Probing Dark Matter with the DELVE Milky Way Satellite Galaxy Census

The satellite dwarf galaxies of the Milky Way are powerful probes of small-scale dark matter physics. Their number counts are sensitive to dark matter particle properties, enabling tests that distinguish the cold dark matter paradigm from its alternatives. However, the known population of Milky Way satellite galaxies, consisting of $\sim 68$ systems, remains highly incomplete and selection-biased, limiting direct comparisons with theoretical predictions.

One of the most effective ways to address incompleteness in the dwarf galaxy population is to conduct a systematic census over wide-field surveys, thereby producing a sample that can be robustly compared with simulations. In this thesis, I present the deepest systematic census of Milky Way satellite galaxies to date, constructed by combining data from the third data release DECam Local Volume Exploration survey (DELVE DR3) with Pan-STARRS1. The newly constructed DELVE DR3 spans more than $20{,}000\,\deg^2$ of the southern sky in the $griz$ bands to a depth of $g \gtrsim 24$, enabling the detection of increasingly faint satellites.

For this census, I developed a galaxy-detection pipeline to identify stellar overdensities that trace a metal-poor locus in color-magnitude space, producing a sample of 49 Milky Way satellite galaxies. I then characterized the census selection function through injection-and-recovery tests with artificial satellites. Assuming an empirical population model, I inferred from the census the  luminosity function and radial distribution of the total Milky Way satellite population. Finally, I compared the census with predictions from the Symphony simulations using galaxy--halo connection models, placing constraints on warm dark matter particle masses by excluding models that underpredict the number of observable galaxies.

The deeper DELVE DR3 data also led to the discovery of four new Milky Way satellites: Leo VI, Carina IV, Phoenix III, and DELVE 7. Using dwarf galaxies from previous PS1 and DES data, I also set leading constraints on mixed cold and thermal-relic warm dark matter models, as well as mixed cold and sterile-neutrino dark matter models.