Do three-body encounters in galactic nuclei affect compact binary merger rates?

Alessandro Alberto Trani

REVIEW 4 major objections 5 minor 12 references

Do three-body encounters in galactic nuclei affect compact binary merger rates?

T0 review · 4 major / 5 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 Close three-body encounters with the compact-remnant cusp around a supermassive black hole nearly triple the merger fraction of black-hole binaries within 1 Myr, from about 10% to about 31%.

desk verdict A preliminary but suggestive simulation study claiming encounters triple BBH mergers around SgrA*, with a control arm that makes the factor-of-3 provisional. read the letter →

arxiv 1908.07535 v1 pith:G7KXDZH7 submitted 2019-08-20 astro-ph.HE astro-ph.GA

Alessandro Alberto Trani This is my paper · ORCID

classification astro-ph.HEastro-ph.GA

keywords blackholephysicsgravitationalwavesmethods:numericalbinaries:generalgalaxies:nucleiKozai-Lidovmechanismthree-bodyencountersgalacticcenter

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

Three-body encounters with the cusp of compact remnants around a supermassive black hole (SMBH) are usually omitted from Kozai-Lidov merger-rate calculations. This paper simulates 1500 black-hole binaries orbiting a SgrA*-like SMBH, with and without such encounters, using a hybrid N-body/Monte Carlo code. It finds that encounters nearly triple the fraction of binaries that merge within 1 Myr, from about 10% to about 31%. The key effect is not direct hardening during encounters, but encounters reorienting binaries so that more of them enter Kozai-Lidov cycles that drive gravitational-wave merger. If the result holds, predicted compact-binary merger rates in galactic nuclei should be revised upward by roughly a factor of three.

What carries the argument

The central object is PROMENADE, a hybrid N-body/Monte Carlo code built on the TSUNAMI few-body integrator, which includes post-Newtonian corrections up to 2.5PN order. Each simulation follows one binary on its orbit around the SMBH; after each time step the code computes an encounter probability from the local cusp density and velocity dispersion, and when a Monte Carlo draw triggers an encounter it injects a fourth body, integrates the encounter directly, and then removes it. Running identical initial conditions twice, once with encounters and once without, isolates the effect of encounters on the Kozai-Lidov merger channel.

What would settle it

Run a direct N-body simulation of the same SgrA*-like nucleus with all three-body encounters explicitly integrated and no analytic evaporation cutoff, then compare the merger fraction against a version with encounters disabled; if the encounter-enabled merger fraction is not roughly three times the no-encounter fraction, the central claim would be falsified.

Watch

Extended reading notes

Core claim

The central claim is that, around a SgrA*-like supermassive black hole, close three-body encounters with interloping compact remnants increase the probability that a black-hole binary merges by almost a factor of three within 1 Myr: 31% of binaries merge in the encounter-enabled set versus about 10% in the Kozai-Lidov-only set. The paper argues that encounters act mainly by altering the binary's orientation with respect to its orbit around the SMBH, thereby widening the parameter space in which Kozai-Lidov eccentricity oscillations can drive the binary to gravitational-wave coalescence; it terms this channel 'encounter-assisted Kozai-Lidov merger.' Based on the simulated merger fraction, the paper derives a merger rate of 1.6e-6 per year per Milky Way-like nucleus.

Load-bearing premise

The comparison assumes that stopping the no-encounter runs at the analytic timescale over which encounters would statistically dissolve the binary is a fair stand-in for actually simulating those disruptions; if that proxy is wrong, the measured factor-of-three enhancement would change.

Editorial extensions

If this is right

Galactic-nucleus merger-rate estimates that ignore encounters are understated by roughly a factor of three for SgrA*-like environments.
The predicted Milky Way-like rate of 1.6e-6 yr^-1 per nucleus feeds directly into estimates of the local compact-binary merger rate for gravitational-wave observatories.
Most encounter-assisted mergers occur during isolated evolution after an encounter has reoriented the binary, so the enhancement is best described as a widening of the Kozai-Lidov window rather than a direct-collision channel.
Using the analytic evaporation timescale to stop no-encounter simulations overestimates binary breakups by about one-third compared with explicitly modeled encounters, implying that more binaries survive to merge later than the proxy would suggest.

Reading between the lines

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

If reorientation by encounters is the operative mechanism, the same enhancement should appear around intermediate-mass black holes or in dense AGN disks, and the effect should strengthen as cusp density increases; varying the cusp mass and density slope in the same Monte Carlo setup would test this.
Encounter-assisted mergers should leave a gravitational-wave signature: binaries whose orbit is suddenly reoriented then merge through standard Kozai-Lidov cycles, potentially producing a population with a distinct eccentricity-versus-frequency distribution that could be searched for in LIGO/Virgo data.
Because the evaporation-timescale criterion overestimates breakups, previous KL-only rate estimates may be understated not only by the factor of three reported here but also by the loss of binaries that would survive and merge on longer timescales if encounters were modeled.

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.

Desk Editor's Note

read the letter

Quick take: this short proceedings paper from Trani is a first attempt to include three-body encounters in the Kozai-Lidov merger channel around SgrA*. The headline result—encounters triple the merger fraction, from ~10% to ~31%—is plausible and worth taking seriously, but the comparison is not as clean as it looks. The KL-only control is stopped at an analytic evaporation timescale rather than run for the full 1 Myr, so the 10% is a censored lower bound, not a true 1 Myr baseline. That means the factor of 3 could be inflated.

What's genuinely new: previous models (Antonini & Perets, Hoang et al., Hamers et al.) considered only secular KL evolution. Trani actually inserts encounters via a Monte Carlo prescription in a direct few-body integrator (TSUNAMI) with PN terms. The idea that encounters reorient binaries and widen the KL parameter space—'encounter-assisted KL merger'—is a reasonable mechanism, and the paper explicitly defers a full demonstration to a coming paper. For a proceedings, the 1500 simulations are a reasonable sample, and the presentation is clear. The author also notes that the evaporation criterion overestimates breakups by a third compared to modeled encounters, which is a useful sanity check.

Where the soft spots are: the asymmetric stopping rule is load-bearing. The KL arm is censored at t_ev from Hoang et al. A binary that would have merged via KL between t_ev and 1 Myr is instead counted as a breakup. Given that f_break is 0.851 in KL vs 0.637 in KL+ENC, the KL arm is losing potentially merging systems. The paper labels Table 1 as 'after 1 Myr' but that's only strictly true for the KL+ENC set. This needs to be fixed by running the KL arm for the full Myr, or at least by reporting the distribution of t_ev and the number of systems that would have merged by 1 Myr if not stopped. Second, there are no error bars on the fractions; with 1500 runs, binomial errors are about ±2%, so not huge, but the ratio uncertainty should be stated. Third, no code or data release, which is typical for a proceedings but makes it hard to check the encounter insertion prescription. These are all fixable in the full paper, and none of them make the result implausible.

No circularity: the enhancement is a simulation output, not fitted, and the rate of 1.6e-6/yr is a reasonable product of assumed formation rate and binary fraction, not a surprise.

Bottom line: this is a solid preliminary result that deserves a serious referee. For a journal, I'd want the control arm fixed before accepting the factor-of-3 as a measurement, but as a proceedings paper it's a useful contribution that flags an effect that's likely real. Worth citing as preliminary, and worth bringing to a dynamics group.

Referee Report

4 major / 5 minor

Summary. This short proceedings paper presents PROMENADE, a hybrid N-body/Monte Carlo scheme that follows a black hole binary on an orbit around a SgrA*-like SMBH while stochastically inserting three-body encounters with cusp compact remnants. The authors run 1500 binaries twice, once with encounters (KL+ENC) and once with encounters disabled (KL), and report that KL+ENC produces a merger fraction of 0.311 versus 0.105 for KL, i.e., a factor of about 3 increase within 1 Myr. They attribute this to encounters reorienting binaries and widening the parameter space for Kozai-Lidov-induced mergers, and they derive a Milky Way-like merger rate of 1.6e-6 per year.

Significance. If the factor-of-3 enhancement is real, it materially affects predicted compact binary merger rates in galactic nuclei and would motivate inclusion of encounter-assisted Kozai-Lidov evolution in population synthesis. The paper has genuine strengths: it uses a high-accuracy few-body integrator with 2.5PN corrections, it explicitly models encounters with Monte Carlo sampling of a concrete cusp profile, and it makes a falsifiable quantitative prediction for a SgrA*-like environment. It also usefully flags the limitation that the mechanism attribution is deferred to future work. However, the central quantitative claim rests on an asymmetric comparison between a full 1 Myr integration arm and a control arm stopped at an analytic evaporation timescale, which is a load-bearing issue.

major comments (4)

[Section 2.2 and Table 1] The KL control arm is stopped at the evaporation timescale (Hoang et al. 2018, equation 3), while the KL+ENC arm is integrated to a full 1 Myr. The paper's Table 1 labels both sets as 'after 1 Myr of evolution,' but the KL merger fraction of 0.105 is a censored lower bound, not a 1 Myr outcome. Since the central claim is the ratio 0.311/0.105, this asymmetric stopping rule is load-bearing and the factor-of-3 cannot be considered established until the KL arm is run to 1 Myr or a consistent censoring is applied to both arms.
[Table 1, f_tot_break row] The paper itself shows the evaporation approximation is aggressive: f_tot_break = 0.851 for KL versus 0.637 for KL+ENC, and the text states this gives about one-third more breakups than full encounter modeling. If a substantial fraction of those analytic breakups would instead have merged under pure Kozai-Lidov evolution within 1 Myr, the KL baseline is artificially low and the enhancement factor is inflated. The authors should quantify this sensitivity, for example by recomputing the KL merger fraction without the evaporation cutoff or by modeling evaporation with a realistic prescription in both arms.
[Section 3, final paragraph] The proposed mechanism, 'encounter-assisted Kozai-Lidov merger,' is explicitly deferred: the text states 'we believe that the encounters are triggering Kozai-Lidov-induced mergers' and 'This will be thoroughly demonstrated in our coming work.' As written, the paper presents the factor-of-3 as the main result but leaves the mechanistic explanation as a conjecture. For the central claim to be fully supported, the paper should at least provide a quantitative diagnostic, such as the distribution of binary orientations before and after encounters, rather than deferring the demonstration entirely.
[Section 2.2 and Table 1 (statistical precision)] The 1500 simulations are run without reported statistical uncertainties, convergence checks, or dependence on the random seed. For the KL+ENC and KL sets, the Poisson errors on the merger fractions are roughly 4% and 8%, respectively, but the factor-of-3 ratio could still be affected by systematic setup choices; a simple error bar on each fraction and a convergence statement (e.g., results for 500 vs 1500 runs) would substantially strengthen the claim.

minor comments (5)

[Section 2.1] The sentence 'we have developed a new hybrid N-body/Monte Carlo code' is followed by a citation '(see Trani et al. 2019; Trani et al. 2019 and Mikkola and Tanikawa 1999)' that lists two Trani et al. references with the same author list and year; please disambiguate with labels such as Trani et al. 2019a,b.
[Section 2.2 and Figure 1] The text around the left panel of Figure 1 contains a garbled fragment ('a inn = R Hill Γ enc = 1/t gw 10 au ...') that appears to be misplaced figure content; the caption and surrounding text should be cleaned so that the described distances and rates are readable.
[Section 3] There is a typo in the sentence 'the using the simple evaporation timescale criterion'; it should read 'using the simple evaporation timescale criterion'.
[Section 2.2] The sentence 'The masses of the two black holes are independently drawn form a log-uniform distribution' contains 'form' instead of 'from'.
[Section 4] The rate calculation Γ_MW = Γ_BH * f_BH * f_tot_merg assumes a black hole formation rate Γ_BH = 1e-4 per year and binary fraction f_BH = 0.05, but the text does not cite sources for these values; adding references would help the reader assess the rate's reliability.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: the factor-of-3 merger enhancement and the absolute rate are simulation outputs produced from stated initial conditions and externally cited encounter-rate formulas, not refits of the claimed result.

full rationale

The central claim is a comparison between two simulation arms (KL+ENC and KL) run from identical initial conditions. Encounter probabilities are computed from the local density and velocity dispersion using equation 9 of Leigh et al. (2016), and the initial conditions (SMBH mass, binary parameter distributions, cusp profiles) are explicitly stated assumptions; no parameter is fitted to the merger fraction. The factor-of-3 and the resulting rate of 1.6e-6/yr are therefore genuine simulation outputs rather than predictions forced by construction. The self-citations to Trani et al. (2019) supply the TSUNAMI integrator and the Monte Carlo prescription for inserting the fourth body; those papers do not assert the factor-of-3 result, so the self-citations are not load-bearing for the central claim. The label 'encounter-assisted Kozai-Lidov merger' is an interpretation of the simulation outcomes, and the paper explicitly defers its full demonstration ('This will be thoroughly demonstrated in our coming work'), which is an incompleteness, not a circular step. There is a legitimate fairness caveat: in Section 2.2 the KL arm is stopped at the analytic evaporation timescale (Hoang et al. 2018, eq. 3) rather than integrated for the full 1 Myr, so the 10% baseline may be censored relative to the KL+ENC arm; however, an asymmetric stopping rule biases the comparison but does not make the derivation equivalent to its inputs. None of the seven circularity patterns is exhibited with the required specific reduction.

Assumptions & free parameters 2 free parameters · 5 assumptions · 0 invented entities

The central claim rests on the assumed encounter rate formula, the cusp density model, the evaporation-timescale control, the PN treatment, and the unstated initial stability of the binaries. These are all inputs from prior literature or modeling choices; none are established by this paper itself. The factor-of-3 is an output of these assumptions, and the absolute rate additionally depends on the assumed black hole formation rate and binary fraction.

free parameters (2)

Black hole formation rate = 1e-4 yr^-1
Assumed from prior estimates to convert merger fraction to a physical rate. The final rate scales linearly with it and is therefore not an independent prediction.
Black hole binary fraction = 0.05
Assumed from prior estimates; the final rate scales linearly with it.

assumptions (5)

domain assumption Encounter rate is computed from local density and velocity dispersion using equation 9 of Leigh et al. 2016.
The Monte Carlo probability of encounters depends on this rate, and the factor-of-3 result depends on the encounter rate.
domain assumption The cusp of compact remnants follows a power-law density profile with index -11/4 and total mass 4e4 Msun between 50 au and 0.2 pc, from Alexander and Hopman 2009.
This cusp model sets the local density and the mass and orbit distributions of the interloping bodies.
domain assumption The KL-only benchmark uses the evaporation timescale criterion from equation 3 of Hoang et al. 2018 to terminate integrations.
The comparison between KL+ENC and KL, and therefore the factor-of-3 enhancement, relies on this proxy for encounter-driven breakups.
domain assumption Post-Newtonian corrections up to 2.5PN in TSUNAMI are sufficient for determining mergers.
Merger detection depends on the PN treatment; 2.5PN accounts for leading-order radiation reaction.
domain assumption The initial binary semimajor axes, up to 50 au, are stable against the SMBH tidal field for the sampled outer orbits.
The paper does not explicitly exclude binaries whose Hill radius may be exceeded at small outer semimajor axes; this could affect the merger fraction.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Do three-body encounters in galactic nuclei affect compact binary merger rates?." pith.science (2026). https://pith.science/paper/G7KXDZH7

@misc{pith2026190807535,
  author       = {Pith},
  title        = {Pith review of: Do three-body encounters in galactic nuclei affect compact binary merger rates?},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/G7KXDZH7}},
  note         = {Machine review of arXiv:1908.07535}
}

read the original abstract

High-density cusps of compact remnants are expected to form around supermassive black holes (SMBHs) in galactic nuclei via dynamical friction and two-body relaxation. Due to the high density, binaries in orbit around the SMBH can frequently undergo close encounters with compact remnants from the cusp. This can affect the gravitational wave merger rate of compact binaries in galactic nuclei. We investigated this process by means of high accuracy few-body simulations, performed with a novel Monte Carlo approach. We find that, around a SgrA*-like SMBH, three-body encounters increase the number of mergers by a factor of 3. This occurs because close encounters can reorient binaries with respect to their orbital plane around the SMBH, increasing the number of Kozai-Lidov induced mergers. We obtain a binary black hole merger rate of 1.6x10^-6/yr per Milky Way-like nucleus.

Figures

Figures reproduced from arXiv: 1908.07535 by the authors.

**Figure 1.** Left panel: encounter rates as a function of distance from the SMBH for binaries with semimajor axis ainn = 10 au (thick solid black line), 1 au (thick dashed black line) and 0.1 au (thick dotted black line). The yellow and grey lines are the contributions to the total encounter rate from the stars and from the compact remnants from the cusp, respectively. The dot-dashed red line is the upper limit to the encounter … view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

12 extracted references · 12 canonical work pages

[1]

and Hopman , C.: 2009, 697 , 1861

Alexander , T. and Hopman , C.: 2009, 697 , 1861

work page 2009
[2]

and Perets , H

Antonini , F. and Perets , H. B.: 2012, 757(1) , 27

work page 2012
[3]

R., Ghez , A

Do , T., Lu , J. R., Ghez , A. M., Morris , M. R., Yelda , S., Martinez , G. D., Wright , S. A., and Matthews , K.: 2013, 764 , 154

work page 2013
[4]

M., Eisenhauer , F., Sari , R., Waisberg , I., Habibi , M., Pfuhl , O., George , E., Dexter , J., von Fellenberg , S., Ott , T., and Genzel , R.: 2017, 837 , 30

Gillessen , S., Plewa , P. M., Eisenhauer , F., Sari , R., Waisberg , I., Habibi , M., Pfuhl , O., George , E., Dexter , J., von Fellenberg , S., Ott , T., and Genzel , R.: 2017, 837 , 30

work page 2017
[5]

J., Mori , K., Bauer , F

Hailey , C. J., Mori , K., Bauer , F. E., Berkowitz , M. E., Hong , J., and Hord , B. J.: 2018, 556 , 70

work page 2018
[6]

S., Bar-Or , B., Petrovich , C., and Antonini , F.: 2018, 865 , 2

Hamers , A. S., Bar-Or , B., Petrovich , C., and Antonini , F.: 2018, 865 , 2

work page 2018
[7]

A., and Dosopoulou , F.: 2018, 856 , 140

Hoang , B.-M., Naoz , S., Kocsis , B., Rasio , F. A., and Dosopoulou , F.: 2018, 856 , 140

work page 2018
[8]

Leigh, N. W. C., Antonini, F., Stone, N. C., Shara, M. M., and Merritt, D.: 2016, Monthly Notices of the Royal Astronomical Society 463(2) , 1605

work page 2016

Show all 12 references

[9]

and Tanikawa , K.: 1999, 310 , 745

Mikkola , S. and Tanikawa , K.: 1999, 310 , 745

1999
[10]

Sch \"o del , R., Eckart , A., Alexander , T., Merritt , D., Genzel , R., Sternberg , A., Meyer , L., Kul , F., Moultaka , J., Ott , T., and Straubmeier , C.: 2007, 469 , 125

2007
[11]

A., Fujii, M

Trani, A. A., Fujii, M. S., and Spera, M.: 2019, The Astrophysical Journal 875(1) , 42

2019
[12]

A., Spera , M., Leigh , N

Trani , A. A., Spera , M., Leigh , N. W. C., and Fujii , M. S.: 2019, arXiv e-prints p. arXiv:1904.07879

2019 arXiv

Pith tools

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

[1] [1]

and Hopman , C.: 2009, 697 , 1861

Alexander , T. and Hopman , C.: 2009, 697 , 1861

work page 2009

[2] [2]

and Perets , H

Antonini , F. and Perets , H. B.: 2012, 757(1) , 27

work page 2012

[3] [3]

R., Ghez , A

Do , T., Lu , J. R., Ghez , A. M., Morris , M. R., Yelda , S., Martinez , G. D., Wright , S. A., and Matthews , K.: 2013, 764 , 154

work page 2013

[4] [4]

M., Eisenhauer , F., Sari , R., Waisberg , I., Habibi , M., Pfuhl , O., George , E., Dexter , J., von Fellenberg , S., Ott , T., and Genzel , R.: 2017, 837 , 30

Gillessen , S., Plewa , P. M., Eisenhauer , F., Sari , R., Waisberg , I., Habibi , M., Pfuhl , O., George , E., Dexter , J., von Fellenberg , S., Ott , T., and Genzel , R.: 2017, 837 , 30

work page 2017

[5] [5]

J., Mori , K., Bauer , F

Hailey , C. J., Mori , K., Bauer , F. E., Berkowitz , M. E., Hong , J., and Hord , B. J.: 2018, 556 , 70

work page 2018

[6] [6]

S., Bar-Or , B., Petrovich , C., and Antonini , F.: 2018, 865 , 2

Hamers , A. S., Bar-Or , B., Petrovich , C., and Antonini , F.: 2018, 865 , 2

work page 2018

[7] [7]

A., and Dosopoulou , F.: 2018, 856 , 140

Hoang , B.-M., Naoz , S., Kocsis , B., Rasio , F. A., and Dosopoulou , F.: 2018, 856 , 140

work page 2018

[8] [8]

Leigh, N. W. C., Antonini, F., Stone, N. C., Shara, M. M., and Merritt, D.: 2016, Monthly Notices of the Royal Astronomical Society 463(2) , 1605

work page 2016