Cosmological inference with halo clustering reconstructed from the redshift-space galaxy distribution

Ryuichiro Hada, Teppei Okumura · 2026

Read Cosmological inference with halo clustering reconstructed from the redshift-space galaxy distribution by Ryuichiro Hada, Teppei Okumura (2026) 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-2604-11694.txt.

Introduction

Hada and Okumura study cosmology with halo clustering from galaxy distributions. Small-scale clustering matters for full-shape work, where satellite galaxies feed one-halo terms and Finger-of-God effects. Using DESI-like red galaxy mocks from AbacusSummit at redshift 1.1, they model the reconstructed-halo spectrum with field theory. Reconstruction systematics fit within field theory: a multipole rescaling from large data captures the dominant effect, while counterterms absorb small-scale residuals. The sample yields unbiased constraints, with growth-rate uncertainty falling over 20 percent and stable fits beyond usual limits.

Catalog record

CREATORRyuichiro Hada, Teppei Okumura
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE72787
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2604.11694

Browse the full library · About the Tower of Babel

Full text

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

Cosmological inference with halo clustering reconstructed from the redshift-space galaxy distribution Ryuichiro Hada, Teppei Okumura Abstract Accurate modeling of small-scale redshift-space clustering is crucial for full shape RSD analyses, where satellite galaxies contribute to 1-halo terms and Finger-of-God distortions. We investigate halo reconstruction based on the cylinder grouping (CG) method of Okumura et al. (2017), which selects an effective halo center tracer from the observed galaxy distribution, and how it impacts cosmological parameter inference. Using DESI-like luminous red galaxy mock catalogs from the AbacusSummit simulations at z = 1.1 z=1.1 , we perform effective field theory (EFT)-based full-shape modeling of the power spectrum of the reconstructed-halo sample. We show that the dominant reconstruction-induced systematics can be described and incorporated within the standard EFT framework. In particular, a simple multipole-dependent rescaling inferred directly from the data on large scales captures the dominant effect, while residual small-scale changes are absorbed by the standard counterterm and stochastic sector, without introducing additional reconstruction-specific parameters. The reconstructed-halo sample yields unbiased constraints on cosmological parameters, including the growth rate f ​ σ 8 f\sigma_{8} and Alcock-Paczynski parameters. Compared to the galaxy sample, it enables both improved robustness and increased statistical precision: the inferred f ​ σ 8 f\sigma_{8} remains stable when extending the fit beyond k max ≃ 0.2 ​ h ​ Mpc − 1 k_{\max}\simeq 0.2\,h\,{\rm Mpc}^{-1} , with its uncertainty reduced by more than 20 % 20\% . I Introduction Redshift-space distortions (RSD) ( 20 ; 15 ; 10 ; 16 ) in galaxy clustering provide a powerful route to measuring the growth of cosmic structure, and hence to testing gravity and dark energy ( 14 ; 27 ; 2 , e.g.,) with spectroscopic surveys. Full-shape analyses of anisotropic galaxy clustering have become a standard route to cosmological inference from redshift-space two-point statistics. Early work already exploited the full anisotropic correlation function at the linear-theory level ( 26 ) . More recent studies have extended these analyses using perturbative modeling of the galaxy power spectrum and emulator-based halo-model approaches ( 12 ; 18 ; 29 ; 21 ) (see also ( 1 ) for the latest full-shape RSD results and a comprehensive overview). In practice, however, extracting this information increasingly relies on modeling modes deep into the quasi-nonlinear regime, where the mapping from real to redshift space is affected by nonlinear evolution, nonlinear bias, and, most importantly for galaxy samples, the virial motions of satellite galaxies. Satellite contamination boosts the large-scale clustering amplitude (because satellites preferentially reside in massive, highly biased halos) while simultaneously suppressing the small-scale signal through Finger-of-God (FoG) smearing 19 . These effects complicate the connection between observed galaxy clustering and theoretical predictions, and can limit the robustness of parameter inference when pushing to high k_{\rm max} ( 25 , e.g.,) . A widely used strategy to extend RSD analyses into the quasi-nonlinear regime is the Effective Field Theory (EFT) of large-scale structure and related perturbative frameworks ( 7 ; 30 ; 36 , e.g.,) . In these approaches, the impact of unresolved small-scale physics is parameterized by higher-derivative counterterms and stochastic contributions, allowing the template to remain flexible while preserving a controlled large-scale expansion. In practice, however, this flexibility comes with an important trade-off: as smaller scales are included, the inference can become increasingly sensitive to the nuisance sector, and degeneracies among stochastic and counterterm parameters can degrade the constraining power on cosmological parameters. Moreover, such degeneracies in a high-dimensional nuisance sector can lead to non-trivial projection effects and increased prior/parameterization sensitivity in marginalized constraints, even when the maximum-likelihood region remains close to the true model ( 17 ; 22 ; 13 , e.g.,) . This motivates complementary approaches that reduce the amplitude of problematic small-scale contributions at the level of the tracer field — in particular, constructing tracers that more closely resemble halo centers, for which 1-halo and FoG effects are substantially suppressed. In simulations, host halos can be identified directly, but in observations the halo-center field must be inferred from the galaxy distribution itself. This has motivated a variety of approaches aimed at mitigating satellite-driven nonlinearities and virial motions already at the level of the observed tracer field ( 25 , e.g.,) . Early work explored redshift-space friends-of-friends (FoF) style algorithms and related “FoG compression” schemes that identify high-density structures and replace member galaxies with group representatives, thereby reducing virial motions and the associated small-scale suppression ( 31 , e.g.,) . A complementary line of work instead suppresses the contribution of the most nonlinear regions by applying density-dependent weighting or transformations of the field (e.g., “clipping”), with the goal of extending the range of scales that can be modeled perturbatively ( 33 , e.g.,) . While these methods can improve the robustness of clustering analyses, their performance depends on sample properties (number density, satellite fraction, typical host-halo mass) and on the details of the adopted procedure. In particular, redshift-space grouping and FoG-compression schemes can introduce additional line-of-sight (LOS) anisotropies through their inherently anisotropic geometry. Instead of the FoF method, 28 proposed a similar but simpler cylinder-grouping (CG) method, a practical halo-reconstruction scheme tailored to suppress satellite contamination in redshift space. The algorithm ranks galaxies by a halo-mass proxy and iteratively identifies central–satellite systems by removing galaxies within a cylinder aligned with the LOS, chosen to encompass FoG displacements. A key advantage of the CG approach is that it relies only on the observed galaxy distribution and a simple geometric criterion, making it particularly attractive for high-number-density samples. At the same time, the LOS-aligned cylinder geometry introduces characteristic reconstruction-induced anisotropies: on small scales through an exclusion-like removal of close pairs, and on large scales through an apparent anisotropic modulation that can mimic RSD. 28 provided a simple semi-analytic description of these systematics, modeling the small-scale effect with an exclusion window and the large-scale effect as a multipole-dependent rescaling of the linear power spectrum, and showed that the reconstructed halo power-spectrum multipoles are accurately reproduced over a wide range of scales. The goal of the present work is to take the next step and assess the practical impact of the CG reconstruction on cosmological parameter inference . This direction is particularly relevant in the context of ongoing and upcoming spectroscopic surveys such as the Dark Energy Spectroscopic Instrument ( 11 , DESI,) and the Prime Focus Spectrograph ( 34 , PFS,) . Specifically, we apply the CG method to galaxy mock catalogs built from cosmological N -body simulations and perform RSD fits in an EFT-based framework (using a Lagrangian-PT template). Our focus is not only on whether the reconstructed clustering can be described at the level of the power-spectrum multipoles, but also whether the reconstruction-induced systematics can be handled consistently within a standard likelihood pipeline. To this end, we organize the modeling around the two qualitative reconstruction effects highlighted by 28 : (i) a small-scale exclusion effect associated with cylinder grouping, whose residual impact is treated within the EFT stochastic and counterterm sector, and (ii) a large-scale “directional masking” effect that primarily rescales the clustering amplitude in a multipole-dependent manner, which we correct by applying large-scale ratios measured from the mocks as a multiplicative renormalization of the large-scale deterministic part of the model. This separation provides a simple and practical

Read the complete text