Multiple stellar populations in Schwarzschild modeling and the application to the Fornax dwarf

TL;DR

This paper extends Schwarzschild modeling to include multiple stellar populations, improving dark matter and anisotropy profile constraints in dwarf galaxies like Fornax, through tests on simulated and real data.

Contribution

The authors develop and test an extended Schwarzschild method that incorporates multiple stellar populations and flexible mass-to-light ratios for better galaxy modeling.

Findings

Multiple stellar populations improve constraints on density profiles.

The method reveals differences in anisotropy between populations.

Application to Fornax shows more accurate anisotropy profiles.

Abstract

Dwarf spheroidal (dSph) galaxies are believed to be strongly dark matter dominated and thus are considered perfect objects to study dark matter distribution and test theories of structure formation. They possess resolved, multiple stellar populations that offer new possibilities for modeling. A promising tool for the dynamical modeling of these objects is the Schwarzschild orbit superposition method. In this work we extend our previous implementation of the scheme to include more than one population of stars and a more general form of the mass-to-light ratio function. We tested the improved approach on a nearly spherical, gas-free galaxy formed in the cosmological context from the Illustris simulation. We modeled the binned velocity moments for stars split into two populations by metallicity and demonstrate that in spite of larger sampling errors the increased number of constraints leads to significantly tighter confidence regions on the recovered density and velocity anisotropy profiles. We then applied the method to the Fornax dSph galaxy with stars similarly divided into two populations. In comparison with our earlier work, we find the anisotropy parameter to be slightly increasing, rather than decreasing, with radius and more strongly constrained. We are also able to infer anisotropy for each stellar population separately and find them to be significantly different.