Studies of Baryonic Physics and Triaxiality in Galaxy Clusters

This dissertation concerns the impact of baryonic physics and triaxiality on galaxy clusters. The first and primary portion of this work is to investigate the impact baryonic processes have on galaxy cluster masses, profiles, and weak lensing mass measurements. To that end, we use the cosmo-OWLS suite of hydrodynamic simulations which incorporate a variety of sub-grid baryonic processes. We obtain mass estimates by fitting spherical NFW profiles to mock weak lensing data using MCMC techniques, in particular examining the difference between dark matter-only runs and those including baryons. We find no significant difference in the mass bias when including baryonic physics. The overall masses, however, are suppressed with the introduction of feedback from active galactic nuclei (AGN). For the lowest mass systems for which a reliable mass can be obtained, we find a bias of ≈ −10 percent. The magnitude of the bias tends to decrease as mass increases and is consistent with no bias for the most massive clusters. For the lowest mass clusters, the bias is particularly sensitive to the fit radii and concentration prior, making reliable mass estimates difficult. Additionally, we find that the scatter in mass estimates due to the presence of baryons is less than the scatter between different projections of individual clusters. The next portion of the work principally concerns cluster shape. We calculate the moment of inertia tensors for ≈1100 clusters in the large volume cosmo-OWLS simulations, using all the particles bound to the clusters. From this information, we derive the cluster axes and orientations in 3D. The cluster axis distributions can be used as priors when fitting triaxial models. We also find that, in keeping with previous studies, clusters tend to be prolate. We develop a new pipeline for fitting triaxial mass models to weak lensing, strong lensing, X-ray and eventually to SZ data. This pipeline is applied to synthetic data sets from cosmological simulations, and to the CLASH data set. In particular, a new 2D prior on the shape of clusters is determined by mapping the mass using weak lensing data. The projections of the 3D models must be consistent with these 2D weak lensing priors. The strong lensing signals are incorporated into the pipeline applied to real data by requiring that the projections of the 3D models have 2D cylinder masses that are consistent with the arc features in the clusters.