Model fitting of kink waves in the solar atmosphere: Gaussian damping and time-dependence

TL;DR

This paper introduces robust statistical techniques for comparing models of kink wave damping in the solar atmosphere, demonstrating their effectiveness with observational data and revealing insights into wave physics and time-dependent behavior.

Contribution

It develops and applies statistical model comparison methods, including residual analysis and goodness-of-fit tests, to solar coronal loop oscillations, improving the reliability of model selection.

Findings

Gaussian damping better fits observed wave attenuation than exponential damping.

Evidence of time-dependent wave properties linked to plasma thermodynamics.

Statistical techniques enhance model comparison accuracy in solar wave studies.

Abstract

{Observations of the solar atmosphere have shown that magnetohydrodynamic waves are ubiquitous throughout. Improvements in instrumentation and the techniques used for measurement of the waves now enables subtleties of competing theoretical models to be compared with the observed waves behaviour. Some studies have already begun to undertake this process. However, the techniques employed for model comparison have generally been unsuitable and can lead to erroneous conclusions about the best model. The aim here is to introduce some robust statistical techniques for model comparison to the solar waves community, drawing on the experiences from other areas of astrophysics. In the process, we also aim to investigate the physics of coronal loop oscillations. } {The methodology exploits least-squares fitting to compare models to observational data. We demonstrate that the residuals between the model and observations contain significant information about the ability for the model to describe the observations, and show how they can be assessed using various statistical tests. In particular we discuss the Kolmogorov-Smirnoff one and two sample tests, as well as the runs test. We also highlight the importance of including any observational trend line in the model-fitting process.} {To demonstrate the methodology, an observation of an oscillating coronal loop undergoing standing kink motion is used. The model comparison techniques provide evidence that a Gaussian damping profile provides a better description of the observed wave attenuation than the often used exponential profile. This supports previous analysis from Pascoe et al. (2016). Further, we use the model comparison to provide evidence of time-dependent wave properties of a kink oscillation, attributing the behaviour to the thermodynamic evolution of the local plasma.}