Optimizing BAO measurements with non-linear transformations of Lyman-alpha forest

TL;DR

This paper investigates non-linear transformations of the Lyman-alpha forest to improve Baryon Acoustic Oscillation (BAO) measurements, finding limited but notable enhancements in signal-to-noise ratio and modeling biasing effects.

Contribution

It introduces an analytic model for the biasing of transformed Lyman-alpha forest fields and evaluates the impact of non-linear transformations on BAO signal detection.

Findings

33% improvement in signal-to-noise ratio for Gaussianized field transversely

Model accurately predicts biasing with respect to velocity gradients

Less successful in modeling biasing related to large-scale density fluctuations

Abstract

We explore the effect of applying a non-linear transformation to the Lyman-$\alpha$ forest transmitted flux $F=e^{-\tau}$ and the ability of analytic models to predict the resulting clustering amplitude. Both the large-scale bias of the transformed field (signal) and the amplitude of small scale fluctuations (noise) can be arbitrarily modified, but we were unable to find a transformation that increases significantly the signal-to-noise ratio on large scales using Taylor expansion up to third order. In particular, however, we achieve a 33% improvement in signal to noise for Gaussianized field in transverse direction. On the other hand, we explore an analytic model for the large-scale biasing of the Ly$\alpha$ forest, and present an extension of this model to describe the biasing of the transformed fields. Using hydrodynamic simulations we show that the model works best to describe the biasing with respect to velocity gradients, but is less successful in predicting the biasing with respect to large-scale density fluctuations, especially for very nonlinear transformations.