Dissecting the Thermal SZ Power Spectrum by Halo Mass and Redshift in SPT-SZ Data and Simulations

TL;DR

This study analyzes the tSZ power spectrum's dependence on halo mass and redshift using SPT data and simulations, revealing model preferences and discrepancies that suggest the need for refined gas models or contamination considerations.

Contribution

It provides a detailed comparison of tSZ power contributions across halo masses and redshifts, highlighting the limitations of current gas models in matching observations.

Findings

SPT data favors models with mass-dependent gas fractions and high AGN feedback at low masses.

Discrepancies indicate simulations fail to reproduce the steep mass dependence observed.

No significant redshift evolution observed in the tSZ contribution, contrary to self-similar models.

Abstract

We explore the relationship between the thermal Sunyaev-Zel'dovich (tSZ) power spectrum amplitude and the halo mass and redshift of galaxy clusters in South Pole Telescope (SPT) data, in comparison with three $N$-body simulations combined with semi-analytical gas models of the intra-cluster medium. Specifically, we calculate both the raw and fractional power contribution to the full tSZ power spectrum amplitude at $\ell = 3000$ from clusters as a function of halo mass and redshift. We use nine mass bins in the range $1 \times 10^{14}\ M_\odot\ h^{-1} < M_{500} < 2 \times 10^{15}\ M_\odot\ h^{-1}$, and two redshift bins defined by $0.25 < z < 0.59$ and $0.59 < z < 1.5$. We additionally divide the raw power contribution in each mass bin by the number of clusters in that bin, as a metric for comparison of different gas models. At lower masses, the SPT data prefers a model that includes a mass-dependent bound gas fraction component and relatively high levels of AGN feedback, whereas at higher masses there is a preference for a model with a lower amount of feedback and a complete lack of non-thermal pressure support. The former provides the best fit to the data overall, in regards to all metrics for comparison. Still, discrepancies exist and the data notably exhibits a steep mass-dependence which all of the simulations fail to reproduce. This suggests the need for additional mass- and redshift-dependent adjustments to the gas models of each simulation, or the potential presence of contamination in the data at halo masses below the detection threshold of SPT-SZ. Furthermore, the data does not demonstrate significant redshift evolution in the per-cluster tSZ power spectrum contribution, in contrast to self-similar model predictions.