Speaker
Description
Gaia has transformed galactic dynamics, yet even DR3 samples only a small and highly selective fraction of the Milky Way's stars.
Turning these data into a complete picture of the Galaxy's mass distribution therefore requires methods that can handle incomplete data and propagate uncertainties consistently.
Information field theory (IFT) offers such a framework.
It treats unknown quantities, such as stellar densities or gravitational potentials, as fields and infers them within a Bayesian framework with flexible, physically motivated priors.
Its numerical implementation, the Python package NIFTy, makes this inference feasible at the scale of Gaia data sets and provides posterior uncertainties.
We apply this approach to two complementary probes of the potential.
The first is a sample of stars near the Sun with full 6D phase-space information, complete for a particular stellar population.
Their kinematics trace the local gravitational potential and thus the local dark matter density.
We present our estimate and discuss its systematic uncertainties.
The second is a stellar stream, a coherent structure of stars stripped from a globular cluster or dwarf galaxy, whose trajectory is shaped by the Galaxy's large-scale potential.
Using mock data generated in a realistic Milky-Way-like potential, we demonstrate that this method can constrain parameters such as the axis ratios and orientation of the potential.
Finally, we discuss how the larger samples, improved astrometry, and epoch data of Gaia DR4 and DR5 will sharpen both measurements.