Equilibrium axisymmetric halo model for the Milky Way and its implications for direct and indirect dark matter searches

Abstract

We for the first time provide self-consistent axisymmetric phase-space distribution models for the Milky Way's dark matter (DM) halo which are carefully matched against the latest kinematic measurements through Bayesian analysis. By using broad priors on the individual galactic components, we derive conservative estimates for the astrophysical factors entering the interpretation of direct and indirect DM searches. While the resulting DM density profiles are in good agreement with previous studies, implying $\rho_\odot \approx 10^{-2} \, M_\odot / \mathrm{pc}^3$, the presence of baryonic disc leads to significant differences in the local DM velocity distribution in comparison with the standard halo model. For direct detection, this implies roughly 30% stronger cross-section limits at DM masses near detectors maximum sensitivity and up to an order of magnitude weaker limits at the lower end of the mass range. Furthermore, by performing Monte-Carlo simulations for the upcoming DARWIN and DarkSide-20k experiments, we demonstrate that upon successful detection of heavy DM with coupling just below the current limits, the carefully constructed axisymmetric models can eliminate bias and reduce uncertainties by more then 50% in the reconstructed DM coupling and mass, but also help in a more reliable determination of the scattering operator. Furthermore, the velocity anisotropies induced by the baryonic disc can lead to significantly larger annual modulation amplitude and sizable differences in the directional distribution of the expected DM-induced events. For indirect searches, we provide the differential $J$-factors and compute several moments of the relative velocity distribution that are needed for predicting the rate of velocity-dependent annihilations. However, we find that accurate predictions are still hindered by large uncertainties regarding the DM distribution near the galactic center.