Mass Functions, Luminosity Functions, and Completeness Measurements from Clustering Redshifts

TL;DR

This study uses clustering redshifts to derive stellar mass and luminosity functions for SDSS galaxies, enabling completeness measurements for BOSS and revealing galaxy evolution insights up to redshift 2.5.

Contribution

It introduces a method to obtain redshift distributions from photometric data using clustering redshifts, applied to SDSS and BOSS data, extending to higher redshifts.

Findings

Mass functions show little evolution between z=0.2 and 0.8.

Stellar mass completeness is 80% above 10^11.4 M_sun for 0.2<z<0.7.

Technique will be valuable for upcoming large photometric surveys.

Abstract

This paper presents stellar mass functions and i-band luminosity functions for Sloan Digital Sky Survey (SDSS) galaxies at $i < 21$ using clustering redshifts, from which we also compute targeting completeness measurements for the Baryon Oscillation Spectroscopic Survey (BOSS). Clustering redshifts is a method of obtaining the redshift distribution of a sample of galaxies with only photometric information by measuring the angular crosscorrelation with a spectroscopic sample in different redshift bins. We construct a spectroscopic sample containing data from the BOSS + eBOSS surveys, allowing us to recover redshift distributions from photometric data out to $z\simeq 2.5$. We produce k-corrected i-band luminosity functions and stellar mass functions by applying clustering redshifts to SDSS DR8 galaxies in small bins of colour and magnitude. There is little evolution in the mass function between $0.2 < z < 0.8$, implying the most massive galaxies form most of their mass before $z = 0.8$. These mass functions are used to produce stellar mass completeness estimates for the Baryon Oscillation Spectroscopic Survey (BOSS), giving a stellar mass completeness of $80\%$ above $M_{\star} > 10^{11.4}$ between $0.2 < z < 0.7$, with completeness falling significantly at redshifts higher than 0.7, and at lower masses. Large photometric datasets will be available in the near future (DECaLS, DES, Euclid), so this, and similar techniques will become increasingly useful in order to fully utilise this data.