Scaling properties of a complete X-ray selected galaxy group sample^⋆

Argelander Institute for Astronomy (AIfA), University of Bonn, Auf dem Hügel 71, 53121 Bonn, Germany
e-mail: lorenzo@astro.uni-bonn.de

Received: 7 April 2014
Accepted: 11 September 2014

Abstract

Context. Upcoming X-ray surveys like eROSITA require precise calibration between X-ray observables and mass down to the low-mass regime to set tight constraints on the fundamental cosmological parameters. Since an individual mass measurement is only possible for relatively few objects, it is crucial to have reliable and clearly understood scaling relations that relate the total mass to easily observable quantities.

Aims. The main goal of this work is to constrain the galaxy group scaling relations corrected for selection effects, and to quantify the influence of non-gravitational physics at the low-mass regime.

Methods. We analyzed XMM-Newton observations for a complete sample of galaxy groups selected from the ROSAT All-Sky Survey and compared the derived scaling properties with a galaxy cluster sample. To investigate the role played by the different non-gravitational processes we then compared the observational data with the predictions of hydrodynamical simulations.

Results. After applying the correction for selection effects (e.g., Malmquist bias), the L_X-M relation is steeper than the observed one. Its slope (1.66 ± 0.22) is also steeper than the value obtained by using the more massive systems of the HIFLUGCS sample. This behavior can be explained by a gradual change of the true L_X-M relation, which should be taken into account when converting the observational parameters into masses. The other observed scaling relations (not corrected for selection biases) do not show any break, although the comparison with the simulations suggests that feedback processes play an important role in the formation and evolution of galaxy groups. Thanks to our master sample of 82 objects spanning two order of magnitude in mass, we tightly constrain the dependence of the gas mass fraction on the total mass, finding a difference of almost a factor of two between groups and clusters. We also found that the use of different AtomDB versions in the calculation of the group properties (e.g., temperature and density) yields a gas fraction of up to 20% lower than an older version.