Reducing the gravitational lensing scatter of Type Ia supernovae without introducing any extra bias

TL;DR

This paper demonstrates that correcting for gravitational lensing using foreground galaxy observations can significantly reduce brightness scatter in Type Ia supernovae, improving cosmological measurements without bias.

Contribution

It introduces a simulation-based method to estimate and correct individual supernova magnifications, reducing lensing scatter without adding bias.

Findings

Lensing correction bias is negligible for current and future surveys.

Lensing scatter can be reduced by approximately a factor of 2.

Total magnitude scatter could be reduced by 4-6%, improving cosmological parameter accuracy.

Abstract

Magnification and de-magnification due to gravitational lensing will contribute to the brightness scatter of Type Ia supernovae (SNe Ia). The purpose of this paper is to investigate the possibility to decrease this scatter by correcting individual SNe Ia using observations of galaxies in the foreground, without introducing any extra bias. We simulate a large number of SN Ia lines of sight populated by galaxies. For each line of sight the true magnification factor and an estimate thereof are calculated. The estimated magnification factor corresponds to what an observer would infer from a survey like SNLS. Using the simulated data we investigate the possibility to estimate the magnification of individual supernovae with enough precision to be able to correct their brightness for gravitational lensing with negligible bias. Our simulations show that the bias arising from gravitational lensing corrections of individual SNe Ia is negligible for current and next generation surveys and that the scatter from lensing can be reduced by approximately a factor 2. The total scatter in the SN Ia magnitudes could be reduced by 4% for an intrinsic dispersion of 0.13 mag. For an intrinsic dispersion of 0.09 mag, which might not be to unrealistic for future surveys, the total scatter could be reduced by 6%. This will reduce the errors on cosmological parameters derived from supernova data by 4-8%. The prospect of correcting for lensing is thus very good.