Category: Thesis

Supervisor: Luca Graziani
co-supervisor: Raffaella Schneider

Fig: Cosmological simulations with our code dustyGadget (Graziani+2020) can be run on large cosmological volumes with comoving size of 40 – 50 h^-1 cMpc (Di Cesare+23), but also on zoom-in simulations centered on selected regions. The simulated galaxies are then post-processed with the photoionization code Cloudy to compute nebular emission lines, and with the radiative transfer code SKIRT to compute dust extinction and continuum emission. An example of a synthetic spectrum of a simulated galaxy at z ≈ 12 obtained assuming a mixture of Pop II stellar binaries and Pop III stars is shown in the rightmost panel, with observed JWST photometric data (pink) and ALMA upper limit (orange). Through this procedure, we will be able to produce synthetic observables for the simulated galaxies at z  ≥  6 – 10 to be compared with the large wealth of JWST data that is being assembled.

The launch of the JWST in December 2021 has enabled an unprecedented epochal leap forward in our understanding of cosmic dawn. As of today – more than 10 galaxies with spectroscopic redshifts z > 10 have been detected, the most distant of which at z ≈ 13.2 (Curtis-Lake+23). This corresponds to a Hubble time of only 320 Myr, less than 3% of the current age of the Universe, and it is expected that the redshift frontier will be pushed to even earlier epochs in the coming months. A swarm of z > 10 – 20 candidates has been photometrically identified, providing the first constraints on the cosmic star formation rate out to z ≈ 16.5 (Harikane+23). The physical properties of these early galaxies in terms of stellar mass, star formation history, chemical enrichment, stellar populations, have revealed unexpected findings, even sparking speculations that these observations might be in tension with the LCDM model. JWST observations provide a new and unique “stress test” for the current generation of models. The conditions in these early galaxies are expected to be much different from those prevailing in the nearby Universe or at intermediate redshifts, that most galaxy formation models have used for calibration and validation up until now. The interstellar gas is likely to have been much denser, at lower metallicity, and more turbulent. The main aim of this thesis project is to investigate the physical processes in the first galaxies, specifically those formed at z > 10 using hydrodynamical simulations run with our dustyGadget code (Graziani+20). These simulations predict star formation, chemical evolution histories, stellar population properties (including Pop III stars, Venditti+23) for a significant sample of galaxies at z > 6 – 10 (Di Cesare+23). Post-processing the simulation outputs with public radiative transfer models (CLOUDY, SKIRT) allows us to generate synthetic galaxy spectra that will be compared to JWST and ALMA observations, in collaboration with the group at INAF/OAR.

References

Curtis-Lake+23, Nat. Ast. 7, 622
Di Cesare+23, MNRAS, 519, 4632
Graziani+20, MNRAS, 494, 1071
Harikane+23, arXiv:2303.11946
Venditti+23, MNRAS, 522, 3809

Supervisor: Raffaella Schneider
co-supervisors: Rosa Valiante, Tommaso Zana

A big question that remains unsolved pertains to the existence of hundreds of quasars at 6 ≤ z ≤ 7.64, which show that 10^9 − 10^10 Msun super massive black holes (SMBHs) can form in ∼ 700 Myr, by efficiently growing mass onto smaller black hole (BH) seeds (see Volonteri+2021; Inayoshi+2020). During the past year the advent of JWST has started to revolutionize this field: several tens of candidate Active Galactic Nuclei (AGNs) at z > 3 have been revealed by deep near-IR photometry, and dozens of AGNs have been identified via their broad emission lines, or as counterparts of X−ray sources. The general common features of these newly discovered AGNs are that these are orders of magnitudes fainter and lighter than quasars, with bolometric luminosities of Lbol ≈ 10^(43.5 – 46) erg/s and black hole masses Mbh ≈ 10^(5.3 – 8) Msun, with the lightest ones that border the mass range for the formation of heavy BH seeds. Intriguingly, the majority of these newly discovered accreting BHs appear to be over-massive compared to the stellar mass of their host galaxies, relative to local MBH − M⋆ relation. We used CAT to test various BH mass growth scenarios against new JWST observations (Schneider+23, see Figure). Our findings suggest that AGNs with properties similar to those observed with JWST can be explained by scenarios where BHs originate from Eddington-limited gas accretion onto heavy BH seeds, or through phases of super-Eddington accretion (Pezzulli+2016, Trinca+2022). However, our understanding of the conditions enabling super-Eddington growth is still limited. We aim to assess the conditions that allow for super-Eddington accretion of BH seeds, its duration, and the feedback it exerts on the host galaxy. High-resolution hydrodynamical simulations will be conducted, focusing on an isolated galaxy circumnuclear disk using the GIZMO code, where a novel BH accretion and momentum-driven feedback model has been recently implemented (Sala+21), which self-consistently account for the evolution of BH mass and spin (Cenci+21). This version will be accessible to our team through an established collaboration with the Institute for Computational Science at the University of Zurich (Pedro Capelo; Lucio Mayer). The results will be used to develop a simulation-informed analytic model to be integrated into CAT.

References:

Cenci+21, MNRAS, 500, 3719
Inayoshi+20, ARA&A, 58, 27
Pezzulli+16, MNRAS,458, 3087
Sala+21, MNRAS, 500, 4788
Schneider+23, MNRAS, 526, 3250
Trinca+22, MNRAS, 511, 616
Volonteri+21, Nat Rev Phys, 3, 732

Supervisor: Raffaella Schneider
co-supervisors: Rosa Valiante, Luca Graziani

Chemical enrichment plays a crucial role in the thermal evolution of star-forming gas clouds. The metal abundance, both in the gas phase and in dust grains, is believed to be a key factor in the transition from a top-heavy Pop III IMF to a Kroupa-like Pop II/I IMF, occurring above a critical metallicity of Zcr=10^(-4±1) Zsun (Schneider+03; Omukai+05; Schneider+06; Schneider+12). Using high-resolution hydrodynamical simulations, we have recently shown that a more gradual transition in the stellar IMF occurs, with a larger fraction of massive stars persisting up to Z ≈ 10^(-2) Zsun (Chon+21). Moreover, the higher temperature of the CMB radiation at z ≥ 10 suppresses cloud fragmentation and reduces the number of low-mass stars in star forming regions with metallicities Z ≥ 10^(-2) Zsun (Schneider+10, Chon+22). As a result, we expect stellar populations with Z ≤ 10^(-2) Zsun or forming at z ≥ 10 to exhibit a mass spectrum consisting of a low-mass Kroupa-like component peaking at m∗ ≈ 0.1 Msun and a top-heavy component peaking at m∗ ≈ 10 Msun, with the mass fraction in the latter increasing with redshift and decreasing with metallicity. A non-universal IMF that depends on metallicity and redshift has far reaching implications for the evolution of the first galaxies. A top-heavy IMF has been recently advocated as a possible explanation for the number density of UV bright galaxies detected by JWST and its apparent lack of evolution between z ≈ 9 and z ≈ 13 – 17 (Trinca+23b), which are in tension with standard model predictions. At z > 10, spectroscopically confirmed galaxies are found to host ≈ 10^(7.2) – 10^9 Msun in stars, with stellar populations having formed half of their mass on timescales 16 – 70 Myr, and sub-solar metallicity (< 0.1 – 0.6 Zsun). As such, these galaxies are massive stars-dominated systems, and are therefore very sensitive to the shape of the high-mass end of the stellar IMF. In addition, stellar winds are believed to be weak at low metallicity, providing less kinetic energy

input to stellar environments. Metallicity also affects the stellar progenitor mass above which massive stars fail to explode and quietly collapse to BHs, with important implications for the timescales on which mechanical feedback due to supernova explosions operate. In this thesis project, we propose to investigate the implications of a non-Universal IMF on the physics of the first galaxies, implementing a parametric form of the IMF in the semi-analytical model CAT (Cosmic Archaeology Tool, Trinca+22). The project foresees comparison with JWST spectroscopic and photometric observations in collaboration with observers at the INAF/OAR.

References:

Chon+21  MNRAS, 508, 4175
Chon+22 MNRAS, 514, 4639
Omukai+05, ApJ, 626, 627
Schneider+03,  Nature, 422, 869
Schneider+06 MNRAS, 369, 1437
Schneider+10 MNRAS, 402, 429
Schneider+12 MNRAS, 419, 1566
Trinca+22  MNRAS, 511, 616
Trinca+23b arxiv:2305.04944

Supervisor: Antonello Calabrò (antonello.calabro@inaf.it)

Star formation rates (SFRs) are critical quantities for the understanding of galaxy evolution. There exist several different methods of estimating the SFR in a galaxy. 
Near-ultraviolet (UV) continuum observations, probing the photospheric emission of massive young (~100 Myr)stars, are a direct tracer of recent star formation on timescales of the order of 100 million years. 
Alternatively, optical and near-IR recombination lines of hydrogen (e.g., H-alpha, Paschen-alpha), originating from ionized gas around massive stars, can be used as SFR indicators. These lines receive most of the contribution from very massive stars (30–40 Msun), hence they trace star-formation over timescales of 3–10 Myr. 
Since the continuum and emission line tracers correspond to star formation on different timescales, their ratio can be used to measure the burstiness of star formation. However, in order to have reliable estimates of star-formation, the above line intensities should be corrected for dust. On the one hand, the observed flux ratios among recombination lines (which have a fixed intrinsic ratio set by the physical conditions of the gas) can be used to constrain the dust attenuation properties in moderately dusty galaxies. On the other hand, the total infrared luminosity (LIR), which traces UV light from massive stars reprocessed by dust, is a useful probe of the total star-formation, hence of the true level of dust attenuation, in very dusty systems where the optical light is completely obscured. The larger the optical depth of the dust, the higher the ratio between infrared and optical/UV based star-formation. 
In this thesis, the student will measure the Halpha, Hbeta, Paschen gamma, Paschen beta, Paschen alpha line fluxes from archival KMOS 3D spectra and new JSWT NIRSPEC data (available starting from December 2022) of star-forming galaxies from redshift 0.5 to 5. Combining all these measurements coming from Balmer and Paschen lines, the student will constrain the dust attenuation properties (also as a function of stellar mass and redshift), and derive dust-corrected, ‘instantaneous’ star-formation rates. Comparing these SFRs to other indicators probing longer timescales (like UV-based SFRs), it will be possible to infer the recent star-formation history (SFH) of the galaxies, and determine whether the star-formation is a smooth process (changing little with time), or it is dominated by frequent short-lived bursts. Finally, these SFRs will be compared to those inferred from the far-infrared luminosity (which probes dust obscured star-forming activity), checking whether we are missing a substantial fraction of the total star-formation of the galaxies with optical/UV lines, and we will look for systems with extreme obscurations.

Supervisor: Antonello Calabrò (antonello.calabro@inaf.it)

One of the major results of the Hubble Space Telescope (HST) has been that galaxy morphologies evolve with cosmic time, with classical elliptical and spiral galaxies dominating below z ∼ 1 and irregular, clumpy, and merging galaxies being more and more common at higher redshifts (e.g., Lee et al. 2013). However, it is still debated when in the universe’s evolution the earliest irregular, clumpy progenitors give way to the regular morphological types we observe today.
Cosmological surface brightness dimming, limited angular resolution, and poor sampling of the Wide Field Camera 3 (WFC3) on board of HST have limited so far a clear characterization of galaxy morphology at high redshift and its evolution across cosmic epochs. Moreover, at z > 3, the optical rest frame shifts beyond the reach of WFC3, forcing one to rely on the rest frame UV light, which is dominated by young stars and thus probes the location of star forming regions rather than the morphology of the bulk of more mature stars.
Now we are in a crucial phase of history. Thanks to the launch of JWST, we can overcome the above limitations by virtue of its superior angular resolution and longer wavelength coverage (up to 30 microns).
In this thesis, the student will analyse the new images coming from Early Release Science and Cycle 1 accepted JWST programs, namely GLASS, CEERS, and PRIMER. First, he/she will assemble the images available in different photometric bands for the star-forming galaxies selected and observed by these programs in the redshift range between 1 and 6. With these images, he/she will derive quantitative morphological parameters, using estimators that are widely tested on HST images, and then identify merging systems and clumpy galaxies. The student will finally investigate how the morphology, merger and clumpy fraction, vary as a function of cosmic epoch and other galaxy properties, such as the stellar mass, star-formation rate, and dust attenuation. This way, we will identify when and what are the main physical mechanisms responsible for the big morphological transformation occurring in galaxies across cosmic time.

Useful references: Treu, Calbrò et al. https://ui.adsabs.harvard.edu/abs/2022arXiv220713527T/abstract

Supervisor: Laura Pentericci

The search for Lyman continuum emitters, i.e.. galaxies from which ionizing radiation can escape into the IGM, is still a challenging tasks. So far only a few LyC emitters have been detected in a solid way. Most other measurements rely on staking analysis, either with deep imaging or spectroscopy.

We will exploit the ultradeep spectra obtained by VANDELS to determine the average amount of LyC emission at z=4.5., by stacking deep spectra of around 100 galaxies in the redshift range 4.3-5 where the IGM trasmission is still >10%, and the 910 A region (LyC emission) enters the VANDELS observed spectral range.

We will also investigate if this fraction varies with physical characteristics of the galaxies (by performing separate stacks for mass/luminosity bins) and also as a function of the properties of the Lyman alpha emission line, as also predicted by models and observed at lower redshift.

Supervisor: Raffaella Schneider

Surveys of metal-poor stars in the halo and in the dwarf satellites of the Milky Way can provide interesting constraints on early chemical enrichment and low-metallicity star formation. A proper comparison between theoretical models and observation data must account for the complex interplay between radiative and chemical feedback effects that shape the evolution of the first stars and their transition to more evolved stellar populations. The aim of the proposed project is to investigate how first star formation and early chemical evolution is imprinted into the low-metallicity tail of the metallicity distribution function observed i the Galactic halo and in nearby dwarf spheroidal galaxies. To this aim we will use our semi-analytical models for early galaxy and black hole formation.

Supervisor: Emiliano Merlin (emiliano.merlin@inaf.it)

Modified Newtonian Dynamics (MOND) is an alternative theory of gravity that does not require Dark Matter to explain the observed properties of galaxies, in particular their rotation curves. A prediction of the theory is that the internal kinematics of galaxies should be affected by the large scale gravitational field (besides standard tidal forces), thus violating the General Relativity (RG) Strong Equivalence Principle, and this has apparently been observed in a recent study (Chae+2020). No such effect is predicted in standard GR cosmology. The candidate will analyze the outputs from hydrodynamical cosmological simulations (IllustrisTNG, Eagle and/or others) to verify that the external field effect is indeed not present (thus further highlighting a tension with observed data) or, if unexpectedly present, which physical process causes the effect thus mimicking a MOND-like behaviour in a classical framework.