular extinction with lower limits (AV) of 4.2 (E) and 3.9 mag (W) as well as even larger values (5.6 (E) and 10.0 (W)) by spectral energy distribution (SED) fitting analysis. The extinction-corrected Paα lower limits of the SFRs are 383 and 230 M⊙ yr−1 for the E and W galaxies, respectively. This represents 50% of the SFRs derived from the [C II]158 μm line and infrared light for the E galaxy and as low as 6% for the W galaxy. The MIRIM observations reveal a clumpy stellar structure, with each clump having 3–5×109 M⊙ mass in stars, leading to a total stellar mass of 2.0 and 1.5×1010 M⊙ for the E and W galaxies, respectively. The specific star formation (sSFR) in the stellar clumps ranges from 25 to 59 Gyr−1, assuming a star formation with a 50–100 Myr constant rate. This sSFR is three to ten times larger than the values measured in galaxies of similar stellar mass at redshifts 6–8. Thus, SPT0311-58 clearly stands out as a starburst system when compared with typical massive star-forming galaxies at similar high redshifts. The overall gas mass fraction is Mgas/M* ∼ 3, similar to that of z ∼ 4.5–6 star-forming galaxies, suggesting a flattening of the gas mass fraction in massive starbursts up to redshift 7. The kinematics of the ionised gas in the E galaxy agrees with the known [C II] gas kinematics, indicating a physical association between the ionised gas and the cold ionised or neutral gas clumps. The situation in the W galaxy is more complex, as it appears to be a velocity offset by about +700 km s−1 in the Paα relative to the [C II] emitting gas. The nature of this offset and its reality are not fully established and require further investigation. The observed properties of SPT0311-58, such as the clumpy distribution at sub(kpc) scales and the very high average extinction, are similar to those observed in low- and intermediate-z luminous (E galaxy) and ultra-luminous (W galaxy) infrared galaxies, even though SPT0311-58 is observed only ∼800 Myr after the Big Bang. Such massive, heavily obscured clumpy starburst systems as SPT0311-58 likely represent the early phases in the formation of a massive high-redshift bulge, spheroids and/or luminous quasars. This study demonstrates that MIRI and JWST are, for the first time, able to explore the rest-frame near-infrared stellar and ionised gas structure of these galaxies, even during the Epoch of Reionization.

Galaxy evolution is complicated. Throughout their lifetimes, galaxies are subject to an amalgamation of astrophysical and cosmological processes that direct the growth of their stellar masses, the transformation of their morphologies, and the cessation of their star formation. The variable action of these processes begets a diverse population of galaxies, which exhibit a variety of brightnesses, colours, shapes, and sizes, among myriad other features. Many of these features are bimodally distributed, which has led to the general acceptance of a simple empirical paradigm of galaxy evolution. However, connecting this diversity among galaxies with the array of processes that are involved in their evolution, and constraining the relative influences of each of these processes, requires that several features are analysed simultaneously. This has been enabled by the recent advent of machine learning techniques, which are capable of extracting scientifically useful information from complicated, multi-dimensional datasets, to astronomy and astrophysics. Unsupervised machine learning techniques, free from the requirement for pre-labelled training data, are especially well suited to the exploration of the data structures of galaxy samples in multi-dimensional feature spaces. This thesis assesses the use of clustering, an unsupervised machine learning technique, for the research of galaxy evolution. Clustering is first tested on a well-characterised sample of galaxies from the GAMA survey. Galaxies are represented in five dimensions by a set of intrinsic astrophysical features. Use of a unique cluster evaluation framework enables the robust identification of reproducible and astrophysically meaningful clustering structures via the k-means method. Outcomes consisting of two, three, five, and six clusters are deemed stable, and form a hierarchical structure that agrees well with established notions of the galaxy bimodality. The two- and three-cluster outcomes are dominated in their structures by the stellar masses, colours, and star formation activity of galaxies, with Sersic indices and half-light radii becoming important for the five- and six-cluster outcomes. Clusters also exhibit broad correspondence with detailed morphological classifications, and it is suggested that the inclusion of additional morphological features might improve this correspondence further. The five- and six-cluster outcomes indicate the differential role of environment in the evolution of galaxies with intermediate colours. This cluster evaluation framework is then applied for the validation of the cosmological, hydrodynamical EAGLE simulations against the GAMA survey. Outcomes consisting of seven and five clusters respectively, determined using the same five features for both samples, are selected for analysis. These outcomes produce an agreement score of Vₐ = 0.76, indicating broad, overall agreement, but differences in their substructures. These differences include discrepancies in the growth of the central bulges of galaxies along the star-forming main sequence, an over-abundance of low-mass, bulge-dominated, star-forming galaxies in the EAGLE sample, and a subpopulation of high-mass, disc-dominated, star-forming galaxies in the EAGLE sample that is not present in the GAMA sample. These differences are attributed to the resolution of EAGLE, and to an active galactic nucleus feedback prescription that is not sufficiently effective in EAGLE. Finally, clustering is used to compare samples of galaxies at low (z ~ 0.06; GSWLC-2) and intermediate (z ~ 0.67; VIPERS) redshifts, in order to examine the evolution of subpopulations of galaxies. Galaxies are clustered in a nine-dimensional feature space defined by a series of ultraviolet-through-near-infrared colours using the Subspace Expectation-Maximisation algorithm, which includes iterative dimensionality reduction. The algorithm models both samples using seven clusters: four containing mostly star-forming galaxies, and three containing mostly passive galaxies. Both sets of star-forming clusters form clear morphological sequences, capturing the gradual internally-driven growth of galaxy bulges at both epochs. At high stellar masses, this growth is linked with quenching. However, it is only at low redshifts that additional, environmental processes appear to be involved in the evolution of low-mass passive galaxies. The results of this thesis demonstrate the utility of clustering as a method with which to analyse the large galaxy samples that are anticipated from next-generation surveys, and with which to facilitate the multi-dimensional comparison of cosmological galaxy simulations with observations. Clustering is robustly able to identify astrophysically meaningful substructures in complex, multi-dimensional feature spaces, and these substructures may readily be interpreted with respect to the evolutionary contexts of the galaxies that they encompass.

Recent observations with JWST have uncovered unexpectedly high cosmic star formation activity in the early Universe, mere hundreds of millions of years after the big bang. These observations are often understood to reflect an evolutionary shift in star formation efficiency (SFE) caused by changing galactic conditions during these early epochs. We present FIREbox$^{\it HR}$, a high-resolution, cosmological hydrodynamical simulation from the Feedback in Realistic Environments (FIRE) project, which offers insights into the SFE of galaxies during the first billion years of cosmic time. FIREbox$^{\it HR}$ re-simulates the cosmic volume ($L=22.1$ cMpc) of the original FIREbox run with eight times higher mass resolution ($m_{\rm b}\sim {}7800\, M_\odot$), but with identical physics, down to $z\sim {}6$. FIREbox$^{\it HR}$ predicts ultraviolet (UV) luminosity functions in good agreement with available observational data. The simulation also successfully reproduces the observed cosmic UV luminosity density at $z\sim {}6{\!-\!}14$, demonstrating that relatively high star formation activity in the early Universe is a natural outcome of the baryonic processes encoded in the FIRE-2 model. According to FIREbox$^{\it HR}$, the SFE–halo mass relation for intermediate mass haloes ($M_{\rm halo}\sim {}10^9{\!-\!}10^{11}\, {\rm M}_\odot$) does not significantly evolve with redshift and is only weakly mass-dependent. These properties of the SFE–halo mass relation lead to a larger contribution from lower mass haloes at higher z, driving the gradual evolution of the observed cosmic UV luminosity density. A theoretical model based on the SFE–halo mass relation inferred from FIREbox$^{\it HR}$ allows us to explore implications for galaxy evolution. Future observations of UV faint galaxies at $z\gt 12$ will provide an opportunity to further test these predictions and deepen our understanding of star formation during Cosmic Dawn.

Context. With its sensitivity in the rest-frame optical, the James Webb Space Telescope (JWST) has uncovered active galactic nuclei (AGN), which comprise intrinsically faint and heavily reddened sources, well into the first billion years of the Universe, at z ∼ 4 − 11. Aims. We revisit the AGN contribution to reionization given the high number densities associated with these objects. Methods. We used the DELPHI semi-analytic model, which we base-lined against the latest high-redshift datasets from the JWST and the Atacama Large millimetre Array (ALMA) to model early star-forming galaxies and AGN. We calculated the escape fractions of ionizing radiation from star formation and AGN and included the impact of reionization feeback in suppressing the baryonic content of low-mass galaxies in ionized regions. This model was validated against the key observables for star-forming galaxies, AGN, and reionization. Results. In our fiducial model, reionization reaches its mid-point at z ∼ 6.9 and ends by z ∼ 5.9. Low stellar mass (M* ≲ 109 M⊙) star-forming galaxies are found to be the key drivers of the reionization process. They provide about 77% of the total photon budget. Despite their high numbers, high accretion rates, and higher escape fractions than star-forming galaxies at z ∼ 5, AGN only provide about 23% of the total reionization budget, which is dominated by black holes in high stellar mass systems (with M* ≳ 109 M⊙). This is because AGN number densities become relevant only at z ≲ 7, and as a result, AGN contribute as much as galaxies as late as z ∼ 6.2, when reionization is already in its end stages. Finally, we find that even contrasting models of the AGN ionizing photon escape fraction (increasing or decreasing with stellar mass) do not qualitatively change our results.

We present the mid-infrared (MIR) morphologies for 64 star-forming galaxies (SFGs) at 0.2 < z < 2.5 with stellar mass M * > 109 M ⊙ using James Webb Space Telescope (JWST) Mid-Infrared Instrument (MIRI) observations from the Cosmic Evolution Early Release Science survey. The MIRI bands span the MIR (7.7–21 μm), enabling us to measure the effective radii (R eff) and Sérsic indexes of these SFGs at rest-frame 6.2 and 7.7 μm, which contains strong emission from Polycyclic aromatic hydrocarbon (PAH) features, a well-established tracer of star formation in galaxies. We define a “PAH band” as the MIRI bandpass that contains these features at the redshift of the galaxy. We then compare the galaxy morphologies in the PAH bands to those in the rest-frame near-ultraviolet (NUV) using Hubble Space Telescope (HST) Advanced Camera for Surveys (ACS)/F435W or ACS/F606W and optical/near-IR using HST WFC3/F160W imaging from UVCANDELS and CANDELS. The R eff of galaxies in the PAH band are slightly smaller (∼10%) than those in F160W for galaxies with M * ≳ 109.5 M ⊙ at z ≤ 1.2, but the PAH band and F160W have similar fractions of light within 1 kpc. In contrast, the R eff of galaxies in the NUV band are larger, with lower fractions of light within 1 kpc compared to F160W for galaxies at z ≤ 1.2. Using the MIRI data to estimate the SFRIR surface density, we find that the correlation between the SFRIR surface density and stellar mass has a steeper slope than that of the SFRUV surface density and stellar mass, suggesting more massive galaxies having increasing amounts of obscured fraction of star formation in their inner regions. This paper demonstrates how the high-angular resolution data from JWST/MIRI can reveal new information about the morphology of obscured star formation.

Context. Feedback from massive stars plays a crucial role in regulating the growth of young star-forming galaxies (SFGs) and in shaping their interstellar medium (ISM). This feedback contributes to the removal and mixing of metals via galactic outflows and to the clearance of neutral gas, which facilitates the escape of ionizing photons. Aims. Our goal is to study the impact of stellar feedback on the chemical abundances of the ISM in a sample of SFGs with strong emission lines at z ∼ 3. Methods. We selected 35 low-mass SFGs (7.9 &lt; log(M⋆/M⊙) &lt; 10.3) from deep spectroscopic surveys based on their CIII]λ1908 emission. We used new follow-up near-infrared (NIR) observations to examine their rest-optical emission lines and to identify ionized outflow signatures through broad emission line wings detected after Gaussian modeling of [OIII]λλ4959,5007 profiles. We characterized the gas-phase metallicity and carbon-to-oxygen (C/O) abundance of the galaxies using a Te-based method via the OIII]λ1666/[OIII]λ5007 ratio and photoionization models. Results. We find line ratios and rest-frame equivalent widths (EWs) characteristic of high-ionization conditions powered by massive stars. Our sample displays a mean rest-frame EW([OIII]λ5007) of ∼560 Å, while about 15% of the SFGs show EW([OIII]λλ4959,5007) &gt; 1000 Å and EW(CIII]) &gt; 5 Å, closely resembling those now seen in epoch of reionization (EoR) galaxies with the James Webb Space Telescope. We find high Te values, which imply low gas-phase metallicities 12+log(O/H) ∼ 7.5–8.5 (mean of 17% solar) and C/O abundances from 23% to 128% solar, with no apparent increasing trend with metallicity. Our sample follows the mass-metallicity relation at z ∼ 3, with some galaxies showing lower gas-phase metallicities. This results in significant deviations from the fundamental metallicity relation. From our [OIII]λλ4959,5007 line profile modeling, we find that 65% of our sample shows an outflow component, which is found both blue- or redshifted relative to the ionized gas systemic velocity, and the mean maximum velocities are vmax ∼ 280 km s−1. We find a weak correlation between vmax and the star formation rate surface density (ΣSFR) of vmax = (2.41 ± 0.03) × ΣSFR(0.06 ± 0.03). Moreover, we find that the mass-loading factor μ of our galaxy sample is typically lower than in more massive galaxies from the literature, but it is higher than in typical local dwarf galaxies. In the stellar mass range covered by our sample, we find that μ increases with ΣSFR. This suggests that for a given stellar mass, denser starbursts in low-mass galaxies produce stronger outflows. Our results complement the picture drawn by similar studies at lower redshift, suggesting that the removal of ionized gas in low-mass SFGs driven by stellar feedback is regulated by their stellar mass and by the strength and concentration of their star formation, that is, ΣSFR.

The shape of the faint-end of the high-z galaxy luminosity function (LF) informs early star formation and reionization physics during the Cosmic Dawn and Epoch of Reionization. Until recently, based on the strong gravitational lensing cluster deep surveys, the Hubble Frontier Fields (HFF) has found a potential turnover in the ultraviolet (UV) LF at $\mathit{ z}$ ∼ 6. In this paper, we analyse the contribution of extremely faint galaxies with the magnitude larger than the turnover magnitude in LF to cosmic reionization. We apply the measurement from HFF to our suppressed star formation efficiency model, including three free parameters: halo mass threshold Mt, curvature parameter β, and a UV conversion factor lUV. According to our fit of 68 per cent confidence level, the high-redshift star formation in haloes smaller than $M_t=1.82^{+2.86}_{-1.08}\times 10^{10} \, \rm M_{\odot }$ is found to be dampened. The turnover magnitude $\rm \gtrsim -13.99-2.45$, correspondingly the halo mass $\lesssim (4.57+20.03)\times 10^{9} \, \rm M_{\odot }$. We find that the absorption trough in the global 21-cm signal is sensitive to our SFE model parameters. Together with (β, lUV) = ($2.17^{+2.42}_{-1.72}$, $9.33^{+0.43}_{-0.42} \, \rm ~erg~yr ~s^{-1}\, M_{\odot }^{-1})$, the trough locates at ∼$134^{+10}_{-17}$$\rm MHz$ with an amplitude of ∼$-237^{-6}_{+7}$$\rm mK$, compared to (106 MHz, -212 mK) in the absence of turnover. Besides, we find that the star formation of faint galaxies has also an impact on the 21-cm power spectra. The best-fitting peak power decreases by$\sim 4{{\ \rm per\ cent}}$ and shifts towards smaller scales from $0.88 \, h\, \rm Mpc^{-1}$ to $0.91 \, h\, \rm Mpc^{-1}$. According to our calculation, such impact is distinguishable with the forthcoming Square Kilometre Array.

Cosmological hydrodynamical simulations have become an important theoretical tool for understanding the formation and evolution of the first galaxies during cosmic dawn, between redshifts 5 and 15. I will introduce the FirstLight database of about 300 zoom-in simulations with a resolution of 10 parsecs. This database agrees well with observed UV luminosity functions and stellar mass functions. I will discuss the origin and evolution of the star-forming main sequence of galaxies and the main drivers of the star formation histories at these early epochs. I will show simulated SEDs from UV to IR, including stellar and nebular emission. The rest-frame UV spectra show steep slopes and a high production efficiency of Lyman continuum photons. These properties are consistent with young stellar populations with low metallicities. Simulated recombination lines allow us to link the physical conditions of the gas around these stellar populations with observables, like equivalent widths in OIII or Hα or BPT diagrams at high-z. These simulations are making predictions that will be tested for the first time in future deep fields with the James Webb Space Telescope (JWST). I will finally discuss preliminary results involving JWST mock fields and predictions for ALMA observations by post-processing FirstLight snapshots with Powderday radiative transfer code.

In this work, we investigate the implications of the Integrated Galaxy-wide stellar Initial Mass Function (IGIMF) approach in the framework of the semi-analytic model GAEA (GAlaxy Evolution and Assembly), which features a detailed treatment of chemical enrichment and stellar feedback. The IGIMF provides an analytic description of the dependence of the stellar IMF shape on the rate of star formation in galaxies. We find that our model with a universal IMF predicts a rather flat [$\alpha$/Fe]-stellar mass relation. The model assuming the IGIMF, instead, is able to reproduce the observed increase of $\alpha$-enhancement with stellar mass, in agreement with previous studies. This is mainly due to the fact that massive galaxies are characterized by larger star formation rates at high-redshift, leading to stronger $\alpha$-enhancement with respect to low-mass galaxies. At the same time, the IGIMF hypothesis does not affect significantly the trend for shorter star formation timescales for more massive galaxies. We argue that in the IGIMF scenario the [$\alpha$/Fe] ratios are good tracers of the highest star formation events. The final stellar masses and mass-to-light-ratio of our model massive galaxies are larger than those estimated from the synthetic photometry assuming a universal IMF, providing a self-consistent interpretation of similar recent results, based on dynamical analysis of local early type galaxies.