redMaPPer. I. ALGORITHM AND SDSS DR8 CATALOG

Abstract

We describe redMaPPer, a new red-sequence cluster finder specifically\ndesigned to make optimal use of ongoing and near-future large photometric\nsurveys. The algorithm has multiple attractive features: (1) It can iteratively\nself-train the red-sequence model based on minimal spectroscopic training\nsample, an important feature for high redshift surveys; (2) It can handle\ncomplex masks with varying depth; (3) It produces cluster-appropriate random\npoints to enable large-scale structure studies; (4) All clusters are assigned a\nfull redshift probability distribution P(z); (5) Similarly, clusters can have\nmultiple candidate central galaxies, each with corresponding centering\nprobabilities; (6) The algorithm is parallel and numerically efficient: it can\nrun a Dark Energy Survey-like catalog in ~500 CPU hours; (7) The algorithm\nexhibits excellent photometric redshift performance, the richness estimates are\ntightly correlated with external mass proxies, and the completeness and purity\nof the corresponding catalogs is superb. We apply the redMaPPer algorithm to\n~10,000 deg^2 of SDSS DR8 data, and present the resulting catalog of ~25,000\nclusters over the redshift range 0.08<z<0.55. The redMaPPer photometric\nredshifts are nearly Gaussian, with a scatter \\sigma_z ~ 0.006 at z~0.1,\nincreasing to \\sigma_z~0.02 at z~0.5 due to increased photometric noise near\nthe survey limit. The median value for |\\Delta z|/(1+z) for the full sample is\n0.006. The incidence of projection effects is low (<=5%). Detailed performance\ncomparisons of the redMaPPer DR8 cluster catalog to X-ray and SZ catalogs are\npresented in a companion paper (Rozo & Rykoff 2014).\n