Seen and Unseen Tidal Caustics in the Andromeda Galaxy

Abstract

Indirect detection of high-energy particles from dark matter interactions is a promising avenue for learning more about dark matter, but is hampered by the frequent coincidence of high-energy astrophysical sources of such particles with putative high-density regions of dark matter. We calculate the boost factor and gamma-ray flux from dark matter associated with two shell-like caustics of luminous tidal debris recently discovered around the Andromeda galaxy, under the assumption that dark matter is its own supersymmetric antiparticle. These shell features could be a good candidate for indirect detection of dark matter via gamma rays because they are located far from the primary confusion sources at the galaxy's center, and because the shapes of the shells indicate that most of the mass has piled up near apocenter. Using a numerical estimator specifically calibrated to estimate densities in N-body representations with sharp features and a previously determined N-body model of the shells, we find that the largest boost factors do occur in the shells but are only a few percent. We also find that the gamma-ray flux is an order of magnitude too low to be detected with Fermi for likely dark matter parameters, and about 2 orders of magnitude less than the signal that would have come from the dwarf galaxy that produces the shells in the N-body model. We further show that the radial density profiles and relative radial spacing of the shells, in either dark or luminous matter, is relatively insensitive to the details of the potential of the host galaxy but depends in a predictable way on the velocity dispersion of the progenitor galaxy.

Enhancements to velocity-dependent dark matter interactions from tidal streams and shells in the Andromeda galaxy

“…The proposed test involves Fermi imaging of very similar substructures (at least in terms of velocity dispersion) to those invoked locally for the PAMELA excess. Without the Sommerfeld enhancement, the material is insufficiently dense to produce a detectable signal (Sanderson & Bertschinger 2010), but with the enhancement signals can be boosted by a factor of up to 10 4 –10 5 at velocities comparable to those achieved in the tidal debris in this example. These boost factors are similar to those expected from local dwarf galaxies (Robertson & Zentner 2009).…”

“…The integrated squared density, weighted by a factor of either 1/ v or 1/ v 2 , was estimated using the optimal procedure identified in Sanderson & Bertschinger (2010), with the addition of estimates for the mean velocity (used to account for interactions between shell and halo dark matter particles) and the velocity dispersion for material in the shell.…”

“…Finally, the two components of Γ involving the N ‐body model are computed using the estimators and where the Riemann sum is over the volumes V i making up the target volume V , and and are estimated at the centre x i of each Riemann volume as in Sanderson & Bertschinger (2010). The halo density ρ h is evaluated at the centre of each Riemann volume for consistency with the positional accuracy of the density estimates.…”

“…The N ‐body model of the debris uses about 1.3 × 10 5 particles to represent the entire tidal structure, of which about 4 × 10 4 end up in Caustic 1 and 2 × 10 4 end up in Caustic 2. This level of resolution is sufficient to resolve the density distribution of the material if a suitable estimator with a suitably small smoothing number ( N s = 10) is used (Sanderson & Bertschinger 2010), but this does not guarantee that the velocity structure of the material is adequately resolved. To maximize the resolution of the velocity structure, we would like to use the smallest possible smoothing number to estimate moments of the velocity distribution as well.…”

“…We follow the notation of Fornengo, Pieri & Scopel (2004) as adapted by Sanderson & Bertschinger (2010) to calculate the gamma‐ray signal, d N γ /d t , for Fermi . As in Fornengo et al (2004), we separate the dependence on the phase‐space distribution of material from most of the details of the dark matter particle model: The first term, ϕ C , depends only on the mass and velocity distribution of the dark matter and the velocity dependence of the Sommerfeld effect.…”