Forward-modelling of Simulated Galaxies to Test the Current Theory on Galaxy Formation and Evolution
: The iMaNGA Project

  • Lorenza Nanni

Student thesis: Doctoral Thesis


My research concerns the creation and analysis of a sample of nearby synthetic galaxies (with median redshift 𝑧 ≈ 0.03) based on the state-of-the-art hydrodynamical simulations to test the theoretical model of galaxy formation and evolution with the constraints given by observed data.
The comparison is made with data from the Sloan Digital Sky Survey (SDSS-IV); in particular, I consider galaxies from the Mapping Nearby Galaxies at the Apache point observatory, i.e. MaNGA, which provides spatially resolved spectra of galaxies through the use of Integral Field Spectroscopy (IFS).
We develop a method to generate mock galaxies from the cosmological magnetohy- drodynamic simulations IllustrisTNG, which are suited to compare with IFS observations of galaxies from the SDSS-IV/MaNGA survey, thanks to the high spatial resolution achieved.
With this method, we include the same instrumental effects and adopt the same procedures as in the acquisition and analysis of MaNGA data. Furthermore, we generate the spectra for the simulated galaxies using new stellar population models based on the MaNGA stellar library (MaStar). In this way, the mock datacubes have the same spatial sampling, cover the same wavelength range (3600-10300 Å), and share the same spectral resolution (𝑅 ≈ 1800) and flux calibration of real MaNGA galaxy spectra. We analyse the correspondent mock MaNGA-like datacubes with the official MaNGA Data Analysis Pipeline and the full-spectral-fitting code FIREFLY, which has been used for the observed spectra in many previous research works.
We also generate a mock MaNGA catalogue, that is, the iMaNGA sample, consisting of 1,500 unique galaxies from the TNG50 cosmological simulations falling into the SDSS-IV/MaNGA sample selection boundaries, which are defined in the redshift and i-band absolute magnitude space. We discuss the general characteristics of this catalogue in terms of morphology, kinematics, and stellar population properties. In particular, we present and discuss a comparison of the general characteristics of the iMaNGA catalogue with the MaNGA sample. Although the iMaNGA sample recovers the MaNGA sample trends well in terms of the stellar mass vs. angular size and spatial resolution relations, a discrepancy in the compactness of the galaxies at low mass is noticed, with simulated galaxies being more compact than their corresponding observed galaxies. The Sérsic index vs. angular size relation, instead, is not reproduced by the simulations, mostly caused by a lack of high-mass elliptical galaxies in TNG50. We also investigate our ability to recover the ‘intrinsic’ galaxy characteristics - i.e. as defined in the simulations
- by analysing the synthetic spectra. We demonstrate that the ‘intrinsic’ and ‘recovered’ stellar kinematics, stellar age and metallicity are consistent: the residuals are compatible with zero within the 1−𝜎 level, for all the Voronoi tassels in the iMaNGA sample. We also compare ‘intrinsic’ and ‘recovered’ star formation histories, which show a strong resemblance. Therefore, our mock generation and spectral fitting processes do not distort ‘intrinsic’ galaxy properties, hence we can use these results to test the theoretical predictions with the observational findings.
Once the methodology to construct mock MaNGA datacubes and the mock MaNGA catalogue are tested, we carry out a comparison against the results obtained exploiting MaNGA data. In particular, we focus on stellar populations, investigating the interplay among morphology, stellar mass, stellar metallicity, stellar surface mass density, galactocentric distance, and environment. As in MaNGA, we find that the metallicity is globally driven by the galaxy mass and morphology. Indeed, the most metal-rich stellar populations are in the most massive galaxies, ETGs in particular. Radial profiles of the stellar surface mass density (Σ∗) in the mass-morphology plane in iMaNGA reproduce those of MaNGA, but with somehow steeper trends in iMaNGA, in particular for low-mass galaxies, i.e. for 𝑀∗ ≤ 109.50 𝑀⊙. However, while the simulated and observed catalogues both exhibit negative stellar [Z/H] gradients at high stellar mass, these deviate from observations at lower and intermediate stellar masses, i.e. for 𝑀∗ ≤ 1010 𝑀⊙, with only observed gradients becoming flat or positive. When fixing Σ∗, for low- and intermediate-mass galaxies in MaNGA, [Z/H] increases with the radius, whereas in iMaNGA the trends are flat or negative. Hence, while for low- and intermediate-mass galaxies in MaNGA there are signs of additional metallicity drivers acting together with mass enhancing the metallicity content in the outskirts, iMaNGA shows the opposite trend. Additionally, we investigate whether the correlation between the metallicity gradients with the total stellar mass observed in MaNGA is recovered in iMaNGA. While both in MaNGA and iMaNGA spiral galaxies metallicity gradients become more negative at higher total stellar mass, iMaNGA does not show a correlation between total stellar mass and metallicity gradients in ellipticals, which is instead observed in MaNGA. On the other hand, neither MaNGA nor iMaNGA shows significant correlations between the metallicity gradients and the galaxy environment.
We can conclude that TNG50 is capable of representing the full picture for the
stellar metallicity drivers for high-stellar mass galaxies (i.e. at M∗ ≥ 1010M⊙), while the interplay among stellar surface mass density, stellar metallicity and galactic distance is not fully captured for low- and intermediate-mass galaxies.
Date of Award5 Mar 2024
Original languageEnglish
Awarding Institution
  • University of Portsmouth
SupervisorDaniel Thomas (Supervisor), Claudia Maraston (Supervisor) & Daniel James Whalen (Supervisor)

Cite this