Decaying dark matter: the case for a deep X-ray observation of Draco

Recent studies of M31, the Galactic Centre (GC), and galaxy clusters have made tentative detections of an X-ray line at ∼3.5 keV that could be produced by decaying dark matter. We use high-resolution simulations of the Aquarius project to predict the likely amplitude of the X-ray decay flux observed in the GC relative to that observed in M31, and also of the GC relative to other parts of the Milky Way halo and to dwarf spheroidal galaxies. We show that the reported detections from M31 and the GC are compatible with each other, and with upper limits arising from high galactic latitude observations, and imply a decay time τ ∼ 1028 s. We argue that this interpretation can be tested with deep observations of dwarf spheroidal galaxies: in 95 per cent of our mock observations, a 1.3 Ms pointed observation of Draco with XMM–Newton will enable us to discover or rule out at the 3σ level an X-ray feature from dark matter decay at 3.5 keV, for decay times τ < 0.8 × 1028 s.