Cosmological Forecasts from the Baryon Acoustic Oscillations in 21cm Intensity Mapping

TL;DR

This paper develops a template-based Fisher forecast for extracting the BAO shift parameter from 21cm intensity mapping with the BINGO telescope and propagating these constraints to cosmological parameters. By modeling the 21cm angular power spectrum $C_l$ through a template $C_l = B_0 C_{ ext{temp}}(l/\alpha) + A_0 + A_1 l + A_2/l^2$ and relating $\alpha$ to $d_A(z)/r_s$, the authors forecast the precision on $\alpha$ and then transform this into constraints on $H_0$, $w_0$, and $w_a$ using Planck priors to break degeneracies. They find that Planck+BINGO BAO yields strong constraints on the Hubble parameter and dark-energy equation of state, with $\sigma_h$ as low as $0.0055$ in $\Lambda$CDM and $\sigma_{w_0}$, $\sigma_{w_a}$ improving notably in $w$CDM and CPL parameterizations; however, the full angular power spectrum provides even tighter constraints at the expense of potential systematics. The work highlights 21cm BAO as a robust, complementary cosmological probe to CMB data, while acknowledging that full APS analyses currently outperform BAO-only approaches and that upcoming surveys like DESI may deliver stronger overall constraints.

Abstract

In this work we use a template method to extract the scale associated with the Baryon Acoustic Oscillation (BAO) signal in 21cm neutral hydrogen intensity maps. We then forecast the constraints on the standard deviations of cosmological parameters using a Fisher matrix analysis. In order to test this method, we choose the survey configuration for the BINGO telescope. We then estimate the constraints on the BAO shift parameter $α$, which we extract from the 21cm angular power spectrum (APS). In addition, we translate those results into constraints on the final cosmological parameters. As BAO data alone can only constrain the product of the Hubble constant and the sound horizon $H_0r_s$, degeneracies between the variables mean that we can't get useful constraints with BAO data alone. We break these degeneracies by combining the 21cm intensity mapping BAO results with the Cosmic Microwave Background (CMB) covariances obtained by the Planck satellite. In particular, we find that the best forecasts we can get with this combination are on the standard deviations of the Hubble parameter $σ_h$, and the dark energy parameters $σ_{w_0}$ and $σ_{w_a}$. We find $σ_h = 0.0055\;(0.8\%)$ in the $Λ$CDM model. For the $w$CDM model, we find $σ_h = 0.020\;(2.9\%)$ and $σ_{w_0} = 0.075\;(7.5\%)$. In the CPL parameterization, we find $σ_h = 0.029\;(4.4\%)$, $σ_{w_0} = 0.40\;(40\%)$, and $σ_{w_a} = 1.7$. Finally, we observe that using the full APS provides stronger constraints than the BAO only, however, it is more susceptible to systematic effects.

Figures (4)

• Figure 1: Angular power spectra plotted over 30 redshift bins from $z = 0.13$ (leftmost blue line) to $z = 0.44$ (rightmost red line).
• Figure 2: (Top panel) Comparison of $C_l^{\text{temp}}$ (green solid), $C_l^{\text{nw}}$ (blue dotted), and $C_l^{\text{lin}}$ (yellow dashed). The difference between $C_l^{\text{temp}}$ and $C_l^{\text{lin}}$ is the nonlinear damping which suppresses the oscillations at higher multipoles. (Bottom panel) BAO scale showing the percentage deviation of the full and linear APS from the no-wiggle APS.
• Figure 3: Marginalized $68\%$ and $95\%$ Confidence Ellipses for $w_0$ and $h$ in the $w$CDM model.
• Figure 4: Marginalized $68\%$ and $95\%$ Confidence Ellipses for $w_0$ and $w_a$ in the CPL model.