Constraints on Cosmological Parameters from Strong Gravitational Lensing by Galaxy Clusters

TL;DR

This paper explores how strong gravitational lensing observations of galaxy clusters can constrain cosmological parameters, emphasizing the growth of critical lines with source redshift and developing numerical methods for analysis.

Contribution

It introduces a numerical approach using the Metropolis-Hastings algorithm and elliptical NFW profiles to analyze lensing data for cosmological insights.

Findings

Individual cluster observations provide no constraints without mass distribution knowledge

Using a fixed lens model improves uncertainties in geometric optimization

Parametric models can introduce significant biases in the analysis

Abstract

We investigate how observations of strong lensing can be used to infer cosmological parameters, in particular the equation of state of dark energy. We focus on the growth of the critical lines of lensing clusters with the source redshift as this behaviour depends on the distance-redshift relation and is therefore cosmologically sensitive. Purely analytical approaches are generally insufficient because they rely on axisymmetric mass distributions and thus cannot take irregular critical curves into account. We devise a numerical method based on the Metropolis-Hastings algorithm: an elliptical generalization of the NFW density profile is used to fit a lens model to an observed configuration of giant luminous arcs while simultaneously optimizing the geometry. A semi-analytic method, which derives geometric parameters from critical points, is discussed as a faster alternative. We test the approaches on mock observations of gravitational lensing by a numerically simulated cluster. We find that no constraints can be derived from observations of individual clusters if no knowledge of the underlying mass distribution is assumed. Uncertainties are improved if a fixed lens model is used for a purely geometrical optimization, but the choice of a parametric model may produce strong biases.