Cosmic void exclusion models and their impact on the distance scale measurements from large scale structure

TL;DR

This paper develops and tests two numerical models for cosmic void clustering that improve the robustness and accuracy of cosmological distance measurements from large-scale structure data, especially in BAO studies.

Contribution

The paper introduces two new numerical models for void exclusion effects that outperform traditional models in robustness and consistency for cosmological parameter estimation.

Findings

Models provide unbiased Alcock-Paczynski parameter estimates.

Models are resilient against systematic effects.

Bayesian evidence favors the numerical models over the parabolic model.

Abstract

Baryonic Acoustic Oscillations (BAOs) studies based on the clustering of voids and matter tracers provide important constraints on cosmological parameters related to the expansion of the Universe. However, modelling the void exclusion effect is an important challenge for fully exploiting the potential of this kind of analyses. We thus develop two numerical methods to describe the clustering of cosmic voids. Neither model requires additional cosmological information beyond that assumed within the galaxy de-wiggled model. The models consist in power spectra whose performance we assess in comparison to a parabolic model on Patchy cubic and light-cone mocks. Moreover, we test their robustness against systematic effects and the reconstruction technique. The void model power spectra and the parabolic model with a fixed parameter provide strongly correlated values for the Alcock-Paczynski ($\alpha$) parameter, for boxes and light-cones likewise. The resulting $\alpha$ values -- for all three models -- are unbiased and their uncertainties are correctly estimated. However, the numerical models show less variation with the fitting range compared to the parabolic one. The Bayesian evidence suggests that the numerical techniques are often favoured compared to the parabolic model. Moreover, the void model power spectra computed on boxes can describe the void clustering from light-cones as well as from boxes. The same void model power spectra can be used for the study of pre- and post-reconstructed data-sets. Lastly, the two numerical techniques are resilient against the studied systematic effects. Consequently, using either of the two new void models, one can more robustly measure cosmological parameters.