The global structure of magnetic fields and gas in simulated Milky Way-analogue galaxies

Abstract

We simulate an isolated, magnetised Milky Way-like disc galaxy using a self-consistent model of unresolved star formation and feedback, evolving the system until it reaches statistical steady state. We show that the quasi-steady-state structure is distinctly layered in galactocentric height 𝑧, with an innermost region having comparable gas and magnetic pressures (plasma beta 𝛽 ∼ 1), an outermost region having dominant gas pressures (𝛽 1), and an intermediate region between 300 pc |𝑧| 3 kpc that is dynamically dominated by magnetic fields (𝛽 1). We find field strengths, gas surface densities, and star formation rates that agree well with those observed both in the Galactic centre and in the Solar neighbourhood. The most significant dynamical effect of magnetic fields on the global properties of the disc is a reduction of the star formation rate by a factor of 1.5-2 with respect to an unmagnetised control simulation. At fixed star formation rate, there is no significant difference in the mass outflow rates or profiles between the magnetised and non-magnetised simulations. Our results for the global structure of the magnetic field have significant implications for models of cosmic ray-driven winds and cosmic-ray propagation in the Galaxy, and can be tested against observations with the forthcoming Square Kilometre Array and other facilities. Finally, we report the discovery of a physical error in the implementation of neutral gas heating and cooling in the popular code, which may lead to qualitatively incorrect phase structures if not corrected.

Photochemistry and Heating/Cooling of the Multiphase Interstellar Medium with UV Radiative Transfer for Magnetohydrodynamic Simulations

“…For the maximum/minimum pressure of the WNM/CNM branch, we find P (min) /k B = 94 cm −3 K, P (max) /k B = 278 cm −3 K with n (max) = 0.04 cm −3 , n (min) = 0.36 cm −3 and T (max) = 6350 K, T (min) = 238 K. 31 At the observed solar neighborhood pressure, all gas would be in the cold phase. As pointed out by Wibking & Krumholz (2022), the reason for this low PE heating rate is the very low electron abundance (two orders of magnitude below our fiducial model at n = 1 cm −3 ) This (unphysically) low x e increases the grain charging parameter, which in turn dramatically reduces the efficiency of PE heating (see Section 4.2) compared to our fiducial model (bottom-left panel).…”

“…Similarly, the FIRE-2 model also includes heating and ionization by the UVB and PE heating from grains. However, as pointed out by Wibking & Krumholz (2022), since the equilibrium electron abundance is based on UVB with an attenuation factor that exponentially decreases with density above n ∼ 10 −2 cm −3 (Appendix B in Hopkins et al 2018), the electron abundance drops very sharply for n 10 −2 cm −3 . As a result, both the PE and PI heating efficiencies are significantly reduced and the equilibrium pressure is nearly two orders of magnitude lower than that in our model at n = 1 cm −3 .…”

“…The FIRE-2 cooling at Z = 1 is obtained from the Appendix A in Wibking & Krumholz (2022), and also reproduced from their source code available online 28 . The source of ionization is the UVB (Haardt & Madau 2012), while the source of heating is PI by the UVB plus PE heating (Wolfire et al 2003, with χ 0 = 1).…”

Spirals in galaxies

“…Although spiral-driven turbulence may well have contributed to the promising magnetic field amplification reported by Pakmor et al (2017), their simulations included too many physical processes to be able to isolate the role of non-axisymmetric gravitational forces arising from spiral arm evolution. The simulations by Khoperskov & Khrapov (2018) included self-gravity of the magnetized gas only, but adopted the gravitational field of an imposed, steadily rotating spiral potential, which crucially omits the evolving gravitational field that is important to driving turbulence by radial migration, while Wibking & Krumholz (2021) employed a sub-maximum disk that developed multi-arm spirals that are unable to drive large radial excursions ( §7.1.3). Wielen (1977), and others, pointed out long ago that the random motions of older disk stars in the Milky Way are greater than those of younger ones.…”

Cosmic Ray Interstellar Propagation Tool using Itô Calculus (criptic): software for simultaneous calculation of cosmic ray transport and observational signatures

“…In future work we intend to use to post-process MHD simulations of Milky Way-like galactic discs (e.g., Wibking & Krumholz 2022), in order to compare the results produced by different CR transport models with observable quantities such as the 𝛾-ray spectral index as a function of height above the galactic plane. We also intend to post-process CR hydrodynamics simulations in order to predict detailed observables from them.…”