Measures of Galaxy Environment - III. Difficulties in identifying proto-clusters at z ~ 2

TL;DR

This paper investigates the challenges of measuring galaxy environments at high redshift, especially proto-clusters at z ~ 2, highlighting the sensitivity of measurements to various factors and the implications for identifying and tracking proto-clusters over cosmic time.

Contribution

It demonstrates how environment metrics at high redshift are affected by observational limitations and viewing angles, and assesses the likelihood of proto-clusters remaining overdense at later times.

Findings

Environment measurements are highly sensitive to scale, metric, and viewing angle.

Proto-clusters at z ~ 2 have a high probability (> 90%) of remaining overdense at z = 0.

Different methodologies and data types require careful comparison due to their impact on environment assessment.

Abstract

Galaxy environment is frequently discussed, but inconsistently defined. It is especially difficult to measure at high redshift where only photometric redshifts are available. With a focus on early forming proto-clusters, we use a semi-analytical model of galaxy formation to show how the environment measurement around high redshift galaxies is sensitive to both scale and metric, as well as to cluster viewing angle, evolutionary state, and the availability of either spectroscopic or photometric data. We use two types of environment metrics (nearest neighbour and fixed aperture) at a range of scales on simulated high-z clusters to see how "observed" overdensities compare to "real" overdensities. We also "observationally" identify z = 2 proto-cluster candidates in our model and track the growth histories of their parent halos through time, considering in particular their final state at z = 0. Although the measured environment of early forming clusters is critically dependent on all of the above effects (and in particular the viewing angle), we show that such clusters are very likely (< 90%) to remain overdense at z = 0, although many will no longer be among the most massive. Object to object comparisons using different methodologies and different data, however, require much more caution.