Systematic Problems With Using Dark Matter Simulations to Model Stellar Halos

Abstract

The limits of available computing power have forced models for the structure of stellar halos to adopt one or both of the following simplifying assumptions: (1) stellar mass can be "painted" onto dark matter (DM) particles in progenitor satellites; (2) pure DM simulations that do not form a luminous galaxy can be used. We estimate the magnitude of the systematic errors introduced by these assumptions using a controlled set of stellar halo models where we independently vary whether we look at star particles or painted DM particles, and whether we use a simulation in which a baryonic disk galaxy forms or a matching pure DM simulation that does not form a baryonic disk. We find that the "painting" simplification reduces the halo concentration and internal structure, predominantly because painted DM particles have different kinematics from star particles even when both are buried deep in the potential well of the satellite. The simplification of using pure DM simulations reduces the concentration further, but increases the internal structure, and results in a more prolate stellar halo. These differences can be a factor of 1.5-7 in concentration (as measured by the half-mass radius) and 2-7 in internal density structure. Given this level of systematic uncertainty, one should be wary of overinterpreting differences between observations and the current generation of stellar halo models based on DM-only simulations when such differences are less than an order of magnitude.

Creating mock catalogues of stellar haloes from cosmological simulations

“…To explore the consequences of such limitations, Bailin et al (2014) compared tagged dark matter particles in hydrodynamical simulations to star particles formed self-consistently in the same simulations. They found moderate disagreement in the distributions of halo stars represented by these two sets of particles, and concluded that tightly bound dark matter is not a reliable proxy for the dynamics of stars in satellite galaxies.…”

“…They found moderate disagreement in the distributions of halo stars represented by these two sets of particles, and concluded that tightly bound dark matter is not a reliable proxy for the dynamics of stars in satellite galaxies. However, there is a fundamental difference between the tagging method tested by Bailin et al (2014) and that used by C10. Bailin et al (2014) carried out a single tagging procedure for each satellite at the time of its infall into the main halo, whereas C10 tag each stellar population at the time of its formation.…”

“…However, there is a fundamental difference between the tagging method tested by Bailin et al (2014) and that used by C10. Bailin et al (2014) carried out a single tagging procedure for each satellite at the time of its infall into the main halo, whereas C10 tag each stellar population at the time of its formation. Star particles undergo considerable diffusion in energy over time, particularly if their host halo is disturbed, and hence would not be expected to occupy the same region of phase space as tightly bound dark matter chosen long after they form.…”

“…Dark matter particles with similar binding energy chosen at the time of star formation, on the other hand, will undergo similar diffusion to star particles. This diffusion is therefore captured in the C10 approach, but not in the method used by Bailin et al (2014). Two forthcoming papers will test the tagging technique of C10 in a similar way (Le Bret et al, in preparation; Cooper et al, in preparation).…”

Origins and Interpretation of Tidal Debris

“…While this agreement is encouraging, when Helmi et al (2011) repeated the Bell et al (2008) analysis on the Cooper et al (2010) model stellar halos (which were derived by "painting" stars in the "Aquarius" self-consistent dark matter simulations), they found systematically larger deviations from a smooth background than the prior work at a level that was inconsistent with the observations. Recent work by Bailin et al (2014) contrasting simulations with and without Galactic disk components suggest that this inconsistency with both observations and the work might be attributed to the Aquarius simulations lacking the extra potential structure due to the disk.…”