Galactic outflow rates in the EAGLE simulations

Abstract

We present measurements of galactic outflow rates from the EAGLE suite of cosmological simulations. We find that gas is removed from the interstellar medium (ISM) of central galaxies with a dimensionless mass loading factor that scales approximately with circular velocity as V −3/2 c in the low-mass regime where stellar feedback dominates. Feedback from active galactic nuclei (AGN) causes an upturn in the mass loading for halo masses > 10 12 M . We find that more gas outflows through the halo virial radius than is removed from the ISM of galaxies, particularly at low redshift, implying substantial mass loading within the circumgalactic medium. Outflow velocities span a wide range at a given halo mass/redshift, and on average increase positively with redshift and halo mass up to M 200 ∼ 10 12 M . We present a number of like-for-like comparisons to outflow rates from other recent cosmological hydrodynamical simulations, and show that comparing the propagation of galactic winds as a function of radius reveals substantial discrepancies between different models. Relative to some other simulations, EAGLE favours a scenario for stellar feedback where agreement with the galaxy stellar mass function is achieved by removing smaller amounts of gas from the ISM, but with galactic winds that then propagate and entrain ambient gas out to larger radii.

L-GALAXIES 2020: The evolution of radial metallicity profiles and global metallicities in disc galaxies

“…The reheating mass-loading factors as a function of mass in L-G 2020 are in good agreement with those found by Mitchell et al (2020) for the EAGLE hydrodynamical model. We note that, due to the incorporation of (a) delayed mass return from stars, and (b) non-zero direct enrichment of the CGM, the true mass-loading factors in L-G 2020 are no longer simply equivalent to the disc parameter.…”

“…On the modelling side, degeneracies in the physical prescriptions implemented, and the tendency to focus on only a ★ E-mail: robyates@mpa-garching.mpg.de single galactic component or redshift, has hampered our ability to draw robust, definitive conclusions (e.g. Fu et al 2012;Somerville & Davé 2015;Collacchioni et al 2018;Mitchell et al 2020). Therefore, in order to make further progress in the field of galactic chemical evolution (GCE), we must look to combine accurate observational data from a range of phases and spatial scales with detailed theoretical models that allow the simultaneous modelling of diverse galaxy populations.…”

The Keck Baryonic Structure Survey: using foreground/background galaxy pairs to trace the structure and kinematics of circumgalactic neutral hydrogen at z ∼ 2

“…Additionally, although various hydrodynamical simulations of galaxy formation differ substantially in their spatial resolution and feedback implementation, our best-fitting outflow velocity of ∼600 km s −1 is consistent with the upper range of gas velocities found for simulated galaxies with similar halo mass and redshift. For example, in the FIRE zoom-in simulations, the 95th percentile gasphase velocity is ∼ 600 km s −1 at 0.25R vir (r 25 pkpc) (Muratov et al 2015); for larger volume simulations, which depend to a greater extent on 'sub-grid' treatment of feedback physics, the level of agreement depends on the simulation suite: in the IllustrisTNG simulations, the 95th percentile gas-phase velocity is ∼650 km s −1 at r = 10 pkpc (Nelson et al 2019), whereas in the EAGLE simulations, the 90 percentile gas velocity between 0.1 and 0.2 r vir (equivalent to r ∼ 10-20 pkpc for the galaxies in our observed sample) is smaller, ∼350 km s −1 (Mitchell et al 2020). It is important to note that the comparison of our observations with the FIRE simulation in Section 4.1 is more direct, since it was confined to neutral H in the simulation box, while the results quoted above pertain to all outflowing gas, regardless of physical state.…”

Predictions for the angular dependence of gas mass flow rate and metallicity in the circumgalactic medium

“…This signature of gas outflow rate modulated by azimuthal angle is similar to the findings in Nelson et al (2019b) for TNG50 at higher redshift. We remark that Mitchell et al (2020b) compare outflow rates between TNG and EAGLE and report notable similarities and/or differences depending on stellar mass and distance from the galaxy. Specifically, while the outflow rates we measure at M = 10 10.5 ± 0.2 M are qualitatively similar between the two models, significant differences exist at other masses (and radii) due fundamentally to the distinct physical models invoked for galactic outflows.…”

“…While gas accretes on to galaxies from the cosmic web filaments, outflowing gas preferentially leaves the galaxy following the path of least resistance, along its minor axis. Where they compete, galactic winds prevent infall of material from regions above and below the disc place (Brook et al 2011;Mitchell et al 2020b;DeFelippis et al 2020). As a result, inflowing gas is almost co-planar with the major axis of the galaxy (Stewart et al 2011;Shen et al 2012;.…”