The mass of the dark matter halo of the Milky Way can be estimated by fitting analytical models to the phase-space distribution of dynamical tracers. We test this approach using realistic mock stellar halos constructed from the Aquarius N-body simula
tions of dark matter halos in the $Lambda$CDM cosmology. We extend the standard treatment to include a Navarro-Frenk-White (NFW) potential and use a maximum likelihood method to recover the parameters describing the simulated halos from the positions and velocities of their mock halo stars. We find that the estimate of halo mass is highly correlated with the estimate of halo concentration. The best-fit halo masses within the virial radius, $R_{200}$, are biased, ranging from a 40% underestimate to a 5% overestimate in the best case (when the tangential velocities of the tracers are included). There are several sources of bias. Deviations from dynamical equilibrium can potentially cause significant bias; deviations from spherical symmetry are relatively less important. Fits to stars at different galactocentric radii can give different mass estimates. By contrast, the model gives good constraints on the mass within the half-mass radius of tracers even when restricted to tracers within 60kpc. The recovered velocity anisotropies of tracers, $beta$, are biased systematically, but this does not affect other parameters if tangential velocity data are used as constraints.
While various codes exist to systematically and robustly find haloes and subhaloes in cosmological simulations (Knebe et al., 2011, Onions et al., 2012), this is the first work to introduce and rigorously test codes that find tidal debris (streams an
d other unbound substructure) in fully cosmological simulations of structure formation. We use one tracking and three non-tracking codes to identify substructure (bound and unbound) in a Milky Way type simulation from the Aquarius suite (Springel et al., 2008) and post-process their output with a common pipeline to determine the properties of these substructures in a uniform way. By using output from a fully cosmological simulation, we also take a step beyond previous studies of tidal debris that have used simple toy models. We find that both tracking and non-tracking codes agree well on the identification of subhaloes and more importantly, the {em unbound tidal features} associated with them. The distributions of basic properties of the total substructure distribution (mass, velocity dispersion, position) are recovered with a scatter of $sim20%$. Using the tracking code as our reference, we show that the non-tracking codes identify complex tidal debris with purities of $sim40%$. Analysing the results of the substructure finders, we find that the general distribution of {em substructures} differ significantly from the distribution of bound {em subhaloes}. Most importantly, both bound and unbound {em substructures} together constitute $sim18%$ of the host halo mass, which is a factor of $sim2$ higher than the fraction in self-bound {em subhaloes}. However, this result is restricted by the remaining challenge to cleanly define when an unbound structure has become part of the host halo. Nevertheless, the more general substructure distribution provides a more complete picture of a halos accretion history.
We report evidence for extended gamma-ray emission from the Virgo, Fornax and Coma clusters based on a maximum-likelihood analysis of the 3-year Fermi-LAT data. For all three clusters, excess emission is observed within three degrees of the center, p
eaking at the GeV scale. This emission cannot be accounted for by known Fermi sources or by the galactic and extragalactic backgrounds. If interpreted as annihilation emission from supersymmetric dark matter (DM) particles, the data prefer models with a particle mass in the range 20-60 GeV annihilating into the b-bbar channel, or 2-10 GeV and >1 TeV annihilating into mu-mu final states. Our results are consistent with those obtained by Hooper and Linden from a recent analysis of Fermi-LAT data in the region of the Galactic Centre. An extended DM annihilation profile dominated by emission from substructures is preferred over a simple point source model. The significance of DM detection is 4.4 sigma in Virgo and lower in the other two clusters. We also consider the possibility that the excess emission arises from cosmic ray (CR) induced gamma-rays, and infer a CR level within a factor of three of that expected from analytical models. However, the significance of a CR component is lower than the significance of a DM component, and there is no need for such a CR component in the presence of a DM component in the preferred DM mass range. We also set flux and cross-section upper limits for DM annihilation into the b-bbar and mu-mu channels in all three clusters.
