Hierarchical analysis of the quiet Sun magnetism

TL;DR

This paper introduces a hierarchical probabilistic approach to analyze quiet Sun magnetism, overcoming biases of traditional histogram methods by accounting for noise and degeneracies, revealing very weak, mostly horizontal magnetic fields.

Contribution

The study applies a novel hierarchical Bayesian method to infer magnetic field properties from Hinode data, improving accuracy over standard histogram analyses.

Findings

Magnetic fields are very weak, below 275 G with 95% credibility.

Slight preference for horizontal magnetic fields.

Distribution close to quasi-isotropic.

Abstract

Standard statistical analysis of the magnetic properties of the quiet Sun rely on simple histograms of quantities inferred from maximum-likelihood estimations. Because of the inherent degeneracies, either intrinsic or induced by the noise, this approach is not optimal and can lead to highly biased results. We carry out a meta-analysis of the magnetism of the quiet Sun from Hinode observations using a hierarchical probabilistic method. This model allows us to infer the statistical properties of the magnetic field vector over the observed field-of-view consistently taking into account the uncertainties in each pixel due to noise and degeneracies. Our results point out that the magnetic fields are very weak, below 275 G with 95% credibility, with a slight preference for horizontal fields, although the distribution is not far from a quasi-isotropic distribution.