Siblings, friends and acquaintances: Testing galaxy association methods

In order to constrain the limitations of association methods applied to galaxy surveys, we analysed the catalogue of haloes at z = 0 of a cosmological simulation, trying to reproduce the limitations dealt with by an observational survey. We focused on the percolation method, usually called the friends of friends method, commonly used in the literature. The analysis was carried out on the dark matter cosmological simulation MDPL2, from the MultiDark project. The results point to a large fraction of contaminants for massive haloes in high-density environments. Thresholds in the association parameters and the subsequent analysis of observational properties can mitigate the occurrence of fake positives. The use of tests for substructures can also be efficient in particular cases.