Pith. sign in

REVIEW 2 major objections 4 minor 27 references

The relation of galaxies and dark matter haloes to the filamentary cosmic web

T0 review · 2 major / 4 minor · reviewed 2026-08-11 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read The cosmic web shapes galaxies mostly by selecting the haloes they live in.

desk verdict A careful, well-scoped study of one SAM and one web finder, whose central 'halo mass bias' claim is plausible but slightly overreaches its own evidence; still deserves a serious referee. read the letter →

arxiv 2412.03438 v2 pith:6H3MA2XC submitted 2024-12-04 astro-ph.CO astro-ph.GA

Navdha , Philipp Busch , Simon D. M. White This is my paper · ORCID
classification astro-ph.COastro-ph.GA MSC 85A40 PACS 98.80.-k98.65.-r
keywords cosmicweblarge-scalestructuredarkmatterhaloessemi-analyticgalaxyformationMillenniumSimulationspecificstarratebacksplashgalaxiesVoronoitessellation
topics Dark Matter
open problems Dark Matter
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper asks whether the filamentary cosmic web directly shapes galaxies or merely marks where heavier haloes live. Using the Millennium Simulation and a web defined as the unique percolating region with density above $5.25\langle\rho\rangle$, the authors find that the web holds an increasing fraction of haloes and galaxies with mass. Low-mass web galaxies are overwhelmingly satellites or backsplash galaxies inside more massive haloes, and once central/satellite and backsplash status are accounted for, the star-formation-rate distributions look almost the same inside and outside the web. The paper concludes that, within the semi-analytic galaxy formation model adopted, web versus non-web differences are almost entirely a consequence of the different halo mass distributions in the two environments, not of a direct effect of the filament itself.

What carries the argument

The carrying object is the BW20 cosmic web definition: a Voronoi-tessellation density field built from the dark matter particles, thresholded at $5.25\langle\rho\rangle$ to leave one percolating structure. This supplies a binary in/out web classification and, through a KDTree, a distance-to-web measure, enabling comparisons of mass functions and star-formation distributions as functions of web distance. The second load-bearing element is the galaxy classification into centrals, satellites, and backsplash galaxies, where backsplash galaxies are present-day centrals that were once satellites inside a more massive halo and later exited; these are found by following merger-tree main progenitor branches. Splitting the sample by these types is what separates the halo-mass effect from any direct web effect.

What would settle it

Run a hydrodynamical cosmological simulation at comparable resolution and compare specific star-formation-rate distributions at fixed halo mass for star-forming central galaxies inside and outside the web; if filament gas quenches galaxies beyond the virial radius, the curves will separate. Repeating the analysis with a numerically specified backsplash threshold, for example a factor-two drop in halo mass, would show whether the residual central-galaxy difference depends on the loose definition.

Watch

Extended reading notes

Core claim

The central claim is that the relation between galaxies and the cosmic web is a halo-mass selection effect. The web, defined by the BW20 prescription (a Voronoi-tessellated density field thresholded at $5.25\langle\rho\rangle$), contains 35% of the mass in 0.62% of the volume; halo web fractions rise from 26% at $10^{11}\,M_\odot$ to over 90% above $10^{14}\,M_\odot$. Galaxy web fractions rise from 45% at $10^{8.5}\,M_\odot$ to 90% at $10^{11.5}\,M_\odot$, and this excess at low stellar mass appears because most low-mass web galaxies are satellites of massive haloes. The paper shows that for galaxies of fixed stellar mass, specific star-formation-rate distributions of central, satellite, and backsplash galaxies are nearly the same in and outside the web; the strong web/non-web difference in the overall population is driven by the two-to-three times larger satellite fraction and ten times larger backsplash fraction in the web. For non-web objects, stellar mass functions and sSFR distributions are almost independent of distance to the web. The conclusion follows that differences between web and non-web galaxy populations are almost entirely due to the difference in halo mass distribution between the environments.

Load-bearing premise

The conclusions rest on the assumption that galaxy properties are set by halo mass-assembly history, with no gas stripping or evaporation from the filament environment acting outside virial radii; they also rely on a qualitative criterion for identifying backsplash galaxies as centrals whose halo mass "dropped significantly".

Editorial extensions

If this is right

  • If the claim holds, surveys that count low-mass galaxies in filaments are largely seeing the satellite population of massive haloes, not a population quenched by the filament itself.
  • Outside the web, halo and galaxy mass functions and star-formation distributions are nearly independent of distance to the web, so environmental gradients should be confined to the web boundary in this model.
  • Web haloes of a given mass contain only 10 to 20 per cent more satellites and modestly more backsplash galaxies than non-web haloes, so most of the web's galaxy population bias comes from the halo mass function, not halo content.
  • The tenfold higher backsplash fraction among web centrals is the main residual star-formation difference, suggesting past accretion into a more massive system, not current web location, suppresses star formation.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • A hydrodynamical simulation would include ram-pressure and evaporation from gas in filaments, which could add quenching beyond virial radii; comparing sSFR at fixed halo mass in such a simulation is a direct test that could overturn this conclusion for real galaxies.
  • The paper's web is defined from dark matter only; connecting it to galaxy-defined webs would reveal how observational filament classifications change the inferred web/non-web contrasts.
  • Because the conclusion comes from one semi-analytic model, a controlled comparison across galaxy formation models with the same web definition would show how much of the halo-mass interpretation is model-dependent.
  • The absence of a direct web effect on star formation suggests that filament orientation effects are better sought in galaxy shapes and spins, which this analysis does not address.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

2 major / 4 minor

Summary. This paper uses the Millennium Simulation and the Busch & White (2020) cosmic web definition, the unique percolating object above 5.25 times the mean matter density, to study how the fractions of dark matter haloes and galaxies inside versus outside the web depend on mass, how the mass functions of non-web objects depend on distance from the web, and whether specific star-formation rate (sSFR) distributions differ between web and non-web environments. It finds that web fractions increase with mass for both haloes and galaxies, that non-web mass functions are nearly independent of distance from the web, and that the lower sSFRs of web galaxies are largely associated with higher satellite and backsplash fractions in the web. The paper concludes that, for the Guo et al. (2011) semi-analytic model, differences between web and non-web galaxy populations are almost entirely due to the difference in halo mass distribution between the two environments.

Significance. If the central conclusion holds, it implies that the cosmic web's influence on galaxy star formation is largely indirect, acting through the halo mass distribution, with only modest additional effects at fixed halo mass. The paper has several strengths: it uses a large-volume, high-resolution simulation with negligible counting noise in the main statistics; the web definition is taken from prior work rather than fitted to the galaxy properties analyzed, so circularity is not a concern; and the authors openly identify the semi-analytic model's lack of hydrodynamical filament physics as a limitation for extrapolation to real galaxies. The analysis pipeline is transparent and reproducible in principle from public data. However, the headline claim that differences are 'almost entirely' due to the halo mass distribution is not quantitatively demonstrated, because no mass-matched or reweighted comparison isolates the contribution of the mass distribution from residual fixed-mass environmental effects.

major comments (2)
  1. [Section 3.3 and Abstract] The central claim that differences between web and non-web galaxy populations are 'almost entirely due to the difference in halo mass distribution' is not directly tested. Figure 6 shows that at fixed present-day halo mass, web haloes contain 10-20% more satellites and up to about a factor of three more backsplash galaxies than non-web haloes, and the text states that these effects 'account for only a small fraction of the stellar population shifts seen in Fig. 4' without showing the calculation. A quantitative decomposition, such as a halo-mass-matched comparison or a reweighting that fixes the halo mass distribution, is needed to separate the contribution of the mass distribution itself from these residual environmental effects. Without this, the abstract's 'almost entirely' outruns the evidence presented.
  2. [Section 2.2, backsplash identification] The definition of backsplash galaxies relies on the qualitative statement that the virial mass of the halo 'dropped significantly' after the object passed through a type 1 phase, but no numerical threshold is given. Since the right panels of Fig. 4 and the conclusion that sSFR distributions match once backsplash fractions are accounted for depend directly on this classification, the threshold should be specified and its sensitivity tested. If the fraction of objects classified as backsplash varies strongly with the chosen threshold, the robustness of the central conclusion would need to be re-evaluated.
minor comments (4)
  1. [Section 4, Discussion] The paper states that the qualitative conclusions would remain valid at a lower threshold such as rho_th/<rho> = 1, but this is not demonstrated quantitatively for the sSFR comparisons. A short alternative-threshold test would strengthen the claim that the results are not an artifact of the specific threshold choice.
  2. [Section 1, Introduction] There is a typo in 'correpond' in the first paragraph; it should be 'correspond'.
  3. [References] The reference to Regos et al. (2024) lacks a volume and page/article number (only 'ApJ' is given), and in the text it appears as 'Reg¨os et al.' with an odd diacritic; this should be cleaned up.
  4. [Figure 4, caption] The caption description is clear, but in the main text the phrase 'the peak at∼ −12' is missing a space before the equals sign; minor formatting fixes throughout would improve readability.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the BW20 web definition is fixed from dark matter density alone, and the galaxy-web comparisons are independent measurements within the Guo et al. model.

full rationale

The cosmic web is defined by Busch & White (2020, with White a coauthor here) from a Voronoi-tessellated dark-matter density field at a fixed threshold of 5.25 times the cosmic mean, chosen from a percolation transition; it does not use any galaxy property analyzed in this paper. Galaxy membership in the web is inherited from the associated simulation particle, so the quantities whose web dependence is studied (halo mass, stellar mass, sSFR, satellite/backsplash status) do not enter the web definition. The central result that web/non-web sSFR differences largely disappear for fixed central/satellite/backsplash class, and that residual population differences are driven by higher satellite and backsplash fractions in the web, is a direct measurement in the Guo et al. (2011) model, not an algebraic identity. Fig. 6 does show fixed-halo-mass excesses of satellites and backsplash galaxies in the web, and the statement that these account for only a small fraction of the population shifts is an interpretive estimate rather than a derived decomposition; this is a strength-of-evidence concern, not circularity. The only notable self-citation is to BW20 for the web construction, which is prior, independent of the target galaxy statistics, and applicable without reference to the present conclusions; it does not reduce the paper's results to its inputs. The paper also explicitly acknowledges its limitations (no filament hydrodynamics in the SAM; qualitative backsplash criterion), which further supports that the analysis is not circular.

Assumptions & free parameters 3 free parameters · 5 assumptions · 0 invented entities

No new particles or forces; the only adjustable inputs are inherited from prior work (web threshold, persistence ratio) plus an unquantified backsplash definition. The central claim is conditional on the SAM and web definition.

free parameters (3)
  • Web density threshold rho_th / <rho> = 5.25
    Chosen in Busch & White (2020) at the percolation transition, not fitted to the galaxy properties studied here; the paper shows quantitative web fractions depend strongly on this choice.
  • Persistence ratio for TLT peak suppression = 10
    Taken from BW20 to suppress discreteness noise in the tessellation; not varied or fitted here.
  • Backsplash identification mass-drop threshold = not specified
    The paper defines backsplash as centrals whose halo virial mass 'dropped significantly' after a type-1 phase, without giving a numerical criterion; the resulting backsplash fractions are central to the sSFR decomposition.
assumptions (5)
  • domain assumption The Millennium Simulation's Lambda-CDM cosmology and particle distribution adequately represent cosmic structure at z=0.
    The entire analysis is built on this simulation without comparing web statistics to observations.
  • domain assumption The BW20 Voronoi-tessellation density field and percolation threshold identify the cosmic web as a physical object.
    The web definition is taken as given from Busch & White (2020); all results are conditional on it.
  • domain assumption Galaxy properties in the Guo et al. (2011) semi-analytic model depend only on halo merger-tree history, not on accretion morphology.
    This is the model assumption that makes the halo-mass explanation natural; the paper notes it may fail due to assembly bias.
  • domain assumption Backsplash galaxies can be reliably identified by following main progenitor branches and detecting a past type-1 phase with a significant virial mass drop.
    The backsplash classification drives the sSFR decomposition in Fig. 4, but the mass-drop criterion is not quantified.
  • domain assumption Assigning each galaxy to a specific simulation particle and using its web membership as the galaxy's web membership is valid.
    This bypasses issues of halo boundary but assumes the particle choice is representative.

how reviews work

0 comments
Cite this review

Pith. "Pith review of The relation of galaxies and dark matter haloes to the filamentary cosmic web." pith.science (2026). https://pith.science/paper/6H3MA2XC

@misc{pith2026241203438,
  author       = {Pith},
  title        = {Pith review of: The relation of galaxies and dark matter haloes to the filamentary cosmic web},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/6H3MA2XC}},
  note         = {Machine review of arXiv:2412.03438}
}
abstract

We use the Millennium Simulation to study the relation of galaxies and dark matter haloes to the cosmic web. We define the web as the unique, fully connected, percolating object with (unsmoothed) matter density everywhere exceeding 5.25 times the cosmic mean. This object contains 35\% of all cosmic mass but occupies only 0.62\% of all cosmic volume. It contains 26\% of dark matter haloes of mass $10^{11}M_\odot$, rising to 50\% at $10^{12.7}M_\odot$, and to $>90\%$ above $10^{14}M_\odot$. In contrast, it contains 45\% of all galaxies of stellar mass $10^{8.5}M_\odot$, rising to 50\% at $10^{10}M_\odot$, to 60\% at $10^{11}M_\odot$ and to 90\% at $10^{11.5}M_\odot$. This difference arises because a large fraction of all satellite and backsplash galaxies are part of the cosmic web. Indeed, more than 50\% of web galaxies are satellites for stellar masses below that of the Milky Way, rising to about 70\% below $10^{10}M_\odot$, whereas centrals substantially outnumber satellites in the non-web population at all stellar masses. As a result, web galaxies have systematically lower specific star-formation rates (sSFR's) than non-web galaxies. For the latter, the distributions of stellar mass and sSFR are almost independent of web distance. Furthermore, for both central and satellite galaxies, the sSFR distributions at given stellar mass are very similar in and outside the web, once differences in backsplash fraction are accounted for. For the galaxy formation model considered here, differences between web and non-web galaxy populations are almost entirely due to the difference in halo mass distribution between the two environments.

Figures

Figures reproduced from arXiv: 2412.03438 by the authors.

Figure 1
Figure 1. A 20h−1Mpc thick slice along the X-coordinate showing all galaxies with stellar mass M∗ > 109.5M⊙. Those which are part of the cosmic web are coloured red, the rest black. The blue shading in the right half of the plot indicates the minimum distance d from each point in the midplane to a particle in the web (the so-called EDT). BW20 chose this slice to include the global maximum of the EDT field (in the bottom right… view at source ↗
Figure 2
Figure 2. Mass functions and web fractions for dark matter haloes (1st row), central and satellite galaxies (2nd row), and all galaxies (3rd row). In the mass function panels on the left, the overall population is shown in blue and the subpopulation belonging to the web in green. The sharp drop in the stellar mass functions at low mass reflects the limited mass resolution of the MS; the MS agrees reasonably well with the high… view at source ↗
Figure 3
Figure 3. Normalized cumulative distributions of distance to the cosmic web for galaxies and haloes as a function of their mass, as well as for random dark matter particles and for random points in space. In the left panel the curves are for all objects in each class, so that their value at d → 0 corresponds to the fraction which are in the web (see also [PITH_FULL_IMAGE:figures/full_fig_p007_3.png] view at source ↗
Figures from the paper (2 more)
Figure 4
Figure 4. Figure 4: Normalized sSFR distributions in mass ranges indicated by a legend above each panel for galaxies in and outside the cosmic web (CW). The left panels show results for all galaxies, splitting non-web objects into two groups according to distance from the web. The central…
Figure 5
Figure 5. Figure 5: Mean value of sSFR along with the standard error in the mean for the non-web galaxy population plotted as a function of distance to the cosmic web, split into four mass ranges. Values of the sSFR averaged over all distances for each mass range are shown as grey horizon…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

27 extracted references · 8 canonical work pages

  1. [1]

    Abbott T. M. C., et al., 2018, @doi [Phys. Rev. D] 10.1103/PhysRevD.98.043526 , 98, 043526

    doi:10.1103/physrevd.98.043526 2018

  2. [3]

    Binggeli B., 1982, , https://ui.adsabs.harvard.edu/abs/1982A&A...107..338B 107, 338

    work page 1982

  3. [4]

    R., Kofman L., Pogosyan D., 1996, @doi [ ] 10.1038/380603a0 , https://ui.adsabs.harvard.edu/abs/1996Natur.380..603B 380, 603

    Bond J. R., Kofman L., Pogosyan D., 1996, @doi [ ] 10.1038/380603a0 , https://ui.adsabs.harvard.edu/abs/1996Natur.380..603B 380, 603

    doi:10.1038/380603a0 1996

  4. [5]

    Boylan-Kolchin M., Springel V., White S. D. M., Jenkins A., Lemson G., 2009, @doi [ ] 10.1111/j.1365-2966.2009.15191.x , 398, 1150

    arXiv 2009

  5. [6]

    Busch P., White S. D. M., 2020, @doi [ ] 10.1093/mnras/staa572 , 493, 5693

    work page doi:10.1093/mnras/staa572 2020

  6. [7]

    Codis S., Pichon C., Devriendt J., Slyz A., Pogosyan D., Dubois Y., Sousbie T., 2012, @doi [ ] 10.1111/j.1365-2966.2012.21636.x , https://ui.adsabs.harvard.edu/abs/2012MNRAS.427.3320C 427, 3320

    work page arXiv 2012

  7. [10]

    J., et al., 2006, @doi [ ] 10.1111/j.1365-2966.2005.09675.x , https://ui.adsabs.harvard.edu/abs/2006MNRAS.365...11C 365, 11

    Croton D. J., et al., 2006, @doi [ ] 10.1111/j.1365-2966.2005.09675.x , https://ui.adsabs.harvard.edu/abs/2006MNRAS.365...11C 365, 11

    work page arXiv 2006

  8. [11]

    S., White S

    Davis M., Efstathiou G., Frenk C. S., White S. D. M., 1985, @doi [ ] 10.1086/163168 , https://ui.adsabs.harvard.edu/abs/1985ApJ...292..371D 292, 371

    doi:10.1086/163168 1985

Show all 27 references
  1. [12]

    De Lucia G., Blaizot J., 2007, @doi [ ] 10.1111/j.1365-2966.2006.11287.x , https://ui.adsabs.harvard.edu/abs/2007MNRAS.375....2D 375, 2

    2007

  2. [13]

    M., et al., 2023, @doi [ ] 10.1093/mnras/stad1781 , https://ui.adsabs.harvard.edu/abs/2023MNRAS.523.5899D 523, 5899

    Delgado A. M., et al., 2023, @doi [ ] 10.1093/mnras/stad1781 , https://ui.adsabs.harvard.edu/abs/2023MNRAS.523.5899D 523, 5899

    2023 doi

  3. [14]

    Gao L., Springel V., White S. D. M., 2005, @doi [ ] 10.1111/j.1745-3933.2005.00084.x , https://ui.adsabs.harvard.edu/abs/2005MNRAS.363L..66G 363, L66

    2005

  4. [16]

    Guo Q., et al., 2011, @doi [ ] 10.1111/j.1365-2966.2010.18114.x , https://ui.adsabs.harvard.edu/abs/2011MNRAS.413..101G 413, 101

    2011

  5. [17]

    I., et al., 2018, @doi [ ] 10.1093/mnras/stx1976 , https://ui.adsabs.harvard.edu/abs/2018MNRAS.473.1195L 473, 1195

    Libeskind N. I., et al., 2018, @doi [ ] 10.1093/mnras/stx1976 , https://ui.adsabs.harvard.edu/abs/2018MNRAS.473.1195L 473, 1195

    2018 doi

  6. [18]

    F., Abadi M

    Navarro J. F., Abadi M. G., Steinmetz M., 2004, @doi [ ] 10.1086/424902 , https://ui.adsabs.harvard.edu/abs/2004ApJ...613L..41N 613, L41

    2004 doi

  7. [19]

    C., Gnedin N

    Neyrinck M. C., Gnedin N. Y., Hamilton A. J. S., 2005, @doi [ ] 10.1111/j.1365-2966.2004.08505.x , https://ui.adsabs.harvard.edu/abs/2005MNRAS.356.1222N 356, 1222

    2005

  8. [20]

    Planck Collaboration et al., 2020, @doi [ ] 10.1051/0004-6361/201833910 , https://ui.adsabs.harvard.edu/abs/2020A&A...641A...6P 641, A6

    2020 doi

  9. [21]

    Reg o s et al., 2024, @doi [Preprint at arXiv] arXiv:2407.02574

    2024 arXiv

  10. [22]

    N., et al., 2003, @doi [ ] 10.1086/377226 , https://ui.adsabs.harvard.edu/abs/2003ApJS..148..175S 148, 175

    Spergel D. N., et al., 2003, @doi [ ] 10.1086/377226 , https://ui.adsabs.harvard.edu/abs/2003ApJS..148..175S 148, 175

    2003 doi

  11. [23]

    Springel V., White S. D. M., Tormen G., Kauffmann G., 2001, @doi [ ] 10.1046/j.1365-8711.2001.04912.x , 328, 726

    2001

  12. [24]

    Springel V., et al., 2005, @doi [ ] 10.1038/nature03597 , https://ui.adsabs.harvard.edu/abs/2005Natur.435..629S 435, 629

    2005 doi

  13. [25]

    S., White S

    Springel V., Frenk C. S., White S. D. M., 2006, @doi [ ] 10.1038/nature04805 , https://ui.adsabs.harvard.edu/abs/2006Natur.440.1137S 440, 1137

    2006 doi

  14. [26]

    Stoughton C., et al., 2002, @doi [ ] 10.1086/324741 , https://ui.adsabs.harvard.edu/abs/2002AJ....123..485S 123, 485

    2002 doi

  15. [27]

    White S. D. M., 1996, in Schaeffer R., Silk J., Spiro M., Zinn-Justin J., eds, Cosmology and Large Scale Structure. p. 349

    1996

  16. [28]

    White S. D. M., Frenk C. S., Davis M., Efstathiou G., 1987, @doi [ ] 10.1086/164990 , https://ui.adsabs.harvard.edu/abs/1987ApJ...313..505W 313, 505

    1987 doi

  17. [29]

    Zakharova D., Vulcani B., De Lucia G., Xie L., Hirschmann M., Fontanot F., 2023, @doi [ ] 10.1093/mnras/stad2562 , https://ui.adsabs.harvard.edu/abs/2023MNRAS.525.4079Z 525, 4079

    2023 doi

  18. [30]

    J., Huchra J

    de Lapparent V., Geller M. J., Huchra J. P., 1986, @doi [ ] 10.1086/184625 , https://ui.adsabs.harvard.edu/abs/1986ApJ...302L...1D 302, L1

    1986 doi

  19. [31]

    308, @doi 10.1017/S174392131601098X

    van de Weygaert R., Shandarin S., Saar E., Einasto J., eds, 2016, The Zeldovich Universe: Genesis and Growth of the Cosmic Web IAU Symposium Vol. 308, @doi 10.1017/S174392131601098X

    2016 doi

Pith tools

Reviewed August 11, 2026 · model on record in the stance chip above.