Speaker
Description
In this talk, I will explore the interfacing of simulations, observations, and machine learning techniques to constrain the distribution of Dark Matter for indirect detection. I will focus on the Galactic Center and dwarf galaxies as laboratories for Dark Matter detection. For the Galactic Center, I will show how we can combine the theoretical understanding of adiabatic contraction with bursty feedback to predict the Dark Matter density profile in the inner Milky Way, and the resulting range of indirect-detection signals from the Galactic Center. For dwarf galaxies, I will present GraphNPE, a novel Graph Neural Network methodology that facilitates the accurate extraction of Dark Matter density profiles, validated against realistic FIRE-2 simulations and applied to two dwarf galaxies, Boötes I and Draco. I will conclude by arguing that the Milky Way and galactic dynamics more broadly offer a complementary probe of Dark Matter, showing direct examples of how stellar dynamics can be integrated with our understanding of Dark Matter in the Galaxy and its connection to detection experiments.