Gaussianizing the non-Gaussian lensing convergence field II: the applicability to noisy data

TL;DR

This paper investigates the effectiveness of Gaussianizing the lensing convergence field in noisy data, finding that Wiener filtering enhances Gaussianization by significantly reducing non-Gaussian features despite noise degradation.

Contribution

It extends previous work by analyzing the Gaussianization method's applicability to noisy lensing data and demonstrates the effectiveness of Wiener filtering in this context.

Findings

Shape measurement noise degrades Gaussianization performance, especially in shallower surveys.

Wiener filtering effectively reduces noise impact, enabling Gaussianization to suppress higher-order statistics.

Gaussianized, Wiener-filtered maps can significantly diminish skewness, kurtosis, and higher-order cumulants.

Abstract

In paper I (Yu et al. [1]), we show through N-body simulation that a local monotonic Gaussian transformation can significantly reduce non-Gaussianity in a noise-free lensing convergence field. This makes the Gaussianization a promising theoretical tool to understand high-order lensing statistics. Here we present a study of its applicability in lensing data analysis, in particular when shape measurement noise is presented in lensing convergence maps. (i) We find that shape measurement noise significantly degrades the Gaussianization performance and the degradation increases for shallower surveys. (ii) The Wiener filter is efficient in reducing the impact of shape measurement noise. The Gaussianization of the Wiener-filtered lensing maps is able to suppress skewness, kurtosis, and the 5th- and 6th-order cumulants by a factor of 10 or more. It also works efficiently to reduce the bispectrum to zero.