AVISM: Algorithm for Void Identification in coSMology

TL;DR

AVISM introduces a novel void finder that jointly utilizes density and velocity divergence on a grid to identify expanding voids in both simulated and observed large-scale structures. It reconstructs voids by growing and merging cubes, then leverages a multi-level grid hierarchy to reveal void substructure, producing complex, non-spherical boundaries without prior shape assumptions. The method is validated on mock tests, diverse cosmological simulations with multiple tracers, and real survey data, and it is benchmarked against DIVE and ZOBOV, showing robust performance in terms of void statistics and cross-matcher consistency, especially for large voids. AVISM demonstrates strong computational efficiency, scalability, and compatibility with external field reconstructions for surveys, making it a practical tool for studying voids and their cosmological implications across simulations and observations.

Abstract

Cosmic voids are key elements in our understanding of the large-scale structure of the Universe. They are crucial to constrain cosmological parameters, understand the structure formation and evolution of our Universe, and they could also be pristine laboratories for studying galaxy formation without all the hassle due to environmental effects. Thus, the ability to accurately and consistently identify voids, both in numerical simulations and in observations, becomes mandatory. We present Algorithm for Void Identification in coSMology (AVISM), a new void finder for analysing both cosmological simulation outputs and observational galaxy catalogues. In the first case, the code should handle raw particle or cell data, dark matter halos or synthetic galaxy catalogues. In the case of observational data, the code should be coupled with external tools providing with the required dynamical information to apply the algorithm. A set of numerical tests designed to assess the code's capabilities are carried out. AVISM's performance is also compared, both statistically and on a one-to-one basis, with the DIVE and ZOBOV state-of-the-art void finders using as input a dark matter halo catalogue from a large-volume cosmological simulation. An application to a galaxy survey is provided to demonstrate the code's ability to handle real data. We have designed a new void finder algorithm that combines geometrical and dynamical information to identify void regions plus a hierarchical merging process to reconstruct the whole 3D structure of the void. The outcome of this process is a void catalogue with complex boundaries without assuming a prior shape. This process can be repeated at different levels of resolution using finer grids, leading to a list of voids-in-voids and a proper description of void substructure.

Figures (12)

• Figure 1: Sketch of the void-finding procedure in an idealised 2D case. Top panel a) shows the complete set of volume-ordered cubes $\{C_i\}_{i=1}^{N_C}$ covering a region susceptible to be a void. Bottom panel b) displays how the algorithm is able to correctly group the cubes to produce well-separated voids illustrated in different colours.
• Figure 2: Relative errors between the mock void sample and the one obtained by AVISM as defined in the main text for four quantities: centre offset (top right), effective radius (top left), ellipticity (bottom left), and inverse porosity (bottom right). Colours stand for results for small (blue), intermediate (gold), and large (red) voids. The text within each panel displays the mean error of the considered quantity for each population.
• Figure 3: Distribution of voids obtained by AVISM when applied to a $z = 0$ snapshot of the TNG300-2 simulation on four different matter tracers: dark matter particles, gas particles, halos and galaxies. The images show, for each tracer, all voids intersecting a thin slice of 302.6 $\mathrm{Mpc}$ side and $\approx 10 \; \mathrm{Mpc}$ depth, together with the integrated density field, for which a colour-bar is displayed. Voids matching another from the reference catalogue (using halos as matter tracers) with DSC coefficient larger (smaller) than 0.4 are displayed using the same colour and continuous (dotted) lines.
• Figure 4: Top panel: void size function for the different void catalogues obtained by AVISM when run on four different numerical tracers of the TNG300-2 cosmological simulation from the IllustrisTNG suite. Bottom panel: relative difference with respect to the reference void size function obtained using halos as tracers.
• Figure 5: Slice of a zoom in on a region centred at a $R_e \approx 40 \; \mathrm{Mpc}$ void at $\ell = 0$ (dark blue solid line) together with its biggest sub-void at $\ell = 1$ (light blue dash-dotted line) and a substructure of that sub-void at $\ell = 2$ (white dashed line). The slice is $5 \; \mathrm{Mpc}$ depth. The colour palette displays the integrated density contrast. The analysis was performed on a snapshot of a MASCLET simulation at $z = 0$. More substructures are obtained for the same void and its sub-voids; however, only one of each kind is shown for the sake of clarity.