Cosmology without window functions. II. Cubic estimators for the galaxy bispectrum

Abstract

The large-scale structure (LSS) is determined by the nonlinear structure growth of the Universe and this information is encoded in higher-order statistics such as the bispectrum (3-point correlation function). Robust inference of parameters from the galaxy bispectrum is much more challenging than for the power spectrum (2-point-function). One major problem is the computationally intensive and expensive handling of the convolution of the measured data with ``window functions'' which describe survey filters. However, the author develops an estimator for the ``unwindowed'' bispectrum that will allow detailed Markov chain Monte Carlo analysis and full parameter exploration of the plethora of LSS data to come in this decade.

Topics & Concepts