We introduce the Illustris Project, a series of large-scale hydrodynamical simulations of galaxy formation. The highest resolution simulation, Illustris-1, covers a volume of (106.5 Mpc) 3 , has a dark mass resolution of 6.26 × 10 6 M , and an initial baryonic matter mass resolution of 1.26 × 10 6 M . At z = 0 gravitational forces are softened on scales of 710 pc, and the smallest hydrodynamical gas cells have an extent of 48 pc. We follow the dynamical evolution of 2 × 1820 3 resolution elements and in addition passively evolve 1820 3 Monte Carlo tracer particles reaching a total particle count of more than 18 billion. The galaxy formation model includes: primordial and metal-line cooling with self-shielding corrections, stellar evolution, stellar feedback, gas recycling, chemical enrichment, supermassive black hole growth, and feedback from active galactic nuclei. Here we describe the simulation suite, and contrast basic predictions of our model for the present day galaxy population with observations of the local universe. At z = 0 our simulation volume contains about 40, 000 well-resolved galaxies covering a diverse range of morphologies and colours including early-type, late-type and irregular galaxies. The simulation reproduces reasonably well the cosmic star formation rate density, the galaxy luminosity function, and baryon conversion efficiency at z = 0. It also qualitatively captures the impact of galaxy environment on the red fractions of galaxies. The internal velocity structure of selected well-resolved disk galaxies obeys the stellar and baryonic Tully-Fisher relation together with flat circular velocity curves. In the well-resolved regime the simulation reproduces the observed mix of early-type and late-type galaxies. Our model predicts a halo mass dependent impact of baryonic effects on the halo mass function and the masses of haloes caused by feedback from supernova and active galactic nuclei.

We introduce an updated physical model to simulate the formation and evolution of galaxies in cosmological, large-scale gravity+magnetohydrodynamical simulations with the moving mesh code AREPO. The overall framework builds upon the successes of the Illustris galaxy formation model, and includes prescriptions for star formation, stellar evolution, chemical enrichment, primordial and metal-line cooling of the gas, stellar feedback with galactic outflows, and black hole formation, growth and multi-mode feedback. In this paper we give a comprehensive description of the physical and numerical advances which form the core of the IllustrisTNG (The Next Generation) framework. We focus on the revised implementation of the galactic winds, of which we modify the directionality, velocity, thermal content, and energy scalings, and explore its effects on the galaxy population. As described in earlier works, the model also includes a new black hole driven kinetic feedback at low accretion rates, magnetohydrodynamics, and improvements to the numerical scheme. Using a suite of (25 Mpc h −1 ) 3 cosmological boxes we assess the outcome of the new model at our fiducial resolution. The presence of a self-consistently amplified magnetic field is shown to have an important impact on the stellar content of 10 12 M haloes and above. Finally, we demonstrate that the new galactic winds promise to solve key problems identified in Illustris in matching observational constraints and affecting the stellar content and sizes of the low mass end of the galaxy population.

The IllustrisTNG project is a new suite of cosmological magneto-hydrodynamical simulations of galaxy formation performed with the AREPO code and updated models for feedback physics. Here we introduce the first two simulations of the series, TNG100 and TNG300, and quantify the stellar mass content of about 4000 massive galaxy groups and clusters (10 13 M 200c /M 10 15 ) at recent times (z 1). The richest clusters have half of their total stellar mass bound to satellite galaxies, with the other half being associated with the central galaxy and the diffuse intra-cluster light. Haloes more massive than about 5 × 10 14 M have more diffuse stellar mass outside 100 kpc than within 100 kpc, with power-law slopes of the radial mass density distribution as shallow as the dark matter's ( −3.5 α 3D−3). Total halo mass is a very good predictor of stellar mass, and vice versa: at z = 0, the 3D stellar mass measured within 30 kpc scales as ∝ (M 500c ) 0.49 with a ∼ 0.12 dex scatter. This is possibly too steep in comparison to the available observational constraints, even though the abundance of TNG less massive galaxies ( 10 11 M in stars) is in good agreement with the measured galaxy stellar mass functions at recent epochs. The 3D sizes of massive galaxies fall too on a tight (∼0.16 dex scatter) power-law relation with halo mass, with r stars 0.5 ∝ (M 200c ) 0.53 . Even more fundamentally, halo mass alone is a good predictor for the whole stellar mass profiles beyond the inner few kpc, and we show how on average these can be precisely recovered given a single mass measurement of the galaxy or its halo.

Hydrodynamical simulations of galaxy formation have now reached sufficient volume to make precision predictions for clustering on cosmologically relevant scales. Here we use our new IllustrisTNG simulations to study the non-linear correlation functions and power spectra of baryons, dark matter, galaxies and haloes over an exceptionally large range of scales. We find that baryonic effects increase the clustering of dark matter on small scales and damp the total matter power spectrum on scales up to k ∼ 10 h Mpc −1 by 20%. The non-linear two-point correlation function of the stellar mass is close to a power-law over a wide range of scales and approximately invariant in time from very high redshift to the present. The two-point correlation function of the simulated galaxies agrees well with SDSS at its mean redshift z 0.1, both as a function of stellar mass and when split according to galaxy colour, apart from a mild excess in the clustering of red galaxies in the stellar mass range 10 9 − 10 10 h −2 M . Given this agreement, the TNG simulations can make valuable theoretical predictions for the clustering bias of different galaxy samples. We find that the clustering length of the galaxy auto-correlation function depends strongly on stellar mass and redshift. Its power-law slope γ is nearly invariant with stellar mass, but declines from γ ∼ 1.8 at redshift z = 0 to γ ∼ 1.6 at redshift z ∼ 1, beyond which the slope steepens again. We detect significant scaledependencies in the bias of different observational tracers of large-scale structure, extending well into the range of the baryonic acoustic oscillations and causing nominal (yet fortunately correctable) shifts of the acoustic peaks of around ∼ 5%.

We introduce the first two simulations of the IllustrisTNG project, a next generation of cosmological magnetohydrodynamical simulations, focusing on the optical colors of galaxies. We explore TNG100, a rerun of the original Illustris box, and TNG300, which includes 2×2500 3 resolution elements in a volume twenty times larger. Here we present first results on the galaxy color bimodality at low redshift. Accounting for the attenuation of stellar light by dust, we compare the simulated (g-r) colors of 10 9 < M /M < 10 12.5 galaxies to the observed distribution from the Sloan Digital Sky Survey (SDSS). We find a striking improvement with respect to the original Illustris simulation, as well as excellent quantitative agreement with the observations, with a sharp transition in median color from blue to red at a characteristic M ∼ 10 10.5 M . Investigating the build-up of the color-mass plane and the formation of the red sequence, we demonstrate that the primary driver of galaxy color transition is supermassive blackhole feedback in its low-accretion state. Across the entire population the median color transition timescale ∆t green is ∼ 1.6 Gyr, a value which drops for increasingly massive galaxies. We find signatures of the physical process of quenching: at fixed stellar mass, redder galaxies have lower SFRs, gas fractions, and gas metallicities; their stellar populations are also older and their large-scale interstellar magnetic fields weaker than in bluer galaxies. Finally, we measure the amount of stellar mass growth on the red sequence. Galaxies with M > 10 11 M which redden at z < 1 accumulate on average ∼ 25% of their final z = 0 mass post-reddening; at the same time, ∼ 18% of such massive galaxies acquire half or more of their final stellar mass while on the red sequence.

The Feedback In Realistic Environments (FIRE) project explores feedback in cosmological galaxy formation simulations. Previous FIRE simulations used an identical source code ("FIRE-1") for consistency. Motivated by the development of more accurate numerics -including hydrodynamic solvers, gravitational softening, and supernova coupling algorithms -and exploration of new physics (e.g. magnetic fields), we introduce "FIRE-2", an updated numerical implementation of FIRE physics for the GIZMO code. We run a suite of simulations and compare against FIRE-1: overall, FIRE-2 improvements do not qualitatively change galaxy-scale properties. We pursue an extensive study of numerics versus physics. Details of the star-formation algorithm, cooling physics, and chemistry have weak effects, provided that we include metal-line cooling and star formation occurs at higher-than-mean densities. We present new resolution criteria for high-resolution galaxy simulations. Most galaxy-scale properties are robust to numerics we test, provided: (1) Toomre masses are resolved; (2) feedback coupling ensures conservation, and (3) individual supernovae are time-resolved. Stellar masses and profiles are most robust to resolution, followed by metal abundances and morphologies, followed by properties of winds and circum-galactic media (CGM). Central (∼kpc) mass concentrations in massive (> L * ) galaxies are sensitive to numerics (via trapping/recycling of winds in hot halos). Multiple feedback mechanisms play key roles: supernovae regulate stellar masses/winds; stellar mass-loss fuels late star formation; radiative feedback suppresses accretion onto dwarfs and instantaneous star formation in disks. We provide all initial conditions and numerical algorithms used.

The inefficiency of star formation in massive elliptical galaxies is widely believed to be caused by the interactions of an active galactic nucleus (AGN) with the surrounding gas. Achieving a sufficiently rapid reddening of moderately massive galaxies without expelling too many baryons has however proven difficult for hydrodynamical simulations of galaxy formation, prompting us to explore a new model for the accretion and feedback effects of supermassive black holes. For high accretion rates relative to the Eddington limit, we assume that a fraction of the accreted rest mass energy heats the surrounding gas thermally, similar to the 'quasar mode' in previous work. For low accretion rates, we invoke a new, pure kinetic feedback model that imparts momentum to the surrounding gas in a stochastic manner. These two modes of feedback are motivated both by theoretical conjectures for the existence of different types of accretion flows as well as recent observational evidence for the importance of kinetic AGN winds in quenching galaxies. We find that a large fraction of the injected kinetic energy in this mode thermalizes via shocks in the surrounding gas, thereby providing a distributed heating channel. In cosmological simulations, the resulting model produces red, non star-forming massive elliptical galaxies, and achieves realistic gas fractions, black hole growth histories and thermodynamic profiles in large haloes.

We present a new comprehensive model of the physics of galaxy formation designed for large-scale hydrodynamical simulations of structure formation using the moving mesh code AREPO. Our model includes primordial and metal line cooling with self-shielding corrections, stellar evolution and feedback processes, gas recycling, chemical enrichment, a novel subgrid model for the metal loading of outflows, black hole (BH) seeding, BH growth and merging procedures, quasar-and radio-mode feedback, and a prescription for radiative electromagnetic (EM) feedback from active galactic nuclei (AGN). Our stellar evolution and chemical enrichment scheme follows nine elements (H, He, C, N, O, Ne, Mg, Si, Fe) independently. Stellar feedback is realised through kinetic outflows. The metal mass loading of outflows can be adjusted independently of the wind mass loading. This is required to simultaneously reproduce the stellar mass content of low mass haloes and their gas oxygen abundances. Radiative EM AGN feedback is implemented assuming an average spectral energy distribution and a luminosity-dependent scaling of obscuration effects. This form of feedback suppresses star formation more efficiently than continuous thermal quasar-mode feedback alone, but is less efficient than mechanical radio-mode feedback in regulating star formation in massive haloes. We contrast simulation predictions for different variants of our galaxy formation model with key observations, allowing us to constrain the importance of different modes of feedback and their uncertain efficiency parameters. We identify a fiducial best match model and show that it reproduces, among other things, the cosmic star formation history, the stellar mass function, the stellar mass -halo mass relation, g-, r-, i-, z-band SDSS galaxy luminosity functions, and the Tully-Fisher relation. We can achieve this success only if we invoke very strong forms of stellar and AGN feedback such that star formation is adequately reduced in both low and high mass systems. In particular, the strength of radio-mode feedback needs to be increased significantly compared to previous studies to suppress efficient cooling in massive, metal-enriched haloes.