Stacked CMB lensing and ISW signals around superstructures in the DESI Legacy Survey

TL;DR

This study measures CMB lensing and ISW signals around superstructures in the DESI Legacy Survey, finding general consistency with ΛCDM predictions but noting discrepancies at high redshift and with previous claims of anomalously high ISW signals.

Contribution

It provides new measurements of CMB lensing and ISW signals around superstructures using DESI data and compares them with ΛCDM model predictions, addressing previous anomalies.

Findings

Lensing signals mostly agree with models, except at high redshift in density peaks.

ISW signals are generally consistent with predictions, with no significant detection from voids.

Results contradict earlier claims of anomalously high ISW signals around superstructures.

Abstract

The imprints of large-scale structures on the Cosmic Microwave Background can be studied via the CMB lensing and Integrated Sachs-Wolfe (ISW) signals. In particular, the stacked ISW signal around supervoids has been claimed in several works to be anomalously high. In this study, we find cluster and void superstructures using four tomographic redshift bins with $0<z<0.8$ from the DESI Legacy Survey, and measure the stacked CMB lensing and ISW signals around them. To compare our measurements with $\Lambda$CDM model predictions, we construct a mock catalogue with matched galaxy number density and bias, and apply the same photo-$z$ uncertainty as the data. The consistency between the mock and data is verified via the stacked galaxy density profiles around the superstructures and their quantity. The corresponding lensing convergence and ISW maps are then constructed and compared. The stacked lensing signal agrees with data well except at the highest redshift bin in density peaks, where the mock prediction is significantly higher, by approximately a factor 1.3. The stacked ISW signal is generally consistent with the mock prediction. We do not obtain a significant signal from voids, $A_{\rm ISW}=-0.10\pm0.69$, and the signal from clusters, $A_{\rm ISW}=1.52\pm0.72$, is at best weakly detected. However, these results are strongly inconsistent with previous claims of ISW signals at many times the level of the $\Lambda$CDM prediction. We discuss the comparison of our results with past work in this area, and investigate possible explanations for this discrepancy.