Stellar halos offer a clear path to working out the most dominant satellite to have merged with a galaxy. This is what Richard D’Souza and I claim in a paper just submitted to the Monthly Notices of the Royal Astronomical Society.
Many galaxies have a diffuse halo of stars around them; these stars are thought to be primarily from other satellite galaxies that fall into and merge into the main one (see Bullock et al. 2001 and Bullock & Johnston 2005; see also Bell et al. 2008). Tides tear off the stars from these satellite galaxies and spread them into a diffuse halo. There are uncertainties in this picture; it is possible that stars from the main galaxy (in situ stars) are moved into the diffuse halo by changes in the gravitational potential, and nobody knows which galaxies exactly were tidally shredded to make what halo.
In this paper, we were interested in understanding how the chemical composition (metallicity; a joint probe of how many generations of stars have lived and died in a galaxy, coupled with how well that galaxy keeps its chemical elements rather than spewing them out into intergalactic space) of these stellar halos might be related to the total amount of stars in a stellar halo. Observations (Harmsen et al. 2017) show such a relationship exists, and models predicted such a relationship (Deason et al. 2016), and we wanted to understand how tight such a relationship might be and what factors drive its shape and scatter.
Accreted metallicity of the 6 GHOSTS galaxies in addition to the Milky Way and M31 plotted as a function of their stellar halo mass measured between 10 and 40 kpc. The gray points indicated the accreted stellar metallicity estimated in a wedge along the minor axis for Illustris GHOSTS-like galaxies as a function of the accreted stellar mass measured in an aperture between 10 and 40 kpc. Top: We compare the distribution function of the aperture accreted stellar halo masses of Illustris GHOSTS-like galaxies (grey) with the aperture stellar halo masses of the GHOSTS data (red) measured between 10 and 40 kpc. In contrast, the distribution function of the aperture stellar halo masses (in situ + accreted) of Illustris GHOSTS-like galaxies is not consistent with the observational data.
We use the Illustris hydrodynamical simulations – a powerful theoretical tool for exploring the interplay of physical processes relevant for galaxy formation – to explore this issue. Importantly, if in situ stars are prominent at in halos at the level predicted by Illustris, halos would be too massive and would show too weak a metallicity-mass relation – Illustris dramatically over-predicts in situ stars. Consequently, we focused on accreted stars – stars torn from satellite galaxies through tides. These accreted stars a mass-metallicity relation close to the observed one (see figure, Figure 3 from the paper).
We learned that the metallicity-mass relation of stellar halos exists largely because halos are often dominated by a single disrupted dwarf galaxy – it was massive so drives up the mass of the halo, and it was metal-rich (because large galaxies are more metal rich) and drove up the metallicity of the halo. The metallicity of the halo is a sensitive probe of the extent to which a halo is dominated by the single most massive satellite disruption. This is potentially very important – we should be able to learn about the most massive satellite that merged with the main galaxy by measuring the metallicity of a galaxy’s stellar halo.
We also find that density and metallicity gradients in halos are steeper for earlier accretion/merger times. This conclusion appears to be rather more sensitive to modeling details but nonetheless opens up the intriguing possibility that we can learn about the time of the most important merger by studying the structures of a sample of stellar halos from nearby galaxies.