24–28 Aug 2026
Leiden University
Europe/Zurich timezone

A self-calibration method of imaging systematics using power spectrum and bispectrum

Not scheduled
20m
Gorlaeus gebouw (Leiden University)

Gorlaeus gebouw

Leiden University

Einsteinweg 55, 2333 CC Leiden
Talk Methods / Statistical Inference / Machine Learning

Speaker

Qianran Xia (University of Michigan)

Description

Imaging systematics in galaxy surveys generate large-scale angular selection effects that can contaminate measurements of primordial non-Gaussianity. We develop a analytical framework to propagate two-dimensional imaging templates into three-dimensional power spectrum and bispectrum observables. In this approach, angular templates are embedded into radial survey shells, and their effects on the observed monopoles of power spectrum and bispectrum can be modeled analytically. We use this formalism to quantify the degeneracy between imaging contamination and local-type PNG, and to test whether adding bispectrum information can improve self-calibration beyond the power spectrum alone. This framework provides an estimator-level route for incorporating imaging systematics into joint power spectrum and bispectrum analyses of future large-scale structure surveys.

Author

Qianran Xia (University of Michigan)

Co-author

Dragan Huterer (University of Michigan)

Presentation materials

There are no materials yet.