The Squeezed Bispectrum from CHIME HI Emission and Planck CMB Lensing: Current Sensitivity and Forecasts

TL;DR

This study targets the cross-correlation between CHIME HI line-of-sight variance and Planck CMB lensing to probe the squeezed bispectrum arising from nonlinear gravitational coupling between small-scale HI fluctuations and large-scale lensing modes. By implementing a position-dependent power spectrum approach, the authors forecast and attempt a detection using 94 CHIME nights at 1.0 < z < 1.3, finding the signal currently five times below the noise. Simulations that include nonlinear HI evolution predict a measurable cross-signal, suggesting that a tenfold increase in CHIME data could yield a ~3σ detection with Planck-like lensing and ~4σ with higher-signal SO-like lensing. The work demonstrates a promising path to extract nonlinear coupling information from HI intensity maps via a squeezed bispectrum, motivating larger surveys and cross-survey synergies (e.g., ACT/SO/HERA/SKA) to realize this cosmological probe.

Abstract

Line intensity mapping using atomic hydrogen (HI) has the potential to efficiently map large volumes of the universe if the signal can be successfully separated from overwhelmingly bright radio foreground emission. This motivates cross-correlations, to ascertain the cosmological nature of measured HI fluctuations, and to study their connections with galaxies and the underlying matter density field. However, these same foregrounds render the cross-correlation with projected fields such as the lensing of the cosmic microwave background (CMB) difficult. Indeed, the correlated Fourier modes vary slowly along the line of sight, and are thus most contaminated by the smooth-spectrum radio continuum foregrounds. In this paper, we implement a method that avoids this issue by attempting to measure the non-linear gravitational coupling of the small-scale 21cm power from the Canadian Hydrogen Intensity Mapping Experiment (CHIME) with large-scale Planck CMB lensing. This measurement is a position-dependent power spectrum, i.e. a squeezed integrated bispectrum. Using 94 nights of CHIME data between $1.0 < z < 1.3$ and aggressive foreground filtering, we find that the expected signal is five times smaller than the current noise. We forecast that incorporating the additional nights of CHIME data already collected would enable a signal-to-noise ratio of 3, without any further improvements in filtering for foreground cleaning.

Figures (7)

• Figure 1: We estimate the cross-correlation of the lensing convergence and the variance of the line-of-sight neutral hydrogen fluctuations $\langle {\rm HI}\xspace^2\rangle$ fields using a tiled flat-sky approach. Noise-free non-Gaussian simulations of the expected signal from SkyLineskyline_2023 were used to produce the maps shown here.
• Figure 2: Gaussian simulations of the 21$\,$cm signal (blue dots) do not capture the three-point correlation function, and thus predict a null cross-power spectrum of CMB lensing and 21$\,$cm variance along the line of sight (Eq \ref{['eq:xcorr-th']}). Nonlinear simulations such as SkyLine, based on N-body simulations, are thus needed to predict our signal (orange dots).
• Figure 3: Map of line of sight variance estimated from the filtered CHIME data. The vertical axis is $\sin$ of the angle from zenith, a coordinate that corresponds to the Fourier conjugate of the telescope's North-South baseline grid. The locations of bright points sources are masked in a region around their meridian crossing, when they are directly overhead CHIME (small ellipses). For the brightest sources, we additionally mask the tracks they follow as they transit through the far side-lobes of the telescope beam (curved tracks), and all declinations at their transit time (vertical bars). The top and bottom of the map have been smoothly tapered. Residual contamination can be seen near the plane of the Galaxy and along tracks from sources outside the field that are not masked.
• Figure 4: The measured correlation function for Planck lensing and CHIME variance maps (blue points with error bars) is consistent with noise. Indeed, the chi-squared with respect to the null hypothesis (i.e. no signal model) is $\chi_\nu^2=0.90$, with a PTE of $0.51$. This includes the non-zero correlation between data points.
• Figure 5: The correlation matrix between the samples in Figure \ref{['fig:data-corrfunc']} shows that they are not independent.