{"id":"8c900d33-2784-4e3d-8de9-4c3644b5e01d","arxiv_id":"2603.01352","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":10,"one_line_summary":"CO(11-10)/CO(7-6) excitation and kinematic modeling imply a ~2e8 Msun black hole accreting at several times the Eddington limit inside the obscured nucleus of W2305-0039 at z=3.111.","lead":"Ultra-sharp ALMA images of a dusty galaxy at z=3.1 reveal a compact patch of very hot carbon monoxide, pointing to X-ray heating by a buried black hole. Gas motions imply a black hole of about 200 million suns, and the galaxy shines roughly 20 times brighter than that black hole should allow.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The reported dynamical black-hole mass may be prior-dominated: with the sphere of influence unresolved and the velocity gradient weak, the PV model likely cannot uniquely separate a point mass from the extended disk, so the λ_Edd≥4 lower bound is not robust.","rationale":"The reader's weakest_assumption identifies the unresolved sphere of influence and model degeneracy as the key vulnerability. I agree that this is the most load-bearing concern because the super-Eddington quantitative claim λ_Edd≥4 depends directly on the upper bound of MBH, and the upper bound is model-dependent. I considered whether the SED-based L_AGN could be a larger concern, but the paper adopts a conservative systematic error, and even a 0.3-dex reduction in L_AGN would still yield λ_Edd>1. The XDR-vs-PDR argument supports the AGN interpretation but is not necessary for the Eddington-ratio calculation. The mock-recovery test is the decisive check: if the PV modeling cannot recover injected MBH values, then the reported mass and Eddington ratio are not physically meaningful. The reader already assigned CONDITIONAL; my analysis reinforces this without changing the verdict, hence UNCHANGED.","tokens_in":16497,"tokens_out":9689,"duration_ms":99677,"concrete_test":"Run a mock-recovery experiment: generate synthetic ALMA PV cubes with the same uv-coverage, beam, noise correlation, and assumed disk parameters as 2024.1.01175.S, but with input black-hole masses spanning log(MBH/Msun) = 7.0, 8.3, and 9.5. Fit each mock with the same KinMS+emcee pipeline and priors. If the recovered 68% credible intervals on MBH overlap substantially or do not track the input values, the data are not sensitive to MBH and the quoted constraint is prior-dominated, invalidating the λ_Edd≥4 claim. Also re-fit the real data without the [CI]-based priors on M_disk and R_e_disk and with a Hernquist stellar cusp added; if the 68% upper bound on MBH exceeds 10^9.1, the λ_Edd≥4 lower bound fails.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim that W2305−0039 accretes at λ_Edd≥4 rests on the dynamical MBH constraint of log(MBH/Msun)=8.3^{+0.7}_{-0.6} from Section 4.2. This constraint is not secure. The synthesized beam (230 pc) is roughly 20× the expected sphere of influence (~10 pc for MBH~2e8 Msun and σ~277 km/s), so the PV diagram is sensitive to the total mass distribution within the beam, not to a central point mass. The CO(11–10) kinematics are dispersion-dominated (σ=277 km/s), and the ordered rotation is described as a 'weak velocity gradient' (Section 4.1). In such a regime, a point-mass contribution is easily degenerate with a compact exponential disk, a stellar cusp, or non-circular motions. The model assumes a two-component axisymmetric rotator (point mass + exponential disk) with Gaussian priors on M_disk and R_e_disk derived from [CI], and no stellar mass term. If a stellar cusp or a different mass distribution is present, the fitted 'MBH' is a generic central mass concentration, and the upper bound that drives the λ_Edd lower bound is not a reliable upper limit on the black-hole mass. The paper's own inclination stress test shifts the 84th-percentile upper bound to log MBH~9.4, decreasing λ_Edd to ~2—still super-Eddington but no longer ≥4. Relaxing disk-mass priors or adding a stellar component could push this upper bound even higher, potentially eroding even the stronger super-Eddington claim. Thus the quantitative headline λ_Edd≳4 is not robust unless the PV model demonstrably recovers input MBH values in controlled tests.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper presents ALMA 0.03\" (~230 pc) observations of the Hot DOG W2305−0039 at z=3.111, targeting CO(7–6) and CO(11–10), plus [CI] and continuum. Visibility-plane modeling shows the CO(11–10) emission is compact (Re≈173 pc) and the CO(11–10)/CO(7–6) ratio rises above unity in the central kiloparsec; the authors argue this ratio can be reproduced by XDR models but not by PDR models. Forward modeling of the CO(11–10) position–velocity diagram with an axisymmetric disk plus a central point mass yields log(MBH/M⊙)=8.3^{+0.7}_{−0.6} and σgas≈277 km/s. Combining this with L_AGN≈10^47.7 erg/s from SED decomposition leads to λ_Edd≳4, interpreted as super-Eddington accretion in a heavily obscured nucleus.","tokens_in":16979,"tokens_out":6820,"duration_ms":67371,"significance":"If the dynamical black-hole mass and the Eddington ratio are correct, this would be a rare direct, dust-insensitive constraint on a hyperluminous obscured AGN at z~3, supporting theories of rapid obscured black-hole growth. The paper has clear strengths: visibility-plane fitting to separate compact and extended components, explicit treatment of correlated noise in the PV modeling, use of public tools (KinMS, emcee), and an honest stress test varying the inclination prior. These are valuable methodological contributions. However, the central λ_Edd≳4 claim rests on a dynamical MBH estimate that is largely prior-driven and not uniquely determined by the data, given that the sphere of influence is unresolved by a factor of ~20 and the kinematics are dispersion-dominated with a weak velocity gradient. The quoted uncertainties do not include the dominant systematic terms.","major_comments":[{"comment":"The black-hole mass estimate is not a unique dynamical measurement because the sphere of influence is unresolved. The synthesized beam is ~230 pc (0.03\"), whereas for log MBH≈8.3 and σgas≈277 km/s the expected sphere of influence is ~10 pc. The PV diagram is therefore sensitive to the total mass within the beam, not to a central point mass. The model's point-mass term is degenerate with a compact exponential disk, a stellar cusp, or non-circular motions, especially since the kinematics are dispersion-dominated and the ordered rotation is described as a weak gradient (§4.1). The quoted asymmetric errors are statistical only; no systematic term accounts for the mass-distribution uncertainty. The result should be phrased as a central mass concentration, and the MBH/λ_Edd claims need to be correspondingly weakened.","section":"Section 4.2, Fig. 5"},{"comment":"The disk-mass prior is informative and not conservative. The Gaussian prior on M_disk is centered on the [CI]-inferred gas mass (§3.4), which assumes LTE, optically thin [CI], Tex=100 K, and a carbon abundance of 8×10^-5. This prior—and the absence of any stellar component—directly controls the fitted MBH: if the true central mass distribution is more massive (e.g., a stellar cusp of ~10^9–10^10 M⊙ or a higher gas mass), MBH would be correspondingly lower. The posterior for log Mdisk ≈9.7±0.2 is essentially the prior, not a data-driven constraint. The inclination stress test explores only one of the model uncertainties; relaxing the disk-mass prior or adding a stellar component could push the MBH upper bound above log MBH≈9.4. The conclusion that λ_Edd≳4 is therefore not robust until these alternative assumptions are tested.","section":"Section 4.2, Fig. 7"},{"comment":"The λ_Edd calculation uses only the statistical uncertainty on MBH and the adopted ±0.1 dex on L_AGN. Sun et al. (2024) quote a fractional systematic error of 0.33 dex in the BayeSED framework, which the paper acknowledges but does not propagate into the Eddington-ratio lower bound. Under the paper's own inclination stress test, log MBH≈9.4 gives λ_Edd≈2, and with additional systematic uncertainties in L_AGN and in the mass model, the lower bound could fall further. The abstract's headline 'λ_Edd≳4' is thus not supported; the strongest defensible statement at this point is that the source is plausibly super-Eddington (λ_Edd≳1–2), not that it exceeds 4.","section":"Section 5"},{"comment":"The excitation diagnostic compares only PDR and XDR models from galaxySLED. Mechanical heating by turbulence, shocks, or cosmic rays is not included, yet the nuclear gas is dispersion-dominated (σgas≈277 km/s) and likely hosts shocks, making such heating plausible. The statement that XDR models 'successfully reproduce' the ratio while PDR models 'fail' is therefore an overclaim: the ratio is consistent with XDR but does not uniquely require X-ray irradiation. This does not directly undermine the dynamical MBH inference, but it weakens the AGN-heating interpretation used to connect the CO excitation to the accretion state.","section":"Section 3.3"}],"minor_comments":[{"comment":"The 1D posterior label '8.3+0.7 □0.6' appears to have a missing minus sign; it should read '8.3^{+0.7}_{-0.6}'.","section":"Figure 7"},{"comment":"The circularized Re for CO(11–10) is quoted as 173±30 pc; make explicit that this is the circularized radius derived from the axis ratio q=0.76.","section":"Abstract / Table 2"},{"comment":"The statement that the extended-disk gas mass is a lower limit depends on the assumed abundance; the same caveat should be stated symmetrically for the compact-disk mass estimate.","section":"Section 3.4"},{"comment":"The claim that elevated velocity dispersion across multiple beams rules out beam smearing as the primary cause should be supported by a beam-convolved model comparison, since beam smearing affects the velocity gradient and dispersion in a coupled way.","section":"Section 4.1"}],"recommendation":"major_revision","confidential_remarks":"The manuscript is well-written, uses appropriate public tools, and the data are of high quality, but the headline MBH/λ_Edd result exceeds what the data can currently support. The reader's stress-test concern is on target: the MBH estimate is prior-dominated, the sphere of influence is unresolved, and the quoted uncertainties ignore the dominant systematics. I would recommend major revision with a request to reframe the MBH as a central mass concentration, to test alternative mass distributions (including a stellar component and a wider disk-mass prior), and to soften the Eddington-ratio claim to a robust super-Eddington/quasi-super-Eddington statement rather than λ_Edd≳4."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The data are the real thing. These are the first 0.03″ CO(11–10) and CO(7–6) maps of a Hot DOG, and the visibility-plane work is careful: the compact CO(11–10) component (Re≈173 pc), the even more concentrated 1.0 mm continuum (Re≈77 pc), and the steep central rise in CO(11–10)/CO(7–6) all hang together. The PDR/XDR comparison is persuasive—extreme PDR models cannot reach ratio ~1, while XDR models can with FX≳1 erg s−1 cm−2. That part of the paper deserves to be published.\n\nThe soft spot is exactly where the reader puts it: the black-hole mass. The paper itself admits the velocity gradient is weak and the kinematics are dispersion-dominated, and the beam is ~20× the nominal sphere of influence. So the PV model is really measuring a point-mass contribution inside the beam, not resolving the black hole. The model uses Gaussian priors on disk mass and radius from [CI], includes no stellar mass term, and the quoted errors are statistical only. The MBH–inclination degeneracy is acknowledged and the stress test without the sin(i) prior raises the 84th percentile to log MBH≈9.4—which still gives λ_Edd≈2, but that test does not touch the disk-mass priors or the missing stellar component. The λ_Edd≥4 lower bound is therefore not robust; the weaker claim of super-Eddington accretion is defensible but still assumption-dependent. The XDR interpretation also omits a shock-heating baseline, though that is a minor caveat given the contrast with PDR models.\n\nThe paper is honest about its limitations—it explicitly flags the weak gradient and looks ahead to long-baseline ALMA for resolving the sphere of influence. The reasoning is clear and the source is rare and important. What it needs is a systematic-error treatment: propagate the [CI] abundance and excitation-temperature uncertainties, run a model with a stellar cusp, and test whether the posterior recovers known input MBH values in controlled kinematics.\n\nRecommendation: send it to a serious referee. The excitation result and the dataset justify the time, even if the dynamical-MBH headline will need revision in revision.","headline":"Genuinely new high-resolution ALMA data and a compelling XDR excitation signature; the dynamical black-hole mass is plausible but not secure enough to carry the λ_Edd≳4 headline.","tokens_in":17489,"tokens_out":2076,"would_cite":true,"duration_ms":24155,"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":"A black hole in a dusty galaxy at z=3.1 is swallowing matter more than four times faster than the Eddington limit, a new dynamical measurement suggests.","keywords":["high-redshift galaxies","dust-obscured galaxies","hot dust-obscured galaxies","supermassive black holes","Eddington ratio","molecular gas kinematics","CO line excitation","X-ray-dominated regions"],"falsifier":"A sub-0.005-arcsecond (~40 pc) observation of CO(11-10) or another dense-gas tracer that resolves the nuclear rotation curve: if the velocity field shows that the central mass is not point-like (e.g., a spatially extended stellar cusp) or is dominated by outflows/non-circular motions, the derived black-hole mass and the super-Eddington conclusion would be falsified.","tokens_in":16379,"feed_emoji":"🕳️","tokens_out":11413,"duration_ms":97527,"temperature":0.7,"pith_summary":"The paper tries to establish that the black hole at the center of W2305-0039, a hyperluminous dust-obscured galaxy at z=3.111, is accreting matter at a rate well above the Eddington limit. It does this in two steps: first, high-resolution CO line observations show that the gas within ~500 pc is excited by X-rays from an obscured active nucleus; second, forward modeling of the CO kinematics yields a dynamical black-hole mass of log(M/Msun)=8.3. Combined with the infrared-derived AGN luminosity, this implies an Eddington ratio of at least 4, meaning the black hole doubles its mass in a few million years. A sympathetic reader would care because this is one of the first dust-insensitive, dynamical black-hole mass measurements in an obscured high-redshift galaxy, and it directly supports the idea that super-Eddington accretion powers the fastest growth of early supermassive black holes.","feed_headline":"Black hole at z=3.1 grows at 4× the Eddington limit","feed_subtitle":"CO gas motions weigh a buried black hole and reveal accretion far above the Eddington limit.","key_machinery":"Two coupled measurements carry the argument. First, the radial CO(11-10)/CO(7-6) luminosity ratio, derived from visibility-plane two-disk fits, serves as an excitation diagnostic: X-ray-dominated region (XDR) calculations reproduce the steep rise within ~500 pc, while photodissociation-region models fail, identifying the central engine as an X-ray-luminous AGN. Second, a forward kinematic model of the position-velocity diagram treats the CO(11-10) gas as an axisymmetric rotating disk — an exponential surface-density distribution plus a central point mass — with [CI]-based Gaussian priors on the disk mass and size, and an MCMC fit to the observed PV diagram yields the black-hole mass and intr","core_discovery":"The central claim is that the molecular gas in the nucleus of W2305-0039 is being irradiated by a deeply obscured active galactic nucleus, and that the same gas moves in a dispersion-dominated disk whose weak ordered rotation carries the dynamical signature of a central point mass of about 2×10^8 solar masses. Because the galaxy's AGN luminosity is about 6×10^47 erg/s, the resulting Eddington ratio is ≥4, so the black hole is growing far above the classical Eddington limit. The evidence for the AGN is the CO(11-10)/CO(7-6) line ratio, which exceeds unity within 500 pc and can only be reproduced by X-ray-dominated region models; the evidence for the black-hole mass is a forward model of the p","pith_inferences":["If this measurement survives higher-resolution follow-up, single-epoch virial black-hole masses for hot dust-obscured galaxies may need systematic downward revision: the dynamical mass here is ~0.5-1 dex below typical virial estimates, which would shift inferred Eddington-ratio distributions toward even more extreme values.","The same two-step recipe — high-J CO excitation ratio plus PV forward modeling — can be applied to fainter or lower-redshift obscured AGN where the sphere of influence is better resolved, making this a general black-hole-mass method rather than a single-object curiosity.","A decisive but untested alternative is that the central point mass is not a black hole but an unresolved stellar cusp or a nuclear star cluster; the current 230 pc beam cannot separate these, so the super-Eddington conclusion would collapse if a stellar mass component were shown to dominate the central potential.","The XDR interpretation itself could be probed by mapping additional CO transitions: a full high-J CO SLED spanning J=6 to J=13, in combination with the existing ratio, would either confirm or falsify the AGN-heating scenario independent of the kinematics."],"forward_implications":["If the derived black-hole mass is correct, W2305-0039 is accreting at λ_Edd ≥ 4, with a mass-doubling time of only a few million years — fast enough to build a 10^9 solar-mass black hole by z≈6.","The high intrinsic gas velocity dispersion (~280 km/s) implies the nuclear disk is pressure- and turbulence-supported, consistent with the feedback expected from super-critical accretion.","The steep CO(11-10)/CO(7-6) ratio in the central 500 pc demonstrates that high-J CO emission is a practical, dust-insensitive diagnostic of buried AGN activity, locating the nucleus even when optical/UV tracers are obscured.","The stream-like CO(7-6) residuals after subtracting the two-disk model suggest that gas inflow from a larger-scale reservoir can sustain the compact nuclear disk for tens of millions of years at the present accretion rate.","Resolving the black-hole sphere of influence with longer-baseline observations would turn the current point-mass constraint into a unique black-hole mass measurement and directly test the super-Eddington interpretation."],"fun_headline_variants":["ALMA's 230-pc view catches black hole eating beyond Eddington","Super-Eddington black hole found in dusty z=3.1 galaxy","CO gas motions weigh black hole at 10^8.3 Msun","Obscured AGN's X-rays light up CO, hint at fast growth","Dynamical mass of black hole shows 4× Eddington accretion"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The kinematic model assumes the CO(11-10) gas forms an axisymmetric rotating disk in equilibrium, and that the fitted central point mass is the black hole — but with a ~230 pc beam that is about 20 times larger than the expected sphere of influence, any non-circular motion, an unresolved stellar cusp, or a different mass distribution could mimic the black-hole term.","fun_headline_variants_meta":{"raw":{"variants":["ALMA's 230-pc view catches black hole eating beyond Eddington","Super-Eddington black hole found in dusty z=3.1 galaxy","CO gas motions weigh black hole at 10^8.3 Msun","Obscured AGN's X-rays light up CO, hint at fast growth","Dynamical mass of black hole shows 4× Eddington accretion"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000448,"raw_usage":{"total_tokens":2166,"prompt_tokens":881,"completion_tokens":1285,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":625,"completion_tokens_details":{"reasoning_tokens":1182}},"tokens_in":625,"tokens_out":1285,"duration_ms":12973,"temperature":1.0,"reasoning_tokens":1182,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-02T19:39:30.803783+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A sub-0.005-arcsecond (~40 pc) observation of CO(11-10) or another dense-gas tracer that resolves the nuclear rotation curve: if the velocity field shows that the central mass is not point-like (e.g., a spatially extended stellar cusp) or is dominated by outflows/non-circular motions, the derived black-hole mass and the super-Eddington conclusion would be falsified.","supporting_citations":[],"review_version":1}