IMAGINE: Testing a Bayesian pipeline for Galactic Magnetic Field model optimization

TL;DR

This paper evaluates the early development and debugging of the IMAGINE Bayesian pipeline for Galactic magnetic field modeling, highlighting challenges and lessons learned in handling complex parametric models.

Contribution

It provides insights into the difficulties faced in Bayesian analysis of high-dimensional Galactic magnetic field models and documents the development process of the IMAGINE pipeline.

Findings

Identified the need for caution with parameter dependencies and model errors.

Resolved issues in the pipeline that improved its robustness.

Documented lessons learned during pipeline development.

Abstract

This work contains the details and results of my master's project on testing the IMAGINE pipeline for Galactic magnetic field estimation. The project was carried out from early 2016 to early 2017. For it, an unpublished early development version of the IMAGINE pipeline was tested and debugged. The thesis reports about the kind of difficulties faced when dealing with high dimensional complex parametric Galactic magnetic field models. It was found that such models require extra caution to allow for dependencies between parameters and model implementation errors, which need to be taken into account when performing a Bayesian analysis. These findings, reported here in this thesis, helped to resolve such issues in the later, now published version of the IMAGINE pipeline. The thesis therefore documents the genesis of the pipeline and lessons learned during this process. This document contains original text of the master thesis for reference. Parts of its content therefore do not reflect the current state of the IMAGINE pipeline.