Angular spectra of the intrinsic galaxy ellipticity field, their observability and their impact on lensing in tomographic surveys

Abstract

This paper describes intrinsic ellipticity correlations between galaxies, their statistical properties, their observability with future surveys and their interference with weak gravitational lensing measurements. Using an angular-momentum-based, quadratic intrinsic alignment model we derive correlation functions of the ellipticity components and project them to yield the four non-zero angular ellipticity spectra |$C^\epsilon _E(\ell )$|⁠, |$C^\epsilon _B(\ell )$|⁠, |$C^\epsilon _C(\ell )$| and |$C^\epsilon _S(\ell )$| in their generalization to tomographic surveys. For a Euclid-like survey, these spectra would have amplitudes smaller than the weak lensing effect on non-linear structures, but would constitute an important systematics. Computing estimation biases for cosmological parameters derived from an alignment-contaminated survey suggests biases of +5σw for the dark energy equation of state parameter w, |$-20\sigma _{\Omega _{\rm m}}$| for the matter density Ωm and |$-12\sigma _{\sigma _8}$| for the spectrum normalization σ8. Intrinsic alignments yield a signal that is easily observable with a survey similar to Euclid: while not independent, significances for estimates of each of the four spectra reach values of tens of σ if weak lensing and shape noise are considered as noise sources, which suggests relative uncertainties on alignment parameters at the percent level, implying that galaxy alignment mechanisms can be investigated by future surveys.