Galaxies on FIRE (Feedback In Realistic Environments): Stellar Feedback Explains Cosmologically Inefficient Star Formation

TL;DR

The FIRE simulations test whether explicit, physically motivated stellar feedback can account for cosmologically inefficient star formation without tuning. By resolving a multi-phase ISM and implementing feedback from radiation, SNe, winds, and photo-heating tied to stellar evolution, the authors reproduce the observed stellar mass–halo mass relation and Kennicutt-Schmidt law across a wide mass and redshift range. The results show that feedback physics—not numerical details—drives galaxy masses and star formation histories, producing bursty SF in dwarfs and quasi-equilibrium SF in larger systems, while many previous sub-grid models fail in dwarfs or high redshift. The study highlights the critical role of radiative feedback in GMC disruption and calls for inclusion of additional physics (e.g., AGN, magnetic fields) to fully capture quenching in massive halos and CGM/IGM interactions.

Abstract

We present a series of high-resolution cosmological simulations of galaxy formation to z=0, spanning halo masses ~10^8-10^13 M_sun, and stellar masses ~10^4-10^11. Our simulations include fully explicit treatment of both the multi-phase ISM (molecular through hot) and stellar feedback. The stellar feedback inputs (energy, momentum, mass, and metal fluxes) are taken directly from stellar population models. These sources of stellar feedback, with zero adjusted parameters, reproduce the observed relation between stellar and halo mass up to M_halo~10^12 M_sun (including dwarfs, satellites, MW-mass disks, and small groups). By extension, this leads to reasonable agreement with the stellar mass function for M_star<10^11 M_sun. We predict weak redshift evolution in the M_star-M_halo relation, consistent with current constraints to z>6. We find that the M_star-M_halo relation is insensitive to numerical details, but is sensitive to the feedback physics. Simulations with only supernova feedback fail to reproduce the observed stellar masses, particularly in dwarf and high-redshift galaxies: radiative feedback (photo-heating and radiation pressure) is necessary to disrupt 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 in the simulation, and is instead determined by self-regulation via stellar feedback. Feedback reduces star formation rates considerably and produces a reservoir of gas that leads to rising late-time star formation histories significantly different from the halo accretion history. Feedback also produces large short-timescale variability in galactic SFRs, especially in dwarfs. Many of these properties are not captured by common 'sub-grid' galactic wind models.

Figures (13)

• Figure 1: Gas in a representative simulation of a Milky Way-mass halo ( m12i in Table \ref{['tbl:sims']}). Image shows the projected gas density, log-weighted ($\sim4\,$dex stretch). Magenta shows cold molecular/atomic gas ($T<1000\,$K). Green shows warm ionized gas ($10^{4}\lesssim T \lesssim 10^{5}\,$K). Red shows hot gas ($T\gtrsim 10^{6}\,$K).$^{\ref{['foot:gas.coloring']}}$ Each image shows a box centered on the main galaxy. Left: Box $200\,$kpc (physical) on a side at high redshift. The galaxy has undergone a violent starburst, leading to strong outflows of hot and warm gas that have blown away much of the surrounding IGM (even outside the galaxy). Note that the "filamentary" structure of cool gas in the IGM is clearly affected by the outflows. Right: Near present-day, with a $\sim50\,$kpc box. A more relaxed, well-ordered disk has formed, with molecular gas tracing spiral structure, and a halo enriched by diffuse hot outflows.
• Figure 2: Stars in the m12i simulation at $z\sim0$, in a box $50\,$kpc on a side near present-time. Image is a mock $u/g/r$ composite. The disk is approximately face-on, and the spiral structure is visible. (The image uses STARBURST99 to determine the SED of each star particle given its known age and metallicity, then ray-traces the line-of-sight flux following hopkins:lifetimes.letter, attenuating with a MW-like reddening curve with constant dust-to-metals ratio for the abundances at each point.)
• Figure 3: Gas, as Fig. \ref{['fig:demo.image.1']}, for a dwarf galaxy ( m10 in Table \ref{['tbl:sims']}). Top:$40\,$kpc (physical) box, at high redshift. Bottom:$20\,$kpc box at intermediate redshift. Strong outflows are still present, though they are more spherical, because the galaxy halo is itself small and embedded within a much larger filament.
• Figure 4: Galaxy stellar mass-halo mass relation at $z=0$. Top: $M_{\ast}(M_{\rm halo})$. Bottom:$M_{\ast}$ relative to the Universal baryon budget of the halo ($f_{b}\,M_{\rm halo}$). Each simulation (points) from Table \ref{['tbl:sims']} is shown; large point denotes the most massive halo in each box. We compare the relation if all baryons became stars ($M_{\ast}=f_{b}\,M_{\rm halo}$; dotted) and the observationally inferred relationship as determined in moster:2013.abundance.matching.sfhs and behroozi:2012.abundance.matching.sfhs (dashed lines denote extrapolation beyond the observed range).$^{\ref{['foot:moster.behroozi.halo.defns']}}$ The agreement with observations is excellent at $M_{\rm halo}\lesssim10^{13}\,M_{\sun}$, including dwarf though MW-mass galaxies. We stress that there are zero adjusted parameters here: stellar feedback, with known mechanisms taken from stellar population models, is sufficient to explain galaxy stellar masses at/below $\sim L_{\ast}$.
• Figure 5: $M_{\ast}-M_{\rm halo}$ relation as Fig. \ref{['fig:mg.mh.z0']} (points follow the legend therein), at different redshifts. Observational constraints are also shown at each redshift (each pair of lines shows the $\pm1\,\sigma$ fit to the observations at that redshift). With no tuned parameters, the simulations predict $M_{\ast}-M_{\rm halo}$ and, by extension, the stellar MF and galaxy clustering, at all $z$. Redshift evolution in $M_{\ast}-M_{\rm halo}$ is weak, with the sense that low-mass dwarfs become higher-$M_{\ast}$, leading to a steeper faint-end galaxy MF, in agreement with constraints from reionization kuhlen:2012.reionization.escape.fractions.