Gravitational wave cosmology : an introduction

Gr\'egoire Pierra; Simone Mastrogiovanni

REVIEW 4 major objections 4 minor 1 cited by

Gravitational wave cosmology : an introduction

T0 review · 4 major / 4 minor · reviewed 2026-08-06 · deepseek-v4-flash

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

Pith's one-line read This chapter argues that gravitational waves from compact binary coalescences can act as standard sirens—sources whose distance is read directly from the waveform—and lays out bright, dark, and spectral siren routes to measuring the…

desk verdict A useful but typo-heavy textbook chapter that re-exposes standard GW cosmology; no new research, and the bright-siren normalization worry is a prose slip, not a formula error. read the letter →

arxiv 2507.10597 v1 pith:TLV4CJ3S submitted 2025-07-12 gr-qc astro-ph.COastro-ph.IM

Grégoire Pierra , Simone Mastrogiovanni This is my paper

classification gr-qcastro-ph.COastro-ph.IM

keywords gravitationalwavecosmologystandardsirensbrightdarkspectralhierarchicalBayesianinferenceselectioneffectsHubbleconstant

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 chapter is a review of how gravitational-wave observations can be used as standard sirens: compact binary coalescences let us read the luminosity distance directly from the waveform's amplitude, and once a redshift is attached, the event becomes a point on the distance–redshift curve that fixes cosmological parameters such as the Hubble constant. The text walks through three ways to attach the redshift: a bright electromagnetic counterpart, a galaxy catalog, and the shape of the black-hole mass distribution. It also lays out the hierarchical Bayesian framework needed to combine many noisy events and to correct for the fact that detectors only see loud, nearby, favorably oriented mergers. If the framework holds, gravitational-wave cosmology gives a measurement of cosmic expansion that does not rely on the distance ladder, and can be compared with supernova and cosmic-microwave-background values.

What carries the argument

The central machinery is the hierarchical Bayesian likelihood (Eqs. 1.50 and 1.54), which combines per-event gravitational-wave likelihoods with a population rate model and a detection probability $p(\mathrm{DET}=1|\theta)$. The cosmological content enters through the luminosity distance $d_L(z)$, the redshifted chirp mass $M_d=(1+z)M_s$, and the change of variables between source-frame and detector-frame rates, whose Jacobian involves $d_L/dz$ and $\mathrm{d}V_c/\mathrm{d}z$. The key identity is that the waveform's phase measures the redshifted chirp mass while its amplitude measures $d_L$, so redshift must be supplied externally or statistically; the chapter's three methods are the ways of supplying it.

What would settle it

Simulate a bright-siren catalog with a known $H_0$ and a merger-rate model whose comoving-volume factor varies as $H_0^{-3}$; if applying Eq. 1.84 shifts the recovered $H_0$ away from the injected value, the constant-rate assumption behind the derivation fails.

Watch

Extended reading notes

Core claim

The chapter's central claim is that the gravitational-wave signal from a compact binary carries two cosmologically useful quantities: the luminosity distance, from the amplitude, and the detector-frame chirp mass, from the phase. Because the same observed waveform could come from a light nearby binary or a heavier distant one, the redshift cannot be read from the signal alone. The chapter argues that the redshift ambiguity can be broken in three ways—an electromagnetic counterpart, a galaxy catalog, or a model of the source-frame mass distribution—and that in each case a hierarchical Bayesian likelihood, built from the merger rate and detector selection effects, converts the events into constraints on cosmological parameters. It emphasizes that the mass model always enters the galaxy-catalog method as a weight, and that selection effects, such as the $H_0^3$ scaling of the detectable volume, must be modeled to avoid biased inferences.

Load-bearing premise

The load-bearing simplification is that, in the bright-siren derivation, the merger-rate factor can be treated as a constant when varying the Hubble constant; in a full treatment that factor depends on $H_0$ through the volume of space sampled, roughly as the inverse cube of $H_0$.

Editorial extensions

If this is right

The bright-siren route, demonstrated by GW170817, yields a direct, ladder-free $H_0$ measurement, though such events are rare.
Dark sirens with galaxy catalogs extract cosmological information from many events; the selection term scales as $H_0^3$ and so must be modeled.
Spectral sirens use the source-frame mass distribution as a ruler and can be applied to events without counterparts, but their $H_0$ posteriors shift with the assumed mass model.
The hierarchical likelihood, with a Poisson term for expected detections, gives a unified formula that covers all three methods and naturally handles selection effects.
As more detections accumulate, standard-siren cosmology can provide an independent cross-check of the distance ladder and cosmic-microwave-background-based expansion history.

Reading between the lines

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

Editorial inference: the chapter's simplified bright-siren likelihood (Eq. 1.84) is a quick falsification target for full-rate simulations that retain the $\mathrm{d}V_c/\mathrm{d}z \propto H_0^{-3}$ scaling; if the simplified posterior shifts, the constant-rate assumption, not the standard-siren idea, is the part to replace.
Editorial inference: because spectral sirens infer redshift entirely from an assumed source-frame mass distribution, a population that evolves with redshift—for instance, a mass peak that shifts in height or location—would masquerade as cosmology; the paper's Eq. 1.60 shows the degeneracy but does not develop redshift-dependent mass models.
Editorial inference: the galaxy-catalog and spectral-siren methods are unified through the same merger-rate likelihood; a natural next step, left implicit, is to fit $H_0$ jointly with the exponent linking host probability to galaxy luminosity, using real catalogs and a flux-limited completeness correction.

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.

Desk Editor's Note

read the letter

Punchline: this is a well-intentioned textbook chapter, not a research paper. It has no new results, and it has several equation typos, but the overall framework is standard and the presentation is clear enough to be useful.

What's actually new: nothing, and that's fine. The chapter is an introduction for newcomers, covering bright, dark, and spectral sirens within a hierarchical Bayesian framework. It does that well. The derivations follow the literature (Schutz, Mandel et al., Gair et al.) and the chapter is honest about the assumptions. The inclusion of worked examples with ICAROGW and the Gair et al. code is a nice touch; students can reproduce the figures.

Soft spots: the typos are real and should be fixed. Eq. 1.23 drops the speed of light from dL(z). Eq. 1.34 writes the detector-frame phase as Psi_s(fd/(1+z)) instead of Psi_s(fd(1+z)). Eqs. 1.42-1.43 use xi in the Gaussian posterior exponent where they need mu. For a pedagogical chapter these matter, because a student copying the equations will learn the wrong thing. They're small fixes, though.

The reader's biggest worry was the bright-siren normalization. I checked: the sentence after Eq. 1.83 says the rate factor is a normalization constant with respect to H0, which is wrong because dVc/dz scales as H0^-3. But Eq. 1.84 already carries the compensating H0^-3 denominator. So it's a prose error, not a formula error. The stress-test note is right.

I'd send this to a referee, not because it breaks new ground, but because a pedagogical chapter with this many typos needs a careful reader before it goes into a book. The core content is sound; the presentation just needs tidying. My sense is that the authors would welcome that.

Referee Report

4 major / 4 minor

Summary. This book chapter is a pedagogical introduction to gravitational-wave (GW) cosmology with standard sirens. It derives the cosmological luminosity distance, discusses the redshifted GW waveform from compact binary coalescences, introduces the hierarchical Bayesian inference framework with selection effects, and presents three methods for cosmological inference: spectral sirens based on the source mass distribution, dark sirens cross-correlated with galaxy catalogs, and bright sirens with electromagnetic counterparts. The chapter includes worked examples using public data products and references to publicly available codes (ICAROGW and the Gair et al. 2023 galaxy-catalog codes).

Significance. The manuscript is a tutorial rather than a new research contribution, and its value lies in the clarity and correctness of the presentation. If the displayed equations are corrected, it will serve as a useful entry point for graduate students and researchers entering GW cosmology: it gives a rate-based derivation of the hierarchical likelihood, an explicit treatment of selection biases, and reproduces the standard likelihood expressions for galaxy-catalog dark sirens (Eq. 1.77) and low-redshift bright sirens (Eq. 1.84). The central claim that bright, dark, and spectral siren methods provide valid means to infer cosmological parameters from GW detections is standard and is supported by the final expressions, with the caveats noted below. The chapter is also commendable for pointing readers to reproducible code and for clearly emphasizing the mass-redshift degeneracy and the role of selection effects.

major comments (4)

[1.2.1, Eq. (1.23)] The luminosity distance integral is missing the speed of light c in the numerator. It should read d_L(z) = (1+z) ∫_0^z c dz' / [H0 sqrt(Ωm(1+z')^3 + ΩΛ)]. As printed, Eq. (1.23) has incorrect dimensions and is inconsistent with Eq. (1.17), which contains cz/H0, and with Eq. (1.25), which contains c/H0. Since this formula is the foundation for all subsequent distance-redshift calculations, it must be corrected.
[1.2.2, Eq. (1.34)] The detector-frame phase is written as Ψ_d(f_d) = Ψ_s(f_d/(1+z)). Because the source-frame frequency corresponding to a detector frequency f_d is f_s = f_d(1+z), the correct relation is Ψ_d(f_d) = Ψ_s(f_d(1+z)). The subsequent line, Eq. (1.35), uses the correct substitution, so the sign error is confined to Eq. (1.34), but as displayed it teaches the wrong redshift mapping for the GW phase.
[1.3.1, Eqs. (1.42) and (1.43)] The exponent in the simplified Gaussian likelihood and posterior is written with x_i rather than µ. For example, Eq. (1.42) contains (x_i - \barµ)^2, and Eq. (1.43) repeats the same error. The correct statement is that, up to a µ-independent constant, L({x}|µ,σ) ∝ exp[-(µ - \barµ)^2/(2\barσ^2)], with \barµ = N^{-1} Σ_i x_i and \barσ = σ/√N. As printed, the expressions depend on a single data point and are dimensionally and statistically incorrect.
[1.6, text after Eq. (1.83)] The sentence claiming that the rate factor R0 ψ(z)/(1+z) dVc/dz acts as a normalization constant with respect to a varying H0 is not correct, because dVc/dz depends on H0 (approximately ∝ H0^{-3} at low redshift). The final bright-siren likelihood in Eq. (1.84) is the standard low-redshift expression and carries the expected H0^{-3} dependence through its denominator, so the displayed result is correct; however, the explanatory text should be revised to describe how the rate factor is absorbed (for instance, through the selection denominator) rather than asserting that it is H0-independent. This matters because the sentence is the justification for dropping population uncertainties in the simplified bright-siren analysis.

minor comments (4)

[1.2.1, Eq. (1.25)] Eq. (1.25) uses z' in the denominator Ωm(1+z')^3 + ΩΛ after the right-hand side should be a function of z, or alternatively the expression should be written as an integral over z'. Please correct the variable mismatch.
[1.2.2, Eq. (1.33)] The sentence following Eq. (1.33) states that in the limit z → ∞ the detector spectrum is concentrated at low frequencies with a large amplitude. This is misleading: while the frequency is redshifted downward, the amplitude scaling from the (1+z) prefactor combined with the f^{-7/6} spectral index does not support an unqualified 'large amplitude' statement. Please qualify or rephrase.
[1.3.2, Eqs. (1.57)–(1.59)] The Jacobian transformation for the detector-frame rate is compactly written but easy to misread. Please define explicitly the matrix J_{d→s} and its entries, and verify the powers of (1+z) in Eq. (1.59), since the displayed chain of factors is not transparent as it stands.
[1.9, Figure captions and labels] In Figure 1.2, the label 'Flat CDM' should read 'Flat ΛCDM' to match the model described in the text. There are also several typographical slips in the references, such as missing volume/page information for Palmese and Mastrogiovanni (2025) and Vitale et al. (2020).

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the chapter is a self-contained tutorial that re-derives standard hierarchical Bayesian and siren formulas from stated assumptions, with self-citations used only for software, tutorials, and reviews, none of which is load-bearing.

full rationale

The chapter is a pedagogical introduction rather than a paper claiming a new derivation or prediction. Its central equations (the hierarchical likelihood, Eq. 1.54; the galaxy-catalog likelihood, Eq. 1.77; and the bright-siren likelihood, Eq. 1.84) are re-derived within the text from Bayes' theorem, the CBC merger-rate parameterization, and the stated assumptions about detection probabilities and redshift information. The self-citations to ICAROGW, the Gair et al. 2023 codes, and the Palmese-Mastrogiovanni review are used only as pointers to software, tutorials, and literature for further reading; they are not invoked to justify a load-bearing premise or to forbid alternative approaches. The mass-model assumptions for spectral sirens are stated explicitly as modeling choices, not derived from the target result. The only questionable passage is the claim after Eq. 1.83 that the rate factor R0 psi(z)/(1+z) dVc/dz acts as a normalization constant with respect to H0, since dVc/dz itself depends on H0; however, this is an internal consistency/correctness issue in the explanatory text, not a circular step, and Eq. 1.84 already carries the compensating H0^{-3} dependence. There is no fitted input renamed as a prediction, no uniqueness argument imported from the authors' prior work, and no known result merely relabeled as organization. The chapter is therefore best scored 0 for circularity.

Assumptions & free parameters 0 free parameters · 6 assumptions · 0 invented entities

The chapter is a review, so the central claim is pedagogical rather than a new scientific result. It relies on standard cosmological assumptions (FLRW, Friedmann equations) and explicitly stated simplifications for each siren method. No new entities or parameters are introduced; all model parameters are inputs from prior literature or illustrative choices.

assumptions (6)

domain assumption The universe is homogeneous and isotropic on large scales, justifying the FLRW metric.
Invoked in Sec. 1.2.1 to derive the luminosity distance and Friedmann equations.
domain assumption The universe is flat and dominated by matter and dark energy, so the first Friedmann equation has only Omega_m(1+z)^3 + Omega_Lambda.
Stated in Sec. 1.2.1, Eq. 1.21; standard simplification.
domain assumption GW amplitude yields luminosity distance directly without calibration.
Standard siren property, Sec. 1.2.2; the chapter adopts this.
domain assumption Source-frame mass distribution is independent of redshift.
Explicitly flagged as a key assumption in Sec. 1.4 before Eq. 1.60.
domain assumption In the dark siren toy model, the galaxy catalog is complete, redshift and sky positions are perfect, and galaxies are equally likely hosts.
Listed as assumptions in Sec. 1.5 before Eq. 1.75.
domain assumption In the bright siren example, the EM counterpart is perfectly localized in redshift, EM provides no inclination information, and GW masses are independent of distance and sky position.
Assumptions stated in Sec. 1.6 before Eq. 1.83.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Gravitational wave cosmology : an introduction." pith.science (2026). https://pith.science/paper/TLV4CJ3S

@misc{pith2026250710597,
  author       = {Pith},
  title        = {Pith review of: Gravitational wave cosmology : an introduction},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/TLV4CJ3S}},
  note         = {Machine review of arXiv:2507.10597}
}

read the original abstract

This chapter introduces gravitational wave cosmology, focusing on the use of gravitational waves as standard sirens to probe the expansion history of the Universe. It presents and explains the methodologies behind bright and dark siren analyses, including their respective data requirements and underlying assumptions. Particular attention is given to the theoretical foundations of these approaches, the statistical frameworks used to interpret gravitational-wave events, and the treatment of selection effects. Examples and applications are provided for each method, with the aim of offering a clear and accessible introduction to the tools and concepts enabling cosmological inference from gravitational-wave observations.

Figures

Figures reproduced from arXiv: 2507.10597 by the authors.

**Figure 1.12.** However, while this treatment gives a rapid idea of the estimation of the [PITH_FULL_IMAGE:figures/full_fig_p029_1_12.png] view at source ↗

Discussion (0). Sign in to comment.

Forward citations

Cited by 1 Pith paper

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

Inferring cosmological parameters from galaxy and dark sirens cross-correlation
astro-ph.CO 2025-10 conditional novelty 4.0 of 10

A full-likelihood forecast shows dark-siren×galaxy cross-correlations with 3G detectors and Euclid could constrain H0 at 0.7% and complement galaxy clustering on other parameters.

Reference graph

Works this paper leans on

4 extracted references · 4 canonical work pages · cited by 1 Pith paper

[1]

This is a reasonable assumption, as light pulses are generated by physical processes occurring on timescales significantly smaller than cosmic times 6 GRAVIT A TIONAL WAVES happens today, with a(td) = 1 and a(ts) < 1, then their ratio is greater than one, and the photons are redshifted. It follows that we can define a relation between the scale factor and...

work page 2022
[2]

d” when we refer to detector quantities and “ s

Here, remember that we assume that the Universe is flat, i.e. Ωk = 0. 8 GRAVIT A TIONAL WAVES which represents the energy density required for a spatially flat Universe. All den- sity parameters are then expressed relative to this critical density. Remembering the expression of the comoving distance in Eq. 1.17 and Friedmann’s Eq. 1.21, we can rewrite the...

work page 2023
[3]

histogram

the reader can check the tutorials for the actual values of the population parameters Gravitational Wave Cosmology : an introduction 25 0.0 0.5 1.0 1.5 2.0 2.5 3.0 Galaxy number density profile True redshift Interpolant dVc/dz True redshift Interpolant dVc/dz 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 z 0.0 0.5 1.0 1.5 2.0 2.5 3.0 Over(under)density 0.0 0.2 0.4 0.6 ...

work page 2023
[4]

icarogw : A python package for inference of population properties of noisy, heterogeneous and incomplete ob- servations

https://github.com/simone-mastrogiovanni/hitchhiker_guide_dark_sirens 30 GRAVIT A TIONAL WAVES Posterior [km 1 s Mpc] ( dL/dL = 10%) LOS1 - 1 deg2 LOS1 - 5 deg2 LOS2 - 1 deg2 LOS2 - 5 deg2 Truth Posterior [km 1 s Mpc] ( dL/dL = 20%) 40 50 60 70 80 90 100 110 120 H0[km s 1 Mpc 1] Posterior [km 1 s Mpc] ( dL/dL = 30%) FIGURE 1.11 – From top to bottom Hubble...

work page 2017

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.

[1] [1]

This is a reasonable assumption, as light pulses are generated by physical processes occurring on timescales significantly smaller than cosmic times 6 GRAVIT A TIONAL WAVES happens today, with a(td) = 1 and a(ts) < 1, then their ratio is greater than one, and the photons are redshifted. It follows that we can define a relation between the scale factor and...

work page 2022

[2] [2]

d” when we refer to detector quantities and “ s

Here, remember that we assume that the Universe is flat, i.e. Ωk = 0. 8 GRAVIT A TIONAL WAVES which represents the energy density required for a spatially flat Universe. All den- sity parameters are then expressed relative to this critical density. Remembering the expression of the comoving distance in Eq. 1.17 and Friedmann’s Eq. 1.21, we can rewrite the...

work page 2023

[3] [3]

histogram

the reader can check the tutorials for the actual values of the population parameters Gravitational Wave Cosmology : an introduction 25 0.0 0.5 1.0 1.5 2.0 2.5 3.0 Galaxy number density profile True redshift Interpolant dVc/dz True redshift Interpolant dVc/dz 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 z 0.0 0.5 1.0 1.5 2.0 2.5 3.0 Over(under)density 0.0 0.2 0.4 0.6 ...

work page 2023

[4] [4]

icarogw : A python package for inference of population properties of noisy, heterogeneous and incomplete ob- servations

https://github.com/simone-mastrogiovanni/hitchhiker_guide_dark_sirens 30 GRAVIT A TIONAL WAVES Posterior [km 1 s Mpc] ( dL/dL = 10%) LOS1 - 1 deg2 LOS1 - 5 deg2 LOS2 - 1 deg2 LOS2 - 5 deg2 Truth Posterior [km 1 s Mpc] ( dL/dL = 20%) 40 50 60 70 80 90 100 110 120 H0[km s 1 Mpc 1] Posterior [km 1 s Mpc] ( dL/dL = 30%) FIGURE 1.11 – From top to bottom Hubble...

work page 2017