The Imprint of the Baryon Acoustic Oscillations (BAO) in the Cross-correlation of the Redshifted HI 21-cm Signal and the Ly-alpha Forest

TL;DR

This paper proposes using the cross-correlation of the Ly-alpha forest and 21-cm emission to detect baryon acoustic oscillations, offering a promising method less affected by foregrounds and sampling issues.

Contribution

The authors develop a theoretical framework to predict the cross-correlation signal and its variance, and estimate observational parameters needed for BAO detection.

Findings

Detection of BAO is feasible with current and upcoming surveys.

Specific noise levels are required for BOSS and BIGBOSS to detect BAO.

Multiple radio arrays can achieve the necessary observations within a few years.

Abstract

The cross-correlation of the Ly-alpha forest and redshifted 21-cm emission has recently been proposed as an observational tool for mapping out the large-scale structures in the post-reionization era z < 6. This has a significant advantage as the problems of continuum subtraction and foreground removal are expected to be considerably less severe in comparison to the respective auto-correlation signals. Further, the effect of discrete quasar sampling is less severe for the cross-correlation in comparison to the Ly-alpha forest auto-correlation signal. In this paper we explore the possibility of using the cross-correlation signal to detect the baryon acoustic oscillation (BAO). To this end, we have developed a theoretical formalism to calculate the expected cross-correlation signal and its variance. We have used this to predict the expected signal, and estimate the range of observational parameters where a detection is possible. For the Ly-$\alpha$ forest, we have considered BOSS and BIGBOSS which are expected have a quasar density of 16 deg^{-2} and 64 deg^{-2} respectively. A radio interferometric array that covers the redshift range z=2 to 3 using antennas of size 2 m * 2 m, is well suited for the 21-cm observations. It is required to observe 25 independent fields of view, which corresponds to the entire angular extent of BOSS. We find that it is necessary to achieve a noise level of 1.1 * 10^{-5} K^2 and 6.25 * 10^{-6} mK^2 per field of view in the 21-cm observations to detect the angular and radial BAO respectively with BOSS. The corresponding figures are 3.3 * 10^{-5} mK^2 and 1.7 * 10^{-5} mK^2 for BIGBOSS. Four to five independent radio interferometric arrays, each containing 400 antennas uniformly sampling all the baselines within 50 m will be able to carry out these observations in the span of a few years.