Speaker: Andreea Font (LJMU)

Date: Wednesday 29 May 2024

Time: 15:00

Location: 3.29

Disentangling the formation histories of Milky Way-type galaxies with cosmological simulations and machine learning

Over the past few years, there have been major advances in our understanding of the assembly history of the Milky Way, thanks to exquisite data from dedicated Galactic surveys (such as Gaia, GALAH or APOGEE) and more realistic zoomed-in cosmological simulations (e.g., ARTEMIS and Auriga). Currently, most of the puzzle of the Milky Way’s merger history has been re-assembled, however there are still some pieces that are missing, particularly with regards to the number and accretion times of dwarf galaxies that merged with our Galaxy in the very early stages of its formation.

For a complete picture of the merger history of the Milky Way we need to be able to disentangle the information about these early accretion events from the chemo-dynamical structure of the stellar debris in the halo. This is, however, a challenging task, because much of this information is hidden in regions of the Galaxy which are dominated by the stellar disc or by the in-situ stellar halo. In this talk, I will describe how we employ a combination of hydrodynamical cosmological simulations (the ARTEMIS suite) and machine learning techniques (e.g., emulators, artificial neural networks, decision trees) to constrain the full merger history of the Milky Way. I will highlight the areas in which machine learning can make meaningful progress and outline the main challenges that still lie ahead in this area.