pith. sign in

arxiv: 2605.27509 · v1 · pith:ZWVA7ZRJnew · submitted 2026-05-26 · 🌌 astro-ph.GA

Reconstructing the Globular Cluster Initial Mass Function from Present-Day Globular Cluster Systems

Elizabeth Moreno-Hilario , Hui Li , Luis A. Martinez-Medina This is my paper

Pith reviewed 2026-06-29 16:46 UTC · model grok-4.3

classification 🌌 astro-ph.GA
keywords globular clustersinitial mass functiondynamical evolutionmass losshalo massN-body simulationsgalaxy formation
0
0 comments X

The pith

Reversing dynamical mass loss shows globular cluster initial mass functions follow power laws whose slopes steepen with host halo mass.

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

The paper develops a technique to recover the original masses of globular clusters by inverting the mass they have lost through dynamical interactions over time. The technique relies on an environment-specific mass-loss prescription calibrated directly from N-body simulations rather than on any assumed shape for the initial distribution. When applied to observed cluster systems in galaxies with halo masses from 10^9 to 10^12 solar masses, the reconstructed initial distributions are shifted upward in mass and display power-law tails at the high-mass end. The slopes of those tails become systematically steeper in more massive host halos. This result matters because it indicates that the well-known turnover in present-day cluster mass functions is produced by later evolution acting on an initial distribution that itself varies with galactic environment.

Core claim

The recovered GCIMFs are systematically shifted towards higher masses and exhibit a power-law behavior at the high-mass end. The inferred slopes vary across galaxies and show a strong correlation with host halo mass, with more massive galaxies exhibiting steeper high-mass slopes. Our results suggest that the slope of the GCIMF depends on the galactic properties and provides a direct empirical link between present-day globular cluster systems and their high-redshift progenitors.

What carries the argument

An environment-dependent mass-loss model calibrated by direct N-body simulations in time-dependent tidal fields, which maps observed present-day cluster masses back to progenitor masses without assuming any functional form for the GCIMF.

If this is right

  • The high-mass end of the GCIMF is a power law whose slope correlates with host halo mass.
  • More massive galaxies host GCIMFs with steeper high-mass power-law slopes.
  • The near-universal turnover mass observed today in globular cluster systems arises from dynamical evolution acting on these initial power-law distributions.
  • The method supplies an empirical bridge between present-day cluster populations and the conditions under which clusters formed at high redshift.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • If the slope-halo mass relation holds, hydrodynamic simulations of galaxy formation at early times should produce cluster populations whose initial mass functions also vary systematically with halo mass.
  • The reconstruction technique can be applied to statistical samples of galaxies where only global cluster properties are measured, expanding the set of systems for which progenitor masses can be estimated.
  • The dependence on galactic properties implies that the physics setting the upper end of the cluster mass spectrum is not universal but responds to the depth of the galactic potential or the ambient density at formation.

Load-bearing premise

The mass-loss rates given by the N-body calibrated model in varying tidal fields correctly describe the actual dynamical evolution that globular clusters experience inside real galaxies.

What would settle it

Repeating the reconstruction on an independent sample of galaxies with directly measured cluster masses and finding no correlation between the recovered high-mass slope and halo mass would falsify the central claim.

Figures

Figures reproduced from arXiv: 2605.27509 by Elizabeth Moreno-Hilario, Hui Li, Luis A. Martinez-Medina.

Figure 1
Figure 1. Figure 1: Present-day GCMFs of galaxies with in￾dividual GC mass measurements. The gray shaded region marks the expected turnover mass range 1 × 105 M⊙ ≤ MTO ≤ 2.6 × 105 M⊙ for all the galaxies studied in A. Jord´an et al. (2007). Despite the wide range of host galaxy masses, the distributions exhibit the characteristic near-universal log-normal form. tion 4 we present representative reconstructed GCIMFs and charact… view at source ↗
Figure 3
Figure 3. Figure 3: Calibration of the average fractional mass-loss rate as a function of initial cluster mass, M0, for the host dwarf galaxies from Paper I. Symbols indicate results from direct N-body simulations across multiple sets, including set 1 (circles), set 2 (diamonds), and set 3 (stars) from Pa￾per I, as well as the latest higher-resolution set 4 (pentagons) with N0 = 2 × 105 particles. Solid lines are power-law fi… view at source ↗
Figure 4
Figure 4. Figure 4: Reconstructed GCIMF (12 Gyr ago) for the Milky Way (yellow histogram) compared to the present-day GCMF (teal histogram) of 167 GCs. The dashed line shows the best-fit high-mass power-law slope, yielding β = 1.89 ± 0.18 in dN/d log M. 4.1. High-Mass regime slopes of the reconstructed GCIMF Since the main goal of this work is to characterize the GCIMF (or at least its high-mass end) for a large number of gal… view at source ↗
Figure 5
Figure 5. Figure 5: High-mass slope β of the reconstructed GCIMF as a function of halo mass for the galaxy sam￾ple analyzed in this work. The dashed line shows the best-fitting linear relation. The best fit parameters are β = (0.699±0.038) log(Mh)−(6.698±0.428), with a scatter of σ = 0.368. function histograms that, while technically fit-able, con￾tained very few clusters per bin (e.g., three bins with a single GC each), maki… view at source ↗
read the original abstract

The near-universal turnover mass of the present-day globular cluster mass function (GCMF), $M_{\rm TO} \sim 2 \times 10^5\ {\rm M_\odot}$, is a well established observational feature across galaxies of different types and masses, providing an important empirical benchmark for understanding the globular cluster initial mass function (GCIMF). Competing explanations of this property invoke either dynamical evolution from an initial power-law distribution or an imprint of cluster formation physics. We address this problem by reconstructing the high-mass regime of the GCIMF by inverting the mass loss due to dynamical evolution of present-day globular cluster systems across a wide range of host galaxy masses. Our method is based on an environment-dependent mass-loss model calibrated by direct $N$-body simulations in time-dependent tidal fields, enabling a mapping between observed cluster masses and their progenitor values without assuming a priori a functional form for the GCIMF. We apply our method to galaxies spanning halo masses of $\sim10^{9}$ - $10^{12}\ {\rm M_\odot}$, combining systems with individually measured globular cluster masses as well as large statistical samples constructed from observed global properties. The recovered GCIMFs are systematically shifted towards higher masses and exhibit a power-law behavior at the high-mass end. The inferred slopes vary across galaxies and show a strong correlation with host halo mass, with more massive galaxies exhibiting steeper high-mass slopes. Our results suggest that the slope of the GCIMF depends on the galactic properties and provides a direct empirical link between present-day globular cluster systems and their high-redshift progenitors.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

1 major / 1 minor

Summary. The paper claims to reconstruct the high-mass regime of the globular cluster initial mass function (GCIMF) by inverting dynamical mass loss from observed present-day globular cluster systems. The inversion employs an environment-dependent mass-loss model calibrated directly on N-body simulations in time-dependent tidal fields, without assuming a functional form for the GCIMF a priori. Applied to galaxies spanning halo masses ~10^9–10^12 M_⊙ (using both individually measured clusters and statistical samples from global properties), the recovered GCIMFs are shifted to higher masses, exhibit power-law behavior at the high-mass end, and display slopes that correlate strongly with host halo mass (steeper in more massive galaxies). The results are presented as providing an empirical link between present-day systems and high-redshift progenitors.

Significance. If the mapping is robust, the work offers a direct empirical constraint on the GCIMF across a wide galaxy mass range and suggests environment-dependent formation physics. A notable strength is the avoidance of an assumed GCIMF functional form, relying instead on N-body calibrated mass-loss rates. The combination of resolved and statistical samples broadens applicability. However, the central correlation result requires explicit validation against model-induced artifacts to be considered definitive.

major comments (1)
  1. [Abstract/method description] Abstract/method description: The environment-dependent mass-loss model (calibrated on N-body simulations in time-dependent tidal fields) is applied on a per-galaxy basis. Tidal strength scales with host galaxy properties that correlate with halo mass over 10^9–10^12 M_⊙. This differential mapping can induce steeper recovered high-mass power-law slopes in more massive halos even if the true GCIMF is identical across systems. No test isolating the model’s environment dependence from the data is described, which is load-bearing for the reported slope-halo mass correlation.
minor comments (1)
  1. [Abstract] Abstract: No error bars, validation tests against known clusters, or data exclusion criteria are provided for the mass-loss inversion, limiting assessment of robustness.

Simulated Author's Rebuttal

1 responses · 0 unresolved

We thank the referee for their thorough review and valuable comments on our manuscript. We address the major comment below.

read point-by-point responses

  1. Referee: The environment-dependent mass-loss model (calibrated on N-body simulations in time-dependent tidal fields) is applied on a per-galaxy basis. Tidal strength scales with host galaxy properties that correlate with halo mass over 10^9–10^12 M_⊙. This differential mapping can induce steeper recovered high-mass power-law slopes in more massive halos even if the true GCIMF is identical across systems. No test isolating the model’s environment dependence from the data is described, which is load-bearing for the reported slope-halo mass correlation.

    Authors: We agree this is an important point requiring explicit validation. The mass-loss model is physically calibrated from N-body simulations in time-dependent tides, and the observed present-day GCMFs are near-universal across the halo mass range studied. Applying stronger mass loss in more massive halos to match the same present-day masses therefore requires higher initial masses, with the mass-dependent mapping affecting the recovered high-mass slope. To isolate the model’s environment dependence, we will add a forward-modeling test in the revised manuscript: we assume a universal GCIMF, apply the environment-dependent mass-loss rates for different halo masses, and compare the predicted present-day distributions against the observed (near-universal) GCMFs. This will demonstrate that a single GCIMF cannot reproduce the data and that the recovered slope variation is required by the observations under the calibrated model. revision: yes

Circularity Check

0 steps flagged

No significant circularity; mass-loss model calibrated externally via N-body simulations

full rationale

The paper's core derivation inverts present-day GCMF to GCIMF using an environment-dependent mass-loss model calibrated exclusively by direct N-body simulations in time-dependent tidal fields. This calibration is independent of the target observational sample and does not assume a functional form for the GCIMF. The reported power-law slopes and their correlation with host halo mass emerge as outputs when the mapping is applied across galaxies spanning 10^9–10^12 M_⊙. No equations reduce by construction to fitted inputs, no self-citations are load-bearing for the uniqueness of the method, and no ansatz is smuggled via prior work. The derivation remains self-contained against the external simulation benchmark.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 0 invented entities

Abstract-only review yields no explicit free parameters, axioms, or invented entities; the mass-loss model is stated to be calibrated externally by N-body simulations.

pith-pipeline@v0.9.1-grok · 5825 in / 1103 out tokens · 36803 ms · 2026-06-29T16:46:44.247899+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

43 extracted references · 42 canonical work pages · 1 internal anchor

  1. [1]

    D., Vanzella, E., et al

    Adamo, A., Bradley, L. D., Vanzella, E., et al. 2024, Nature, 632, 513, doi: 10.1038/s41586-024-07703-7

    work page doi:10.1038/s41586-024-07703-7 2024

  2. [2]

    P., Mac Low, M.-M., Agertz, O., Renaud, F., & Li, H

    Andersson, E. P., Mac Low, M.-M., Agertz, O., Renaud, F., & Li, H. 2024, A&A, 681, A28, doi: 10.1051/0004-6361/202347792

    work page doi:10.1051/0004-6361/202347792 2024

  3. [3]

    2018, MNRAS, 478, 1520, doi: 10.1093/mnras/sty1057

    Baumgardt, H., & Hilker, M. 2018, MNRAS, 478, 1520, doi: 10.1093/mnras/sty1057

    work page doi:10.1093/mnras/sty1057 2018

  4. [4]

    2023, The Galactic Globular Cluster Database,, https: //people.smp.uq.edu.au/HolgerBaumgardt/globular/

    Baumgardt, H., Sollima, A., Hilker, M., & et al. 2023, The Galactic Globular Cluster Database,, https: //people.smp.uq.edu.au/HolgerBaumgardt/globular/

    2023

  5. [5]

    H., Hearin, A

    Behroozi, P., Wechsler, R. H., Hearin, A. P., & Conroy, C. 2019, MNRAS, 488, 3143, doi: 10.1093/mnras/stz1182

    work page doi:10.1093/mnras/stz1182 2019

  6. [6]

    F., McIntosh, D

    Bell, E. F., McIntosh, D. H., Katz, N., & Weinberg, M. D. 2003, ApJS, 149, 289, doi: 10.1086/378847

    work page internal anchor Pith review doi:10.1086/378847 2003

  7. [7]

    P., & Strader, J

    Brodie, J. P., & Strader, J. 2006, ARA&A, 44, 193, doi: 10.1146/annurev.astro.44.051905.092441

    work page doi:10.1146/annurev.astro.44.051905.092441 2006

  8. [8]

    Caldwell, N., & Romanowsky, A. J. 2016, ApJ, 824, 42, doi: 10.3847/0004-637X/824/1/42

    work page doi:10.3847/0004-637x/824/1/42 2016

  9. [9]

    Chen, Y., & Gnedin, O. Y. 2023, MNRAS, 522, 5638, doi: 10.1093/mnras/stad1328

    work page doi:10.1093/mnras/stad1328 2023

  10. [10]

    2026, arXiv e-prints, arXiv:2601.16281, doi: 10.48550/arXiv.2601.16281 de Boer, T

    Claeyssens, A., Adamo, A., Kokorev, V., et al. 2026, arXiv e-prints, arXiv:2601.16281, doi: 10.48550/arXiv.2601.16281 de Boer, T. J. L., & Fraser, M. 2016, A&A, 590, A35, doi: 10.1051/0004-6361/201527580

    work page doi:10.48550/arxiv.2601.16281 2026

  11. [11]

    Dornan, V., & Harris, W. E. 2026, ApJ, 998, 41, doi: 10.3847/1538-4357/ae2742

    work page doi:10.3847/1538-4357/ae2742 2026

  12. [12]

    M., & Zhang, Q

    Fall, S. M., & Zhang, Q. 2001, ApJ, 561, 751, doi: 10.1086/323358

    work page doi:10.1086/323358 2001

  13. [13]

    A., Weinberg, M

    Fardal, M. A., Weinberg, M. D., Babul, A., et al. 2013, MNRAS, 434, 2779, doi: 10.1093/mnras/stt1121

    work page doi:10.1093/mnras/stt1121 2013

  14. [14]

    A., Read, J

    Forbes, D. A., Read, J. I., Gieles, M., & Collins, M. L. M. 2018, MNRAS, 481, 5592, doi: 10.1093/mnras/sty2584

    work page doi:10.1093/mnras/sty2584 2018

  15. [15]

    Garcia, F. A. B., Ricotti, M., Sugimura, K., & Park, J. 2023, MNRAS, 522, 2495, doi: 10.1093/mnras/stad1092

    work page doi:10.1093/mnras/stad1092 2023

  16. [16]

    Gieles, M., & Gnedin, O. Y. 2023, MNRAS, 522, 5340, doi: 10.1093/mnras/stad1287

    work page doi:10.1093/mnras/stad1287 2023

  17. [17]

    Y., & Ostriker, J

    Gnedin, O. Y., & Ostriker, J. P. 1997, ApJ, 474, 223, doi: 10.1086/303441

    work page doi:10.1086/303441 1997

  18. [18]

    E., Harris, G

    Harris, W. E., Harris, G. L., & Hudson, M. J. 2015, ApJ, 806, 36, doi: 10.1088/0004-637X/806/1/36

    work page doi:10.1088/0004-637x/806/1/36 2015

  19. [19]

    E., Harris, G

    Harris, W. E., Harris, G. L. H., & Alessi, M. 2013, ApJ, 772, 82, doi: 10.1088/0004-637X/772/2/82

    work page doi:10.1088/0004-637x/772/2/82 2013

  20. [20]

    Hunter, J. D. 2007, Computing in Science and Engineering, 9, 90, doi: 10.1109/MCSE.2007.55 Jord´ an, A., McLaughlin, D. E., Cˆ ot´ e, P., et al. 2007, ApJS, 171, 101, doi: 10.1086/516840 Jord´ an, A., Peng, E. W., Blakeslee, J. P., et al. 2009, ApJS, 180, 54, doi: 10.1088/0067-0049/180/1/54

    work page doi:10.1109/mcse.2007.55 2007

  21. [21]

    Karim, N., Collins, M. L. M., Forbes, D. A., & Read, J. I. 2024, MNRAS, 530, 4936, doi: 10.1093/mnras/stae611

    work page doi:10.1093/mnras/stae611 2024

  22. [22]

    Kimm, T., Cen, R., Rosdahl, J., & Yi, S. K. 2016, ApJ, 823, 52, doi: 10.3847/0004-637X/823/1/52

    work page doi:10.3847/0004-637x/823/1/52 2016

  23. [23]

    Kruijssen, J. M. D. 2015, MNRAS, 454, 1658, doi: 10.1093/mnras/stv2026

    work page doi:10.1093/mnras/stv2026 2015

  24. [24]

    Li, H., & Gnedin, O. Y. 2019, MNRAS, 486, 4030, doi: 10.1093/mnras/stz1114

    work page doi:10.1093/mnras/stz1114 2019

  25. [25]

    McCrady, N., & Graham, J. R. 2007, ApJ, 663, 844, doi: 10.1086/518357

    work page doi:10.1086/518357 2007

  26. [26]

    McLaughlin, D. E. 1994, PASP, 106, 47, doi: 10.1086/133340

    work page doi:10.1086/133340 1994

  27. [27]

    E., & Pudritz, R

    McLaughlin, D. E., & Pudritz, R. E. 1996, ApJ, 457, 578, doi: 10.1086/176754

    work page doi:10.1086/176754 1996

  28. [28]

    A., Li, H., Souza, S

    Moreno-Hilario, E., Martinez-Medina, L. A., Li, H., Souza, S. O., & P´ erez-Villegas, A. 2024, MNRAS, 527, 2765, doi: 10.1093/mnras/stad3306

    work page doi:10.1093/mnras/stad3306 2024

  29. [29]

    G., Desprez, G., et al

    Mowla, L., Iyer, K. G., Desprez, G., et al. 2022, ApJL, 937, L35, doi: 10.3847/2041-8213/ac90ca

    work page doi:10.3847/2041-8213/ac90ca 2022

  30. [30]

    Pace, A. B. 2024, arXiv e-prints, arXiv:2411.07424, doi: 10.48550/arXiv.2411.07424

    work page doi:10.48550/arxiv.2411.07424 2024

  31. [31]

    2007, MNRAS, 378, 910, doi: 10.1111/j.1365-2966.2007.11817.x 2, 17, 21

    Parmentier, G., & Gilmore, G. 2007, MNRAS, 377, 352, doi: 10.1111/j.1365-2966.2007.11611.x

    work page doi:10.1111/j.1365-2966.2007.11611.x 2007

  32. [32]

    W., Jord´ an, A., Cˆ ot´ e, P., et al

    Peng, E. W., Jord´ an, A., Cˆ ot´ e, P., et al. 2008, ApJ, 681, 197, doi: 10.1086/587951

    work page doi:10.1086/587951 2008

  33. [33]

    Pfeffer, J., Kruijssen, J. M. D., Crain, R. A., & Bastian, N. 2018, MNRAS, 475, 4309, doi: 10.1093/mnras/stx3124

    work page doi:10.1093/mnras/stx3124 2018

  34. [34]

    2019, A&A, 621, A56, doi: 10.1051/0004-6361/201833355

    Posti, L., & Helmi, A. 2019, A&A, 621, A56, doi: 10.1051/0004-6361/201833355

    work page doi:10.1051/0004-6361/201833355 2019

  35. [35]

    2012, Ap&SS, 341, 195, doi: 10.1007/s10509-012-0986-9

    Rejkuba, M. 2012, Ap&SS, 341, 195, doi: 10.1007/s10509-012-0986-9

    work page doi:10.1007/s10509-012-0986-9 2012

  36. [36]

    Y., & Wurster, J

    Rieder, S., Dobbs, C., Bending, T., Liow, K. Y., & Wurster, J. 2022, MNRAS, 509, 6155, doi: 10.1093/mnras/stab3425

    work page doi:10.1093/mnras/stab3425 2022

  37. [37]

    L., Hafen, Z., Grudi´ c, M

    Rodriguez, C. L., Hafen, Z., Grudi´ c, M. Y., et al. 2023, MNRAS, 521, 124, doi: 10.1093/mnras/stad578 The pandas development Team. 2025, pandas-dev/pandas: Pandas, v2.3.1 Zenodo, doi: 10.5281/zenodo.3509134

    work page doi:10.1093/mnras/stad578 2023

  38. [38]

    Y., & Chies-Santos, A

    Tudorica, A., Georgiev, I. Y., & Chies-Santos, A. L. 2015, A&A, 581, A84, doi: 10.1051/0004-6361/201525615 van der Walt, S., Colbert, S. C., & Varoquaux, G. 2011, Computing in Science and Engineering, 13, 22, doi: 10.1109/MCSE.2011.37

    work page doi:10.1051/0004-6361/201525615 2015

  39. [39]

    2023, ApJ, 945, 53, doi: 10.3847/1538-4357/acb59a

    Vanzella, E., Claeyssens, A., Welch, B., et al. 2023, ApJ, 945, 53, doi: 10.3847/1538-4357/acb59a

    work page doi:10.3847/1538-4357/acb59a 2023

  40. [40]

    2026, A&A, 705, A171, doi: 10.1051/0004-6361/202556570

    Vanzella, E., Messa, M., Adamo, A., et al. 2026, A&A, 705, A171, doi: 10.1051/0004-6361/202556570

    work page doi:10.1051/0004-6361/202556570 2026

  41. [41]

    2001, MNRAS, 322, 247, doi: 10.1046/j.1365-8711.2001.04072.x 14E

    Vesperini, E. 2001, MNRAS, 322, 247, doi: 10.1046/j.1365-8711.2001.04072.x 14E. Moreno-Hilario et al

    work page doi:10.1046/j.1365-8711.2001.04072.x 2001

  42. [42]

    Jean Zinn-Justin and Ulrich D

    Virtanen, P., Gommers, R., Oliphant, T. E., et al. 2020, Nature Methods, 17, 261, doi: 10.1038/s41592-019-0686-2 Wes McKinney. 2010, in Proceedings of the 9th Python in Science Conference, ed. St´ efan van der Walt & Jarrod Millman, 56 – 61, doi: 10.25080/Majora-92bf1922-00a

    work page doi:10.1038/s41592-019-0686-2 2020

  43. [43]

    Zhang, Q., & Fall, S. M. 1999, ApJL, 527, L81, doi: 10.1086/312412

    work page doi:10.1086/312412 1999