Magnetic reconnection, plasmoids and numerical resolution

TL;DR

This paper investigates how numerical resolution affects the observation of plasmoid instability in magnetic reconnection simulations, questioning its role in explaining fast reconnection rates.

Contribution

It demonstrates that plasmoid instability is highly sensitive to numerical resolution, challenging its proposed role in fast magnetic reconnection.

Findings

Well-resolved simulations show no plasmoid instability at high Lundquist numbers.

Under-resolved simulations exhibit larger plasmoids, indicating resolution dependence.

Results question the validity of plasmoid instability as a universal mechanism for fast reconnection.

Abstract

Explaining fast magnetic reconnection in electrically conducting plasmas has been a theoretical challenge in plasma physics since its first description by Eugene N. Parker. In the recent years the observed reconnection rate has been shown by numerical simulations to be explained by the plasmoid instability that appears in highly conductive plasmas. In this work we show that the plasmoid instability is very sensitive to the numerical resolution used. It is shown that well resolved runs display no plasmoid instability even at Lundquist number as large as $5\cdot10^5$ achieved at resolutions of $32\,768^2$ grid points. On the contrary in simulations that are under-resolved below a threshold, the plasmoid instability manifests itself with the formation of larger plasmoids the larger the under-resolving is. The present results thus question the description of the plasmoid instability as a mechanism for fast magnetic reconnection.