A new void finder based on cosmic dynamics

Cosmic voids are large empty regions of the Universe, constituting a fundamental component of the Cosmic Web. Voids point out the scale at which density fluctuations have decoupled from the Hubble flow and formed overdense regions, such as galaxy clusters. Because of their low-density nature, voids are powerful probes to underline several crucial aspects of the Universe. Voids are formed from underdensities in the primordial field of density perturbations and may provide crucial information on the large scale structures. The main goal of this thesis is to implement a new Void Finder code based on a dynamical criterion The idea is to use tracers as test particles, reconstructing their orbits from the actual clustered configuration to a homogeneous and isotropic pristine distribution. This back-in-time evolution in the displacement field can be performed adopting two different approaches. The former assumes the Zel'dovich approximation, while the latter exploits the two-point correlation function. Since in both the approaches, the particle path depends on the initial random seeds, our Void Finder algorithm performs several reconstructions and then averages the obtained fields by using a Gaussian weighting. Then it computes the divergence field, that represents, by definition, the density at each point. The algorithm consider only the local minima with negative divergence, which identify the subvoids. Through a watershed method we can then ``fill'' the near subvoids in order to identify a local void. The new dynamical Void Finder implemented in this work has been included in the CosmoBolognaLib libraries. We compared this code to the LZVF algorithm, finding a good agreement in the detection of voids. We will extend this work, applying our algorithm to observed datasets, such as the VIPERS catalogues. Our Void Finder is currently competing in the new Euclid Void Challenge, whose primary goal is to compare different methods to detect cosmic voids.