Systematic effects on dark energy from 3D weak shear

TL;DR

This paper investigates how systematic uncertainties in multi-band wide field surveys affect dark energy parameter estimation using 3D weak lensing methods, highlighting the importance of priors on systematics.

Contribution

It extends the Fisher matrix framework to jointly analyze photometric redshift, shear distortion, and intrinsic alignment systematics in 3D weak lensing.

Findings

Photometric redshift uncertainties significantly reduce the dark energy Figure of Merit.

Shear distortion systematics have a smaller impact on the Figure of Merit.

Including priors on systematics can largely recover the loss in constraining power.

Abstract

We present an investigation into the potential effect of systematics inherent in multi-band wide field surveys on the dark energy equation of state determination for two 3D weak lensing methods. The weak lensing methods are a geometric shear-ratio method and 3D cosmic shear. The analysis here uses an extension of the Fisher matrix framework to jointly include photometric redshift systematics, shear distortion systematics and intrinsic alignments. We present results for DUNE and Pan-STARRS surveys. We show that assuming systematic parameters are fixed, but possibly biased, results in potentially large biases in dark energy parameters. We quantify any potential bias by defining a Bias Figure of Merit. We also show the effect on the dark energy Figure of Merit of marginalising over each systematic parameter individually. We find that the largest effect on the Figure of Merit comes from uncertainty in the photometric redshift systematic parameters. These can reduce the Figure of Merit by up to a factor of 2 to 4 in both 3D weak lensing methods, if no informative prior on the systematic parameters is applied. Shear distortion systematics have a smaller overall effect. Intrinsic alignment effects can reduce the Figure of Merit by up to a further factor of 2. This, however, is a worst case scenario. By including prior information on systematic parameters the Figure of Merit can be recovered to a large extent. We conclude that, as a rule of thumb, given a realistic current understanding of intrinsic alignments and photometric redshifts, then including all three primary systematic effects reduces the Figure of Merit by at most a factor of 2, but that in reality this factor should be much less. [abridged]