Towards simulating a realistic data analysis with an optimised angular power spectrum of spectroscopic galaxy surveys

TL;DR

This paper evaluates a novel, computationally efficient method for analyzing galaxy survey data by combining thick and thin redshift bins, demonstrating its effectiveness through simulated data analysis.

Contribution

It introduces and tests a new binning approach that reduces computation while improving analysis accuracy compared to standard tomography methods.

Findings

The new method significantly reduces computational time.

It provides more accurate results than traditional tomography.

The approach is validated with synthetic data analysis.

Abstract

The angular power spectrum is a natural tool to analyse the observed galaxy number count fluctuations. In a standard analysis, the angular galaxy distribution is sliced into concentric redshift bins and all correlations of its harmonic coefficients between bin pairs are considered---a procedure referred to as `tomography'. However, the unparalleled quality of data from oncoming spectroscopic galaxy surveys for cosmology will render this method computationally unfeasible, given the increasing number of bins. Here, we put to test against synthetic data a novel method proposed in a previous study to save computational time. According to this method, the whole galaxy redshift distribution is subdivided into thick bins, neglecting the cross-bin correlations among them; each of the thick bin is, however, further subdivided into thinner bins, considering in this case all the cross-bin correlations. We create a simulated data set that we then analyse in a Bayesian framework. We confirm that the newly proposed method saves computational time and gives results that surpass those of the standard approach.