Effects of completeness and purity on cluster dark energy constraints

Abstract

The statistical properties of galaxy clusters can only be used for cosmological purposes if observational effects related to cluster detection are accurately characterized. These effects include the selection function associated to cluster finder algorithms and survey strategy. The importance of the selection becomes apparent when different cluster finders are applied to the same galaxy catalog, producing different cluster samples. We consider parametrized functional forms for the observable-mass relation, its scatter as well as the completeness and purity of cluster samples, and study how prior knowledge on these function parameters affects dark energy constraints derived from cluster statistics. Under the assumption that completeness and purity reach 50 % at masses around 10^{13.5} Msun/h, we find that self-calibration of selection parameters in current and upcoming cluster surveys is possible, while still allowing for competitive dark energy constraints. We consider a fiducial survey with specifications similar to those of the Dark Energy Survey (DES) with 5000 deg^2, maximum redshift of zmax ~ 1.0 and threshold observed mass M_{th} ~ 10^{13.8} Msun/h, such that completeness and purity ~ 60 % - 80 % at masses around M_{th}. Perfect knowledge of all selection parameters allows for constraining a constant dark energy equation of state to sigma(w)=0.033. Employing a joint fit including self-calibration of the effective selection degrades constraints to sigma(w)=0.046. External calibrations at the level of 1 % in the parameters of the observable-mass relation and completeness/purity functions are necessary to improve the joint constraints to sigma(w)=0.041. In the lack of knowledge of selection parameters, future experiments probing larger areas and greater depths suffer from stronger relative degradations on dark energy constraints compared to current surveys.