Updating the MACHO fraction of the Milky Way dark halo with improved mass models

Recent interest in primordial black holes as a possible dark matter candidate has motivated the reanalysis of previous methods for constraining massive astrophysical compact objects in the Milky Way halo and beyond. In order to derive these constraints, a model for the dark matter distribution around the Milky Way must be used. Previous microlensing searches have assumed a semi-isothermal density sphere for this task. We show this model is no longer consistent with data from the Milky Way rotation curve, and test two replacement models, namely Navarro–Frenk–White (NFW) and power law. The power-law model is the most flexible as it can break spherical symmetry, and best fits the data. Thus, we recommend the power-law model as a replacement, although it still lacks the flexibility to fully encapsulate all possible shapes of the Milky Way halo. We then use the power-law model to rederive some previous microlensing constraints in the literature, while propagating the primary halo-shape uncertainties through to our final constraints. Our analysis reveals that the microlensing constraints towards the Large Magellanic Cloud weaken somewhat for massive astrophysical compact halo object (MACHO) masses around 10 M_⊙ when this uncertainty is taken into account, but the constraints tighten at lower masses. Exploring some of the simplifying assumptions of previous constraints we also study the effect of wide mass distributions of compact halo objects, as well as the effect of spatial clustering on microlensing constraints. We find that both effects induce a shift in the constraints towards smaller masses, and can effectively remove the microlensing constraints from M ∼ 1–10 M_⊙ for certain MACHO populations.