How well is the local Large Scale Structure of the Universe known? CosmicFlows vs. Biteau's Galaxy Catalog with Cloning

Knowledge of the actual density distribution of matter in the local universe is needed for a variety of purposes, for instance as a baseline model for ultrahigh energy cosmic ray sources in the continuum limit and for predicting the diffuse dark matter annihilation signal. Determining the local mass density and velocity distribution is the aim of the CosmicFlows project. An alternate approach is based on catalogs of galaxies, supplemented with some scheme for filling in unseen galaxies. Here we compare the density field proposed by J. Biteau (2021) with the quasi-linear density field of CosmicFlows2 (Y. Hoffman et al. 2018) and the mean posterior field of CosmicFlows4 (A. Valade 2026). We find factor-of-two level differences in some regions and even larger discrepancies toward the Galactic center zone of avoidance (ZoA), defined by absolute longitude less than 30 degrees and absolute latitude less than 20 degrees, as filled by Biteau using cloning. Within 11 Mpc the density field is well determined by the Local Volume catalog (I. D. Karachentsev et al. 2018), which is directly incorporated by Biteau, and is preferred over the CosmicFlows modeling. At larger distances Biteau (2021) is also preferred for the direction and integrated mass of non-obscured structures, but the radial mass distribution is less robust due to line-of-sight peculiar velocities. Within the ZoA the galaxies of Biteau (2021) are entirely fictitious and their use is ill advised except where they do not contribute to the observable of interest. Moreover, as shown by the South Pole Wall, discovered by CosmicFlows but absent from galaxy catalogs, obscuration can be significant even outside the ZoA. Unexpectedly, we find that the angular positions of structures in CosmicFlows are sometimes misaligned with their true positions as seen in galaxy catalogs outside the ZoA.

💡 Research Summary

This paper presents a systematic, quantitative comparison of three recent reconstructions of the local large‑scale matter density field: CosmicFlows‑2 (CF2), CosmicFlows‑4 (CF4), and the semi‑empirical galaxy catalog of Biteau 2021 (B21), which fills the Galactic zone of avoidance (ZoA) by “cloning” galaxies across the obscured region. The authors motivate the study by noting that accurate knowledge of the nearby density field is essential for applications ranging from ultra‑high‑energy cosmic‑ray source modeling to predictions of diffuse dark‑matter annihilation signals.

Methodology
The authors divide space into concentric spherical shells of 1.43 Mpc thickness, spanning 0–350 Mpc. Each shell is pixelated with HEALPix at NSIDE = 32 (≈3.36 deg²). CF2 voxel data are taken from previous works; CF4, which contains 55 877 galaxies, is provided on a 512³ Cartesian grid, and the density of the nearest grid point is assigned to each shell‑pixel (the grid spacing matches the shell thickness). For B21, each galaxy is represented by a Gaussian radial profile with smoothing scales Rs = 5 Mpc and 10 Mpc (the same values used by Biteau for visualisation). The fraction of a galaxy’s mass that falls into a given shell is obtained by integrating the Gaussian over the shell thickness, and the resulting mass is placed in the corresponding HEALPix pixel. Within 0–10 Mpc the authors apply a 2‑D nested‑pixel smoothing (redistributing mass over a 3‑ring neighbourhood, equivalent to an angular smoothing of ≈10.5°) to avoid artificial mass leakage caused by the cloning procedure when the smoothing radius exceeds the distance.

Results – Radial Over‑Density
Figure 1 shows the mean overdensity (ρ/⟨ρ⟩) in 10 Mpc thick shells. All three reconstructions display a pronounced peak near 70 Mpc, corresponding to the Great Attractor, but the absolute amplitudes differ: CF4 is systematically higher than CF2 by ≈20 % in the 50–100 Mpc range. B21 exhibits an overall excess of 10–30 % relative to the CosmicFlows models, even when the ZoA is excluded, indicating that the cloning algorithm injects extra mass into the average budget. Between 270 and 340 Mpc the B21 overdensity rises sharply; the authors attribute this to the cloned galaxies populating the ZoA and artificially inflating the density.

Results – Sky Maps
Figure 2 presents logarithmic density contrast maps for each model in several distance shells. Major clusters (Virgo, Hydra‑Centaurus, Perseus‑Pisces, Coma) appear in all three, but notable differences emerge:

• CF4 resolves finer sub‑structures (Antlia, NGC 5846, Fornax‑Eridanus) and shows larger variance than CF2.
• B21, when smoothed with a 10 Mpc kernel, reproduces many of the same large‑scale features as CF4 beyond 40 Mpc, yet displays an artificial concentration of mass along the super‑galactic plane in the 0–20 Mpc shell—absent in the CosmicFlows maps.
• In the 20–60 Mpc shells, B21 shows spurious structures near the Galactic centre that are direct artefacts of the cloning rule (mirroring galaxies across |b| = 20°). The cloned region is outlined by green lines in the figures.
• The Pavo‑Indus Cloud appears more compact and symmetric in B21 than in CF4, especially in the 40–60 Mpc and 80–130 Mpc shells. This is again a cloning effect, because the cloud lies close to the ZoA boundary and its mirrored counterpart is added on both sides of the plane.
• The South Pole Wall, discovered by CosmicFlows, is visible in CF4 (and to a lesser extent in CF2) but completely missing from B21, illustrating that velocity‑based reconstructions can recover matter hidden behind dust and gas clouds.

Angular Mis‑Alignments
A detailed voxel‑by‑voxel comparison (Appendix Fig. 6) reveals that CF2 and CF4 sometimes disagree on the angular position of structures by several degrees. The authors suggest that distance estimates based purely on redshift, especially for clusters with large peculiar velocities, can shift the inferred location of overdensities, leading to apparent mis‑alignments.

Interpretation and Recommendations
The authors conclude that:

• Within ≈11 Mpc the Local Volume catalog (Karachentsev et al. 2018) incorporated directly into B21 provides the most reliable density field, because distances are measured with high‑precision methods (TRGB, Cepheids, etc.).
• Beyond this range, for regions not obscured by the Milky Way, B21’s galaxy‑based approach yields a reasonable estimate of the direction and total mass of prominent structures, but the radial mass distribution remains uncertain due to line‑of‑sight peculiar velocities.
• In the ZoA, the cloned galaxies are entirely fictitious; their inclusion is only justified when the observable of interest is insensitive to the mass in that region (e.g., when modelling UHECR propagation that avoids the Galactic plane).
• CosmicFlows‑4, with its velocity‑based reconstruction, remains the most trustworthy all‑sky model because it can infer matter in obscured zones (e.g., the South Pole Wall) and does not rely on ad‑hoc cloning.
• Future work should either improve the cloning scheme (e.g., by using statistical models of the hidden galaxy population) or, preferably, rely on velocity‑based reconstructions complemented by deeper infrared surveys to reduce the size of the ZoA.

Overall, the paper provides a thorough, method‑ical assessment of the strengths and weaknesses of galaxy‑catalog versus velocity‑field approaches, highlighting where each excels and where caution is required.