Magnetic field evolution and reversals in spiral galaxies

TL;DR

This study uses 3D SPMHD simulations to investigate magnetic field evolution and reversals in spiral galaxies, highlighting the conditions for reversals and the effects of spiral arm strength on magnetic amplification.

Contribution

It demonstrates the occurrence of magnetic field reversals linked to spiral arm dynamics and assesses the impact of simulation resolution and numerical methods on magnetic field behavior.

Findings

Reversals occur when velocity jumps exceed ~20 km/s at spiral shocks.

Reversals are more frequent with stronger spiral potentials.

Magnetic field amplification is partly numerical, as shown by Athena comparisons.

Abstract

We study the evolution of galactic magnetic fields using 3D smoothed particle magnetohydrodynamics (SPMHD) simulations of galaxies with an imposed spiral potential. We consider the appearance of reversals of the field, and amplification of the field. We find magnetic field reversals occur when the velocity jump across the spiral shock is above $\approx$20km s$^{-1}$, occurring where the velocity change is highest, typically at the inner Lindblad resonance (ILR) in our models. Reversals also occur at corotation, where the direction of the velocity field reverses in the co-rotating frame of a spiral arm. They occur earlier with a stronger amplitude spiral potential, and later or not at all with weaker or no spiral arms. The presence of a reversal at a radii of around 4--6 kpc in our fiducial model is consistent with a reversal identified in the Milky Way, though we caution that alternative Galaxy models could give a similar reversal. We find that relatively high resolution, a few million particles in SPMHD, is required to produce consistent behaviour of the magnetic field. Amplification of the magnetic field occurs in the models, and while some may be genuinely attributable to differential rotation or spiral arms, some may be a numerical artefact. We check our results using Athena, finding reversals but less amplification of the field, suggesting that some of the amplification of the field with SPMHD is numerical.