• astroparticle physics;
  • dark matter;
  • gamma-rays: galaxies: clusters


Clusters of galaxies are potentially important targets for indirect searches for dark matter (DM) annihilation. Here we reassess the detection prospects for annihilation in massive haloes, based on a statistical investigation of 1743 clusters in the new Meta-Catalogue of X-ray Clusters (MCXC). We derive a new limit for the extragalactic DM annihilation background of at least 20 per cent of that originating from the Galaxy for an integration angle of 0°.1. The number of clusters scales as a power law with their brightness (boosted by DM substructures), suggesting that stacking may provide a significant improvement over a single target analysis. The mean angle containing 80 per cent of the DM signal for the sample (assuming a Navarro–Frenk–White DM profile) is ∼0°.15 (excluding the contribution from the point spread function of any instrument), indicating that instruments with this angular resolution or better would be optimal for a cluster annihilation search based on stacking. A detailed study based on the Fermi-LAT performance and position-dependent background suggests that stacking may result in a factor of ∼2 improvement in sensitivity, depending on the source selection criteria. Based on the expected performance of Cherenkov Telescope Array, we find no improvement with stacking, due to the requirement for pointed observations. We note that several potentially important targets – Opiuchius, A2199, A3627 (Norma) and CIZA J1324.7−5736 – may be disfavoured due to a poor contrast with respect to the Galactic DM signal. The use of the homogenized MCXC meta-catalogue provides a robust ranking of the targets, although the absolute value of their signal depends on the exact DM substructure content. For conservative assumptions, we find that galaxy clusters (with or without stacking) can probe 〈σv〉 down to 10−25–10−24 cm3 s−1 for DM masses in the range 10–100 GeV. For more favourable substructure configurations, 〈σv〉 ∼ 10−26 cm3 s−1 may be reached.