We present a series of high-resolution cosmological simulations 1 of galaxy formation to z = 0, spanning halo masses ∼ 10 8 − 10 13 M , and stellar masses ∼ 10 4 − 10 11 M . Our simulations include fully explicit treatment of the multi-phase ISM & stellar feedback. The stellar feedback inputs (energy, momentum, mass, and metal fluxes) are taken directly from stellar population models. These sources of feedback, with zero adjusted parameters, reproduce the observed relation between stellar and halo mass up to M halo ∼ 10 12 M . We predict weak redshift evolution in the M * − M halo relation, consistent with current constraints to z > 6. We find that the M * − M halo relation is insensitive to numerical details, but is sensitive to feedback physics. Simulations with only supernova feedback fail to reproduce observed stellar masses, particularly in dwarf and high-redshift galaxies: radiative feedback (photo-heating and radiation pressure) is necessary to destroy GMCs and enable efficient coupling of later supernovae to the gas. Star formation rates agree well with the observed Kennicutt relation at all redshifts. The galaxy-averaged Kennicutt relation is very different from the numerically imposed law for converting gas into stars, and is determined by self-regulation via stellar feedback. Feedback reduces star formation rates and produces reservoirs of gas that lead to rising late-time star formation histories, significantly different from halo accretion histories. Feedback also produces large short-timescale variability in galactic SFRs, especially in dwarfs. These properties are not captured by common "sub-grid" wind models.

We develop a model for the cosmological role of mergers in the evolution of starbursts, quasars, and spheroidal galaxies. By combining theoretically well-constrained halo and subhalo mass functions as a function of redshift and environment with empirical halo occupation models, we can estimate where galaxies of given properties live at a particular epoch. This allows us to calculate, in an a priori cosmological manner, where major galaxy-galaxy mergers occur and what kinds of galaxies merge, at all redshifts. We compare this with the observed mass functions, clustering, fractions as a function of halo and galaxy mass, and small-scale environments of mergers, and show that this approach yields robust estimates in good agreement with observations, and can be extended to predict detailed properties of mergers. Making the simple ansatz that major, gas-rich mergers cause quasar activity (but not strictly assuming they are the only triggering mechanism), we demonstrate that this model naturally reproduces the observed rise and fall of the quasar luminosity density from z = 0 − 6, as well as quasar luminosity functions, fractions, host galaxy colors, and clustering as a function of redshift and luminosity. The recent observed excess of quasar clustering on small scales at z ∼ 0.2 − 2.5 is a natural prediction of our model, as mergers will preferentially occur in regions with excess small-scale galaxy overdensities. In fact, we demonstrate that quasar environments at all observed redshifts correspond closely to the empirically determined small group scale, where major mergers of ∼ L * gas-rich galaxies will be most efficient. We contrast this with a secular model in which quasar activity is driven by bars or other disk instabilities, and show that while these modes of fueling probably dominate the high-Eddington ratio population at Seyfert luminosities (significant at z = 0), the constraints from quasar clustering, observed pseudobulge populations, and disk mass functions suggest that they are a small contributor to the z 1 quasar luminosity density, which is dominated by massive BHs in predominantly classical spheroids formed in mergers. Similarly, lowluminosity Seyferts do not show a clustering excess on small scales, in agreement with the natural prediction of secular models, but bright quasars at all redshifts do so. We also compare recent observations of the colors of quasar host galaxies, and show that these correspond to the colors of recent merger remnants, in the transition region between the blue cloud and the red sequence, and are distinct from the colors of systems with observed bars or strong disk instabilities. Even the most extreme secular models, in which all bulge (and therefore BH) formation proceeds via disk instability, are forced to assume that this instability acts before the (dynamically inevitable) mergers, and therefore predict a history for the quasar luminosity density which is shifted to earlier times, in disagreement with observations. Our model provides a powerful means to predict the abunda...

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.

We study the formation of galaxies in a large volume (50 h−1 Mpc, 2 × 2883 particles) cosmological simulation, evolved using the entropy and energy‐conserving smoothed particle hydrodynamics (SPH) code gadget‐2. Most of the baryonic mass in galaxies of all masses is originally acquired through filamentary ‘cold mode’ accretion of gas that was never shock heated to its halo virial temperature, confirming the key feature of our earlier results obtained with a different SPH code. Atmospheres of hot, virialized gas develop in haloes above 2–3 × 1011 M⊙, a transition mass that is nearly constant from z= 3 to 0. Cold accretion persists in haloes above the transition mass, especially at z≥ 2. It dominates the growth of galaxies in low‐mass haloes at all times, and it is the main driver of the cosmic star formation history. Our results suggest that the cooling of shock‐heated virialized gas, which has been the focus of many analytic models of galaxy growth spanning more than three decades, might be a relatively minor element of galaxy formation. At high redshifts, satellite galaxies have gas accretion rates similar to central galaxies of the same baryonic mass, but at z < 1 the accretion rates of low‐mass satellites are well below those of comparable central galaxies. Relative to our earlier simulations, the gadget‐2 simulations predict much lower rates of ‘hot mode’ accretion from the virialized gas component. Hot accretion rates compete with cold accretion rates near the transition mass, but only at z≤ 1. Hot accretion is inefficient in haloes above ∼5 × 1012 M⊙, with typical rates lower than 1 M⊙ yr−1 at z≤ 1, even though our simulation does not include active galactic nuclei (AGN) heating or other forms of ‘preventive’ feedback. Instead, the accretion rates are low because the inner density profiles of hot gas in these haloes are shallow, with long associated cooling times. The cooling recipes typically used in semi‐analytic models can overestimate the accretion rates in these haloes by orders of magnitude, so these models may overemphasize the role of preventive feedback in producing observed galaxy masses and colours. A fraction of the massive haloes develop cuspy profiles and significant cooling rates between z= 1 and 0, a redshift trend similar to the observed trend in the frequency of cooling flow clusters.