{"id":"b91e6f87-c8b4-41bf-9bb9-ccfe5fb38d0a","arxiv_id":"2412.01925","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":5,"one_line_summary":"Accreting black hole binaries in AGN disks contract and merge in thin disks but expand in thick ones, yielding merger times of 1e5 to 1e7 years and rates of 0.2 to 5 per Gpc^3 per year.","lead":"This paper models how pairs of black holes that are actively eating gas inside the disks of active galaxies either shrink and merge or grow apart, depending on how thick the disk is. It finds that in thin disks the pairs merge within a few million years and estimates that this channel could contribute a few percent to a significant fraction of the gravitational wave mergers LIGO sees.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Central contraction/expansion boundary rests on the unverified dominance of the (2,1) outer Lindblad resonance; if ILR/corotation or a positive gravitational torque dominates, h_crit disappears.","rationale":"The paper's strongest claim is the existence of a critical aspect ratio h_crit separating expansion and contraction. The derivation is analytically coherent: Eqs. (33)–(34) follow from the torque balance and mass-variation terms, and the GW phase uses standard Peters equations. The weakest point is not the algebra but the physical assertion that the (2,1) OLR dominates the gravitational torque, so T_grav = −T_visc. This is exactly the assumption the reader flagged. The paper itself recognizes the risk in Section 2.2, citing Chen et al. (2020) for a case where the ILR can dominate. Since the sign of da/dt—and hence the existence of h_crit—turns on this balance, it is the most load-bearing concern. The proposed test (recomputing torque ratios with a simulated cavity density profile) would settle whether OLR dominance actually holds in the thin-disc regime. Because the qualitative trend (thick discs expand, thin discs contract) has independent support from Tiede et al. (2020) and Heath & Nixon (2020), the concern does not reject the paper; it reinforces the reader's CONDITIONAL verdict. The quantitative rates should not be used until the torque balance is checked against a numerical density profile.","tokens_in":18888,"tokens_out":7321,"duration_ms":193976,"concrete_test":"Take the time-averaged surface density profile from a published high-resolution 2D hydro simulation of a circumbinary disc with h=0.01, q=0.5 (e.g., Tiede et al. 2020), evaluate the torque expressions (16)–(17) at the actual resonance radii r_CR, r_ILR, r_OLR, and recompute the ratios T_OLR/T_ILR and T_OLR/T_CR. If either ratio is below unity, the OLR-dominance assumption fails and the model's h_crit is not predictive; if both exceed unity, the concern is mitigated for this parameter point.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central result—contraction below h_crit and expansion above—is governed by the sign of da/dt in Eq. (33). In the disc-driven regime the model sets T_grav ≈ T_OLR = −T_visc, so the binary always loses angular momentum to the disc and only the accretion torque T_acc can reverse the sign. This identification is the load-bearing step. T_OLR dominance is asserted from Eqs. (18)–(19) assuming a monotonic power-law density profile, but the paper's own caveat in Section 2.2 notes that a steep surface-density gradient can shift higher-order resonances so that the ILR dominates (Chen et al. 2020), and recent simulations find a positive gravitational torque on the binary (Muñoz et al. 2019; Duffell et al. 2020). If T_grav is not equal to −T_visc—or is positive—then thin discs need not contract, and the h_crit boundary and the derived merger rates (0.2–5 Gpc^-3 yr^-1) are not robust. The rest of the model (mass-transfer terms, GW inspiral) is internally consistent, but it inherits this unverified sign.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper develops a simple analytic model for the orbital evolution of stellar-mass black hole binaries embedded in AGN discs, including mass accretion onto the binary. From energy and angular momentum conservation with a presumed (m,l)=(2,1) outer Lindblad resonance torque balanced by the viscous torque, the authors derive coupled disc-plus-GW evolution equations for the semi-major axis and eccentricity (Eqs. 33-38), integrate them, and find a critical disc aspect ratio h_crit (roughly 0.04-0.16 depending on the accretion profile) separating orbital expansion in thicker discs from contraction and merger in thinner discs. They then compute merger timescales tau ~ 10^5-10^7 yr and GW merger rates R ~ 0.2-5 Gpc^-3 yr^-1. The quantitative predictions are forward-looking, and the caveats about the torque sign and the thin-disc accretion suppression are acknowledged in Sections 2.2 and 2.4.","tokens_in":19152,"tokens_out":17252,"duration_ms":163862,"significance":"If the central assumption on the sign of the gravitational torque is correct, the paper provides a useful transparent framework for a contested question and a falsifiable prediction: the existence of a critical aspect ratio and a non-monotonic merger-time dependence. It improves on earlier work by including mass accretion terms consistently in both the torque balance and the orbital evolution equations, and by following the coupled disc+GW evolution through merger. The rate estimates are explicit and can be compared with LVK observations. The main limitation is that the two load-bearing ingredients, OLR dominance (so that T_grav = -T_visc) and the 10 h_cbd accretion suppression, are empirical or assumed rather than derived, and the paper's own discussion cites simulations finding a positive gravitational torque. Because the model is analytic and the equations are explicit, these assumptions could in principle be tested in a revision; the current manuscript, however, does not quantify how its central conclusions depend on them.","major_comments":[{"comment":"The identification T_grav ≈ T_OLR = -T_visc is the load-bearing step in Eq. (33). It is introduced in Section 2.2 from the (m,l) = (2,1) outer Lindblad resonance dominance argument based on Eqs. (18)-(19), but the accompanying cautionary note states that a steep surface-density gradient can make the ILR dominate (Chen et al. 2020), and Section 7 itself cites simulations that find a positive net gravitational torque on the binary. If the gravitational torque is not equal to -T_visc, or if it is positive, then LJa can be positive even for thin discs, and the h_crit boundary and the rates in Sections 4-6 do not follow. The manuscript should quantify the range of T_grav/T_visc for which a critical aspect ratio still exists and state the conditions under which the central claim survives.","section":"Section 2.2 (Eqs. 18-19) and Section 3 (Eq. 33)"},{"comment":"The existence and value of h_crit depend on the thin-disc accretion suppression relation Ẍ_b = 10 h_cbd Ẍ_cbd (Eq. 32, from Ragusa et al. 2016). With this relation, T_visc and the positive accretion and mass terms in Eq. (33) scale as h_cbd^2 and h_cbd^3, respectively, so the crossing at h_crit is natural. If the suppression factor were instead constant (e.g., Ẍ_b = Ẍ_cbd), both sides of Eq. (33) would scale as h_cbd^2 and the contraction/expansion boundary would not depend on the aspect ratio, so the paper's central distinction between thin and thick discs would disappear. A sensitivity study with no suppression and with modified suppression factors, together with a justification for applying the SPH-based relation to AGN discs, is required before the h_crit values in Table 1 can be considered robust.","section":"Section 2.4 (Eqs. 26 and 32)"}],"minor_comments":[{"comment":"The symbol p is used both for the surface-density power-law index in Eq. (17) and for the radial accretion-rate index in Eq. (25); these are different quantities and the notation should be changed to avoid confusion.","section":"Section 2.2 and Section 2.4"},{"comment":"The merger time is defined as \"the point where the numerical solution approaches the abscissa axis\"; please specify a quantitative stopping criterion (e.g., a = 6GM/c^2 or a separation of a few Schwarzschild radii), because the quoted timescales depend on the cutoff.","section":"Section 5"},{"comment":"There are typographical issues: \"T able 1\" appears before Table 1, and \"L VK\" appears in Sections 7 and 8; these should read \"LIGO/Virgo/KAGRA\" or \"LVK\".","section":"Pages 7 and 11"},{"comment":"The caption of Fig. 1 does not define h_cbd, h_crit, or h_tr; please add definitions directly in the caption, since these quantities are central to the figure.","section":"Figure 1"},{"comment":"The discussion of reduced eccentricity growth mentions a factor-of-ten reduction but gives no corresponding merger-rate estimate; a brief rate estimate for that case would complete the parameter scan.","section":"Section 7"}],"recommendation":"major_revision","confidential_remarks":"The paper is within the scope of MNRAS and the derivation is transparent. My main concern is not internal consistency but the external validity of the sign of the gravitational torque and the accretion-suppression scaling; both are acknowledged by the authors but not tested. A revision that adds a sensitivity analysis over these two ingredients would make the central claim much stronger. I would not recommend rejection because the model can be repaired within its stated scope."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Short version: this is a useful analytic framework with a clear caveat at its load-bearing joint, and the authors know it. The coupled disc+GW evolution for accreting binaries is genuinely new relative to Hayasaki (2009), who included the accretion torque but dropped the mass-variation terms in the orbital element equations. Deriving Eqs. (7)–(8) from energy and angular momentum conservation is clean, and the final rates (0.2–5 Gpc^-3 yr^-1) are forward predictions, not fits to the observed rate. The qualitative result that thin discs contract and thick discs expand matches the Tiede et al. (2020) and Heath & Nixon (2020) simulations, and the hcrit values (0.04–0.16) sit in the right ballpark. Credit where due: the model is transparent, the parameter dependence is spelled out, and the merger timescales of 10^5–10^7 yr are a concrete quantitative output.\n\nThe soft spot is exactly where the stress-test note points. Everything hinges on Tgrav ≈ TOLR = −Tvisc. The authors justify OLR dominance through the torque ratios in Eqs. (18)–(19), but those ratios assume the corotation and ILR sites sit in a depleted cavity. The paper's own Section 2.2 cautionary note admits a steep surface-density gradient can shift higher-order resonances so that ILR dominates (Chen et al. 2020), and the recent simulations finding positive gravitational torques (Muñoz et al. 2019; Duffell et al. 2020) are cited but not incorporated. If the gravitational torque is not simply −Tvisc, the hcrit boundary and the derived rates lose their quantitative grounding. This is not a fatal flaw in the sense of an internal inconsistency; it is an unverified physical assumption. The authors are upfront about it, which is more than many papers manage, but it means the central number—0.2–5 Gpc^-3 yr^-1—should be read as conditional on the resonance picture.\n\nThe 10 hcbd accretion-suppression factor from Ragusa et al. (2016) is another empirical input with a factor-of-a-few uncertainty, and the rate normalization (nAGN, NBH, fd, fb) spans the usual order-of-magnitude range, so the rate estimate is not tight. But the paper does not oversell it; they call it conservative.\n\nWho is this for? People working on AGN disc channels for GW sources and anyone comparing analytic torque-balance models to hydrodynamic simulations. The paper deserves a serious referee: it is internally consistent, engages the debate honestly, and gives the community a simple model to test against simulations. My recommendation is to send it to review, with the request that the referee push for a parameter scan or a direct comparison with the positive-torque simulations in the discussion. The structure is sound; the caveat is already on the page.","headline":"Useful analytic framework whose central contraction/expansion boundary hangs on the (2,1) OLR dominance assumption—a load-bearing step the authors themselves flag, so the rates should be read as conditional.","tokens_in":19704,"tokens_out":2107,"would_cite":true,"duration_ms":21735,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"This paper claims that the disc aspect ratio of an AGN disc determines whether an accreting binary black hole contracts toward a gravitational-wave merger (thin discs) or expands away from it (thick discs).","keywords":["binary black holes","AGN discs","gravitational waves","disc-binary interaction","Lindblad resonances","accretion torque","merger timescales","merger rates"],"falsifier":"A hydrodynamical simulation with a self-consistently evolving binary orbit that scans disc aspect ratios from $h=0.01$ to $h=0.2$ would settle it: the time-averaged semi-major axis derivative must change sign near $h_{\\rm crit}\\sim0.04$–$0.16$, remaining negative below that value.","tokens_in":18616,"feed_emoji":"🌌","tokens_out":9128,"duration_ms":75505,"temperature":0.7,"pith_summary":"The paper tries to settle whether binary black holes embedded in the discs of active galactic nuclei shrink toward a gravitational-wave merger or drift outward. Its answer is that the disc aspect ratio, the disc height divided by radius, sets the outcome: below a critical value $h_{\\rm crit}\\sim0.04$–$0.16$ the binary contracts and merges, while above it the binary expands and no merger occurs. Because real AGN discs are often thin, the paper argues that a large fraction of such binaries can reach the gravitational-wave-driven regime, with merger timescales of $10^5$–$10^7$ years. It also estimates the resulting merger rate density, roughly $\\mathcal{R}\\sim(0.2$–$5)\\,\\mathrm{Gpc}^{-3}\\,\\mathrm{yr}^{-1}$, which would be a non-negligible part of the observed gravitational-wave event rate.","feed_headline":"Thin discs squeeze black-hole binaries toward merger","feed_subtitle":"A torque-balance model finds a critical disc height: thinner discs shrink the orbit; thicker discs expand it.","key_machinery":"The carrying object is the torque-balance equation $\\dot{L}_b = T_{\\rm grav}+T_{\\rm acc}$, with $T_{\\rm grav}$ approximated by the $(m,l)=(2,1)$ outer Lindblad resonance torque, taken as $T_{\\rm grav}\\approx T_{\\rm OLR}=-T_{\\rm visc}$ at the cavity inner edge, and $T_{\\rm acc}$ parametrized from accretion onto the mini-discs around each black hole. The disc aspect ratio $h=H/r$ enters because the binary accretion rate $\\dot{M}_b$ scales with the circumbinary disc aspect ratio, with thin discs suppressing accretion, so the positive accretion torque grows with $h$ while the negative viscous torque does not; the balance point defines $h_{\\rm crit}$. These torques feed a coupled set of 'disc+GW'-driven evolution equations for the semi-major axis $a$ and eccentricity $e$, which are integrated from the disc-dominated regime at large separations into the gravitational-wave-driven regime at small separations.","core_discovery":"The paper claims that the long-running disagreement over whether accreting binaries in AGN discs contract or expand is resolved by the disc aspect ratio: below a critical value $h_{\\rm crit}$ the negative viscous torque dominates over the positive accretion torque, so the binary shrinks into the gravitational-wave regime, whereas above $h_{\\rm crit}$ the accreted angular momentum wins and the binary expands and never merges. It further claims that contraction is usually accompanied by eccentricity growth in the disc-driven phase, which accelerates the subsequent gravitational-wave inspiral, and that this makes accreting binaries capable of merging faster than non-accreting ones, with a non-monotonic dependence on disc thickness. Quantitatively, the paper derives merger timescales of $\\tau_{\\rm merger}\\sim10^5$–$10^7$ years and a gravitational-wave merger rate density of roughly $0.2$–$5\\,\\mathrm{Gpc}^{-3}\\,\\mathrm{yr}^{-1}$ for this channel, which it presents as a conservative estimate of a few to tens of percent of the observed rate.","pith_inferences":["If the critical-aspect-ratio picture is right, gravitational-wave events from AGN discs should preferentially come from thin-disc environments, so the host AGN population should be biased toward radiatively efficient, low-aspect-ratio discs; this is a testable demographic prediction.","The same torque-balance machinery could be applied to unequal-mass or extreme-mass-ratio binaries, where the $(2,1)$ outer Lindblad resonance dominance is less secure and $h_{\\rm crit}$ would shift; scanning mass ratio in simulations would map that shift.","A discriminating observational test is the eccentricity of detected mergers: this model predicts noticeable eccentricity in the gravitational-wave band for the disc channel, so a clean measurement of eccentricity at low frequency would constrain the channel's contribution.","Because the paper finds that $h_{\\rm crit}$ depends on the accretion-rate profile power law $p$, the merger-rate estimate of $0.2$–$5\\,\\mathrm{Gpc}^{-3}\\,\\mathrm{yr}^{-1}$ should be read as sensitive to the outflow strength in real AGN discs."],"forward_implications":["Below the critical aspect ratio, accreting BBHs in AGN discs contract and reach the gravitational-wave-driven regime, while above it they expand and never merge, so the channel's contribution is confined to thin discs.","Typical merger timescales of $10^5$–$10^7$ years are short enough to occur within an AGN disc lifetime and several orders of magnitude shorter than a purely gravitational-wave inspiral in the field.","Accretion does not always slow the inspiral: because it can pump eccentricity through the dominant outer Lindblad resonance, accreting binaries can merge faster than non-accreting ones, with the fastest mergers occurring at a moderate aspect ratio $h_{\\rm tr}$ below $h_{\\rm crit}$.","The implied gravitational-wave merger rate density of roughly $0.2$–$5\\,\\mathrm{Gpc}^{-3}\\,\\mathrm{yr}^{-1}$ would be a non-negligible fraction of the observed binary black hole merger rate.","Even if eccentricity growth is strongly reduced, accreting binaries can still shrink and merge in thin discs, so the qualitative contraction result does not depend on the eccentricity-driving assumption."],"supporting_citations":[{"why":"First SPH simulations showing binary semi-major axis decay coupled with eccentricity growth, the contraction baseline the model builds on.","marker":"Artymowicz et al. 1991"},{"why":"Establishes that the (2,1) outer Lindblad resonance sets the cavity inner edge and dominates the disc-binary torque, the paper's key structural assumption.","marker":"Lubow & Artymowicz 2000"},{"why":"Linear resonance theory supplying the Lindblad and corotation torque expressions used to justify $T_{\\rm grav}\\approx T_{\\rm OLR}$.","marker":"Goldreich & Tremaine 1980"},{"why":"Provides the alpha-prescription for the viscous torque and surface density used to set $T_{\\rm visc}$.","marker":"Shakura & Sunyaev 1973"},{"why":"Supplies the gravitational-wave-driven equations for $\\dot{a}$ and $\\dot{e}$ that govern the small-separation regime in the coupled evolution.","marker":"Peters 1964"},{"why":"Earlier analytic model using the same accretion-torque parametrization, extended here by including the accretion mass terms in the orbital evolution equations.","marker":"Hayasaki 2009"},{"why":"Hydrodynamic simulations finding that accreting binaries gain net angular momentum and expand, the recent expansion claim the model must reconcile.","marker":"Muñoz et al. 2019"},{"why":"Hydrodynamic result that the net torque on fixed-orbit binaries is almost always positive, providing the strongest opposing baseline.","marker":"Duffell et al. 2020"},{"why":"2D isothermal simulations finding a transition from outspiral in thick discs to inspiral in thin discs, giving the $h\\sim0.04$ comparison value for $h_{\\rm crit}$.","marker":"Tiede et al. 2020"},{"why":"3D live-binary simulations showing contraction in thin discs and a parameter-dependent critical aspect ratio near 0.1, supporting the paper's range for $h_{\\rm crit}$.","marker":"Heath & Nixon 2020"},{"why":"3D SPH result that accretion onto the binary is suppressed in thin discs, used as the $\\dot{M}_b=10h_{\\rm cbd}\\dot{M}_{\\rm cbd}$ scaling that makes thin discs contract.","marker":"Ragusa et al. 2016"}],"fun_headline_variants":["Thin discs drive black-hole binaries into faster mergers","Disc thickness controls black-hole binary contraction or expansion","Critical disc height decides if black-hole binaries merge","Thin AGN discs accelerate black-hole binary mergers","Black-hole binary expansion vs contraction: disc height is key"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the $(2,1)$ outer Lindblad resonance dominates the disc-binary torque, so the net disc torque is just the viscous torque plus the accretion torque, with the paper's own caution that a different resonance balance or cavity density profile could change the torque direction.","fun_headline_variants_meta":{"raw":{"variants":["Thin discs drive black-hole binaries into faster mergers","Disc thickness controls black-hole binary contraction or expansion","Critical disc height decides if black-hole binaries merge","Thin AGN discs accelerate black-hole binary mergers","Black-hole binary expansion vs contraction: disc height is key"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000206,"raw_usage":{"total_tokens":1455,"prompt_tokens":1060,"completion_tokens":395,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":676,"completion_tokens_details":{"reasoning_tokens":320}},"tokens_in":676,"tokens_out":395,"duration_ms":3992,"temperature":1.0,"reasoning_tokens":320,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-12T00:01:49.305644+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A hydrodynamical simulation with a self-consistently evolving binary orbit that scans disc aspect ratios from $h=0.01$ to $h=0.2$ would settle it: the time-averaged semi-major axis derivative must change sign near $h_{\\rm crit}\\sim0.04$–$0.16$, remaining negative below that value.","supporting_citations":[{"cited_title":"J., Lubow S","cited_arxiv_id":null,"evidence_quote":"First SPH simulations showing binary semi-major axis decay coupled with eccentricity growth, the contraction baseline the model builds on."},{"cited_title":"H., Artymowicz P., 2000, in Mannings V., Boss A","cited_arxiv_id":null,"evidence_quote":"Establishes that the (2,1) outer Lindblad resonance sets the cavity inner edge and dominates the disc-binary torque, the paper's key structural assumption."},{"cited_title":"C., 1964, Physical Review, 136, 1224","cited_arxiv_id":null,"evidence_quote":"Supplies the gravitational-wave-driven equations for $\\dot{a}$ and $\\dot{e}$ that govern the small-separation regime in the coupled evolution."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Earlier analytic model using the same accretion-torque parametrization, extended here by including the accretion mass terms in the orbital evolution equations."},{"cited_title":"C., D'Orazio D., Derdzinski A., Haiman Z., MacFadyen A., Rosen A","cited_arxiv_id":null,"evidence_quote":"Hydrodynamic result that the net torque on fixed-orbit binaries is almost always positive, providing the strongest opposing baseline."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"2D isothermal simulations finding a transition from outspiral in thick discs to inspiral in thin discs, giving the $h\\sim0.04$ comparison value for $h_{\\rm crit}$."},{"cited_title":"M., Nixon C","cited_arxiv_id":null,"evidence_quote":"3D live-binary simulations showing contraction in thin discs and a parameter-dependent critical aspect ratio near 0.1, supporting the paper's range for $h_{\\rm crit}$."},{"cited_title":"J., 2016, , 460, 1243","cited_arxiv_id":null,"evidence_quote":"3D SPH result that accretion onto the binary is suppressed in thin discs, used as the $\\dot{M}_b=10h_{\\rm cbd}\\dot{M}_{\\rm cbd}$ scaling that makes thin discs contract."}],"review_version":1}