The Impact of Intrinsic Alignments: Cosmological Constraints from a Joint Analysis of Cosmic Shear and Galaxy Survey Data

TL;DR

This paper investigates how intrinsic alignments affect cosmological constraints from cosmic shear data, demonstrating that combined analysis of shear-shear and shear-position correlations can mitigate biases and produce robust parameter estimates.

Contribution

It introduces a joint analysis method combining cosmic shear and intrinsic alignment data, providing updated cosmological constraints that account for intrinsic alignment uncertainties.

Findings

Intrinsic alignments can significantly bias sigma_8 measurements.

Simultaneous shear-shear and shear-position analyses reduce bias.

Constraints remain stable despite additional model parameters.

Abstract

Constraints on cosmology from recent cosmic shear observations are becoming increasingly sophisticated in their treatment of potential systematic effects. Here we present cosmological constraints which include modelling of intrinsic alignments. We demonstrate how the results are changed for three different intrinsic alignment models, and for two different models of the cosmic shear galaxy population. We find that intrinsic alignments can either reduce or increase measurements of the fluctuation amplitude parameter sigma_8 depending on these decisions, and depending on the cosmic shear survey properties. This is due to the interplay between the two types of intrinsic alignment, II and GI. It has been shown that future surveys must make a careful treatment of intrinsic alignments to avoid significant biases, and that simultaneous constraints from shear-shear and shear-position correlation functions can mitigate the effects. For the first time we here combine constraints from cosmic shear surveys (shear-shear correlations) with those from "GI" intrinsic alignment data sets (shear-position correlations). We produce updated constraints on cosmology marginalised over two free parameters in the halo model for intrinsic alignments. We find that the additional freedom is well compensated by the additional information, in that the constraints are very similar indeed to those obtained when intrinsic alignments are ignored, both in terms of best fit values and uncertainties.