The First Billion Years with JWST+ALMA: Massive Galaxies and Active Black Holes at the Observational Frontier
The first billion years of the Universe saw the collapse of the first dark-matter halos, the formation of the first stars and galaxies, and the growth of the first supermassive black holes (SMBHs). The rarest and most luminous systems at this epoch—drawn from the exponential tail of the underlying halo mass function—provide outsized leverage on the physics by which baryons assemble into massive galaxies and SMBHs; but precisely because they are rare, characterizing them demands wide-area surveys and large samples. The James Webb Space Telescope (JWST) has now pushed the observational frontier to this early epoch and revealed unexpected populations of massive galaxies and active galactic nuclei (AGN). In this dissertation, I leverage wide-area JWST imaging in the COSMOS field to examine three extreme populations and their role in early galaxy evolution.
The bright end of the UV luminosity function (UVLF) at cosmic dawn traces the buildup of the first massive galaxies and is sensitive to the efficiency by which they convert gas into stars. I present a large sample of UV-luminous galaxy candidates at z>9 selected over ~0.35 sq. degrees of JWST imaging in the COSMOS field. The wide-area selection provides unique sensitivity to the rarest, most luminous objects: the sample includes 25 sources with intrinsic luminosities M_UV<−21 at z~8–11, comparable to or brighter than GN-z11. The UVLF exhibits a relatively shallow evolution at the bright end out to z~13. Reproducing this evolution through a simple empirical model based on the halo mass function requires a monotonically increasing star-formation efficiency with redshift, reaching ~20% at z>13; high, but not in tension with ΛCDM cosmology.
Through cosmic history, however, most of the growth of massive galaxies is obscured by dust. Characterizing the most intense sites of star formation thus requires a panchromatic approach. To that end, I present high-resolution ALMA [CII] 158μm observations and JWST NIRCam+MIRI imaging of MAMBO-9, a pair of optically dark dusty star-forming galaxies (DSFGs) at z=5.85. The new ALMA data resolve the gas kinematics of both components, reveal a continuous, moderately obscured tidal bridge between them, and—combined with previous CO observations—suggest a high gas fraction >50% and super-solar metallicity. With 39 galaxies spectroscopically confirmed in the surrounding ~25 cMpc region, MAMBO-9 emerges as a likely progenitor of a present-day brightest cluster galaxy at the center of a forming protocluster—extending the canonical picture of submillimeter galaxy formation to earlier times.
Finally, I explore the population of “little red dots” (LRDs): compact, extremely red sources that were largely not anticipated prior to the launch of JWST. While the LRDs were first identified as candidate massive, compact, dust-obscured galaxies at z>7, they may also harbor AGN, challenging interpretations of their demographics and contribution to galaxy formation. From the 0.54 sq. degree COSMOS-Web survey I assemble a sample of 434 LRDs at z~5–9 and fit each under either an AGN-dominated or a compact-starburst interpretation. While the LRDs can be explained by stellar models, their compact sizes would imply extreme central stellar densities, approach those of globular clusters. Under the AGN scenario, the implied bolometric luminosity function sits a factor of ~100 above the UV-selected quasar LF at the same epoch, even as stacks of the available X-ray, mid- and far-IR, submillimeter, and radio data return non-detections across the board. Targeted follow-up observations of individual LRDs probe these interpretations directly. I present NIRSpec observations of an LRD at z=7.04 with broad Hα and high-equivalent-width CIII] and CIV lines in the UV. The line ratios suggest a stellar population dominates the rest-UV, while an AGN powers the optical continuum. I also present ALMA observations of three LRDs at z=3–7, searching for signatures of molecular gas in their host galaxies. One object shows a tentative [CI] detection, with a narrow line width implying a dynamical mass an order of magnitude below the SED-derived stellar mass, favoring a non-stellar origin for the optical continuum. Together, these results establish the LRDs as abundant and likely AGN-dominated, yet stubbornly resistant to classification within the unified model of AGN.
The rarest objects in the first billion years offer the most demanding tests of how galaxies and black holes form, and the work here establishes the physical properties and demographics of the most extreme populations within JWST's reach. Large spectroscopic samples of LRDs and UV-luminous galaxies, now being assembled, provide the natural next step towards characterizing the physical processes at play in the formation of the first galaxies and black holes.
