Among the different approaches that exist to model galaxy evolution, numerical simulations are our best theoretical tool to explicitly couple dark matter and baryons and disentangle the distinct effects of the physical processes involved in galaxies. In particular, cosmological, hydrodynamic simulations are designed to reproduce virtual universes, with volume widths between a few and several hundreds of Mpc. Therefore, they provide a statistical sample of structures that enables a helpful comparison with observational surveys. By modelling cosmic evolution on large scales, they are used within the GalSpec team to study the progress of galaxy and black hole growth and their non-linear interplay.
For this purpose, we mainly rely on the state-of-the-art IllustrisTNG suite of cosmological simulations (Springel+18, Pillepich+18, Dylan+18). The IllustrisTNG simulations consist of three volumes of 50, 100 and 300 Mpc in width. They have varying resolution between a few tens and hundreds of parsecs, and include complex baryon-physical models. These simulations are performed with the Arepo code (Springel+10), which adopts a moving-mesh finite volume refinement treatment to model magneto-hydrodynamics.
In order to understand how feedback affects the ISM, star formation, black hole growth and the inner properties of galaxies, it is needed to reach resolution of a few parsecs, which can not be achieved in large cosmological volumes such as IllustrisTNG. Our best compromise is to make use of the zoom-in technique: we extract the initial conditions of a sample of objects targeted from a cosmological volume, and re-simulate them at a better resolution, with the possibility of testing various physics at a moderate numerical cost.
Most specifically, the GalSpec team conducts an extensive suite of zoom-in simulations from the (100 Mpc)3 TNG volume, reaching an increased resolution of up to a few pc. We have started to target massive quiescent galaxies at z=3.5, as a counterpart for high-z observable objects by surveys with e.g., JWST. By means of this suite of zoom simulations, we want to explore how efficient AGN winds affect the ISM, the CGM and the structural properties of high-z galaxies. One of our goals is to validate AGN feedback model(s) against high-z observables, such as emission lines compared to synthetic observations of ionised outflows from AGN.