Bayesian Reconstruction of the Velocity Distribution of Weakly Interacting Massive Particles from Direct Dark Matter Detection Data

Abstract: In this paper, we extended our earlier work on the reconstruction of the (time-averaged) one-dimensional velocity distribution of Galactic Weakly Interacting Massive Particles (WIMPs) and introduce the Bayesian fitting procedure to the theoretically predicted velocity distribution functions. In this reconstruction process, the (rough) velocity distribution reconstructed by using raw data from direct Dark Matter detection experiments directly, i.e. measured recoil energies, with one or more different target materials, has been used as "reconstructed-input" information. By assuming a fitting velocity distribution function and scanning the parameter space based on the Bayesian analysis, the astronomical characteristic parameters, e.g. the Solar and Earth's Galactic velocities, will be pinned down as the output results. Our Monte-Carlo simulations show that this Bayesian scanning procedure could reconstruct the true (input) WIMP velocity distribution function pretty precisely with negligible systematic deviations of the reconstructed characteristic Solar and Earth's velocities and 1 sigma statistical uncertainties of <~ 20 km/s. Moreover, for the use of an improper fitting velocity distribution function, our reconstruction process could still offer useful information about the shape of the velocity distribution. In addition, by comparing these estimates to theoretical predictions, one could distinguish different (basic) functional forms of the theoretically predicted one-dimensional WIMP velocity distribution function with 2 sigma to 4 sigma confidence levels.