The Disordered Heterogeneous Universe: Galaxy Distribution and Clustering Across Length Scales

Studies of disordered heterogeneous media and galaxy cosmology share a common goal: analyzing the distribution of particles at `microscales' to predict physical properties at `macroscales', whether for a liquid, composite material, or entire Universe. The former theory provides an array of techniques to characterize a wide class of microstructures; in this work, we apply them to the distributions of galaxies. We focus on the lower-order correlation functions, `void' and `particle' nearest-neighbor functions, pair-connectedness functions, percolation properties, and a scalar order metric. Compared to homogeneous Poisson and typical disordered systems, the cosmological simulations exhibit enhanced large-scale clustering and longer tails in the nearest-neighbor functions, due to the presence of quasi-long-range correlations. On large scales, the system appears `hyperuniform', due to primordial density fluctuations, whilst on the smallest scales, the system becomes almost `antihyperuniform', and, via the order metric, is shown to be a highly correlated disordered system. Via a finite scaling analysis, we show that the percolation threshold of the galaxy catalogs is significantly lower than for Poisson realizations; this is consistent with the observation that the galaxy distribution contains larger voids. However, the two sets of simulations share a fractal dimension, implying that they lie in the same universality class. Finally, we consider the ability of large-scale clustering statistics to constrain cosmological parameters using simulation-based inference. Both the nearest-neighbor distribution and pair-connectedness function considerably tighten bounds on the amplitude of cosmological fluctuations at a level equivalent to observing twenty-five times more galaxies. These are a useful alternative to the three-particle correlation, and are computable in much reduced time. (Abridged)