Constraining cosmology with shear peak statistics: tomographic analysis

TL;DR

This paper demonstrates that shear peak statistics, especially when using tomographic analysis, can significantly enhance cosmological parameter constraints from weak lensing data, complementing traditional cluster-based methods.

Contribution

It introduces a tomographic peak counting method directly on shear maps, improving cosmological constraints and highlighting the complementary value of shear peaks and clusters.

Findings

Tomographic analysis improves constraints by 20% over non-tomographic methods.

Peak statistics provide constraints twice as strong as cluster methods when the observable-mass relation is unknown.

Shear peak constraints are orthogonal to cluster constraints, offering complementary information.

Abstract

The abundance of peaks in weak gravitational lensing maps is a potentially powerful cosmological tool, complementary to measurements of the shear power spectrum. We study peaks detected directly in shear maps, rather than convergence maps, an approach that has the advantage of working directly with the observable quantity, the galaxy ellipticity catalog. Using large numbers of numerical simulations to accurately predict the abundance of peaks and their covariance, we quantify the cosmological constraints attainable by a large-area survey similar to that expected from the Euclid mission, focussing on the density parameter, {\Omega}m, and on the power spectrum normalization, {\sigma}8, for illustration. We present a tomographic peak counting method that improves the conditional (marginal) constraints by a factor 1.2 (2) over those from a two-dimensional (i.e., non-tomographic) peak-count analysis. We find that peak statistics provide constraints an order of magnitude less accurate than those from the cluster sample in the ideal situation of a perfectly known observable-mass relation; however, when the scaling relation is not known a priori, the shear-peak constraints are twice as strong and orthogonal to the cluster constraints, highlighting the value of using both clusters and shear-peak statistics.