The Number Densities and Stellar Populations of Massive Galaxies at 3 < z < 6: A Diverse, Rapidly Forming Population in the Early Universe

Abstract

We present the census of massive (log(M${}_{\ast }$/M${}_{\odot }$)$\ge 11$) galaxies at $3<z<6$ identified over the COSMOS/UltraVISTA Ultra-Deep field stripes: consisting of $\approx 100$ and $\approx 20$ high-confidence candidates at $3<z<4$ and $4<z<6$, respectively. The $3<z<4$ population is comprised of post-starburst, UV star-forming and dusty-star forming galaxies in roughly equal fractions, while UV-star-forming galaxies dominate at $4<z<6$ . We account for various sources of biases in SED modelling, finding that the treatment of emission line contamination is essential for understanding the number densities and mass growth histories of massive galaxies at $z>3$. The significant increase in observed number densities at $z\sim 4$ ($>\times$ 5 in $\lesssim 600$ Myrs) implies that this is the epoch at which log(M${}_{\ast }$/M${}_{\odot }$)$\ge 11$ galaxies emerge in significant numbers, with stellar ages ($\approx 500-900$ Myrs) indicating rapid formation epochs as early as $z\sim 7$. Leveraging ancillary multi-wavelength datasets, we perform panchromatic SED modelling to constrain the total star-formation activity of the sample. The star-formation activity of the sample is generally consistent with being on the star-formation main sequence at the considered redshifts, with $\approx 15-25\%$ of the population showing evidence of suppressed star-formation rates, indicating that quenching mechanisms are already at play by $z\sim 4$. We stack available HST imaging, confirming their compact nature ($r_e\lesssim 2.2$ kpc), consistent with expected sizes of high-$z$ star-forming galaxies. Finally, we discuss how our results are in-line with the early formation epochs and short formation timescales inferred from the fossil records of the most massive galaxies in the Universe.

22 pages, 13 figures, 3 tables (31 pages, 16 figures, 4 tables with Appendix). ApJ submitted. Comments welcome!