The Impact of Galaxy Cluster Mergers on Cosmological Parameter Estimation from Surveys of the Sunyaev-Zel'dovich Effect

TL;DR

This study investigates how galaxy cluster mergers affect the accuracy of cosmological parameters derived from Sunyaev-Zel'dovich effect surveys, finding that the integrated Comptonization parameter Y is more reliable than y_max.

Contribution

The paper demonstrates that cluster mergers transiently boost SZ observables and shows that Y remains a robust estimator for cosmological parameters despite these effects.

Findings

Y is less affected by mergers than y_max.

Merger boosts cause significant biases in y_max-based estimates.

Y provides more accurate cosmological constraints in the presence of mergers.

Abstract

Sensitive surveys of the Cosmic Microwave Background will detect thousands of galaxy clusters via the Sunyaev-Zel'dovich (SZ) effect. Two SZ observables, the central or maximum and integrated Comptonization parameters y_max and Y, relate in a simple way to the total cluster mass, which allow the construction of mass functions (MFs) that can be used to estimate cosmological parameters such as Omega_M, sigma_8, and the dark energy parameter w. However, clusters form from the mergers of smaller structures, events that can disrupt the equilibrium of intracluster gas upon which SZ-M relations rely. From a set of N-body/hydrodynamical simulations of binary cluster mergers, we calculate the evolution of Y and y_max over the course of merger events and find that both parameters are transiently "boosted," primarily during the first core passage. We then use a semi-analytic technique developed by Randall et al. (2002) to estimate the effect of merger boosts on the distribution functions YF and yF of Y and y_max, respectively, via cluster merger histories determined from extended Press-Schechter (PS) merger trees. We find that boosts do not induce an overall systematic effect on YFs, and the values of Omega_M, sigma_8, and w were returned to within 2% of values expected from the nonboosted YFs. The boosted yFs are significantly biased, however, causing Omega_M to be underestimated by 15-45%, sigma_8 to be overestimated by 10-25%, and w to be pushed to more negative values by 25-45%. We confirm that the integrated SZ effect, Y, is far more robust to mergers than y_max, as previously reported by Motl et al. (2005) and similarly found for the X-ray equivalent Y_X, and we conclude that Y is the superior choice for constraining cosmological parameters.