Small-scale Dynamo in Supernova-driven Interstellar Turbulence

Frederick A Gent,Nishant K Singh,Maarit J Käpylä,Mordecai-Mark Mac Low

doi:10.3847/2041-8213/abed59

Abstract

Abstract Magnetic fields grow quickly even at early cosmological times, suggesting the action of a small-scale dynamo (SSD) in the interstellar medium of galaxies. Many studies have focused on idealized turbulent driving of the SSD. Here we simulate more realistic supernova-driven turbulence to determine whether it can drive an SSD. Magnetic field growth occurring in our models appears inconsistent with simple tangling of magnetic fields, but consistent with SSD action, reproducing and confirming models by Balsara et al. that did not include physical resistivity η. We vary η, as well as the numerical resolution and supernova rate, , to delineate the regime in which an SSD occurs. For a given we find convergence for SSD growth rate with resolution of a parsec. For , with the solar neighborhood rate, the critical resistivity below which an SSD occurs is , and this increases with the supernova rate. Across the modeled range of 0.5–4 pc resolution we find that for , the SSD saturates at about 5% of kinetic energy equipartition, independent of growth rate. In the range growth rate increases with . SSDs in the supernova-driven interstellar medium commonly exhibit erratic growth.

Highlights

We here study the small-scale dynamo (SSD) in the interstellar medium (ISM)
The large-scale dynamo (LSD) with much longer turnover times generates magnetic fields ordered on kiloparsec scales
To understand the role of physical resistivity η and viscosity ν on the SSD, we need to determine the value at each resolution where they exceed numerical diffusion in strength

Summary

Introduction

We here study the small-scale dynamo (SSD) in the interstellar medium (ISM). SSD acts at small eddy scales of turbulence, driving magnetic field growth at correspondingly short timescales. Any magnetic noise produced by tangling of a large-scale field will grow exponentially if an LSD is present. The limited resolution study of BKMM4 did not allow demonstration of solution convergence They imposed a uniform background field and implemented no physical resistivity or viscosity. A broad resolution and parameter study allows us to show numerical convergence and determine the critical resistivity for excitation of an SSD, which we follow to saturation (Section 4). This provides objective criteria for the action of SSD in simulations (such as Gent et al 2013a; Steinwandel et al 2019; Gressel & Elstner 2020).

Disentangling the Dynamo

Supernova-driven Turbulence Model Design

Resolution and Convergence

Effective Resistivity and Prandtl Number

Tangling of the Imposed Field

Conclusions