We introduce the Simba simulations, the next generation of the Mufasa cosmological galaxy formation simulations run with Gizmo's meshless finite mass hydrodynamics. Simba includes updates to Mufasa's sub-resolution star formation and feedback prescriptions, and introduces black hole growth via the torque-limited accretion model of Anglés-Alcázar et al. (2017a) from cold gas and Bondi accretion from hot gas, along with black hole feedback via kinetic bipolar outflows and X-ray energy. Ejection velocities are taken to be ∼ 10 3 km s −1 at high Eddington ratios, increasing to ∼ 8000 km s −1 at Eddington ratios below 2%, with a constant momentum input of 20L/c. Simba further includes an on-the-fly dust production, growth, and destruction model. Our Simba run with (100h −1 Mpc) 3 and 1024 3 gas elements reproduces numerous observables, including galaxy stellar mass functions at z = 0 − 6, the stellar mass-star formation rate main sequence, H i and H 2 fractions, the mass-metallicity relation at z ≈ 0, 2, star-forming galaxy sizes, hot gas fractions in massive halos, and z = 0 galaxy dust properties. However, Simba also yields an insufficiently sharp truncation of the z = 0 mass function, and too-large sizes for low-mass quenched galaxies. We show that Simba's jet feedback is primarily responsible for quenching massive galaxies.

We examine galaxy star formation rates (SFRs), metallicities, and gas contents predicted by the Mufasa cosmological hydrodynamic simulations, which employ meshless hydrodynamics and novel feedback prescriptions that yield a good match to observed galaxy stellar mass assembly. We combine 50, 25, and 12.5h −1 Mpc boxes with a quarter billion particles each to show that Mufasa broadly reproduces a wide range of relevant observations, including SFR and specific SFR functions, the mass-metallicity relation, H i and H 2 fractions, H i (21 cm) and CO luminosity functions, and cosmic gas density evolution. There are mild but significant discrepancies, such as too many high-SFR galaxies, overly metal-rich and H i-poor galaxies at M * 10 10 M , and sS-FRs that are too low at z ∼ 1 − 2. The H i mass function increases by ×2 out to z ∼ 1 then steepens to higher redshifts, while the CO luminosity function computed using the Narayanan et al. conversion factor shows a rapid increase of CO-bright galaxies out to z ∼ 2 in accord with data. Ω HI and Ω H2 both scale roughly as ∝ (1 + z) 0.7 out to z ∼ 3, comparable to the rise in H i and H 2 fractions. Mufasa galaxies with high SFR at a given M * have lower metallicities and higher H i and H 2 fractions, following observed trends; we make quantitative predictions for how fluctuations in the baryon cycle drive correlated scatter around galaxy scaling relations. Most of these trends are well converged with numerical resolution. These successes highlight Mufasa as a viable platform to study many facets of cosmological galaxy evolution.

In this paper we investigate environment driven gas depletion in satellite galaxies, taking full advantage of the atomic hydrogen (H i) spectral stacking technique to quantify the gas content for the entire gas-poor to -rich regime. We do so using a multi-wavelength sample of 10,600 satellite galaxies, selected according to stellar mass (log M /M 9) and redshift (0.02 z 0.05) from the Sloan Digital Sky Survey, with H i data from the Arecibo Legacy Fast ALFA (ALFALFA) survey. Using key H i-to-stellar mass scaling relations, we present evidence that the gas content of satellite galaxies is, to a significant extent, dependent on the environment in which a galaxy resides. For the first time, we demonstrate that systematic environmental suppression of gas content at both fixed stellar mass and fixed specific star formation rate (sSFR) in satellite galaxies begins in halo masses typical of the group regime (log M h /M < 13.5), well before galaxies reach the cluster environment. We also show that environment driven gas depletion is more closely associated to halo mass than local density. Our results are then compared with state-of-the-art semi-analytic models and hydrodynamical simulations and discussed within this framework, showing that more work is needed if models are to reproduce the observations. We conclude that the observed decrease of gas content in the group and cluster environments cannot be reproduced by starvation of the gas supply alone and invoke fast acting processes such as ram-pressure stripping of cold gas to explain this.