Master Thesis Defense by Liam Mads Eichstedlund de Búrca

Title: Decrypting AGN Spectra

Abstract: The spectra from active galactic nuclei (AGN) are cryptic at best. They are the product of unresolved, highly energetic, and rapidly moving gas near an actively accreting supermassive black hole (SMBH). The resultant emission lines are broadened to such an extent, that many form groups, or complexes, making it impossible to separate and analyse a single emission line individually. However, if we endeavour to understand the dynamics, and evolution of AGN, as well as their influence on their host galaxies, we must be able to study the emission originating from them. Furthermore, we need a common, agreed-upon approach, such that findings of AGN studies are not sensitive to the methods employed.

The goal of my thesis is to develop a complete spectral decomposition method and software, backed by physical reasoning and statistical tests, which reverse engineers UV-optical AGN spectra reliably. My method utilises a stepwise approach, where I fit the power law continuum within statistically determined, continuum-dominated wavelength intervals, and proceed to fit emission lines with the help from machine learning models. I also develop robust methods for detecting narrow absorption lines, as well as computationally efficient methods for calculating errors on parameters. The final code performs comparably with other codes, but in a fraction of the time.

Supervisor:

• Marianne Vestergaard, University of Copenhagen, Niels Bohr Institute

Censor:

• Hans Kjeldsen, Aarhus University