Beyond the Power Spectrum: A New Framework for Non-Stationary Fields with Applications to Light-Cone Effects in Line Intensity Mapping

TL;DR

This paper introduces a novel non-Fourier basis for analyzing non-stationary cosmological fields, improving parameter constraints in large-volume surveys like 21cm intensity mapping by accounting for evolution effects along the line of sight.

Contribution

It presents a new eigen decomposition-inspired basis that captures evolution effects, offering a more optimal summary statistic than the traditional power spectrum.

Findings

Tighter constraints on astrophysical parameters using the new basis.

Demonstration of the basis's effectiveness with realistic HERA forecasts.

Practical guide for applying the method in observational data analysis.

Abstract

Modern cosmological surveys cover extremely large volumes and map fluctuations on scales reaching gigaparsecs. As a result, it is no longer a valid assumption to ignore cosmological evolution along the line of sight from one end of the survey to the other. When extracting cosmological information, the power spectrum becomes suboptimal because it relies on the assumption of translational invariance of the observed field. For example, during the Epoch of Reionization (EoR), the 21cm brightness temperature field on the nearby (low-redshift) end of a large survey volume exhibits statistical properties that differ significantly from those at far (high-redshift) end. To overcome these limitations, we have developed a eigen decomposition-inspired non-Fourier basis that captures evolution effects. Our work demonstrates that using this new basis integrated in a new summary statistic yields tighter constraints on astrophysical parameters compared to the traditional power spectrum. Additionally, we provide an illustrative example and a practical guide for applying this basis in the context of a realistic forecast for interferometers such as the Hydrogen Epoch of Reionization Array (HERA).