Angular BAO Forecasts for the IBIS Medium-Band Survey

TL;DR

This work demonstrates the feasibility of extracting angular BAO signals from high-redshift, medium-band imaging surveys by LAE tracers, using a detailed angular power-spectrum framework with redshift-space distortions, BAO damping, and spline-based broadband marginalization. The authors quantify forecasted precision on the BAO dilation parameter $\alpha$ for IBIS-like surveys across multiple redshift bins and survey configurations, finding sub-percent to a few-percent constraints depending on area, tracer density, and interloper fraction. They show that for a 5000 deg$^2$ IBIS-like survey, $\sigma(\alpha_{\rm BAO})$ can reach as low as $\sim 1.6$–$3\%$, and that Stage-V imaging campaigns (Rubin-like MB and HSC MB) can push angular BAO precision to $\lesssim 1\%$ when combining multiple MB tracers. These results argue for early angular BAO measurements from imaging surveys to calibrate and inform subsequent spectroscopic analyses, with implications extending to high redshift up to $z\sim5$ when extended MB data are available.

Abstract

Ongoing and near-future spectroscopic surveys, such as DESI, DESI-II and Spec-S5, rely on imaging-based selections to construct uniform, three-dimensional tracers of large-scale structure. While spectroscopic data from these surveys constrain the baryonic acoustic oscillation (BAO) feature with high precision, the imaging surveys used for target selection can provide useful information on the angular diameter distance $D_A(z)$. In this work we explore the feasibility of angular BAO measurements for the Intermediate-Band Imaging Survey (IBIS) using recent constraints on clustering from a pilot survey spanning $2.2<z<3.5$. Through Fisher forecasts, we find that a 5000 deg$^2$ survey of LAEs with realistic bias, a tracer density of $2\times 10^{-4}$ (h/Mpc)$^3$ and interloper fraction $f_{\rm int}=10\%$ can constrain the BAO dilation parameter $α$ at $z_{\rm eff}=2.8$ with a precision of 2.6\%, with dependence on the sample properties that is consistent with shot noise-dominated measurements. We then explore medium-band survey specifications for the planned Stage-V Spectroscopic Instrument (Spec-S5) and beyond, demonstrating the potential for precise high-redshift BAO measurements. Our forecasts motivate early measurements of BAO from these imaging surveys, which may inform later spectroscopic analyses.

Figures (7)

• Figure 1: Top: LAE angular auto-power spectra, taking $b(z)=2.0-2.5$ as per Table \ref{['tab:survey_specs']}. The grey shaded region indicates the range of shot noise levels explored in our forecasts (see § \ref{['sec:survey_spec']}). Bottom: residual BAO wiggle components. We restrict our analysis to $\ell > 100$, corresponding to $k\gtrsim 0.015$ at $z=3.0$.
• Figure 2: Forecast uncertainty on $\alpha_{\rm{BAO}}$ for an IBIS-like imaging survey, as a function of number density and survey area. These forecasts assume no interlopers, i.e., $f_{\rm int}=0$, though we consider their impact in § \ref{['sec:interlopers']}.
• Figure 3: Dependence of $\sigma(\alpha_{\textrm{BAO}})$ on maximum angular multipole $\ell_{\rm max}$. To illustrate the scale dependence of the BAO information we show single-parameter forecasts for $\sigma(\alpha_{\rm BAO})$ (i.e., not marginalized over spline parameters). The top panel shows the combined constraints from all five redshift bins, for a range of tracer densities. The bottom panel shows the same but for individual redshift bins and a single tracer density ($\bar{n}_{\rm LAE}=825$ deg$^{-2}$). We assume no interlopers for these comparisons (see § \ref{['sec:interlopers']}).
• Figure 4: Degradation in $\sigma(\alpha_{\rm{BAO}})$ due to interlopers in the target sample, with varying interloper fraction.
• Figure 5: Assumed filter profiles for the planned Stage-V MB imaging (blue curves) and an extended, longer wavelength HSC MB-like configuration (red). The corresponding Lyman-$\alpha$ selection extends to $z=5.2$ (i.e., $\lambda=(1+z)1216$ Å). We assume $dN/dz$ distributions that follow the normalized filter profiles.