Resolving the Metallicity Distribution of the Stellar Halo with the H3 Survey

Abstract

The Galactic stellar halo is predicted to have formed at least partially from the tidal disruption of accreted dwarf galaxies. This assembly history should be detectable in the orbital and chemical properties of stars. The H3 Survey is obtaining spectra for 200,000 stars, and, when combined with Gaia data, is providing detailed orbital and chemical properties of Galactic halo stars. Unlike previous surveys of the halo, the H3 target selection is based solely on magnitude and Gaia parallax; the survey therefore provides a nearly unbiased view of the entire stellar halo at high latitudes. In this paper we present the distribution of stellar metallicities as a function of Galactocentric distance and orbital properties for a sample of 4232 kinematically-selected halo giants to 100 kpc. The stellar halo is relatively metal-rich, [Fe/H] = −1.2, and there is no discernable metallicity gradient over the range 6 < R gal < 100 kpc. However, the halo metallicity distribution is highly structured including distinct metal-rich and metal-poor components at R gal < 10 kpc and R gal > 30 kpc, respectively. The Sagittarius stream dominates the metallicity distribution at 20 − 40 kpc for stars on prograde orbits. The Gaia-Enceladus merger remnant dominates the metallicity distribution for radial orbits to ≈ 30 kpc. Metal-poor stars with [Fe/H]< −2 are a small population of the halo at all distances and orbital categories. We associate the "in-situ" stellar halo with stars displaying thick-disk chemistry on halo-like orbits; such stars are confined to |z| < 10 kpc. The majority of the stellar halo is resolved into discrete features in orbital-chemical space, suggesting that the bulk of the stellar halo formed from the accretion and tidal disruption of dwarf galaxies. The relatively high metallicity of the halo derived in this work is a consequence of the unbiased selection function of halo stars, and, in combination with the recent upward revision of the total stellar halo mass, implies a Galactic halo metallicity that is typical for its mass.

The artemis simulations: stellar haloes of Milky Way-mass galaxies

“…A more recent analysis of the stellar halo using Gaia DR2 data has revealed that approximately half of the total halo population in the Solar neighborhood is likely to have been born in situ (Belokurov et al 2019). A significant population of in situ stars in the inner halo has also been revealed by the H3 survey (Conroy et al 2019). These findings lend strong credence to the early predictions made about the in situ halo using hydrodynamical simulations (Zolotov et al 2009;Font et al 2011).…”

“…Until recently, observational studies have targeted stars that were either metal-poor or kinematically distinct from the disc. With the recent confirmation of the importance of the in situ halo in the Milky Way (Belokurov et al 2019;Conroy et al 2019), and with strong evidence in favour of an important in situ component of the M31 halo (Dorman et al 2013), there is a need to provide accurate predictions for the nature of stellar haloes using cosmological simulations. Our study has provided detailed information about the structure of in situ and accreted components and has made a number of predictions which can be tested in the future.…”

“…Although these values are apparently similar in a visual inspection, quantitatively they are different, as discussed below. The observations include the haloes of six GHOSTS galaxies, for which [Fe/H] 10−40 have been measured along the minor axes (data from Harmsen et al 2017 are shown with filled black circles); we also include the haloes of M31 (red circle) and of the Milky Way (yellow circle), for the latter using the revised values from Deason et al (2011) and Conroy et al (2019). The simulated values lie comfortably within the bounds of the values obtained for the Milky Way and M31.…”

“…For the Milky Way, many halo samples are affected by selection biases, which may influence the metallicity measurements (see Conroy et al 2019). This may explain the discrepancy between the (weak) metallicity gradient found initially (Xue et al 2015;Carollo et al 2007), and a flat distribution around [Fe/H] ≈ −1.2 determined more recently (Conroy et al 2019). In contrast, M31 exhibits a strong metallicity gradient (Gilbert et al 2014;Ibata et al 2014), even when measured along the minor axis of the galaxy.…”

“…Given the difficulty of measuring metallicities across the full extent of the stellar haloes, current observations are less conclusive with regards to the frequency of metallicity gradients. For the Milky Way, many halo samples are affected by selection biases, which may influence the metallicity measurements (see Conroy et al 2019). This may explain the discrepancy between the (weak) metallicity gradient found initially (Xue et al 2015;Carollo et al 2007), and a flat distribution around [Fe/H] ≈ −1.2 determined more recently (Conroy et al 2019).…”