Toward a tomographic analysis of the cross-correlation between Planck CMB lensing and H-ATLAS galaxies

TL;DR

This paper improves the analysis of the cross-correlation between Planck CMB lensing maps and H-ATLAS galaxies, revealing a significant redshift evolution of galaxy bias and a tension with the standard cosmological model.

Contribution

It introduces a tomographic approach to cross-correlation analysis, utilizing improved Planck data and galaxy sample selections for more detailed insights.

Findings

Galaxy bias increases with redshift, from 2.89 to 4.75.

Cross-correlation amplitude exceeds ΛCDM expectations by 45%.

Results are robust against systematic uncertainties.

Abstract

We present an improved and extended analysis of the cross-correlation between the map of the Cosmic Microwave Background (CMB) lensing potential derived from the \emph{Planck} mission data and the high-redshift galaxies detected by the \emph{Herschel} Astrophysical Terahertz Large Area Survey (H-ATLAS) in the photometric redshift range $z_{\rm ph} \ge 1.5$. We compare the results based on the 2013 and 2015 \textit{Planck} datasets, and investigate the impact of different selections of the H-ATLAS galaxy samples. Significant improvements over our previous analysis have been achieved thanks to the higher signal-to-noise ratio of the new CMB lensing map recently released by the \textit{Planck} collaboration. The effective galaxy bias parameter, $b$, for the full galaxy sample, derived from a joint analysis of the cross-power spectrum and of the galaxy auto-power spectrum is found to be $b = 3.54^{+0.15}_{-0.14}$. Furthermore, a first tomographic analysis of the cross-correlation signal is implemented, by splitting the galaxy sample into two redshift intervals: $1.5 \le z_{\rm ph} < 2.1$ and $z_{\rm ph}\ge 2.1$. A statistically significant signal was found for both bins, indicating a substantial increase with redshift of the bias parameter: $b=2.89\pm0.23$ for the lower and $b=4.75^{+0.24}_{-0.25}$ for the higher redshift bin. Consistently with our previous analysis we find that the amplitude of the cross correlation signal is a factor of $1.45^{+0.14}_{-0.13}$ higher than expected from the standard $\Lambda$CDM model for the assumed redshift distribution. The robustness of our results against possible systematic effects has been extensively discussed although the tension is mitigated by passing from 4 to 3$\sigma$.