Probabilistic Super-Resolution of Solar Magnetograms: Generating Many Explanations and Measuring Uncertainties

TL;DR

This paper introduces a Bayesian super-resolution method for solar magnetograms that quantifies uncertainties and generates multiple plausible high-resolution explanations, addressing the need for error estimation in scientific imaging.

Contribution

It presents a novel Bayesian framework that decomposes uncertainties and applies it to super-resolve solar magnetic field images, providing uncertainty measures and multiple explanations.

Findings

Successfully super-resolved solar magnetograms with uncertainty quantification

Generated multiple high-resolution explanations consistent with low-resolution data

Validated the approach on solar magnetic field images

Abstract

Machine learning techniques have been successfully applied to super-resolution tasks on natural images where visually pleasing results are sufficient. However in many scientific domains this is not adequate and estimations of errors and uncertainties are crucial. To address this issue we propose a Bayesian framework that decomposes uncertainties into epistemic and aleatoric uncertainties. We test the validity of our approach by super-resolving images of the Sun's magnetic field and by generating maps measuring the range of possible high resolution explanations compatible with a given low resolution magnetogram.