Pith. sign in

REVIEW 2 major objections 6 minor 4 cited by

Spectral-siren gravitational-wave pipelines can process the Einstein Telescope's expected one-year event volume, and a blinded test shows three independent code implementations recover the same hidden cosmology.

Reviewed by Pith at T0; open to challenge. T0 means a machine referee read the full paper against a public rubric. the ladder, T0–T4 →

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

T0 review · deepseek-v4-flash

2026-08-02 22:07 UTC pith:THJYB6NN

load-bearing objection A serious blinded mock challenge: three independent spectral-siren pipelines agree at 3G scale, but the GPU-scalability claim is extrapolated for pymcpop-gw. the 2 major comments →

arxiv 2602.17756 v2 pith:THJYB6NN submitted 2026-02-19 astro-ph.CO gr-qc

Pushing spectral siren cosmology into the third-generation era: a blinded mock data challenge

Matteo Tagliazucchi , Michele Moresco , Alessandro Agapito , Michele Mancarella , Sarah Ferraiuolo , Simone Mastrogiovanni , Nicola Borghi , Francesco Pannarale
show 1 more author
Daniele Bonacorsi
This is my paper
classification astro-ph.CO gr-qc
keywords gravitational wavesspectral sirensEinstein Telescopecosmological parametershierarchical inferenceblack hole mass distributionmock data challengeGPU acceleration
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved

The pith

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

This paper tests whether current spectral-siren analysis pipelines — which measure cosmology from gravitational waves alone using features in the black-hole mass distribution — are ready for third-generation detectors like Einstein Telescope, whose event rates are three orders of magnitude larger than today's. The authors simulate one year of detector observations, blind the injected cosmology and population parameters, and run three independent numerical implementations of the hierarchical likelihood on catalogs of roughly 7,000 and 12,000 high-signal-to-noise mergers. In the blinded analysis all three codes recover the same cosmological and population parameters, all consistent with the hidden truth. The authors forecast a 2.4% measurement of the expansion rate H(z) at redshift ~1.5, and a mean precision of 2.8% on H(z) across 0.7

Core claim

The central claim is that spectral-siren cosmology is computationally and statistically feasible in the third-generation era. Using a blinded mock catalog built from a flat ΛCDM fiducial model with a power-law-plus-peak mass distribution and a Madau–Dickinson merger-rate evolution, the three pipelines — each computing the same hierarchical likelihood in a different way — produce fully consistent posteriors and recover nearly all fiducial parameters within 68% credible intervals. The key quantitative results are that a single GPU can process up to roughly 10^5 events with thousands of posterior samples each in one or two weeks, with likelihood evaluation time scaling linearly in event number,

What carries the argument

The central object is the hierarchical spectral-siren likelihood, which jointly constrains cosmology and population parameters from gravitational-wave data alone by reweighting per-event posterior samples with a population model that depends on redshift through the assumed cosmology. The three public pipelines implement this likelihood with distinct numerical strategies: a Monte Carlo sum over posterior samples (icarogw), a weighted kernel-density estimate of the per-event gravitational-wave kernel (chimera), and a latent-variable Hamiltonian Monte Carlo that avoids per-event Monte Carlo marginalization (pymcpop-gw). The machinery that carries the performance claim is GPU acceleration combin

Load-bearing premise

The mock population is assumed to have a source-frame black-hole mass distribution that does not evolve with redshift; if the real BBH mass spectrum changes with cosmic time, the clean separation between cosmological and population parameters that the spectral-siren method relies on breaks down, and the forecast precision and cross-code agreement could be optimistic.

What would settle it

Rerun the same blinded mock challenge with an injected mass distribution whose peak position or power-law slope evolves with redshift (for example, a 20% shift in peak position between z=0 and z=3). If the recovered H0 or Ωm,0 then moves by more than the quoted ~10% and ~26% uncertainties, or if the three pipelines diverge, the validation demonstrated here does not generalize to real third-generation data. A direct observational alternative would be an electromagnetic counterpart to an Einstein Telescope merger at z≈1.5, whose independent redshift would test the 2.4% H(z) prediction.

Watch this falsifier. Get emailed when new claim-graph text bears on it.

If this is right

  • With roughly 12,000 high-S/N mergers from one year of Einstein Telescope data, flat-ΛCDM spectral siren analysis measures the expansion rate H(z) at z≈1.5 to 2.4%, and to a mean 2.8% across 0.7<z<1.8.
  • Lowering the S/N threshold from 75 to 60 adds about 5,000 events and improves the H0–Ωm,0 figure of merit by 43%, consistent with the expected √N statistical gain.
  • Single-GPU profiling shows likelihood evaluation time scales linearly with the number of events, so the full one-year Einstein Telescope volume of about 10^5 events falls within one to two weeks of compute; the per-event factorization permits distributed multi-GPU processing for even larger catalogs.
  • Fixing Ωm,0 to its fiducial value sharpens the H0 constraint from 12% to 3% on the S/N>75 catalog, indicating that percent-level H0 from spectral sirens will likely require external information on Ωm,0 from probes such as baryon acoustic oscillations.
  • Events at low luminosity distance near mass-spectrum features drive constraints on H0, Ωm,0, and the peak position; events beyond about 6 Gpc mainly constrain Ωm,0 and the rate-evolution slope.
  • The three pipelines' agreement demonstrates that the spectral-siren hierarchical likelihood is not tied to one numerical implementation, since Monte Carlo, kernel-density, and latent-variable approaches reach the same inferred cosmology.

Where Pith is reading between the lines

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

  • The same blinded protocol could be extended to test redshift-evolving mass models; a natural stress test is injecting a population whose mass-spectrum peak shifts with redshift and asking whether the three pipelines' agreement and unbiased recovery survive.
  • Because the strongest constraint lands at 0.7<z<1.8, where cosmic-chronometer measurements of H(z) also operate, spectral sirens and chronometers could be cross-checked in the same redshift window; the paper notes the overlap but does not quantify a combined constraint.
  • The memory saturation of the kernel-density pipeline near 10^5 events suggests a hybrid analysis strategy: use that pipeline for catalogs up to the size tested, and the Monte Carlo or Hamiltonian pipeline for larger volumes, or split the event set across GPUs to beat the memory limit.
  • If the 2.4% H(z) forecast survives with real data, spectral sirens would become a competitive probe of the expansion history in a regime currently dominated by model-dependent distance measurements; reaching percent-level H0 will likely require extra mass scales, such as neutron-star mergers, or external Ωm,0 priors.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

2 major / 6 minor

Summary. The paper presents a blinded mock data challenge (MDC) for spectral-siren cosmology at third-generation (3G) detector volume. The authors simulate one year of Einstein Telescope observations, producing catalogs of ~68,000 and ~119,000 events above S/N thresholds of 75 and 60, respectively, from a population with Madau–Dickinson merger-rate evolution and a power-law-plus-peak mass distribution. Fiducial hyperparameters are drawn from a GWTC-3-inspired posterior and hidden externally. Three public pipelines with different numerical implementations of the hierarchical likelihood — ICAROGW, CHIMERA, and pymcpop-gw — are used to infer cosmological and population parameters. The paper reports that all pipelines recover the injected values within credible intervals, that GPU-accelerated likelihoods scale roughly linearly to ~10^5 events, and that the S/N>60 catalog yields ~10.5% precision on H0, ~26% on Omega_m,0, and ~2.4% precision on H(z) near z~1.5. It also analyzes which events drive cosmological constraints, finding that low-distance events near mass-population features dominate.

Significance. If the results hold, this is a valuable community resource: a blinded, externally validated test of three independent spectral-siren codes at 3G data volume, with a large injection campaign (5e7 injections), quantitative effective-sample-size checks, and consistent recovery of all injected parameters. The cross-code agreement is a genuine strength, as it tests numerical implementation rather than a single code's internal consistency. The computational benchmarks, while informative, are not uniform across the three codes: the GPU scalability claim is demonstrated for ICAROGW and CHIMERA but only extrapolated for pymcpop-gw. The cosmological forecasts are explicitly conditional on a non-evolving mass distribution, a limitation the paper acknowledges. Overall, the paper provides a useful stress test and a path toward 3G spectral-siren analyses, but its headline performance claim needs to be scoped more carefully.

major comments (2)
  1. [Sec. 4.1, Fig. 2, abstract] The abstract and conclusion claim that 'these pipelines can process the events expected from ET' thanks to GPU acceleration. For pymcpop-gw this statement is not supported by the evidence presented. The paper states that pymcpop-gw is 'unable to produce fully converged chains on GPU for the full data volumes considered in this work'; the converged posterior samples used in the blinded comparison were obtained via CPU runs; and the GPU timing numbers used for the 'linear fit' come from 'two short chains of 30 iterations.' The forecast for 8e4–1.3e5 events is therefore a linear extrapolation of two non-converged points, justified by the scaling of other codes rather than by converged pymcpop-gw runs. This is load-bearing for the central performance claim. I recommend revising the claim to state that GPU-scalability is demonstrated for ICAROGW and CHIMERA, while pymcpop-gw currently require
  2. [Sec. 4.2, Eq. (3)] The cosmological forecasts assume a source-frame mass distribution p(m1,m2|lambda_m) with no redshift evolution. The spectral-siren method relies on population features as rulers; if the real BBH mass spectrum evolves with redshift, as argued in several cited works (Mukherjee 2022; Karathanasis et al. 2023; Pierra et al. 2024; Agarwal et al. 2025), the clean separation between cosmology and population parameters is compromised. The paper flags this as 'a central physical problem beyond the scope of this work', which is honest, but the abstract's forecast precision (e.g., 2.4% on H(z)) is conditional on this static-population assumption. Since the manuscript is a forecast for ET, I recommend either adding a quantitative robustness test with a redshift-evolving mass model (even a simple parametrized shift) or stating more prominently in the abstract and conclusions that the quoted precisio
minor comments (6)
  1. [Sec. 4.1] The wording is contradictory: the text says 'We do not forecast timings for pymcpop-gw but instead plot the linear fit of the two measured points.' Plotting a linear fit of the two measured points is itself a forecast. Please rephrase to clarify what is measured and what is extrapolated.
  2. [Sec. 4.1] 'However, we warrant that more complex astrophysical models...' should read 'we warn' or 'we caution'.
  3. [Eq. (14)] The equation uses 'N_PEP' in the summation index and denominator; this appears to be a typo for 'N_PE'. Please unify notation.
  4. [Fig. 6 caption] In the caption, 'such curves are calculate d for' has a spacing typo ('calculate d').
  5. [Sec. 2.1, Eq. (6)] The derivation of the Monte Carlo sum would benefit from explicitly defining the proposal/prior pi(theta) used for the PE samples; currently the notation is introduced in text but the relation to the single-event prior is not stated in one place.
  6. [Table 1] The units for H0 are written as 'km s−1 Mpc−3'; this should be 'km s−1 Mpc−1'.

Circularity Check

0 steps flagged

No circular derivation: blinded forward simulation, independent cross-checks, and no parameter fitted to the target result.

full rationale

The paper's central claims are a blinded mock data challenge: it simulates ET catalogs from fiducial population/cosmology parameters that are withheld from the analysts, then runs three independently implemented hierarchical-likelihood codes on the simulated events. The 'predictions' (recovered H0, Omega_m, H(z) precision) are outputs of Bayesian inference on those forward-simulated data, not inputs to the generative model; the fiducial values are hidden by an external-safekeeping blinding procedure. No equation defines a target quantity in terms of itself: Eq. (1) is the standard hierarchical likelihood, Eq. (3) is the assumed population model, and the mock generation uses the same model family but this is the intended closed-box test, not a circle. The self-citations to ICAROGW, CHIMERA, and pymcpop-gw describe the codes being benchmarked; their validity rests on the shown agreement of three distinct numerical strategies with the injected truth, which is independent evidence. No authors' uniqueness theorem or ansatz is imported as a load-bearing premise. The main caveats — FIM-based PE, no redshift-evolving mass spectrum, and the pymcpop-gw GPU chains not being fully converged at the largest volumes — are explicitly acknowledged by the paper and are modeling/robustness limitations, not circular reductions of a result to its inputs. Therefore no step satisfying the quoted-reduction criterion is present.

Axiom & Free-Parameter Ledger

4 free parameters · 6 axioms · 0 invented entities

The central claims are empirical (numerical consistency, timing, forecast precision) rather than derivations. The model parameters listed are the simulation/inference targets whose features generate the spectral-siren rulers; the forecast precision is conditional on them. No new entities are introduced.

free parameters (4)
  • Fiducial hyperparameter vector (H0, Ωm,0, α, β, δ_m, m_low, m_high, μ_g, σ_g, λ_g, γ, κ, z_p, R_0) = H0=64.81 km/s/Mpc, Ωm=0.23, μ_g=36.44 M☉, etc. (Fig. 3 inset)
    Drawn from a GWTC-3-informed posterior (with narrowed cosmological priors) to set the mock population. All forecast precisions (H(z) 2.4%, H0 ~10%, Ωm ~26%) are conditional on this model and the mass/redshift features it produces.
  • S/N thresholds (60, 75) = 60, 75
    Hand-chosen to ease computing load; define the two analyzed catalogs (11,896 and 6,843 events) and therefore set the statistical precision of the forecasts.
  • Number of PE samples per event (N_PE=5000) = 5000
    Chosen to satisfy numerical stability criteria (Neff_PE>20 for ICAROGW, Neff_weights>2 for CHIMERA); timing results scale with this choice.
  • Injection scaling factor N_inj = N_obs × 390 = 390
    Chosen to match the S/N>60 catalog's injected-to-detected ratio; used to extrapolate timing to larger event counts.
axioms (6)
  • standard math The hierarchical spectral-siren likelihood of Eq. (1), including selection effects via N_exp, is the correct statistical model
    Foundational to the analysis; cited to Mandel et al. 2019 and Vitale et al. 2020.
  • domain assumption Single-event posterior distributions are well described by the Fisher-matrix Gaussian approximation computed with GWFAST
    All PE sample sets are drawn from FIM covariances; the paper states this is verified for high-S/N events, but non-Gaussianities in real PE would affect forecast precision and cross-code agreement.
  • domain assumption The BBH mass distribution is redshift-independent (power-law + peak, no evolution)
    Stated in Sec. 3 (population model); the paper itself lists redshift-evolving mass spectrum as a possible bias source and defers it to future work. Spectral siren cosmology is degenerate with population evolution.
  • domain assumption ET nominal sensitivity curve (two L-shaped 15-km interferometers, Hild et al. 2011) is the correct detector configuration
    Used for injection/detection; timing and forecast results depend on this sensitivity.
  • domain assumption Madau–Dickinson star-formation history describes the merger rate evolution ψ(z)
    Adopted for mock generation; the recovered z_p measurement is interpreted within this model.
  • domain assumption Blinding procedure was effective (fiducial values deleted from origin server)
    The blinded design is credible by construction, but the paper provides no independent audit trail that the values were not accessible to the analyzers.

pith-pipeline@v1.3.0-alltime-deepseek · 20854 in / 13645 out tokens · 126585 ms · 2026-08-02T22:07:37.232216+00:00 · methodology

0 comments
read the original abstract

Gravitational wave (GW) spectral sirens offer a promising method for measuring cosmological parameters using GW data only - without relying on external redshift information such as electromagnetic counterparts or galaxy catalogs - by exploiting distributional features in the population of GW sources. The advent of third-generation detectors like the Einstein Telescope (ET) will provide catalogs three orders of magnitudes larger than current ones, raising questions about the scalability and robustness of existing inference pipelines. We present a blinded mock data challenge that tests three public pipelines with distinct numerical implementations, namely, $\texttt{ICAROGW}$, $\texttt{CHIMERA}$, and $\texttt{pymcpop-gw}$, on simulated ET observations containing the best $\mathcal{O}(10^4)$ binary black hole mergers that can be observed in 1 year. We assess their computational performance, validate their agreement in a blinded setting, and forecast cosmological constraints. We find that, thanks to GPU acceleration, these pipelines can process the events expected from ET within a manageable timeframe. All pipelines recover consistent cosmological and population parameters. Assuming a flat $\Lambda$CDM model, we measure $H(z)$ at $z\sim1.5$ with 2.4% precision, and achieve a mean precision on $H(z)$ of 2.8% across $0.7<z<1.8$ with a catalog of $\sim 12,000$ high-S/N events. This corresponds to joint constraints of $\sim 10%$ on $H_0$ and $\sim 26%$ on $\Omega_{\rm m,0}$. We also identify the events that contribute mostly to constraining cosmological parameters, showing that low-distance sources near population features drive the constraining power on all cosmological parameters, while higher-distance events primarily constrain $\Omega_{\rm m,0}$. Our results establish a validated, performance-tested framework for spectral siren cosmology in the era of third-generation GW observatories.

Figures

Figures reproduced from arXiv: 2602.17756 by Alessandro Agapito, Daniele Bonacorsi, Francesco Pannarale, Matteo Tagliazucchi, Michele Mancarella, Michele Moresco, Nicola Borghi, Sarah Ferraiuolo, Simone Mastrogiovanni.

Figure 1
Figure 1. Figure 1: Properties of the blinded mock BBH catalogs. Left panel: Reverse cumulative distribution of the S [PITH_FULL_IMAGE:figures/full_fig_p004_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: Scaling of computational times for a single evaluation [PITH_FULL_IMAGE:figures/full_fig_p006_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: Cosmological and population parameter constraints from the S [PITH_FULL_IMAGE:figures/full_fig_p007_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: Comparison of constraints from the S/N > 60 and S/N > 75 catalogs. Left: 2D marginalized posterior for (H0, Ωm,0). In particular, we plot the 68% and 95% credible regions. Center and right: predictive posterior distributions for the primary mass spectrum and the redshift event rate Eq. (13), respectively. Dashed lines indicate the blinded fiducial values and population model. also note that because the cat… view at source ↗
Figure 5
Figure 5. Figure 5: Top: The 1-σ contours of the predictive posterior distri￾bution for H(z)/(1 + z). Bottom: The relative precision on H(z), computed as the width of the 1-σ C.I. divided by twice the me￾dian of H(z). Dashed vertical lines indicate the redshift at which the constraint on H(z) is strongest. (2024) found that with 3G detectors it will be possible to achieve percent-level constraints on the Hubble constant and Ω… view at source ↗
Figure 6
Figure 6. Figure 6: 2D histogram of the events with S/N > 75 in the dL-md,1 plane. Color indicates the mean Pearson correlation coefficient between the marginal likelihood of events in each bin and the hyperparameters H0 (top left), Ωm,0 (top right), µg (bottom left), and γ (bottom right). In each panel we plot the function md,1 · p(md,1) for different values of the hyperparameter considered and at various luminosity-distance… view at source ↗

discussion (0)

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

Forward citations

Cited by 4 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score.

  1. Assessing the Impact of Instrumental Requirements on the Scientific Performance of the Einstein Telescope

    astro-ph.IM 2026-07 accept novelty 6.0

    Degrading the Einstein Telescope's sensitivity in specific frequency bands hurts different science goals in predictable ways, but the mission remains scientifically strong even in the worst modelled cases.

  2. Radio sirens: inferring $H_0$ with binary black holes and neutral hydrogen in the era of the Einstein Telescope and the SKA Observatory

    astro-ph.CO 2026-05 unverdicted novelty 6.0

    Using simulated binary black hole mergers and neutral hydrogen maps, the radio sirens method constrains H0 to 8% precision with 3000 high-SNR events, offering a 90% improvement over standard dark siren analyses.

  3. Gravitational-wave constraints on $H_0$ are robust to (putative) redshift evolution in the binary black hole mass spectrum at current sensitivity

    astro-ph.CO 2026-05 conditional novelty 5.0

    Spectral-siren H0 constraints from GWTC-4.0 binary black holes remain robust when the mass spectrum is permitted to evolve with redshift at current detector sensitivity.

  4. Recovering cosmological parameters from the mock gravitational wave data of the Einstein Telescope

    astro-ph.CO 2026-04 unverdicted novelty 4.0

    Mock Einstein Telescope data recovers the Hubble constant to 1% or matter density to 4% in one year via the intrinsic chirp mass spectrum of stellar-mass black hole binaries.

Reference graph

Works this paper leans on

72 extracted references · 9 linked inside Pith · cited by 4 Pith papers

  1. [1]

    Abac, A. et al. 2025a, [arXiv: 2503.12263]

    Pith/arXiv arXiv

  2. [2]

    Abac, A. G. et al. 2025b, [arXiv: 2509.04348]

    Pith/arXiv arXiv

  3. [3]

    Abac, A. G. et al. 2025c, [arXiv: 2508.18083]

    Pith/arXiv arXiv

  4. [4]

    Abac, A. G. et al. 2025d, [arXiv: 2508.18082]

    Pith/arXiv arXiv

  5. [5]

    2016, Physical Review Let- ters, 116

    Abbott, B., Abbott, R., Abbott, T., & Abernathy, M. 2016, Physical Review Let- ters, 116

    2016

  6. [6]

    P., Abbott, R., Abbott, T

    Abbott, B. P., Abbott, R., Abbott, T. D., et al. 2019, Phys. Rev. X, 9, 031040

    2019

  7. [7]

    Adame, A. G. et al. 2025, JCAP, 02, 021

    2025

  8. [8]

    Agarwal, A. et al. 2025, Astrophys. J., 987, 47

    2025

  9. [9]

    & Rasio, F

    Antonini, F. & Rasio, F. A. 2016, Astrophys. J., 831, 187

    2016

  10. [10]

    2022, Phys

    Belfiglio, A., Luongo, O., & Mancini, S. 2022, Phys. Rev. D, 105, 123523

    2022

  11. [11]

    2024, ApJ, 964, 191

    Borghi, N., Mancarella, M., Moresco, M., et al. 2024, ApJ, 964, 191

    2024

  12. [12]

    2018,http://github.com/ jax-ml/jax

    Bradbury, J., Frostig, R., Hawkins, P., et al. 2018,http://github.com/ jax-ml/jax

    2018

  13. [13]

    Branchesi, M. et al. 2023, JCAP, 07, 068

    2023

  14. [14]

    2025, Phys

    Califano, M., De Martino, I., & Vernieri, D. 2025, Phys. Rev. D, 111, 123535

    2025

  15. [15]

    M., & Gupta, I

    Chen, H.-Y ., Ezquiaga, J. M., & Gupta, I. 2024, CQG, 41, 125004

    2024

  16. [16]

    Chen, H.-Y ., Fishbach, M., & Holz, D. E. 2018, Nature, 562, 545–547 Del Pozzo, W. 2012, Physical Review D, 86

    2018

  17. [17]

    Evans, M. et al. 2021, [arXiv: 2109.09882]

    Pith/arXiv arXiv 2021

  18. [18]

    Ezquiaga, J. M. & Holz, D. E. 2022, PRL, 129, 061102

    2022

  19. [19]

    M., Callister, T

    Farah, A. M., Callister, T. A., Ezquiaga, J. M., Zevin, M., & Holz, D. E. 2025, ApJ, 978, 153

    2025

  20. [20]

    Farr, W. M. 2019, Research Notes of the AAS, 3, 66

    2019

  21. [21]

    2021, JCAP, 2021, 026

    Finke, A., Foffa, S., Iacovelli, F., Maggiore, M., & Mancarella, M. 2021, JCAP, 2021, 026

    2021

  22. [22]

    2025, Class

    Fishbach, M. 2025, Class. Quant. Grav., 42, 055009

    2025

  23. [23]

    Fishbach, M., Gray, R., & Hernandez, I. M. 2019, ApJL, 871, L13

    2019

  24. [24]

    Ford, K. E. S. & McKernan, B. 2022, Mon. Not. Roy. Astron. Soc., 517, 5827

    2022

  25. [25]

    W., Lang, D., & Goodman, J

    Foreman-Mackey, D., Hogg, D. W., Lang, D., & Goodman, J. 2013, Publ. Astron. Soc. Pac., 125, 306

    2013

  26. [26]

    R., Ghosh, A., Gray, R., et al

    Gair, J. R., Ghosh, A., Gray, R., et al. 2023, ApJ, 166, 22

    2023

  27. [27]

    & Fishbach, M

    Gerosa, D. & Fishbach, M. 2021, Nature Astron., 5, 749

    2021

  28. [28]

    M., Qi, H., et al

    Gray, R., Hernandez, I. M., Qi, H., et al. 2020, Phys. Rev. D, 101

    2020

  29. [29]

    2022, MNRAS, 512, 1127

    Gray, R., Messenger, C., & Veitch, J. 2022, MNRAS, 512, 1127

    2022

  30. [30]

    Gray, R. et al. 2023, JCAP, 12, 023

    2023

  31. [31]

    2025, Phys

    Heinzel, J., Mould, M., & Vitale, S. 2025, Phys. Rev. D, 111, L061305

    2025

  32. [32]

    & Vitale, S

    Heinzel, J. & Vitale, S. 2025, [arXiv: 2509.07221]

    Pith/arXiv arXiv 2025

  33. [33]

    Hild, S. et al. 2011, CQG, 28, 094013

    2011

  34. [34]

    Holz, D. E. & Hughes, S. A. 2005, The Astrophysical Journal, 629, 15–22

    2005

  35. [35]

    2026, Sci

    Jin, S.-J., Song, J.-Y ., Sun, T.-Y ., et al. 2026, Sci. China Phys. Mech. Astron., 69, 220401

    2026

  36. [36]

    2023, Mon

    Karathanasis, C., Mukherjee, S., & Mastrogiovanni, S. 2023, Mon. Not. Roy. Astron. Soc., 523, 4539

    2023

  37. [37]

    2018, PRL, 120, 161102

    London, L., Khan, S., Fauchon-Jones, E., et al. 2018, PRL, 120, 161102

    2018

  38. [38]

    & Dickinson, M

    Madau, P. & Dickinson, M. 2014, Ann. Rev. Astron. Astrophys., 52, 415

    2014

  39. [39]

    Maggiore, M., Broeck, C. V . D., Bartolo, N., et al. 2020, JCAP, 2020, 050–050

    2020

  40. [40]

    2022, PRD, 105, 064030

    Mancarella, M., Genoud-Prachex, E., & Maggiore, M. 2022, PRD, 105, 064030

    2022

  41. [41]

    & Gerosa, D

    Mancarella, M. & Gerosa, D. 2025, Phys. Rev. D, 111, 103012

    2025

  42. [42]

    M., & Gair, J

    Mandel, I., Farr, W. M., & Gair, J. R. 2019, MNRAS, 486, 1086

    2019

  43. [43]

    A., & Artale, M

    Mapelli, M., Bouffanais, Y ., Santoliquido, F., Sedda, M. A., & Artale, M. C. 2022, Mon. Not. Roy. Astron. Soc., 511, 5797

    2022

  44. [44]

    2023, PRD, 108, 042002

    Mastrogiovanni, S., Laghi, D., Gray, R., et al. 2023, PRD, 108, 042002

    2023

  45. [45]

    2021, Phys

    Mastrogiovanni, S., Leyde, K., Karathanasis, C., et al. 2021, Phys. Rev. D, 104

    2021

  46. [46]

    2024, A&A, 682, A167

    Mastrogiovanni, S., Pierra, G., Perriès, S., et al. 2024, A&A, 682, A167

    2024

  47. [47]

    2016, JCAP, 05, 014

    Moresco, M., Pozzetti, L., Cimatti, A., et al. 2016, JCAP, 05, 014

    2016

  48. [48]

    Moresco, M. et al. 2022, Living Rev. Relativ., 25, 6

    2022

  49. [49]

    2022, Mon

    Mukherjee, S. 2022, Mon. Not. Roy. Astron. Soc., 515, 5495

    2022

  50. [50]

    2025, [arXiv: 2511.11795]

    Pierra, G., Colombo, A., & Mastrogiovanni, S. 2025, [arXiv: 2511.11795]

    arXiv 2025

  51. [51]

    2024, PRD, 109, 083504

    Pierra, G., Mastrogiovanni, S., Perriès, S., & Mapelli, M. 2024, PRD, 109, 083504

    2024

  52. [52]

    & Papadopoulos, A

    Pierra, G. & Papadopoulos, A. 2026, [arXiv: 2601.03257]

    Pith/arXiv arXiv 2026

  53. [53]

    2010, CQG, 27, 194002

    Punturo, M., Abernathy, M., Acernese, F., et al. 2010, CQG, 27, 194002

    2010

  54. [54]

    Reitze, D. et al. 2019, Bull. Am. Astron. Soc., 51, 035

    2019

  55. [55]

    Riess, A. G. et al. 2022, Astrophys. J. Lett., 934, L7

    2022

  56. [56]

    2024, Astron

    Rinaldi, S., Del Pozzo, W., Mapelli, M., Lorenzo-Medina, A., & Dent, T. 2024, Astron. Astrophys., 684, A204

    2024

  57. [57]

    2025, Phys

    Sadiq, J., Dent, T., & Lorenzo-Medina, A. 2025, Phys. Rev. D, 112, 083028

    2025

  58. [58]

    Schutz, B. F. 1986, Nature, 323, 310

    1986

  59. [59]

    2026, [arXiv: 2601.03347]

    Tagliazucchi, M., Moresco, M., Borghi, N., & Ciapetti, C. 2026, [arXiv: 2601.03347]

    Pith/arXiv arXiv 2026

  60. [60]

    2025, A&A, 702, A244

    Tagliazucchi, M., Moresco, M., Borghi, N., & Fiebig, M. 2025, A&A, 702, A244

    2025

  61. [61]

    & Golomb, J

    Talbot, C. & Golomb, J. 2023, MNRAS, 526, 3495

    2023

  62. [62]

    R., Gair, J

    Taylor, S. R., Gair, J. R., & Mandel, I. 2012, PRD, 85, 023535

    2012

  63. [63]

    2018, Class

    Tiwari, V . 2018, Class. Quant. Grav., 35, 145009

    2018

  64. [64]

    2025, [arXiv: 2510.25579]

    Tiwari, V . 2025, [arXiv: 2510.25579]

    Pith/arXiv arXiv 2025

  65. [65]

    2024, Astron

    Torniamenti, S., Mapelli, M., Périgois, C., et al. 2024, Astron. Astrophys., 688, A148 van Son, L. A. C., de Mink, S. E., Callister, T., et al. 2022, Astrophys. J., 931, 17

    2024

  66. [66]

    & Evans, M

    Vitale, S. & Evans, M. 2017, Phys. Rev. D, 95, 064052

    2017

  67. [67]

    M., & Taylor, S

    Vitale, S., Gerosa, D., Farr, W. M., & Taylor, S. R. 2020, in Handbook of Gravi- tational Wave Astronomy

    2020

  68. [68]

    Wong, K. W. K., Breivik, K., Kremer, K., & Callister, T. 2021, Phys. Rev. D, 103, 083021

    2021

  69. [69]

    Woosley, S. E. & Heger, A. 2021, Astrophys. J. Lett., 912, L31

    2021

  70. [70]

    & Fishbach, M

    Ye, C. & Fishbach, M. 2021, Phys. Rev. D, 104, 043507

    2021

  71. [71]

    Ye, C. S. & Fishbach, M. 2024, Astrophys. J., 967, 62

    2024

  72. [72]

    & Holz, D

    Zevin, M. & Holz, D. E. 2022, Astrophys. J. Lett., 935, L20 Article number, page 11 A&A proofs:manuscript no. main Appendix A: Results withΩ m.0 fixed In this appendix, we present and discuss the results obtained un- der the assumption thatΩ m,0 is held fixed at its fiducial value. In Fig. A.1, we compare the constraints derived from the S/N >75 catalog u...

    2022