Abstract

Contributed Talk - Splinter LargeScale   (MW-2235)

Probing Large-Scale Structure with One-Point Statistics: Hierarchical Counts-in-Cells model for Radio Surveys

Morteza Pashapour-Ahmadabadi, Dominik J. Schwarz
Bielefeld University

Counts-in-cells statistics of extragalactic radio surveys encode the joint imprint of the cosmic web and the astrophysical features of radio galaxies within dark matter haloes. Existing models for pixelised source counts extend the Poisson assumption to capture over-dispersion arising from clustering due to multiple radio components of a single galaxy, and have been successfully applied to current radio continuum surveys. However, these approaches typically treat the mean source density per pixel as fixed, neglecting halo-to-halo variance in radio galaxy occupation predicted by the halo occupation distribution framework. We introduce a compound two-level hierarchical model that explicitly incorporates this additional source of scatter: the number of haloes per pixel is drawn from a distribution describing fluctuations in the underlying halo population, while each halo contributes radio sources according to an over-dispersed count model. We find that it provides a substantially better description of the observed count histograms than all single-level alternatives, with strong statistical evidence in its favour across a range of flux density thresholds and angular resolutions. Embedding this model within a full-sky Bayesian hierarchical framework, we reconstruct the underlying log-normal halo density field across the survey footprint — using only one-point statistics. This result opens a broader pathway to cross-correlate the reconstructed density field with complementary large-scale structure probes for independent constraints on cosmological parameters.