Combining cluster observables and stacked weak lensing to probe dark energy: Self-calibration of systematic uncertainties

TL;DR

This paper presents a combined analysis of cluster counts, correlation functions, and stacked weak lensing to self-calibrate systematic uncertainties, significantly improving dark energy constraints in wide-field surveys.

Contribution

It introduces a novel joint method that self-calibrates key systematics using a single background galaxy population, enhancing cosmological parameter estimation.

Findings

Dark energy FoM improved by a factor of 4

Constraints on primordial non-Gaussianity parameter f_NL~10

Achieves 0.1 redshift accuracy for source calibration

Abstract

We develop a new method of combining cluster observables (number counts and cluster-cluster correlation functions) and stacked weak lensing signals of background galaxy shapes, both of which are available in a wide-field optical imaging survey. Assuming that the clusters have secure redshift estimates, we show that the joint experiment enables a self-calibration of important systematic errors including the source redshift uncertainty and the cluster mass-observable relation, by adopting a single population of background source galaxies for the lensing analysis. It allows us to use the relative strengths of stacked lensing signals at different cluster redshifts for calibrating the source redshift uncertainty, which in turn leads to accurate measurements of the mean cluster mass in each bin. In addition, our formulation of stacked lensing signals in Fourier space simplifies the Fisher matrix calculations, as well as the marginalization over the cluster off-centering effect, the most significant uncertainty in stacked lensing. We show that upcoming wide-field surveys yield stringent constraints on cosmological parameters including dark energy parameters, without any priors on nuisance parameters that model systematic uncertainties. Specifically, the stacked lensing information improves the dark energy FoM by a factor of 4, compared to that from the cluster observables alone. The primordial non-Gaussianity parameter can also be constrained with a level of f_NL~10. In this method, the mean source redshift is well calibrated to an accuracy of 0.1 in redshift, and the mean cluster mass in each bin to 5-10% accuracies, which demonstrates the success of the self-calibration of systematic uncertainties from the joint experiment. (Abridged)