Galactic archaeology in the 21st century : unveiling accretion in the MW disc
Cargando...
Archivos
Fecha
2022
Autores
Profesor/a Guía
Facultad/escuela
Idioma
en
Título de la revista
ISSN de la revista
Título del volumen
Editor
Universidad Andrés Bello
Nombre de Curso
Licencia CC
Licencia CC
Resumen
The evolution of large galaxies such as the Milky Way (MW) in Λ Cold Dark Matter (ΛCDM)
cosmology is driven by interactions with other galaxies. These interactions affect the structure
of the galaxy in long-lasting ways, not only forming the stellar halo, and modifying the size and
characteristics of the stellar disc, but also contributing gas that fuels star formation and depositing
stars directly into the galaxy. This stellar debris, formed outside the galaxy in which they now reside
(a.k.a. ex-situ), retain the chemical fingerprint of the environment in which they formed, leaving
them distinct from the stars that were formed within the primary galaxy (a.k.a. in-situ). If these
ex-situ stars are located beyond the disc, in the stellar halo, the distributions they occupy in space —
based on the kinematics of their progenitor object — may be readily apparent as streams or great
circles surrounding the primary galaxy.
A population of ex-situ stars are also expected to reside in the stellar disc, even at high circularities.
These stars have been found in simulations to be deposited by relatively few massive mergers, but
finding them poses a challenge in this dense region. I propose a method, based on training and
utilizing neural networks, to classify disc stars as in- or ex-situ based on their chemical parameters.
The use of this method is motivated by my research into the effects mergers and accretion have on the
galactic stellar disc of a suite of simulated MW-like galaxies, in which I demonstrate that accretion
events and accreted stars cause observable changes to calculated metallicity gradients, and can cause
disagreements in correlations between observed parameters based on the ages of the stars being used.
Galactic discs form inside-out, starting small and slowly expanding with new star formation at the
outer edges. This imparts a natural negative age gradient. Stars that have been deposited onto
the disc will not necessarily follow this age gradient, which is one of the primary drivers of the
stellar disc’s metallicity gradient. These ex-situ stars drive several age-specific correlations between metallicity gradient and 𝑅
−1
eff and 𝜆
★
gal that vanish entirely once these contaminants are removed. A
chemistry-based method for flagging potentially-accreted stars will allow researchers to remove the
bulk of these from their samples, which will allow them to examine the evolution of the galactic disc
more finely.
Furthermore, these flagged, ex-situ stars can be grouped by their characteristics, for attempts to
discern early progenitors of the primary galaxy with significantly less noise from stars formed in-situ.
I demonstrate that the Galactic Archaeology Neural Network (GANN) recovers usable fractions of
nearly every contributor to the suite of simulated stellar discs. Thus, a catalogue of stars flagged by
GANN will contain most of the merger information that is present in the stellar disc of a galaxy.
Future applications of this method could aid in the discovery of many ancient remnants, and broaden
our understanding of how our galaxy formed.
Notas
Tesis (Doctor en Astrofísica)
Palabras clave
Redes Neurales (Ciencia de la Computación), Galaxias, Formación, Arqueología Galáctica