Abstract Where in the present-day Milky Way should we search for the remnants of its earliest stars? We address this question using the Dorcha (Gaelic for Dark; DUR-uh-khuh) suite: a set of 25 high-resolution, dark-matter-only cosmological zoom-in simulations of Milky Way analogue (MWA) haloes evolved to z = 0. Of these, 15 are isolated and the rest are in pairs, similar to the MW and M31. By identifying and tagging the most bound material in high-redshift (z ≥ 5) progenitor haloes – those likely to host early star formation – we track the present-day phase-space distribution of this ancient component. We find that this material is highly centrally concentrated at z = 0, with 90 – 100 per cent residing within r ≲ 15 h−1 kpc. It exhibits steep density profiles (ρ∝ r−4), low velocity dispersions (σr/σmax ≲ 0. 6), and radially biased orbits (β ≳ 0. 5 for r ≳ 0. 1 R200), consistent with a relaxed, centrally embedded population. These results hold across haloes with diverse formation histories and environments, suggesting that the dynamical signature of early progenitors is robust to later mergers and interactions. Our findings imply that the fossil record of the first generations of stars – including Population III and extremely metal-poor stars – should be sought in the innermost regions of the Milky Way, where they retain distinctive kinematic imprints. While these stellar populations may overlap, we caution that low metallicity does not uniquely identify ancient stars, nor vice versa. The Dorcha suite thus provides a physically motivated baseline for interpreting observations from Galactic Archaeology surveys targeting the bulge and inner halo.
Balu et al. (Wed,) studied this question.