Gravity-Selected Galaxy Clusters: a Tight Mass-Richness relation and an unclear Compton $Y$-richness trend

TL;DR

This study addresses how to reliably infer cluster masses from non-baryon-selected samples by examining gravity-selected clusters. It analyzes a complete sample of 13 clusters at intermediate redshift, measuring richness from red-sequence galaxies, weak-lensing masses, and Compton $Y$ from SZ maps, using two distinct richness apertures. The authors find an exceptionally tight richness–mass relation with a scatter of about $0.05$–$0.06$ dex, while the Compton $Y$–richness relation shows no clear trend, though the mass–richness relation is strong. The results imply that optical richness is a superior, low-scatter mass proxy compared to SZ-derived $Y$ for gravity-selected clusters, with upcoming Euclid data expected to greatly expand the sample and enable robust multi-probe mass calibration.

Abstract

This paper, the third in a series, investigates the scaling relations between optical richness, weak-lensing mass, and Compton $Y$ for a sample of galaxy clusters selected purely by the effect of their gravitational potential on the shapes of background galaxies. This selection method is uncommon, as most cluster samples in the literature are selected based on signals originating from cluster baryons. We analize a complete sample of 13 gravity-selected clusters at intermediate redshifts (with $0.12 \leq z_{phot} \leq 0.40$) with weak-lensing signal-to-noise ratios exceeding 7. We measured cluster richness by counting red-sequence galaxies, identifying two cases of line-of-sight projections in the process, later confirmed by spectroscopic data. Both clusters are sufficiently separated in redshift that contamination in richness can be straighforwardly dealt because the two red sequences do not blend each other. We find an exceptionally tight richness--mass relation using our red-sequence-based richness estimator, with a scatter of just $\sim0.05$ dex, smaller than the intrinsic scatter of Compton Y with mass for the same sample. The lower scatter highlights the effectiveness of richness compared to Compton $Y$. No outliers are found in the richness-mass scaling, whether or not one cluster with a mass likely affected by projection effects is included in the sample. In the Compton $Y$-richness plane, the data do not delineate a clear trend. The limited sample size is not the sole reason for the unclear relation between Compton $Y$ and richness, since the same sample, with identical richness values, exhibits a highly significant and tight mass-richness correlation.

Figures (11)

• Figure 1: True-color ($grz$) HSC images with overlaid contours of the weak-lensing S-map, which serves as a proxy for signal-to-noise, convolved with a Gaussian filter of $\sigma = 2$ arcmin for display purposes. The red and cyan crosses mark the positions of the target cluster and other clusters (present in O26 and O40 panels), respectively. The true-color images are taken from https://www.legacysurvey.org/viewer
• Figure 2: Binned tangential shear profile of the O26 cluster, incorporating the updated identification. The solid line, with yellow shading, represents the mean model and the 68% uncertainty region. The uncertainty in the model also includes intrinsic scatter, while the error bars plotted account only for shape noise and large-scale structure. The dashed cyan line represents the maximum likelihood estimate (MLE), which is biased. The triangle denotes the $2\sigma$ upper limit.
• Figure 3: Expected O40 tangential shear profile if O40 had the fitted mass value ($\log M/M_\odot=14.5$), and of a $\log M/M_\odot=14.7$ cluster located 5.5 arcmin from O40. The figure demonstrates the negligible contribution of the contaminating cluster.
• Figure 4: Expected O26 tangential shear profile if O26 had the fitted mass value ($\log M/M_\odot=14.5$), and of a $\log M/M_\odot=14.5$ cluster located 100 arcsec from it. The figure demonstrates the possible large contribution of the contaminating cluster.
• Figure 5: Color distribution of galaxies within $r_{200,\rm wl}$ (red histogram) and in a control area (normalized to the cluster solid angle, blue histogram) for O32. The vertical lines indicate the expected color range of the red sequence. The peak on the right corresponds to the color expected for the cluster redshift, while the second peak represents a contaminating structure along the line of sight spectroscopically confirmed at $z=0.136$ (see text for details). For O32, the color range was reduced by 0.1 mag on the blue side to exclude this contamination. Only galaxies brighter than $M^e_V=-20$ mag, assuming all galaxies are at the O32 redshift, are considered, which boost the low redshift contamination.