2. GALAXIES IN GASEOUS HALOS

2.1. The Major Problems of Galaxy Evolution

We will motivate and organize our review with four major galaxy evolution problems in which the CGM is implicated (Figure 2). Why do dark matter halos of different masses give rise to galaxies with drastically different star formation and chemical histories (Sections 2.1.1, 2.1.2)? Why do such a small fraction of cosmic baryons and metals reside in the galaxies (Sections 2.1.3, 2.1.4)? The prevailing answers to these questions all feature the regulation of gas flows into and out of galaxies — which necessarily pass into and through the CGM. We initially pose these problems at low redshift, but they all have high-z counterparts, and their solutions require understanding the CGM and the flows that feed it at all cosmic epochs.

Figure 2. Four important problems in galaxy evolution viewed with respect to M⋆. (a) the gas depletion timescale τdep ∼ Mgas / sfr for star-forming galaxies at z ∼ 0, with Mgas from Peeples et al. (2014) and sfr from Whitaker et al. (2012); the shading denotes ± 0.15 dex scatter in sfr. (b) the galaxy bimodality in terms of M⋆ and specific star formation rate (Schiminovich et al., 2010). (c) the galactic baryon fraction, M⋆ / ((Ωb / Ωm) Mhalo) from Behroozi, Conroy & Wechsler (2010), with stars in red and interstellar gas in blue (from Peeples et al. (2014). (d) the “retained metals fraction”, metals for several galactic components relative to all the metals a galaxy has produced Peeples et al. (2014), with stars in red, interstellar gas in blue, and interstellar dust in orange. Vertical bars mark the properties of sub-L*, L*, and super-L* galaxies at logM⋆ / M⊙ = 9.5 (blue), 10.5 (green), and 11.0 (red), respectively.

2.1.1. How do galaxies sustain their star formation? Star-forming galaxies pose a conundrum: their ISM gas can last for only a small fraction of the time they have been forming stars (Figure 2a), implying an external supply of gas that keeps the ISM in a quasi-equilibrium state. The depletion time, τ_dep ∼ M_gas / _sfr changes by only ∼ 2× over the factor of 30 between sub-L^* and super-L^* galaxies. More generally, sub-L^* galaxies generally have extended bursty star formation histories, as opposed to the more continuous star formation found in more massive galaxies, suggesting differences in how and when these galaxies acquire their star forming fuel. As this fuel is from the CGM, we must explain how sub-L^* and L^* galaxies fuel star formation for longer than their τ_dep.

2.1.2. What quenches galaxies and what keeps them that way? How galaxies become and remain passive is one of the largest unsolved problems in galaxy evolution (Figure 2b). Proposed solutions to this problem involve controlling the gas supply, either by shutting off IGM accretion or keeping the CGM hot enough that it cannot cool and enter the ISM. Low-mass galaxies tend to continue forming stars unless they are a satellite of or near a more massive galaxy (Geha et al., 2012). This finding suggests that the central galaxy's gaseous halo strips the satellite with ram pressure or “starves” the satellite of fresh fuel. These ideas have specific testable implications for the physical state of the CGM.

2.1.3. Why do galaxies lack their fair share of baryons? The ΛCDM model predicts that baryons follow gravitationally-dominant dark matter into halos, where the gas dissipates energy as radiation and cools into the center of the halo. Observed galaxies, however, harbor only small share of the halo's expected baryons in their stars and ISM, with M_b≪(Ω_b / Ω_m) M_h (Figure 2c). Even at their most “efficient”, L^* galaxies have converted only ∼ 20% of their halos' baryons into stars (Figure 2c), with values of only about 5-10% in sub-L^* and super-L^* galaxies (Behroozi, Conroy & Wechsler, 2010, McGaugh et al., 2010). There are three basic possibilities: the baryons are in the halo but not yet detected, such as hot or diffuse gas; the baryons have been accreted and then ejected from the halo altogether; or the baryons have been prevented from accreting into the halo in the first place. While reality probably combines aspects of all three, in any combination they strongly suggest that the CGM is an excellent place to look for missing halo baryons in cold or hot gas, or for direct evidence of past ejection.

2.1.4. Where are the metals? While baryons come from outside the halo, metals are sourced locally by stars and the deaths of stars. Star-forming galaxies over ∼ 3 decades in stellar mass retain a surprisingly flat ∼ 20–25% of the metals they have ever produced (Peeples et al., 2014) in their stars, ISM gas, and dust. Metals have clearly been lost to outflows (Tremonti et al., 2004), but how these outflows scale with galaxy mass is unclear. Models that already struggle to reproduce the observed steep mass-metallicity relation (Somerville & Davé, 2015) fail to retain the low, flat fraction of metals produced (e.g., Muratov et al., 2015, Zahid et al., 2012, Oppenheimer et al., 2016b). While “missing baryons” concern accretion and feedback through the outer boundary of the CGM, metals address the disk/halo interface: do they leave the halo altogether, or recycle back into the galaxy's ISM on long timescales as a “halo fountain” (Oppenheimer & Davé, 2008). On what timescales are ejected metals recycled? How metal-enriched is outflowing material relative to the ambient ISM, i.e., what are the entrainment fractions and metal-loading factors? How does dust survive the journey out of galaxies, and what chemical clues does it hide? As we will show, following the metals as “Nature's tracer particles” is a fruitful and revealing route to understanding of the CGM.

2.2. Our Point of View

How galaxies acquire, eject, and recycle their gas are core issues in galaxy evolution, on par with how they evolve in their shapes and how star formation works. To a large extent these gas flows are galaxy evolution. The CGM is a main venue for these flows: it is potentially the galactic fuel tank, waste dump, and recycling center all at the same time. This review will approach the growing body of empirical results and theoretical insights from the direction of these four major questions. Rather than asking, for example, “what are the Mg ii absorbers?”, we will ask “what do the Mg ii absorbers tell us about the mass and kinematics of galactic outflows?”. We will thus favor physical insights and synthesis of discoveries over detailed discussions of methods, compilations of data, or exhaustive cataloging of the literature. We hope that this approach will improve understanding between those who study gas and galaxies (still disparate groups) and more effectively highlight open issues to be pursued in the future.

For the purposes of our discussion, we define the CGM to be bounded at the outside by the virial radius R_vir of a galaxy's dark matter halo, and on the inside by the disk or ISM. Neither boundary is well-defined, and precisely defining when gas passes through one of these boundaries can be either a valuable research contribution or a fruitless semantic exercise depending on circumstances. We focus on the physics of gas that fills out halos without too much attention to these exact definitions.