Pith. sign in

REVIEW 3 major objections 6 minor 19 references

The Impact of metallicity evolution of the universe on the maximum mass of LIGO binary black holes

T0 review · 3 major / 6 minor · reviewed 2026-08-14 · deepseek-v4-flash

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

Pith's one-line read The inferred maximum black hole mass depends on how the universe's gas became metal-rich: rapid enrichment gives $44^{+9}_{-5}\,M_\odot$; modest enrichment gives $52^{+16}_{-9}\,M_\odot$.

desk verdict A useful caveat about LIGO mass-cutoff inference, but the exact 44 vs 52 M_sun numbers depend on an untested single-metallicity assumption. read the letter →

arxiv 1909.01356 v1 pith:H7SDOVWI submitted 2019-09-03 astro-ph.HE astro-ph.GAgr-qc

Mohammadtaher Safarzadeh , Will M. Farr This is my paper · ORCID
classification astro-ph.HEastro-ph.GAgr-qc
keywords gravitationalwavesbinaryblackholesmaximumholemassmetallicityevolutionpair-instabilitysupernovaeLIGOpopulationinference
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

The paper argues that any claim about an upper mass cutoff for stellar-mass black holes, inferred from gravitational-wave detections, is entangled with assumptions about how the metallicity of star-forming gas evolved over cosmic time. Using the ten binary black hole mergers from LIGO's first two observing runs, it fits a model in which the maximum black hole mass at zero metallicity, $M_{\rm BH}^{\rm max}$, is a parameter and the effective cutoff seen by LIGO is set by the redshift evolution of the star-formation-weighted metallicity. Under rapid metal enrichment, $Z/Z_\odot=e^{-\gamma z}$, the inferred maximum is $M_{\rm BH}^{\rm max}=44^{+9}_{-5}\,M_\odot$; under modest enrichment, $Z/Z_\odot=(1+z)^{-\gamma}$, it rises to $52^{+16}_{-9}\,M_\odot$. The result matters because it turns a would-be direct measurement of the black hole mass scale into a measurement that depends on the still-poorly-constrained chemical history of the universe.

What carries the argument

The central object is the metallicity-gated maximum mass relation $m_1^{\rm max}(Z) = (M_{\rm BH}^{\rm max} - 17.5)\, e^{-6.5\,Z(z,\gamma)} + 17.5$, which converts a single parameter, $M_{\rm BH}^{\rm max}$ (the maximum black hole mass at zero metallicity), into a redshift-dependent cutoff once a metallicity evolution model $Z(z,\gamma)$ is chosen. Two evolution prescriptions carry the argument: $Z/Z_\odot = e^{-\gamma z}$ for rapid enrichment and $Z/Z_\odot = (1+z)^{-\gamma}$ for modest enrichment. The machinery works by making the effective observed cutoff track how much cosmic star formation happened at very low metallicity, which is why the posterior on $M_{\rm BH}^{\rm max}$ anti-correlates with $\gamma$ and why the mass cutoff, not the merger rate, is the sensitive probe.

What would settle it

Measure the actual star-formation-weighted metallicity distribution from low to high redshift, including faint low-mass galaxies: if a non-negligible fraction of star formation occurs at $Z/Z_\odot\lesssim10^{-3}$ at redshifts $z\lesssim1$, the modest-evolution model would also form near-maximum black holes recently, and the inferred cutoff would no longer shift from 44 to 52 solar masses with the assumed enrichment history. Alternatively, detecting a gravitational-wave event with a primary mass well above the rapid-model cutoff, at a redshift that rules out a high-metallicity origin, would break the model's mapping.

Watch

Extended reading notes

Core claim

The central quantitative discovery is the history-dependence of the inferred cutoff. The paper ties the maximum allowed primary mass to the metallicity of the star-forming gas through $m_1^{\rm max}= (M_{\rm BH}^{\rm max} - 17.5)\,e^{-6.5\,Z(z,\gamma)} + 17.5$, calibrated to the maximum black hole mass versus metallicity curve of stellar-collapse models, and convolves the resulting formation rate with a power-law delay-time distribution and the LIGO detection probability. Fitting all six population parameters to the ten O1/O2 events, it finds that $M_{\rm BH}^{\rm max}$ is the best-constrained parameter but its value moves with the assumed metallicity history: $44^{+9}_{-5}\,M_\odot$ for rapid, exponential enrichment and $52^{+16}_{-9}\,M_\odot$ for modest, power-law enrichment. The shift is driven by the need to explain the most massive events, GW170729 and GW170823, without stars forming at the very lowest metallicities. The stated conclusion is that inferring $M_{\rm BH}^{\rm max}$ from gravitational-wave data depends on a metal-enrichment history that is not yet severely constrained.

Load-bearing premise

The argument rests on assuming that all star-forming gas at a given redshift shares one star-formation-weighted metallicity, with no scatter; if low-metallicity dwarf galaxies contribute non-negligibly to star formation at late times, the difference between the 44 and 52 solar mass cutoffs shrinks or disappears.

Editorial extensions

If this is right

  • A reported upper mass cutoff in the LIGO black hole population is model-dependent until the metal-enrichment history is pinned down: the same ten events support either a $44^{+9}_{-5}\,M_\odot$ or a $52^{+16}_{-9}\,M_\odot$ cutoff.
  • The two metallicity histories leave the other population parameters (mass-slope indices, delay-time index, and birth efficiency) nearly unchanged, so the mass cutoff is the cleanest observable discriminator between enrichment histories.
  • If star-forming gas spent little time near zero metallicity, the absence of black holes above roughly $45\,M_\odot$ in the first ten events is an expected selection bias rather than direct proof of a pair-instability gap.
  • The inferred median birth efficiency of roughly $2\times10^{-7}\,M_\odot^{-1}$ and the preference for long delay times do not change with the choice of metallicity-evolution prescription.

Reading between the lines

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

  • If low-metallicity dwarf galaxies contribute a non-negligible fraction of star formation at low redshift, the real metal-enrichment history has scatter around the mean track; that scatter would compress the 44 versus 52 solar mass divide and make the inferred cutoff less history-dependent than the two clean models suggest.
  • A direct test is to repeat the same fit on later gravitational-wave catalogs: a primary black hole firmly above the pair-instability gap (roughly $>65\,M_\odot$ after measurement uncertainty) would break the metallicity-gated cutoff unless a non-stellar formation channel is allowed.
  • Inverted, the model becomes a cosmological probe: a precisely measured mass cutoff, combined with an independent delay-time distribution, would constrain the amount of star formation that ever occurred at metallicities below roughly $Z/Z_\odot\sim10^{-3}$.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

3 major / 6 minor

Summary. This paper investigates whether the inferred maximum black hole mass M_BH^max from the first ten LIGO/Virgo binary black hole events depends on the assumed cosmic metallicity evolution. The authors construct a hierarchical Bayesian model in which the BBH mass distribution is a power law with a metallicity-dependent upper cutoff, using two parametrizations for the evolution of the SFR-weighted gas metallicity: a rapid exponential drop and a more gradual power-law drop. They report M_BH^max = 44^{+9}_{-5} M_sun for the rapid model and 52^{+16}_{-9} M_sun for the modest model, and conclude that LIGO-based bounds on M_BH^max are contingent on assumptions about metal enrichment history.

Significance. If correct, the paper makes a useful and timely point: upper-mass-cutoff estimates for stellar-mass black holes from gravitational-wave data are not prior-independent but depend on the astrophysical assumption of how metallicity evolves with redshift. The hierarchical Poisson likelihood structure is a standard and sensible framework, the model is transparently parametrized, and the reported credible intervals make the statistical uncertainties clear. The qualitative anti-correlation between the metallicity-evolution slope and the inferred M_BH^max is physically plausible and is the paper's main contribution. However, the quantitative central claim rests on the assumption of a single SFR-weighted metallicity at each redshift, and the likelihood implementation as written omits some needed details about the sampling prior; these issues need to be addressed before the quantitative bounds can be taken at face value.

major comments (3)
  1. [Section 2.1, Eq. (8), Fig. 1] The model assumes that all star-forming gas at a given redshift has the same SFR-weighted metallicity Z(z,gamma). Because the mapping from Z to the maximum BH mass is exponential and convex, averaging over a realistic distribution of galaxy metallicities would give a systematically larger contribution from the low-metallicity tail than the single-mean calculation. Low-metallicity dwarf galaxies are known to form stars at low redshift, so in the modest-evolution model a non-negligible tail of low-Z star formation could produce ~50 M_sun BHs at low z without requiring M_BH^max to rise from 44 to 52 M_sun. The paper does not test this: Section 4 lists delay-time and formation-channel caveats but does not discuss metallicity dispersion. Since the headline 44 vs 52 M_sun difference is the central quantitative claim, this simplification is load-bearing. Please either include a simple model for metallicity scatter (e.g., a log-normal distribution around Z(z,gamma)) and show the effect on the inferred M_BH^max, or argue explicitly why the SFR-weighted mean is sufficient for this inference.
  2. [Section 2.2, Eq. (14)] The Monte Carlo estimate in Eq. (14) averages the unnormalized rate density dNmerge/dm1dm2dtdz over posterior samples from P(m1,m2,z|d_i) without showing the importance-weighting by the prior used in the LIGO parameter estimation. The correct estimator for the per-event marginal likelihood is proportional to (1/N) sum_j R(m_j,z_j|theta) / pi(m_j,z_j), where pi is the prior used to generate the samples, up to a per-event constant. If the LIGO posterior samples used a non-flat prior (for example, a prior uniform in comoving volume), Eq. (14) is biased and the resulting posterior on theta, including M_BH^max, will be biased. The paper should state the exact prior used for the individual-event posterior samples and either include the appropriate reweighting or demonstrate that the prior is flat over the relevant ranges.
  3. [Section 2.1, Eqs. (1)-(2), Eq. (16)] The normalization constant C(alpha,beta) in Eq. (2) is written as an integral over m1 and m2 without stating the integration limits, while the allowed range m1 < mmax1(Z(z,gamma)) is redshift-dependent. If C is computed with fixed limits up to M_BH^max, then the total formation efficiency at a given z is implicitly proportional to the fraction of the power law below mmax1, which is a non-trivial z dependence that should be stated explicitly. If instead C is recomputed at each z, it should be written as C(alpha,beta,z; M_BH^max,gamma). This ambiguity affects the rate density used in Eq. (14) and the expected number N_eff in Eq. (16), and therefore directly affects the inferred posterior. Please clarify the definition and, if the former convention is used, justify it physically.
minor comments (6)
  1. [Abstract and throughout] The phrase 'We can be biased against observing massive black holes' is awkward; suggest rewording, e.g., 'Observational bounds on massive black holes can be biased...'.
  2. [Figure 2 caption and Figure 3 caption] The signs in the metallicity-evolution formulas are inconsistent with the main text: Figure 2's caption gives Z/Z_sun = e^{gamma z} while the text uses e^{-gamma z}, and Figure 3's caption gives Z/Z_sun = (1+z)^{gamma} while the text uses (1+z)^{-gamma}. Please correct the captions.
  3. [Section 2.1, after Eq. (7)] The sentence defining Mmin < m2 < m1 < mmax1 should specify that C(alpha,beta) and C(kappa) are normalization constants over the ranges implied by these bounds; otherwise the reader cannot tell whether the mass normalization is redshift-dependent.
  4. [Section 3, second paragraph] The statement that M_BH^max 'is very well constrained' is a bit strong given the broad upper tails in Figures 2 and 3; consider 'relatively well constrained compared to the other parameters'.
  5. [References] The reference to Abbott et al. is incomplete (missing title and journal). Several other entries use 'et al.' inconsistently (e.g., Kovetz et al. 2017, Neijssel et al. 2019). Please format consistently.
  6. [Figure 1] In the top panel, the green and red curves labeled M_BH^max=80 and 40 M_sun should be described in the caption as parametrizations of the Belczynski et al. (2010) maximum-mass-vs-metallicity relation, not as data; the current caption is terse.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: M_BH^max is a fitted parameter, and the metallicity-dependent mapping is imported from external stellar models, not derived from the target inference.

full rationale

I traced the derivation from Eq. (1) through Eq. (16). The paper's central quantity M_BH^max is explicitly one of six fitted parameters in theta=(λ_BBH, α, β, γ, κ, M_BH^max) in Eq. (10), constrained by the hierarchical likelihood over the ten LIGO events in Eqs. (13)–(15). It is not defined in terms of the posterior bound it produces; the quoted 44^{+9}_{-5} and 52^{+16}_{-9} M_sun values are posterior constraints obtained under two adopted metallicity-evolution parametrizations, Z/Z_sun = e^{-γz} and Z/Z_sun = (1+z)^{-γ}. The metallicity dependence of the maximum mass in Eq. (8) is an externally motivated phenomenological curve matched to Belczynski et al. (2010), with fixed constants b=6.5 and c=17.5; it is not derived from the LIGO data and does not encode the result. The observed anti-correlation between γ and M_BH^max is a consequence of the likelihood structure, not an algebraic tautology. There are no load-bearing self-citations: the Belczynski et al. references are external stellar-remnant mass models, and no uniqueness theorem or prior result by the present authors is invoked to force the choice. The paper's stated caveats about a single SFR-weighted metallicity at each redshift, constant λ_BBH, and a single formation channel are robustness limitations rather than circular steps. In particular, the absence of metallicity scatter is a physical-modeling concern that could change the numerical difference between the two models, but it does not make the inference equivalent to its inputs. The central claim—that the inferred maximum BH mass depends on the assumed metal-enrichment history—is a conditional sensitivity statement supported by the two fits, and no step reduces by construction to its own input.

Assumptions & free parameters 8 free parameters · 7 assumptions · 0 invented entities

Everything the central claim rests on: six fitted population parameters, fixed shape constants, and strong assumptions about single-channel formation, a delta-function metallicity at each redshift, the Belczynski et al. (2010) mass-metallicity shape, and a simple two-form metallicity evolution. No new physical entities are introduced.

free parameters (8)
  • M_BH^max = 44.67 (+9.43, -5.35) M_sun for exponential model; 52.56 (+16.84, -9.23) M_sun for power-law model
    Central parameter: the maximum black hole mass at zero metallicity; inferred from the ten LIGO events, weakly constrained in the modest-evolution model.
  • gamma = 2.29 (+1.01, -0.84) for exponential model; 2.69 (+0.85, -0.87) for power-law model
    Metallicity-redshift evolution exponent; fit from data and anti-correlated with M_BH^max.
  • alpha = 1.82 (+0.82, -1.06) for exponential model; 1.85 (+0.76, -1.06) for power-law model
    Power-law slope of the primary black hole mass distribution.
  • beta = 1.16 (+0.85, -0.76) for exponential model; 1.08 (+0.76, -0.76) for power-law model
    Power-law slope of the secondary black hole mass distribution.
  • kappa = 0.50 (+0.46, -0.33) for exponential model; 0.49 (+0.53, -0.36) for power-law model
    Power-law index of the delay-time distribution; shallow slope favors long delay times.
  • lambda_BBH = log10 lambda = -6.69 (+0.33, -0.29) for exponential model; -6.69 (+0.35, -0.27) for power-law model
    Overall BBH formation efficiency; assumed constant with redshift.
  • b and c in Eq. (8) = b = 6.5, c = 17.5 M_sun
    Fixed shape constants chosen to match Belczynski et al. (2010) mass-metallicity relation; not fit to LIGO data but chosen by hand.
  • tmin and tmax for delay times = 1 Myr and 10 Gyr
    Integration bounds for the delay-time distribution in Eq. (5); chosen ad hoc.
assumptions (7)
  • domain assumption BBH formation rate factorizes into a non-evolving efficiency lambda_BBH, a power-law mass function, and the Madau-Dickinson star formation rate density psi(z).
    Eq. (1) in Section 2.1; the paper explicitly states lambda_BBH is assumed non-evolving with redshift, despite stellar wind mass loss being metallicity-dependent.
  • domain assumption All star formation at a given redshift occurs at a single SFR-weighted metallicity Z(z, gamma), with no scatter in metallicity at fixed redshift.
    Introduced in Section 2.1 before Eq. (8); this is the weakest load-bearing modeling choice.
  • domain assumption The maximum black hole mass follows the Belczynski et al. (2010) relation shape, extrapolated by Eq. (8) with M_BH^max as an overall scale, and pair-instability physics is represented only through the cutoff scale.
    Eq. (8) and Fig. 1; this ignores the broad pair-instability mass gap and metallicity-dependent wind prescriptions beyond the chosen curve.
  • ad hoc to paper Metallicity evolution is either Z/Z_sun = e^{-gamma z} or (1+z)^{-gamma}, a one-parameter monotonic functional form.
    Bottom panel of Fig. 1; no physical justification for the two functional forms beyond labeling them rapid versus modest.
  • domain assumption Detection probability P_det(m1, m2, z) is known from LIGO sensitivity and is included in Eq. (9), but its implementation is not described.
    Eq. (9); no formula or reference is given for P_det, limiting reproducibility.
  • domain assumption The ten observed BBH events form a complete, independent Poisson sample from a single formation channel.
    Eq. (15); events are assumed to share one channel with no contamination from dynamical assembly.
  • ad hoc to paper Prior bounds on population parameters are 1 < alpha, beta, gamma, kappa < 4 and 30 < M_BH^max/M_sun < 100.
    Section 2.2; prior ranges are chosen, not derived, and the upper bound at 100 M_sun can influence the high-mass tail.

how reviews work

0 comments
Cite this review

Pith. "Pith review of The Impact of metallicity evolution of the universe on the maximum mass of LIGO binary black holes." pith.science (2026). https://pith.science/paper/H7SDOVWI

@misc{pith2026190901356,
  author       = {Pith},
  title        = {Pith review of: The Impact of metallicity evolution of the universe on the maximum mass of LIGO binary black holes},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/H7SDOVWI}},
  note         = {Machine review of arXiv:1909.01356}
}
abstract

We can be biased against observing massive black holes to merge in the local universe as the bounds on the maximum black hole mass ($M_{BH}^{\rm max}$) depends on the assumptions regarding the metallicity evolution of the star forming gas across the cosmic time. We investigate the bounds on the metallicity evolution, mass distribution and delay times of the binary black hole sources based on the ten observed events by LIGO. We parametrize $M_{BH}^{\rm max}$ to be a function of metallicity which itself is modeled to evolve with redshift in either a modest or rapid fashion. Rapid metallicity evolution models predict a stringent bound of $M_{BH}^{\rm max}=44^{+9}_{-5}M_{\odot}$, while the bound on $M_{BH}^{\rm max}$ in the models with modest metallicity evolution is $M_{BH}^{\rm max}=52^{+16}_{-9} M_{\odot}$. Therefore, inferring $M_{BH}^{\rm max}$ from GW data depends on the assumed metal enrichment history of the universe that is not severely constrained at the moment.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

19 extracted references · 16 canonical work pages

  1. [1]

    P., Collaboration, t

    Abbott et al ., B. P., Collaboration, t. V., Abbott, B. P., et al. 2018, 1811.12940

    arXiv 2018

  2. [2]

    L., et al

    Belczynski, K., Bulik, T., Fryer, C. L., et al. 2010, The Astrophysical Journal, 714, 1217

    work page 2010

  3. [3]

    2016, Astronomy & Astrophysics, A97

    Belczynski, K., Heger, A., Gladysz, W., et al. 2016, Astronomy & Astrophysics, A97

    work page 2016

  4. [4]

    Berg, T. A. M., Ellison, S. L., S´ anchez-Ram´ ırez, R., et al. 2016, Monthly Notices of the Royal Astronomical Society, 3021

    work page 2016

  5. [5]

    E., Cantiello, M., et al

    Brott, I., de Mink, S. E., Cantiello, M., et al. 2011, Astronomy &

    work page 2011

  6. [6]

    P., Prochaska, J

    Cen, R., Ostriker, J. P., Prochaska, J. X., & Wolfe, A. M. 2002, The Astrophysical Journal, 598, 741

    work page 2002

  7. [7]

    Fishbach, M., & Holz, D. E. 2017, The Astrophysical Journal Letters, 851, L25

    work page 2017

  8. [8]

    L., Belczynski, K., Wiktorowicz, G., et al

    Fryer, C. L., Belczynski, K., Wiktorowicz, G., et al. 2012, The Astrophysical Journal, 749, 91 5 log(λ) = −6.69+0.33 −0.29 0.6 1.2 1.8 2.4α α = 1.82+0.82 −1.06 0.6 1.2 1.8 2.4β β = 1.16+0.85 −0.76 0.8 1.6 2.4 3.2γ γ = 2.29+1.01 −0.84 0.4 0.8 1.2 1.6 = 0.50+0.46 −0.33 7.2 6.6 6.0 5.4 log(λ) 45 60 75 90 M max BH 0.61.21.82.4α 0.61.21.82.4 β 0.8 1.6 2.4 3.2γ...

    work page 2012

Show all 19 references
  1. [9]

    L., Woosley, S

    Heger, A., Fryer, C. L., Woosley, S. E., Langer, N., & Hartmann, D. H. 2003, The Astrophysical Journal, 591, 288 Kovetz E. D., Cholis I., Breysse P. C., Kamionkowski M., 2017, PhRvD, 95, 103010

    2003

  2. [10]

    2000, Annual Review of Astronomy and Astrophysics, 38, 613

    Kudritzki, R.-P., & Puls, J. 2000, Annual Review of Astronomy and Astrophysics, 38, 613

    2000

  3. [11]

    P., & Fall, S

    Kulkarni, V. P., & Fall, S. M. 2002, The Astrophysical Journal, 580, 732 Leung S.-C., Nomoto K., Blinnikov S., 2019, arXiv, arXiv:1901.11136

    2002 arXiv

  4. [12]

    2014, Annual Review of Astronomy and Astrophysics, 52, 415 Marchant P., Renzo M., Farmer R., Pappas K

    Madau, P., & Dickinson, M. 2014, Annual Review of Astronomy and Astrophysics, 52, 415 Marchant P., Renzo M., Farmer R., Pappas K. M. W., Taam R. E., de Mink S., Kalogera V., 2018, arXiv, arXiv:1810.13412 Neijssel C. J., et al., 2019, arXiv, arXiv:1906.08136

    2014 arXiv

  5. [13]

    J., King, D

    Pettini, M., Smith, L. J., King, D. L., & Hunstead, R. W. 1997, The Astrophysical Journal, 486, 665

    1997

  6. [14]

    Djorgovski, S. G. 2003, The Astrophysical Journal, 595, L9

    2003

  7. [15]

    X., & Wolfe, A

    Prochaska, J. X., & Wolfe, A. M. 2000, The Astrophysical Journal, 533, L5

    2000

  8. [16]

    Roulet, J., & Zaldarriaga, M. 2019, Monthly Notices of the Royal Astronomical Society, 484, 4216 6 log(λ) = −6.69+0.35 −0.27 0.6 1.2 1.8 2.4α α = 1.85+0.76 −1.06 0.6 1.2 1.8 2.4β β = 1.08+0.76 −0.76 0.8 1.6 2.4 3.2γ γ = 2.69+0.85 −0.87 0.4 0.8 1.2 1.6 = 0.49+0.53 −0.36 7.2 6.6...

    2019

  9. [17]

    2018, The Astrophysical Journal, 856, 173

    Talbot, C., & Thrane, E. 2018, The Astrophysical Journal, 856, 173

    2018

  10. [18]

    S., de Koter, A., & Lamers, H

    Vink, J. S., de Koter, A., & Lamers, H. J. G. L. M. 2001, Astronomy & Astrophysics, 369, 574

    2001

  11. [19]

    Woosley, S. E. 2017, The Astrophysical Journal, 836, 244 Yoshida T., Umeda H., Maeda K., Ishii T., 2016, MNRAS, 457, 351 7 100 101 1 + z 10 20 30 40 50 60 70 80 90 100M max BH GW151012 GW151226 GW170104 GW170608 GW170729 GW170809 GW170814 GW170818 GW170823 GW150914 100 101 1 +...

    2017

Pith tools

Reviewed August 14, 2026 · model on record in the stance chip above.