Galaxy proto-clusters as an interface between structure, cluster, and galaxy formation

Proto-clusters, the distant progenitor large-scale structures of present day galaxy clusters, represent a key phase of cluster growth during which most of the galaxies were still rapidly forming stars. They are potentially powerful cosmological probes, and are unique laboratories to study dark matter assembly, the cosmic baryon cycle, and the environmental impact on galaxy evolution. Albeit its pivotal role in understanding cluster formation, only a small and heterogeneous sample of proto-clusters has been observed to date. Theoretical characterizations have also remained relatively unexplored. In this dissertation, I present baseline models, detailed theory predictions, and broad observational applications of proto-clusters using state-of-the-art numerical simulations and deep-wide galaxy surveys. A dual focus of both structure formation and galaxy evolution is given throughout the thesis. To prepare for large statistical studies in upcoming surveys like the Hobby-Eberly Telescope Dark Energy Experiment (HETDEX), the Subaru Prime Focus Spectrograph (PFS) survey, and the Wide Field Infrared Survey Telescope (WFIRST) survey, I develop key machinery to connect the main observables of proto-clusters with dark matter structure formation using simulations as a guide. In Chapter 2 and 3, I present, for the first time, a thorough analysis of the main properties of proto-clusters using ~3000 clusters in a set of cosmological N-body simulations and semi-analytic galaxy models. I characterize the growth of proto-clusters and their core halos in size and mass with cosmic time. I show that the progenitor regions of galaxy clusters can already be identified in galaxy surveys at very early times (at least up to z~5), provided that the galaxy overdensities are measured on a sufficiently large scale (5--30 Mpc comoving) and with sufficient statistics. I present the overdensities in matter, dark matter halos, and galaxies as functions of present-day cluster mass, redshift, bias, and selection window size that can be used to interpret the wide range of structures found in real surveys. A table of proto-cluster candidates selected from the literature is provided, and I discuss their properties in light of our simulation predictions. In Chapter 4 I report the discovery of a large sample of proto-cluster candidates in the 1.62 deg^2 COSMOS/UltraVISTA field traced by optical/infrared selected galaxies with photometric redshifts. By comparing properly smoothed three-dimensional galaxy density maps of the observations and a set of matched simulations incorporating the main observational effects, I found 36 candidate structures at 1.6<z<3.1 with a ~70% purity to have a z=0 virial mass of >10^14} M_sun. With solely photometric redshifts, I successfully rediscover two spectroscopically confirmed structures in this field, suggesting that our algorithm is robust. This work is the first large sample of uniformly selected proto-cluster candidates, providing rich targets for spectroscopic follow-up and subsequent studies of cluster formation. Because of the need of precise galaxy redshifts for density mapping and the prevalence of star formation before quenching, nearly all the proto-clusters known to date were confirmed by spectroscopy of galaxies with strong emission lines. In Chapter 5 I develop a semi-empirical model for Lya escape and generate a set of mock Lya emitter catalogs. This formalism provides a realistic modeling of the galaxy bias, the scatter of the bias, and the stochasticity of the galaxy-dark matter halo connection, which has an enormous potential for studies of the large-scale structure at high redshift. The model suggests that there are two distinct regimes to power a Lya emitter. For massive galaxies, Lya emitters are preferentially less dusty and slightly less metal enriched, while their ages and star formation rates are indistinguishable from other star-forming galaxies of the same mass. In contrast, low mass Lya emitters M_star<10^9 M_sun) are dominated by young objects with recent or ongoing starburst, with a gas phase metallicity diluted by cold accretion. In Chapter 6 I report a newly discovered large-scale structure at z=2.44 in the HETDEX Pilot Survey. On a scale of a few tens of (comoving) Mpc, this structure shows a complex overdensity of Lya emitters, which coincides with broad-band selected galaxies in the COSMOS/UltraVISTA photometric and zCOSMOS spectroscopic catalogs, as well as overdensities of intergalactic gas revealed by Lya absorption maps. I use the mock Lya emitter catalogs constructed in Chapter 5 to predict the cosmic evolution of this structure, and confirm that part of the structure will collapse to form a galaxy cluster with 10^14.5+-0.4 M_sun by z=0. The galaxies inside this structure have a higher median stellar mass than those outside the structure. The overdense region also shows a boost in the number of extended Lya nebulae, and a marginal excess of active galactic nuclei relative to the field, supporting a scenario of accelerated galaxy evolution in cluster progenitors. Building upon the success of finding and characterizing the z=2.44 proto-cluster in the HETDEX Pilot Survey, in Chapter 7 I use the same mock Lya emitter catalogs to evaluate the baseline performance of a large proto-cluster search in the upcoming HETDEX survey. Based on the correlation between galaxy overdensity and the z=0 descendant halo mass calibrated in the simulation, I predict that several hundred 1.9<z<3.5 proto-clusters with z=0 mass of >10^14.5 M_sun will be discovered in the 8.5 Gpc^3 of space surveyed by HETDEX. This sample will open up a rich, new area of statistical investigations of both structure formation and galaxy evolution processes inside dense structures. The future for this field is bright, as we are entering an era of Gpc^3 surveys beyond the local Universe. In Chapter 8 I close this thesis by providing a future outlook. Specific research directions with great potential to blossom are highlighted. With the theoretical and observational advancements laid out in this dissertation, we now have a much more solid foundation of this fascinating subject.