We explore the novel cosmological consequences of the Scalar Field Dark Matter (SFDM) model in the limit of strong repulsive self-interactions, known as the Thomas-Fermi regime (SFDM-TF). The self-interaction provides pressure support that inhibits gravitational collapse on scales smaller than its Jeans length (R_TF), thereby suppressing small-scale structure formation relative to the standard Cold Dark Matter (CDM) paradigm. SFDM-TF was originally intended to solve the cusp-core and too-big-to-fail problems of CDM by adopting R_TF ~ 1 kpc, resulting in the formation of cores of this size in the centers of dark matter halos. However, by performing both a linear transfer function calculation and a set of nonlinear spherical collapse simulations, we find that such a model is far too suppressive in the early universe to be consistent with observations, which likely require R_TF <~ 10 pc. On the other hand, SFDM-TF's reduction in small-scale power is much gentler and retains significantly more power below its characteristic suppression scale than other suppressive dark matter models such as warm dark matter (WDM) and fuzzy dark matter (FDM). As a result, SFDM-TF may be uniquely well-suited to avoid the growing body of constraints facing WDM and FDM from observations of small-scale structure in the Lyman-alpha forest, Milky Way satellites, and perturbations in stellar streams, while still significantly suppressing small-scale power relative to CDM. 

In addition, we study the spatial and temporal variations in the faint end of the UV luminosity function (UVLF) of galaxies during the Epoch of Reionization -- correlated with local overdensity and redshift of reionization -- using the large-scale, high-resolution, radiation-hydrodynamics Cosmic Dawn (CoDa) simulations. Recent observations leveraging high-magnification gravitational lenses to detect faint galaxies at z = 6 have produced conflicting results for the shape of the faint end of the UVLF. While some studies indicate that the faint end follows a power law down to an absolute UV magnitude of M_UV = -12.5, others show that it "rolls-over" well before reaching that faint a magnitude. Though we find substantial variance in the UVLF on the scale of these lensing surveys, our results indicate a strong preference for the latter "roll-over" models, due to reionization photo-heating the gas in the intergalactic medium to a high enough temperature that low-mass dark matter halos (which would otherwise populate the faint end of the UVLF) are unable to accrete this gas and form stars. Furthermore, we find that previous methods used to estimate the uncertainty due to variance in the UVLF underestimate it by a factor of 2-4 on the scales relevant to the data at M_UV = -12.5. This is primarily due to the extremely high magnification required to detect such galaxies, which results in the effective volume searched for them being much smaller than the scales on which these previous estimates are appropriate. Finally, we compare our simulated high-z UVLFs to recent JWST observations.