Weak Lensing Featured Project: Impact of Chromaticity

Weak Lensing Chromaticity
Caption: 

This figure shows how the relative size of the PSF varies with SED due to chromatic seeing in the r band. The data points show the sizes of galactic PSFs relative to the size of a G5v star PSF (left axis) as a function of r-i (color) and redshift (bottom axis). Generically, it can be seen that redder SEDs produce smaller PSFs, as expected for chromatic seeing. The histograms on the left show the (redshift projected) distribution of stellar (blue) and galactic (red) PSF sizes. An error in PSF size caused by estimating a galactic PSF from a nearby star translates directly into a multiplicative bias on shear measurements, which is plotted on the right axis. The two horizontal bars show the multiplicative bias requirements for DES and LSST. SEDs are drawn from the LSST CatSim database, and should form a representative sample of stars and lensed galaxies.

(Authors: Josh Meyers, Pat Burchat, Stanford University)

The unprecedented statistical precision of LSST cosmic shear measurements will require a deeper understanding of potential systematic biases in galaxy shape measurement than has been achieved to date.  In this highlight project, we investigate biases originating from two chromatic effects in the atmosphere: (i) differential chromatic refraction and (ii) wavelength dependence of seeing.  These biases arise when using the point spread function measured with stars to estimate the shape of a galaxy with a different spectral energy distribution (SED) than the star.  For both effects, we present analytic estimates for the shape measurement biases. From simulated images, we also demonstrate an additional “model bias” due to the unknown shape of the point spread function.  We investigate techniques to correct these biases using multi-filter photometry to predict the individual SEDs of stars and galaxies and associated chromatic effects.

In addition to its relevance to LSST, this analysis is applicable to other large-volume ground-based surveys such as DES and Hyper Suprime-Cam, where the same atmospheric biases will manifest.  The wavelength dependence of seeing present for atmospheric turbulence is also qualitatively similar to the wavelength dependence of the diffraction limit, which will affect space-based cosmic shear surveys such as Euclid.  We have integrated the chromatic simulations software used here into the open source project GalSim, used for the GREAT3 shape measurement challenge and available on github.com.  Similarly, all of the specific analysis software for this project, including software for reproducing the tables and figures presented here, is available at  http://darkenergysciencecollaboration.github.io/chroma/.