The Galaxy-Halo Connection
I am interested in the physics that drives galaxy formation and evolution. This consists of a wide array of processes. In our modern understanding, galaxies form within dark matter halos. These host halos presumably play a role in shaping the galaxies that form inside them, but the nature of dark matter makes this relationship particularly hard to divine. However, like trees in a forest, the properties of dark matter halos are tied to their environment. My research uses this connection, which is inherited by galaxies through their host halos, to learn about the nature of dark matter halos and their role in explaining the great diversity of galaxy shapes and sizes observed.
Machine Learning
The use of machine learning in science and in astronomy in particular is sky-rocketing! Machine learning provides a way forward for exploring complicated correlations and enormous datasets previously limited by traditional techniques. This is becoming more and more relevant as next-generation telescopes bring colossal observational catalogs and simulations grow in size. In my research, I use neural networks to learn the complex relationships between environment, galaxy, and dark matter halo. By training neural networks on simulated mock catalogs of galaxies and halos like those from UniverseMachine, my work opens a new window onto the invisible properties of dark matter halos and the galaxy-halo connection.
Cosmology with Roman
I am part of the team developing a cosmological parameters inference pipeline for the Roman High-Latitude Wide Area Survey. I am primarily focused on creating a neural network emulator to determine the effectiveness of the survey at constraining evolving dark energy. I am also involved in a project to create model vector emulators for Roman weak-lensing + clustering that will allow anyone to run a likelihood analysis in a python notebook. Check out the code section for more information on the pipeline and associated activities.
Papers and Other Links
Halo Properties from Observable Measures of Environment: I. Halo and Subhalo Masses: Bowden et al. 2023
Finding environmental measures sensitive to halo properties using neural networks (Talk at KITP, Talk at CCA/IAP)