Stochastic Ram-Pressure Stripping in the Early Life of the Milky WayR. C. Simons (rsimons[at]stsci.edu) and M. S. Peeples (molly[at]stsci.edu)
Galaxies are surrounded by a diffuse reservoir of gas known as their circumgalactic medium (CGM). For a galaxy of the mass of the Milky Way, much more gas will reside in its CGM than in its disk (Tumlinson, Peeples, & Werk 2017). Despite containing an outsized portion of the total mass, a detailed understanding of the structure of the CGM and its role in the assembly and regulation of galaxies remains an open question for galaxy evolution models.
The CGM is a cocoon—a soft barrier between galaxies and their outside environment. As a satellite galaxy passes through the gaseous CGM of another galaxy, it will encounter a headwind known as ʺram pressure.ʺ The strength of this headwind is set by the gas density of the local medium and its relative speed with the galaxy. Both of these will vary with time—as the galaxy speeds up or slows down, as it samples other parts of the medium, and as the medium evolves. If the strength of ram pressure exceeds the internal self-gravity of the satellite galaxy, its interstellar gas will be stripped away (Gunn & Gott 1977). This process is known as ram-pressure stripping and is one of the main mechanisms regulating the gas content of the low-mass satellite populations around galaxies like the Milky Way and their galaxy groups (e.g., Mayer et al. 2006).
However, the strength and role of ram-pressure stripping in such galaxy groups is still poorly understood. In traditional ram-pressure stripping models, the CGM is often treated as a spatially continuous and symmetric fluid, e.g., as gas in hydrostatic equilibrium. However, over the last decade, observations and numerical simulations have revealed the CGM as a rich multiphase medium with intricate asymmetric velocity and density structure (Tumlinson, Peeples, & Werk 2017). To accurately model ram-pressure stripping, one must treat the CGM as such—an inhomogeneous, clumpy, and variable fluid.
To better understand the nature of ram-pressure stripping in the Milky Way, we use state-of-the-art simulations from the FOGGIE (Figuring Out Gas & Galaxies In Enzo) simulation suite (Simons et al. 2020). The simulations are carried out using the Enzo hydrodynamical adaptive mesh refinement code (Brummel-Smith et al. 2019). The FOGGIE suite consists of 6 halos, selected to be Milky Way-like based on their mass and merger history. The simulations adopt a novel refinement technique to reach high—and physically meaningful—resolutions in their circumgalactic gas. Around the center of the main simulated halo, the simulation is forced to a minimum resolution of 1100 comoving parsec. It is allowed to refine up to a factor 4 further, where the gas density is higher and/or if doing so will resolve the cooling length of the gas. This allows for a more realistic treatment of the cooling and fragmentation of the CGM gas, and helps prevent artificial mixing of high-/low-gas metallicities, densities, and velocities.
In Figure 1, we show gas and star projections for 3 of the FOGGIE halos at z = 2. The simulations recover a web of substructure in the CGM that would be suppressed, or missed entirely, in lower-resolution simulations (Corlies et al. 2018; Peeples et al. 2019). The CGM of the simulations typically spans ~5 dex in density and ~1000 km s–1 in velocity. This large dispersion in density and velocity translates into a large dispersion in ram pressure (which goes as density x velocity2). In Simons et al. (2020), we analyze how such substructure in the CGM affects the efficacy of ram-pressure stripping in galactic halos.
First, to assess the strength and stochasticity of ram-pressure stripping in these simulations, we create 100 test particles in radial free-fall through each of the 6 halos at z = 2. The particles start their free-fall at the same distance from the central galaxy. However, they are injected from different angles and so follow different radial trajectories through the halo. In Figure 2, we show the distribution of the total momentum imparted by ram pressure on these test particles (gray points and error bars). The internal scatter of the distributions are high (~1–2 dex). This reflects significant differences in the CGM probed between trajectories—the CGM is not spherically symmetric.
Next, we show a test particle on a free-fall radial orbit through a spherically symmetric model of the FOGGIE host halos (red points). This emulates the assumption that is often made in ram-pressure stripping models. In this model, the integrated impact of ram pressure is generally higher than that of the test particles moving through the true CGM. This is because the spherically averaged density is skewed by the densest regions of the CGM, where most of the mass is concentrated. However, such high densities comprise a small fraction of the total volume, meaning that the spherical-averaged density misrepresents the typical density of the CGM.
Finally, we assess the practical impact of these results for satellite galaxies. The gray-shaded swaths in Figure 2 indicate the amount of imparted momentum needed to remove gas beyond 1 effective radii for toy satellite galaxies at different masses. The stochastic nature of ram pressure (highlighted by the size of the gray error bars) leads to a large range in the practical efficacy of stripping. As an example, in the same Cyclone halo, a satellite with a stellar mass of 109 Msun may be stripped, while one with 107 Msun may not. The efficacy of ram pressure is highly sensitive to the specific path a satellite galaxy takes through the halo.
These results reveal an erratic and stochastic mode of ram-pressure stripping in Milky Way-like halos at high redshift—one that that is not captured by, and in strong disagreement with, the smooth spherically averaged model of the circumgalactic medium currently adopted in most ram pressure simulations.
Brummel-Smith, C., Bryan, G., Butsky, I., et al. 2019, The Journal of Open Source Software, 4, 1636
Corlies, L., Peeples, M. S., Tumlinson, J., et al. 2018, arXive-prints, arXiv:1811.05060
Gunn, J. E., & Gott, J. R., III 1972, ApJ, 176, 1
Mayer, L., Mastropietro, C., Wadsley, J., Stadel, J., & Moore, B. 2006, MNRAS, 369, 1021
Peeples, M. S., Corlies, L., Tumlinson, J., et al. 2019, ApJ, 873, 129
Simons, R. C., Peeples, M. S., Tumlinson, J., et al. 2020, arXiv-prints, arXiv:2004.14394
Tumlinson, J., Peeples, M. S., & Werk, J. K. 2017, ARA&A, 55, 389