skip to main content
OSTI.GOV title logo U.S. Department of Energy
Office of Scientific and Technical Information

Title: BIAS-FREE SHEAR ESTIMATION USING ARTIFICIAL NEURAL NETWORKS

Abstract

Bias due to imperfect shear calibration is the biggest obstacle when constraints on cosmological parameters are to be extracted from large area weak lensing surveys such as Pan-STARRS-3{pi}, DES, or future satellite missions like EUCLID. We demonstrate that bias present in existing shear measurement pipelines (e.g., KSB) can be almost entirely removed by means of neural networks. In this way, bias correction can depend on the properties of the individual galaxy instead of being a single global value. We present a procedure to train neural networks for shear estimation and apply this to subsets of simulated GREAT08 RealNoise data. We also show that circularization of the point-spread function (PSF) before measuring the shear reduces the scatter related to the PSF anisotropy correction and thus leads to improved measurements, particularly on low and medium signal-to-noise data. Our results are competitive with the best performers in the GREAT08 competition, especially for the medium and higher signal-to-noise sets. Expressed in terms of the quality parameter defined by GREAT08, we achieve a Q{approx} 40, 140, and 1300 without and 50, 200, and 1300 with circularization for low, medium, and high signal-to-noise data sets, respectively.

Authors:
; ; ;  [1]
  1. University Observatory Munich, Scheinerstrasse 1, 81679 Muenchen (Germany)
Publication Date:
OSTI Identifier:
21460107
Resource Type:
Journal Article
Journal Name:
Astrophysical Journal
Additional Journal Information:
Journal Volume: 720; Journal Issue: 1; Other Information: DOI: 10.1088/0004-637X/720/1/639; Journal ID: ISSN 0004-637X
Country of Publication:
United States
Language:
English
Subject:
79 ASTROPHYSICS, COSMOLOGY AND ASTRONOMY; ANISOTROPY; CALIBRATION; COSMOLOGY; GALAXIES; NEURAL NETWORKS; SATELLITES; SIGNALS

Citation Formats

Gruen, D, Seitz, S, Koppenhoefer, J, and Riffeser, A. BIAS-FREE SHEAR ESTIMATION USING ARTIFICIAL NEURAL NETWORKS. United States: N. p., 2010. Web. doi:10.1088/0004-637X/720/1/639.
Gruen, D, Seitz, S, Koppenhoefer, J, & Riffeser, A. BIAS-FREE SHEAR ESTIMATION USING ARTIFICIAL NEURAL NETWORKS. United States. doi:10.1088/0004-637X/720/1/639.
Gruen, D, Seitz, S, Koppenhoefer, J, and Riffeser, A. Wed . "BIAS-FREE SHEAR ESTIMATION USING ARTIFICIAL NEURAL NETWORKS". United States. doi:10.1088/0004-637X/720/1/639.
@article{osti_21460107,
title = {BIAS-FREE SHEAR ESTIMATION USING ARTIFICIAL NEURAL NETWORKS},
author = {Gruen, D and Seitz, S and Koppenhoefer, J and Riffeser, A},
abstractNote = {Bias due to imperfect shear calibration is the biggest obstacle when constraints on cosmological parameters are to be extracted from large area weak lensing surveys such as Pan-STARRS-3{pi}, DES, or future satellite missions like EUCLID. We demonstrate that bias present in existing shear measurement pipelines (e.g., KSB) can be almost entirely removed by means of neural networks. In this way, bias correction can depend on the properties of the individual galaxy instead of being a single global value. We present a procedure to train neural networks for shear estimation and apply this to subsets of simulated GREAT08 RealNoise data. We also show that circularization of the point-spread function (PSF) before measuring the shear reduces the scatter related to the PSF anisotropy correction and thus leads to improved measurements, particularly on low and medium signal-to-noise data. Our results are competitive with the best performers in the GREAT08 competition, especially for the medium and higher signal-to-noise sets. Expressed in terms of the quality parameter defined by GREAT08, we achieve a Q{approx} 40, 140, and 1300 without and 50, 200, and 1300 with circularization for low, medium, and high signal-to-noise data sets, respectively.},
doi = {10.1088/0004-637X/720/1/639},
journal = {Astrophysical Journal},
issn = {0004-637X},
number = 1,
volume = 720,
place = {United States},
year = {2010},
month = {9}
}