On the degeneracy between $f\sigma_8$ tension and its Gaussian process forecasting

TL;DR

This paper uses Gaussian processes and MCMC algorithms to reconstruct cosmic structure growth, comparing different hyperparameter estimation methods and analyzing their impact on the $f\sigma_8$ tension.

Contribution

It introduces a detailed comparison of hyperparameter optimization methods in Gaussian process reconstructions of cosmic growth and assesses their influence on the $f\sigma_8$ tension.

Findings

Difference between hyperparameter methods is about 1%

No significant difference with Planck 2018 $\Lambda$CDM results

Reconstruction methods yield consistent $f\sigma_8$ estimates

Abstract

In this paper we reconstruct the growth and evolution of the cosmic structure of the Universe using Markov Chain Monte Carlo algorithms for Gaussian processes [1]. We estimate the difference between the reconstructions that are calculated through a maximization of the kernel hyperparameters and those that are obtained with a complete exploration of the parameter space. We find that the difference between these two approaches is of the order of $1\%$. Furthermore, we compare our results with those obtained by Planck Collaboration 2018 assuming a $\Lambda$CDM model and we do not find a statistically significant difference in the redshift range were the reconstructions of $f\sigma_{8}$ have been made.