The supernova-regulated ISM – I. The multiphase structure

Abstract

We simulate the multi-phase interstellar medium (ISM) randomly heated and stirred by supernovae (SNe), with gravity, differential rotation and other parameters of the solar neighbourhood. Here we describe in detail both numerical and physical aspects of the model, including injection of thermal and kinetic energy by SN explosions, radiative cooling, photoelectric heating and various transport processes. With a three-dimensional domain extending 1 × 1 kpc 2 horizontally and 2 kpc vertically (symmetric about the galactic mid-plane), the model routinely spans gas number densities 10 −5 -10 2 cm −3 , temperatures 10-10 8 K, local velocities up to 10 3 km s −1 (with Mach number up to 25). The working numerical resolution of 4 pc has been selected via simulations of a single expanding SN remnant, where we closely reproduce, at this resolution, analytical solutions for the adiabatic and snowplough regimes. The feedback of the the halo on the disc cannot be captured in our model where the domain only extends to the height of 1 kpc above the mid-plane. We argue that to reliably model the disc-halo connections would require extending the domain horizontally as well as vertically due to the increasing horizontal scale of the gas flows with height.The thermal structure of the modelled ISM is classified by inspection of the joint probability density of the gas number density and temperature. We confirm that most of the complexity can be captured in terms of just three phases, separated by temperature borderlines at about 10 3 K and 5 × 10 5 K. The probability distribution of gas density within each phase is approximately lognormal. We clarify the connection between the fractional volume of a phase and its various proxies, and derive an exact relation between the fractional volume and the filling factors defined in terms of the volume and probabilistic averages. These results are discussed in both observational and computational contexts. The correlation scale of the random flows is calculated from the velocity autocorrelation function; it is of order 100 pc and tends to grow with distance from the mid-plane. We use two distinct parameterizations of radiative cooling to show that the multi-phase structure of the gas is robust, as it does not depend significantly on this choice.

Supernova-regulated ISM. V. Space and Time Correlations

“…Gent et al (2013a) demonstrate that this resolution is sufficient to reproduce the known solutions for expanding SN remnants in the SedovTaylor and momentum-conserving phases.…”

“…It is then natural to use sufficiently realistic numerical models to address these questions before the more difficult observational exploration. We use such simulations, as detailed in Gent (2012) and Gent et al (2013aGent et al ( , 2013b, which have nontrivial magnetic fields generated by dynamo action, to clarify the correlation (and other statistical) properties of the multiphase ISM. In particular, we compare the autocorrelation and crosscorrelation functions of the random (i.e., small-scale; see Section 2.3) velocity and magnetic fields and density fluctuations, as well as the Faraday depth and synchrotron intensity.…”

“…Such transforms are straightforward in infinite or periodic spaces. However, simulations of the ISM are performed in relatively small domains, only containing on the order of 1000 correlation cells, are not simply periodic because of the open (or similar) boundary conditions at the top and bottom of the domain, and are statistically inhomogeneous because of the stratification (e.g., Korpi et al 1999aKorpi et al , 1999bGent et al 2013a). Furthermore, it is difficult to estimate reliably the statistical uncertainty of the Fourier transforms.…”

“…Joung & Mac Low (2006) obtain a gas density spectrum with a peak at 20 pc, whereas most kinetic energy is contained at scales of 20 40 pc -. Gent et al (2013a) calculate l 100 pc 0 = for the random velocity field in the midplane of the galaxy, also from hydrodynamic simulations. In the MHD simulations of de Avillez & Breitschwerdt (2007), l 70 pc 0 = for the random velocity field.…”

Structure distribution and turbulence in self-consistently supernova-driven ISM of multiphase magnetized galactic discs

“…Tasker & Bryan 2006;Dubois & Teyssier 2008;Bournaud et al 2010;Kim et al 2011;Dobbs et al 2011;Tasker 2011;Hopkins et al 2011;Renaud et al 2013), with a well-resolved interstellar medium (ISM) although the spatial resolution is continuously improving. To address this question, an alternative approach has been developed which consists of simulating a small portion of a galactic disc leading to a better spatial resolution (Korpi et al 1999;Slyz et al 2005;de Avillez & Breitschwerdt 2005;Joung & Mac Low 2006;Hill et al 2012;Kim et al 2011Kim et al , 2013Gent et al 2013;Hennebelle & Iffrig 2014;Gatto et al 2015) although at the expense of solving the large galactic scales. Clearly these two approaches are complementary and must be used in parallel.…”