{"id":"d81fb074-ef6d-4a2d-bba3-a208bcd82738","arxiv_id":"2504.18726","paper_version":1,"verdict":"CONDITIONAL","confidence":"HIGH","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":4,"one_line_summary":"Modeling rotation curves of CO gas in ten exoALMA disks yields dynamical disk masses, scale radii, and effective viscosities, plus a gas-to-dust ratio around 400.","lead":"This paper measures the masses and sizes of ten planet-forming disks by modeling how fast their gas rotates around the central star. It finds that the disks are mostly gravitationally stable and that their gas-to-dust ratios appear unusually high, which affects how we think disks evolve and build planets.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The disk-mass claim hinges on the assumed surface-density shape (γ=1); the paper's defense that masses are unaffected is cited to a companion preprint, and the admitted underestimation of uncertainties could change which sources clear the 5% threshold. A γ-marginalized refit is the decisive check.","rationale":"The paper is careful: it propagates thermal-structure uncertainties by drawing from posteriors, cross-checks dynamical masses against independent HD/N2H+/CO-based estimates for DM Tau, HD 34282, and LkCa15, and uses the stratified rotation-curve model from Lodato et al. (2023) and Martire et al. (2024). The identification of the weakest point in the reader's verdict is sound: the fixed γ=1 surface-density shape is the most consequential modelling assumption, because the non-Keplerian signal that carries the disk-mass information is a small residual, and the shape of the pressure-gradient term directly depends on γ. The paper's reply to this objection is a citation to a companion preprint, and Appendix B simultaneously concedes that uncertainties are underestimated. Given that several sources cluster near the 5% detection threshold, this is not a cosmetic issue: an unmodelled γ bias or an underestimated error could change the sample of secure detections. The proposed refit with γ free is a direct, computationally feasible test that would either confirm the robustness claim or expose a systematic bias. I therefore agree with the reader's conditionality and see no reason to move the verdict; the concern is substantial but addressable. The reader's weakest assumption and my load-bearing concern coincide, and the proposed test is the natural way to settle it.","tokens_in":24200,"tokens_out":11273,"duration_ms":114650,"concrete_test":"Re-run the complete DySc pipeline on all ten sources with γ as a free parameter (e.g., uniform prior 0 ≤ γ ≤ 2), using the same data, thermal posterior draws, and priors as in §3.2. For each source, compare the marginalized posterior median and 95% credible interval for Md/M* to Table 1. The γ=1 assumption is non-load-bearing if, for every source claimed above the 5% threshold, the median Md/M* shifts by less than half the reported 1σ uncertainty and the 95% posterior still excludes 0.05 (excluding AA Tau). If any source crosses the 0.05 boundary, or shifts by more than the quoted σMd/Md, the central claim is falsified and the disk masses must be reported with γ-marginalized uncertainties.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim is that the self-gravitating contribution is measured and yields trustworthy Md values. In Eq. (A13) the non-Keplerian part of the rotation curve is the sum of a pressure term, proportional to [γ′ + (2−γ)(R/Rc)^(2−γ) − d log fg/d log R] times (H/R)^2_mid f(R,z), and a self-gravity term v_d^2. With γ fixed to 1, the pressure term's radial shape is fixed. If the true Σ(R) differs (e.g., γ = 0.5 or 1.5), the pressure-gradient profile changes; because both terms are small corrections to Keplerian rotation, the fit can absorb part of that difference into Md or M*. Appendix B explicitly admits this choice 'underestimates the true uncertainties' and 'introduces a potential bias on the scale radius', but asserts that disk and stellar masses are not affected, citing Andrews et al. 2024 (an arXiv preprint, not demonstrated here). Since the sample classification uses Md/M* > 0.05, and several claimed detections (J1615, J1842, SY Cha) sit near 0.07–0.10 with quoted uncertainties of 15–30%, an underestimated systematic of this size could move sources across the detection threshold. Substructures, which drive 50–70 m/s perturbations, are a second source of comparable-magnitude distortion and are tested only for LkCa15. Thus, within the present paper, the robustness of the disk-mass measurement to the assumed surface-density shape is not established.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper models the azimuthally averaged rotation curves of 12CO and 13CO for ten exoALMA disks using the vertically stratified model of Lodato et al. (2023) and Martire et al. (2024), fitting the stellar mass, disk mass, and scale radius from the combined pressure-gradient and self-gravity signatures. The surface density is assumed to follow the Lynden-Bell & Pringle self-similar form with the power-law index fixed to gamma=1. From the fitted parameters the paper derives disk-to-star mass ratios, Toomre Q profiles, gas-to-dust ratios using the continuum dust masses of Curone et al. (2025), flux-radius comparisons, and effective alpha_S values. AA Tau is excluded from the statistical analysis because of diffuse-backside contamination. Seven of the ten sources are claimed to have Md/M* above the 5% detection threshold, and the dynamical masses for DM Tau, HD 34282, and LkCa15 are compared with independent literature estimates.","tokens_in":24527,"tokens_out":8871,"duration_ms":88191,"significance":"If the derived masses are robust, this is a valuable dynamical census: the method is chemistry-free, the sample is well characterized, the DySc code is public, and the propagation of thermal-structure systematics through 100 posterior draws is a clear improvement over earlier work. The few-percent agreement of the dynamical stellar masses with discminer results and the consistency with independent disk-mass estimates for three sources provide genuine supporting evidence. However, the headline quantitative claims, including the average gas-to-dust ratio of about 400 and the number of secure disk-mass detections, currently rest on an assumed surface-density shape and on fiducial masses for systems below the detection threshold, so the conclusions are important but conditional.","major_comments":[{"comment":"The robustness of the disk masses to the assumed surface-density slope gamma is asserted but not demonstrated in this paper. The text fixes gamma=1 and states that this choice \"underestimates the true uncertainties\" and \"introduces a potential bias on the scale radius\", referring to Andrews et al. (2024) for the claim that the disk and stellar masses are unaffected. Since the pressure-gradient term in Eq. (A13) contains gamma both through gamma' and through (2-gamma)(R/Rc)^(2-gamma), changing gamma changes the radial profile of the non-Keplerian correction, and the three fitted parameters (M*, Md, Rc) can partially absorb that change. With the detection classification set at Md/M*>0.05 and with J1615, J1842, and SY Cha at 0.07-0.10 with 15-30% errors, an unquantified systematic of this size is load-bearing. Please add gamma as a free parameter, or at minimum run the full sample with gamma=0.5 and gamma=1.5 and show the resulting distributions of Md and Rc.","section":"Appendix B / Eq. (A13)"},{"comment":"The average gas-to-dust ratio of about 400 is computed using the fiducial best-fit disk masses for all sources except AA Tau, including the three sub-threshold systems J1852, PDS66, and V4046 Sgr, for which Section 4.2 states that Md=0.05M* may be used as an upper limit instead. The individual gas-to-dust values for these systems carry uncertainties of order 100%, for example 364+342/-357 for PDS66. The average should be recomputed for the secure detections alone, and the effect of treating the three sub-threshold masses as upper limits should be reported; otherwise the \"approximately 400\" claim in the abstract is not supported by the data.","section":"Section 4.2.1 / Table 1"},{"comment":"The substructure test is performed on only one disk, LkCa15, even though the sample is dominated by disks that show pressure-modulated substructures and the quoted 50-70 m/s perturbations are comparable in magnitude to the non-Keplerian signals being fitted. To support the statement that substructures do not bias the dynamical masses across the sample, add a similar test for at least one additional source with prominent substructure, or show that the LkCa15 result is representative by comparing the residual amplitudes across the sample.","section":"Section 2.2 / Section 3.2"}],"minor_comments":[{"comment":"The thermal-structure systematic is propagated from the Galloway-Sprietsma et al. (2025) posteriors, but Table 5 lists only the best-fit Dartois parameters without posterior widths; adding the widths would make the systematic treatment reproducible.","section":"Appendix B / Table 5"},{"comment":"The sentence \"We are able to parametrically describe their surface density\" overstates the result, since gamma is fixed to 1 and only a normalization (Md) and a scale (Rc) are fitted; the fitted family is a one-parameter subset of the self-similar solutions.","section":"Section 4.2"},{"comment":"The CO-depletion check based on Rosotti et al. (2025) is acknowledged to be non-independent of the dynamical masses; the text should present the Trapman et al. (2025) inference as the primary independent test in the main message rather than giving both equal weight.","section":"Section 4.3.1"},{"comment":"The assumption that the stellar accretion rate equals the disk accretion rate at Rc is noted in the text but not propagated into the alpha_S uncertainties; since the accretion-rate uncertainty of 0.35 dex dominates the error budget, a sentence quantifying this limitation in Table 3 or the figure caption would help.","section":"Section 4.4"}],"recommendation":"major_revision","confidential_remarks":"The decisive revision is the gamma-marginalized fit or a full-sample stability test over a plausible range of gamma. If the masses shift materially for gamma in [0.5, 1.5], several conclusions, including the non-detection list and the gas-to-dust average, will need revision; I would be comfortable accepting the paper after that check is shown."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Longarini et al. fit the rotation curves of 12CO and 13CO for ten exoALMA disks using the stratified self-gravitating model from Lodato et al. (2023) and Martire et al. (2024), and get dynamical stellar masses, disk masses, and scale radii. Seven of the ten have Md/M* > 5%. That is a real result: first dynamical masses for these disks, a homogeneous sample, and when combined with MAPS it nearly triples the number of such measurements.\n\nThe paper does several things right. The disk-mass cross-checks for DM Tau, HD 34282, and LkCa15 line up with independent estimates. Propagating the thermal-structure systematics by re-running fits on the posterior is proper. They are candid about AA Tau being problematic and about dust masses being underestimated in the gas-to-dust ratio.\n\nThe main soft spot is the fixed surface-density slope, γ=1. They admit this underestimates uncertainties and biases Rc, but the claim that Md and M* are unaffected is cited to Andrews et al. (2024), a preprint. The three sources near the 5% detection threshold (J1615, J1842, SY Cha) have quoted uncertainties of 15–30%, so if γ variations shift masses at that level it could change which sources are counted as detections. A γ-marginalized refit, or at least a sensitivity test on a few disks, would settle this. Substructures are tested on only one source (LkCa15); the 50–70 m/s perturbations they quote are the same order as the non-Keplerian signal they are measuring, so this deserves more than a single check. Also, many of the inputs (temperature structures, emitting layers, CO depletion) come from companion papers with TBD references, which makes independent verification hard today.\n\nOne more thing: the abstract says the gas-to-dust ratio is ~400 and \"not statistically consistent with the standard value of 100, assuming optically thin dust emission.\" The text later says the optically thin assumption underestimates dust mass. So the ratio is likely an upper limit, and the abstract overreaches a bit.\n\nWho should read it: anyone working on disk masses, gravitational stability, or angular momentum transport in protoplanetary disks. It is a careful application of an established method to a valuable sample, and the central claims are plausible. The issues are addressable without changing the method.\n\nRecommendation: yes to peer review. A good referee should push on the γ dependence and the near-threshold sources. I would not desk-reject this; I'd ask for a sensitivity analysis before acceptance.","headline":"Convincing new dynamical disk masses for ten exoALMA disks, though the fixed surface-density slope (γ=1) and unpublished companion papers leave the error budget less airtight than the headline numbers suggest.","tokens_in":25303,"tokens_out":3094,"would_cite":true,"duration_ms":29604,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"A measurable self-gravity signal in CO rotation curves weighs the gas in ten protoplanetary disks and finds most are at least 5% of their star's mass.","keywords":["protoplanetary disks","rotation curves","dynamical disk masses","self-gravity","pressure gradient","gas-to-dust ratio","viscous evolution","exoALMA"],"falsifier":"Measure gas masses for the same ten disks with a tracer that avoids CO chemistry and dust optical-depth assumptions, such as hydrogen deuteride (HD) emission or N2H+ emission. If these independent masses cluster near the dust-based masses with gas-to-dust near 100, rather than matching the dynamical masses, the self-gravity interpretation of the rotation curves would be refuted.","tokens_in":23980,"feed_emoji":"🪐","tokens_out":11022,"duration_ms":97263,"temperature":0.7,"pith_summary":"The paper tries to show that the small departures of a protoplanetary disk's rotation curve from pure Keplerian motion carry enough information to weigh the disk and pin down its scale radius. It models the $^{12}$CO and $^{13}$CO rotation curves of ten disks from the exoALMA sample with a vertically stratified disk model that includes both the pressure gradient and the disk's self-gravity, obtaining dynamical disk masses for all ten. Seven sources land above the $M_d/M_\\star = 0.05$ threshold where the method is reliable. The resulting masses imply an average gas-to-dust ratio of roughly 400 rather than the standard 100, and effective viscosity parameters $\\alpha_S$ spanning $10^{-5}$ to $10^{-2}$.","feed_headline":"Rotation curve fits weigh ten disks, most above 5% of stellar mass","feed_subtitle":"Self-gravity and pressure imprints in the gas motion yield masses, sizes, gas-to-dust near 400, and viscosity.","key_machinery":"The central object is the rotation-curve model for a vertically stratified disk, which decomposes $v_\\phi^2$ into the stellar Keplerian term, a finite-height correction, the pressure-gradient term, and a self-gravity integral $v_d^2$ (Eq. A13). The disk surface density is assumed to follow the self-similar Lynden-Bell & Pringle profile with power-law index fixed at $\\gamma=1$, and the two-dimensional temperature structure follows the Dartois prescription with parameters fixed from companion thermal fits to the same sources. The fitting code varies stellar mass, disk mass, and scale radius simultaneously against the $^{12}$CO and $^{13}$CO rotation curves, propagating thermal-structure uncertainties by repeating each fit 100 times.","core_discovery":"The paper's central claim is that the self-gravitating contribution to the gravitational potential is directly measurable in CO rotation curves, so dynamical disk masses can be derived without assuming anything about chemistry or dust opacity. For the ten exoALMA disks analyzed, the best-fit disk masses range from about $0.04\\,M_\\odot$ to $0.16\\,M_\\odot$, with seven disks above the $5\\%$ disk-to-star mass detection threshold. Combined with the fitted scale radii, these masses show all of the disks to be gravitationally stable (Toomre $Q>1$), imply an averaged gas-to-dust ratio near 400 under the optically thin dust assumption, and, together with accretion rates, give effective $\\alpha_S$ values between $10^{-5}$ and $10^{-2}$.","pith_inferences":["If the low-$\\alpha_S$ end of the inferred range is real, those disks cannot be transporting angular momentum primarily through turbulence, making non-turbulent mechanisms such as magnetically launched winds a more natural explanation; the paper does not draw this conclusion.","A testable extension would be to check whether the fitted scale radius $R_c$ coincides with the outermost dust substructure in each disk, which would directly test the claim that pressure-modulated substructures are responsible for the small dust radii.","If the gas-to-dust ratio near 400 is due to optically thick dust rather than genuinely high gas content, modeling the continuum with optical-depth effects at multiple wavelengths should raise the estimated dust masses and bring the ratio closer to 100."],"forward_implications":["Seven of the ten exoALMA disks have $M_d/M_\\star > 0.05$, so the sample now provides robust dynamical disk masses that straddle the method's detection threshold.","All disks in the sample are gravitationally stable with Toomre $Q>1$, consistent with the absence of prominent spiral structure in these systems.","The average gas-to-dust ratio near 400, combined with optically thin dust masses, implies that either dust masses are systematically underestimated or the disks genuinely have elevated gas-to-dust ratios.","The effective $\\alpha_S$ values span four orders of magnitude, from $10^{-5}$ to $10^{-2}$, indicating that a single viscosity parameter does not describe angular-momentum transport across the sample.","The comparison of dynamical scale radii with flux-based radii suggests that substructures slow radial drift of dust and that CO depletion can reconcile observed CO radii with thermochemical model predictions."],"supporting_citations":[{"why":"Supplies the stratified rotation-curve model with vertical temperature structure that this paper applies to all ten sources.","marker":"Martire et al. (2024)"},{"why":"Provides the underlying self-gravity and pressure-gradient framework for rotation-curve modeling that the stratified model extends.","marker":"Lodato et al. (2023)"},{"why":"Defines the exoALMA sample and supplies the extracted rotation curves that constitute the data being fit.","marker":"Stadler et al. (2025)"},{"why":"Provides the two-dimensional temperature structures used to evaluate the pressure-gradient term and the emitting-layer heights.","marker":"Galloway-Sprietsma et al. (2025)"},{"why":"Supplies the rotation-curve extraction and emitting-layer heights through the kinematic extraction tool.","marker":"Izquierdo et al. (2025)"},{"why":"Establishes the 5% disk-to-star mass threshold that governs which of the derived masses are treated as secure detections.","marker":"Veronesi et al. (2024)"},{"why":"Quantifies the method's uncertainties and shows that the surface-density power-law index mainly biases the scale radius, not the masses.","marker":"Andrews et al. (2024)"},{"why":"Supplies the self-similar surface-density profile that the fits assume for every disk.","marker":"Lynden-Bell & Pringle (1974)"},{"why":"Provides the dust masses and dust continuum radii used for the gas-to-dust ratios and radius comparisons.","marker":"Curone et al. (2025)"},{"why":"Gives the theoretical relation between CO flux radius, disk mass, and scale radius used to infer CO depletion.","marker":"Trapman et al. (2023)"}],"fun_headline_variants":["Self-gravity in curves weighs ten disks, 7 over 5% star mass","Ten disks weighed: most top 5% of stellar mass, gas-to-dust ~400","Rotation curve fits: disk masses for ten, 7 > 5% stellar","Weighing disks with rotation: 7 of 10 exceed 5% star mass","Ten exoALMA disks: self-gravity sizes and masses, all stable"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The surface-density profile is assumed to be smooth and self-similar with its power-law index fixed at $\\gamma=1$, and if the true profile differs, the pressure-gradient and self-gravity contributions shift, biasing the derived scale radius and, to a lesser extent, the masses.","fun_headline_variants_meta":{"raw":{"variants":["Self-gravity in curves weighs ten disks, 7 over 5% star mass","Ten disks weighed: most top 5% of stellar mass, gas-to-dust ~400","Rotation curve fits: disk masses for ten, 7 > 5% stellar","Weighing disks with rotation: 7 of 10 exceed 5% star mass","Ten exoALMA disks: self-gravity sizes and masses, all stable"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000377,"raw_usage":{"total_tokens":2033,"prompt_tokens":1000,"completion_tokens":1033,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":616,"completion_tokens_details":{"reasoning_tokens":921}},"tokens_in":616,"tokens_out":1033,"duration_ms":9469,"temperature":1.0,"reasoning_tokens":921,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-16T10:11:48.956671+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure gas masses for the same ten disks with a tracer that avoids CO chemistry and dust optical-depth assumptions, such as hydrogen deuteride (HD) emission or N2H+ emission. If these independent masses cluster near the dust-based masses with gas-to-dust near 100, rather than matching the dynamical masses, the self-gravity interpretation of the rotation curves would be refuted.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Defines the exoALMA sample and supplies the extracted rotation curves that constitute the data being fit."},{"cited_title":"2025, , TBD","cited_arxiv_id":null,"evidence_quote":"Provides the dust masses and dust continuum radii used for the gas-to-dust ratios and radius comparisons."}],"review_version":1}