The effects of the antenna power pattern uncertainty within a global 21 cm experiment

Abstract

Experimental 21 cm cosmology aims to detect the formation of the first stars during the cosmic dawn and the subsequent epoch of reionization by utilizing the 21 cm hydrogen line transition. While several experiments have published results that begin to constrain the shape of this signal, a definitive detection has yet to be achieved. In this paper, we investigate the influence of uncertain antenna-sky interactions on the possibility of detecting the signal. This paper aims to define the level of accuracy to which a simulated antenna beam pattern is required to agree with the actual observing beam pattern of the antenna to allow for a confident detection of the global 21 cm signal. By utilising singular value decomposition, we construct a set of antenna power patterns that incorporate minor, physically motivated variations. We take the absolute mean averaged difference between the original beam and the perturbed beam averaged over frequency ($ΔD$) to quantifying this difference, identifying the correlation between $ΔD$ and antenna temperature. To analyse the impact of $ΔD$ on making a confident detection, we utilize the REACH Bayesian analysis pipeline and compare the Bayesian evidence $\log \mathcal{Z}$ and root-mean-square error for antenna beams of different $ΔD$ values. Our calculations suggest that achieving an agreement between the original and perturbed antenna power pattern with $ΔD$ better than -35 dB is necessary for confident detection of the global 21 cm signal. Furthermore, we discuss potential methods to achieve the required high level of accuracy within a global 21~cm experiment.