Systematics mitigation for catalogue-based angular power spectra 2025 Systematics mitigation for catalogue-based angular power spectra – 8

Thomas Cornish, David Alonso, Boris Leistedt, Kevin Wolz · 2025

Read Systematics mitigation for catalogue-based angular power spectra 2025 Systematics mitigation for catalogue-based angular power spectra – 8 by Thomas Cornish, David Alonso, Boris Leistedt, Kevin Wolz (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2510-19912.txt.

Introduction

Thomas Cornish and three co-authors extend a young statistical method that computes angular power spectra directly from catalogues of objects on the sky, skipping the usual step of pixelating maps. Their addition is template deprojection, a way to strip out systematic contamination, worked into the catalogue-based formalism. Deprojection removes some real signal too, a bias they correct with a transfer function estimated from simple simulations.

They also derive exact expressions for the bias deprojection introduces into the shot-noise part of the spectrum. On simulated data, including galaxy clustering with realistic survey selection, the corrected method returns unbiased spectra in every scenario tested. Applied to real data it matches the standard map-based results, and the implementation ships in the public NaMaster code.

Catalog record

CREATORThomas Cornish, David Alonso, Boris Leistedt, Kevin Wolz
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE72 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2510-19912

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Systematics mitigation for catalogue-based angular power spectra 2025 Systematics mitigation for catalogue-based angular power spectra – 8 Thomas Cornish, David Alonso, Boris Leistedt, Kevin Wolz Abstract Recent work has developed a formalism for computing angular power spectra directly from catalogues containing field values at discrete positions on the sky, thereby circumventing the need to create pixelised maps of the fields, as well as avoiding aliasing and finite-resolution effects. Here, we adapt this formalism to incorporate template deprojection as a means of mitigating systematic biases in the measured angular power spectra. We also introduce and validate an alternative method of mitigating the so-called ‘deprojection bias’, caused by the loss of modes after deprojection, employing the use of simple simulations to compute a transfer function. We find that this approach performs at least as well as existing methods, and is relatively insensitive to how well one can guess the true power spectrum of the observed field, except at the largest scales ( ℓ ≲ 3 \ell\lesssim 3 ). Additionally, we develop exact expressions for the bias introduced by deprojection in the shot-noise component, which further improves the accuracy of this approach. We test our formalism on simulated datasets, demonstrating its applicability both to discretely sampled fields, and to the special case of galaxy clustering, with the survey selection function defined in terms of a random catalogue or as a continuous sky map. After computing and removing the bias in the shot noise and using a transfer function to correct for the remaining mode loss, our formalism is able to produce unbiased measurements of the angular power spectrum in all scenarios tested here. Finally, we apply our formalism to real data and demonstrate that it produces results consistent with the standard map-based pseudo- C ℓ C_{\ell} formalism. We implement our method in the publicly available code NaMaster . 1 Introduction The accurate measurement of projected fields on the sky has become a cornerstone of modern cosmology, with perhaps the most famous example being the cosmic microwave background (CMB), for which measurements of anisotropies in its temperature and polarisation enable constraints on models of inflation and the primordial density perturbations, as well as on the natures of neutrinos and dark energy ( Planck Collaboration et al. 2020 ; Louis et al. 2025 ; Camphuis et al. 2025 , e.g.) . Our knowledge of the large-scale structure (LSS) in the Universe is also heavily dependent on our ability to measure projected fields on the sky: tomographic measurements of galaxy clustering exploit the fact that galaxies are biased tracers of the underlying dark matter distribution while leveraging photometric redshifts as a means of estimating their radial positions, thereby enabling indirect measurements of the matter density field in ‘slices’ across cosmic time without the need for spectroscopic observations, which are time-consuming and costly across large areas ( Nicola et al. 2020 ; Rodríguez-Monroy et al. 2022 ; Abbott et al. 2022 ; Vakili et al. 2023 ; Sugiyama et al. 2023 ; Yan et al. 2025 , e.g.) . Furthermore, the effects of weak gravitational lensing can be expressed in terms of projected fields, such as the lensing shear and convergence, both of which are sensitive probes of the intervening dark matter distribution ( Heymans et al. 2021 ; Abbott et al. 2022 ; Miyatake et al. 2023 ; Stölzner et al. 2025 ; Faga et al. 2025 ) . These are only two examples of a myriad other projected probes of the LSS, which include other CMB secondary anisotropies, the Cosmic Infrared Background, etc. ( Stompor & Efstathiou 1999 ; Adam et al. 2017 ; Bocquet et al. 2019 ; Zubeldia & Challinor 2019 ; Maniyar et al. 2019 ; Carron et al. 2022 ; Bleem et al. 2022 ; Jego et al. 2023 ; Fang et al. 2024 ; Shaikh et al. 2024 ; Qu et al. 2024 ; Coulton et al. 2024 ; Ge et al. 2025 , e.g.) . It is no surprise, then, that the angular power spectrum (denoted C_{\ell} ), which quantifies the covariance between angular fluctuations in two fields as a function of angular multipole \ell (and encodes all information in the case of Gaussian fields), has become an essential summary statistic in cosmology. Unbiased estimators of the angular power spectrum have been extensively explored, and optimal quadratic minimum variance (QMV) estimators have been developed which are capable of recovering angular power spectra with minimal loss of information ( Tegmark 1997 ) . However, these estimators require the inversion of a pixel-space covariance matrix in order to optimally weight the data, an operation which can be computationally expensive when working with high-resolution maps. The prohibitive computational demand of these estimators often necessitates an alternative, an example of which can be found in the pseudo- C_{\ell} (PCL) algorithm ( Hivon et al. 2002 ; Alonso et al. 2019 ) . The PCL approach simplifies things by assuming a diagonal data covariance matrix – an assumption which holds if the data in question are spatially uncorrelated and/or dominated by white noise ( Leistedt et al. 2013 ) . This simplification significantly improves the speed of computation, making the PCL algorithm an often attractive approach to C_{\ell} estimation. The characterisation and mitigation of systematic effects is becoming a problem of increasing relevance in cosmological surveys: as the hoards of data being collected continually increase in size, the statistical uncertainties have the potential to become comparable to those arising from systematics. A plethora of techniques have been developed to tackle this issue by attempting to quantify the dependence of the desired field on maps of the survey properties or other contaminants thought to be contributing to the observed signal. These range from linear regression techniques such as mode deprojection ( Rybicki & Press 1992 ; Slosar et al. 2004 ; Elsner et al. 2016 ; Elsner et al. 2017 , also called template deprojection;) , template subtraction ( Ross et al. 2011 ; Ho et al. 2012 ) , and the iterative systematics decontamination (ISD) technique employed by the Dark Energy Survey ( Elvin-Poole et al. 2018 ; Rodríguez-Monroy et al. 2022 , DES;) , to more sophisticated methods employing machine learning tools such as artificial neural networks ( Rezaie et al. 2020 ) or self-organising maps ( Johnston et al. 2021 ) . The approaches discussed so far typically involve data in the form of pixelised maps that quantify the value of a projected field at a given position on the sky. However, it is often the case in cosmology that the fields whose angular power spectra we would like to measure are sampled at discrete positions on the sky, corresponding to the positions of sources in a catalogue. The typical course of action to obtain power spectrum measurements in this case is to bin these discrete points into pixels and produce a map of the mean field value in each pixel. However, recent work by Baleato Lizancos & White 2024 ; Wolz et al. 2025 allows one to circumvent this step, through an update to the PCL formalism which facilitates the computation of the PCL directly from catalogue data. This approach naturally avoids any bias introduced by the process of pixelising discretely

Read the complete text