Optimizing high redshift galaxy surveys for environmental information

TL;DR

This paper evaluates how different survey design choices affect the ability to reconstruct galaxy groups at high redshift, using mock data and various metrics to optimize environmental information extraction.

Contribution

It introduces a comprehensive set of metrics to assess group-finding performance and analyzes the impact of survey parameters on high-redshift galaxy environment studies.

Findings

Group-finding algorithms perform broadly similarly at low redshift.

Survey design parameters significantly influence group identification accuracy.

Trade-offs exist between survey depth, sampling rate, and observational costs.

Abstract

We investigate the performance of group finding algorithms that reconstruct galaxy groups from the positional information of tracer galaxies that are observed in redshift surveys carried out with multiplexed spectrographs. We use mock light-cones produced by the L-Galaxies semi-analytic model of galaxy evolution in which the underlying reality is known. We particularly focus on the performance at high redshift, and how this is affected by choices of the mass of the tracer galaxies (largely equivalent to their co-moving number density) and the (assumed random) sampling rate of these tracers. We first however compare two different approaches to group finding as applied at low redshift, and conclude that these are broadly comparable. For simplicity we adopt just one of these, "Friends-of-Friends" (FoF) as the basis for our study at high redshift. We introduce 12 science metrics that are designed to quantify the performance of the group-finder as relevant for a wide range of science investigations with a group catalogue. These metrics examine the quality of the recovered group catalogue, the median halo masses of different richness structures, the scatter in dark matter halo mass and how successful the group-finder classifies singletons, centrals and satellites. We analyze how these metrics vary with the limiting stellar mass and random sampling rate of the tracer galaxies, allowing quantification of the various trade-offs between different possible survey designs. Finally, we look at the impact of these same design parameters on the relative "costs" in observation time of the survey using as an example the potential MOONRISE survey using the MOONS instrument.