Determining the structure of the Galactic disc in the optical in the midplane through photometry alone is complicated due to the presence of dust. Recently, using the kinematic information provided by the Gaia eDR3 data, a new map of the Milky Way's outer disc was presented (Laporte et al. 2021) revealing a hierarchy of substructure in the disc using a peak detection algorithm inspired by those used to identify halo centers in cosmological simulations (Power et al. 2003). Nonetheless this is only the tip of the iceberg as close inspection reveals the presence of multiple kinematic peaks coinciding on the sky at a given position on the Celestial sphere. How long and extended are the various coherent disc stream-like structures, what is their overall kinematics, chemistry, age distribution of their stellar populations? Are all these substructures excited as a result of kinematic perturbations from satellite galaxies or do they represent folds in the disc seen in projection?
In this project you will devise an algorithm to refine the current algorithm and identify substructure across the whole disc using data from the Gaia satellite beyond the volume probed by the radial velocity survey (RVS). You will track every peak and connect them as individual coherent structures using Gaussian Mixture models (e.g. Bovy 2011). One obvious application would be to better characterise the outer disc of the Milky Way by dissecting it in a coherent way, but one could equally apply the algorithm to study other parts of the Galaxy (e.g. the bar) if time allows it.