Lost in the FoG: Pitfalls of Models for Large-Scale Hydrogen Distributions
Calvin K. Osinga, Benedikt Diemer, Francisco Villaescusa-Navarro · 2025
Read Lost in the FoG: Pitfalls of Models for Large-Scale Hydrogen Distributions by Calvin K. Osinga, Benedikt Diemer, Francisco Villaescusa-Navarro (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-2503-01956.txt.
Introduction
Calvin K Osinga, Benedikt Diemer, and Francisco Villaescusa-Navarro test the models used to read large maps of hydrogen in the universe. Their 2025 paper notes that surveys of neutral hydrogen, cross matched with galaxy maps, could serve as strong probes of cosmology. Reading them needs models for how motion distorts the maps and how tracers differ from matter.
Using the IllustrisTNG simulation suite, the team checked model accuracy for hydrogen maps at modest redshifts. Common shortcuts caused errors as large as or larger than measurement uncertainty. Ignoring the fingers-of-God effect, a smearing from fast small scale motion, was the worst offender. Even fuller models still missed by about ten percent on key scales.
Catalog record
| CREATOR | Calvin K. Osinga, Benedikt Diemer, Francisco Villaescusa-Navarro |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 105 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2503-01956
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Lost in the FoG: Pitfalls of Models for Large-Scale Hydrogen Distributions Calvin K. Osinga, Benedikt Diemer, Francisco Villaescusa-Navarro Abstract Large-scale H {\rm H} i surveys and their cross-correlations with galaxy distributions have immense potential as cosmological probes. Interpreting these measurements requires theoretical models that must incorporate redshift-space distortions (RSDs), such as the Kaiser and fingers-of-God (FoG) effect, and differences in the tracer and matter distributions via the tracer bias. These effects are commonly approximated with assumptions that should be tested on simulated distributions. In this work, we use the hydrodynamical simulation suite IllustrisTNG to assess the performance of models of z ≤ 1 z\leq 1 H {\rm H} i auto and H {\rm H} i -galaxy cross-power spectra, finding that the models employed by recent observations introduce errors comparable to or exceeding their measurement uncertainties. In particular, neglecting FoG causes ≳ 10 % \gtrsim 10\% deviations between the modeled and simulated power spectra at k ≳ 0.1 k\gtrsim 0.1 h cMpc -1 , larger than assuming a constant bias which reaches the same error threshold at slightly smaller scales. However, even without these assumptions, models can still err by ∼ 10 % \sim 10\% on observationally relevant scales. These remaining errors arise from multiple RSD damping sources on H {\rm H} i clustering, which are not sufficiently described with a single FoG term. Overall, our results highlight the need for an improved understanding of RSDs to harness the capabilities of future measurements of H {\rm H} i distributions. numpy [ 57 ] , matplotlib [ 64 ] , seaborn [ 156 ] , h5py [ 28 ] , colossus [ 40 ] and pyfftw [ 54 ] I Introduction Cosmology imprints on the large-scale structure of the Universe, allowing measurements of the distribution of matter to probe the nature of dark matter and dark energy and constrain cosmological parameters [ 66 , 46 , 118 , 37 ] . However, we cannot observe the Universe’s large-scale structure directly. We instead rely on surveys of baryonic tracers such as atomic neutral hydrogen ( {\rm H} i ) or galaxies to infer the underlying matter distribution [ 36 ] , a process which requires models of the matter-tracer relationship. These models are simple in the linear regime where tracers faithfully follow the matter distribution, but baryonic physics and nonlinear effects complicate the behavior on small scales, demanding more sophisticated models. Despite the challenge, the stronger constraining power on small scales makes the development of these complex models worthwhile [ 75 , 25 ] . Consequently, a large body of work has been dedicated to improving our understanding of structure in the quasi-linear and nonlinear regime. Notable examples of modeling techniques include Lagrangian perturbation theory [ 9 , 15 , 151 , LPT,] and halo occupation distribution (HOD) models [ 165 , 168 ] . However, both techniques have limitations: LPT does not natively include baryonic effects and HOD requires knowledge and assumptions of how tracers occupy halos. In certain applications, a simpler and more general prescription which assumes parameterized forms for the various nonlinear contributions is preferred. These general prescriptions have interpreted measurements of {\rm H} i and galactic distributions to constrain cosmological parameters. Of the two tracers, galaxies are the traditional probe of large-scale structure [ 27 , e.g.,] , but {\rm H} i has arisen as a promising candidate due to its ability to cheaply and quickly observe large volumes of space via 21cm intensity mapping [ 11 , 6 , 77 ] . However, 21cm intensity mapping experiments still struggle with low signal-to-noise ratios, arising from galactic foregrounds and systematics [ 91 ] , making the interpretation of {\rm H} i auto power spectra challenging [ 107 , hereafter 107 ] . These noise sources are absent in galaxy surveys, so the noise can be reduced by cross-correlating {\rm H} i and galaxies. These cross-correlations have constrained the cosmological abundance of {\rm H} i ( \Omega_{\mathrm{HI}} ) at low redshifts [ 22 , 90 , 142 , 2 , z\leq 1 ,] . In particular, L. Wolz et al. [163, hereafter W22 ] , S. Cunnington et al. [32, C22 ] , CHIME Collaboration et al. [24, CHIME23 ] , and I. P. Carucci et al. [18, C24 ] have produced \Omega_{\mathrm{HI}} constraints competitive to other {\rm H} i observation techniques. As measurement and data processing techniques improve [ 31 , 16 , 113 ] , {\rm H} i -galaxy cross-correlations will probe other cosmological quantities in the near future [ 93 ] . However, the resulting constraints depend on how well the adopted models capture the relationship between observed tracer and matter distributions. We can evaluate these models’ performance by comparing their predictions to known {\rm H} i , galaxy, and matter distributions in simulations. Cosmological hydrodynamical simulations now produce realistic low-redshift {\rm H} i and galaxy distributions in sufficiently large volumes to be useful for this purpose [ 4 , 96 , 42 , 140 , 34 ] . In particular, C. K. Osinga et al. [101] showed that the {\rm H} i auto- and cross-correlations from IllustrisTNG agree closely with relevant observations at z\leq 1 . We leverage the agreement between IllustrisTNG and observations to gain insight into the performance of large-scale {\rm H} i models in two phases. First, we report IllustrisTNG’s predictions for the various model components and assess how well their simplified forms in the general model agree with the actual values. Second, we evaluate the accuracy of popular large-scale prescriptions by comparing their predictions to the correlations computed directly from IllustrisTNG. The paper is structured as follows. We first describe the data from IllustrisTNG in Section II . Next, we explore general model designs in Section III . Then, we describe predictions for each model ingredient individually in Section IV and the models as a whole in Section V . We discuss directions for future work in Section VI , before concluding in Section VII . For brevity, some figures are referenced but not included; these are provided in the author’s website 1 1 1 www.calvinosinga.com/papers/hi_cosmo/sup_analysis . II Simulation data We use the IllustrisTNG suite of cosmological magneto-hydrodynamics simulations [ 96 , 110 , 139 , 95 , 88 , 97 ] . This suite offers simulations in three different box sizes, 35 h -1 cMpc, 75 h -1 cMpc, and 205 h -1 cMpc, at varying resolutions, using the Planck Collaboration et al. [112] cosmological parameters ( \Omega_{\rm{m}}=0.3089 , \Omega_{\rm{b}}=0.0486 , h=0.6774 , \sigma_{\rm{s}}=0.8159 ). The simulation is evolved using the AREPO code [ 137 ] , which employs a tree-PM method for gravity calculations and a Voronoi mesh-based Godunov scheme for magneto-hydrodynamics. The IllustrisTNG simulations incorporate sub-grid models to simulate unresolved processes like star formation, gas cooling, and AGN [ 152 , 158 ] . The models are calibrated against a selection of observational data to accurately represent the galaxy population at low redshifts [ 111 ] . Dark matter haloes are identified using the Friends-of-Friends (FoF) algorithm [ 33 ] , and their internal substructures, or “subhalos”, are cataloged using the Subfind algorithm [ 138 ] . Our main analyses focus on the largest box, 205 h -1 cMpc