{"id":"4acb2b6f-811d-42ba-98f9-057adf18ceb5","arxiv_id":"1909.02031","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":6,"one_line_summary":"CO rovibrational line ratios in Herbig disks are explained by two distinct inner-disk structures: low gas-to-dust surfaces for compact emission, and large gas cavities with high gas-to-dust cavity walls for extended emission.","lead":"CO gas in the inner disks of young stars called Herbig Ae systems behaves in a way that depends on where it sits: close to the star it is warm and dim in a particular line ratio, farther out it is cool and bright. The authors use physical models to argue these two behaviors trace different inner-disk structures: normal dusty surfaces in one group, and big gas cavities with dust traps in the other.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The low-v2/v1 branch depends on an unmodeled thermal dissociation of CO at the inner rim; Appendix F concedes DALI likely overestimates rim CO, so the dichotomy's physical basis is not yet self-consistently established.","rationale":"The reader's weakest_assumption identifies exactly the same load-bearing step: the assumed thermal dissociation of CO at the inner rim, which is not reproduced by DALI and is flagged as uncertain in Appendix F. My independent reading of Secs. 4.2, 4.3 and Appendix F confirms that the fiducial models fail in a direction that is corrected by subtracting the rim, and that the correction is justified by chemistry the model itself does not compute. This is an addressable but real gap: the low-v2/v1 branch of the central dichotomy is not yet self-consistently established. The high-ratio branch also involves post-hoc choices (gas-to-dust > 10000, rounded edge, reduced scale height), but the inner-rim CO abundance is the most directly load-bearing assumption because it determines whether the best-fitting low-ratio models are physical or imposed. A targeted DALI/chemistry rerun with the high-temperature network, as proposed, would settle whether the assumption lands. The CONDITIONAL verdict therefore remains appropriate; no adjustment is needed.","tokens_in":36051,"tokens_out":3547,"duration_ms":41923,"concrete_test":"Run a DALI model with Rin = 0.4 AU and gas-to-dust = 100 in which the inner-rim chemistry is recomputed with the high-temperature network discussed in Appendix F (CO + H -> C + OH, OH + H -> O + H2, collisional H2 dissociation), with gas heating balanced self-consistently, then ray-trace v1 P(10) and v2 P(4) and measure v2/v1 from the rim component alone without manual subtraction. If the rim still produces v2/v1 > 0.2, the Sec. 4.3 subtraction is unjustified and the dichotomy needs a different physical origin; if the rim is CO-poor (N_CO < 1e16 cm^-2) and gives v2/v1 < 0.2, the central interpretation survives.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The paper's central dichotomy requires that CO is absent at and inside the dust sublimation rim in the low-v2/v1 sources. In Sec. 4.2, the fiducial DALI models fail to reproduce the observed low ratios and overproduce broad line flux; in Sec. 4.3 the authors rescue the match by manually subtracting the inner-rim contribution to the spectra. The physical justification, given in Sec. 5.1.1, is that gas near the sublimation radius is heated above 3000 K and kept atomic. But this is not a prediction of the model: Appendix F states that 'the CO abundance in the inner disk rim is probably overestimated by DALI' and that the gas temperature and chemistry there are 'very uncertain.' Fig. F.1 shows that the DALI kinetic network produces the molecular-to-atomic transition more slowly and at higher temperatures than equilibrium chemistry. If CO at the rim actually survives at columns large enough to emit, as the fiducial models suggest, then the rim subtraction is an imposed constraint rather than an inference, and the low-v2/v1 branch could instead be produced by other excitation or temperature structures. The high-ratio branch also depends on large gas-to-dust ratios and a cool cavity wall, but the weakest load-bearing step is the unsupported removal of the inner rim.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies CO rovibrational line emission in Herbig Ae disks, focusing on the observed dichotomy between the vibrational flux ratio v2/v1 and the inferred CO emitting radius RCO. It combines analytic LTE slab calculations, non-LTE RADEX slab models, and a grid of DALI thermo-chemical disk models with varying inner radius and gas-to-dust ratio, and compares simulated line profiles, fluxes, and radii directly to observed data. The authors conclude that broad CO lines with low v2/v1 originate from a warm (400–1300 K) inner disk surface with gas-to-dust ratio below 1000 and CO column below 1e18 cm^-2, with CO absent at the inner dust rim because of thermal dissociation; narrow lines with high v2/v1 require an inner cavity of at least 5 AU and a cool (100–300 K), high-column (N_CO > 1e18 cm^-2) molecular reservoir at a cavity wall with gas-to-dust ratio above 10000. The paper interprets these structures in terms of dust trapping and giant planets and provides simulated ELT-METIS images as observational predictions.","tokens_in":36331,"tokens_out":7068,"duration_ms":71845,"significance":"If the interpretation is correct, this paper would connect CO rovibrational excitation to inner-disk gas-to-dust structure, dust trapping, and cavity formation in Herbig systems, and would provide a useful observational diagnostic for disk substructures inside 10 AU. The methodological strengths are the clean analytic and RADEX slab analysis, the explicit comparison of model line profiles to the observed stacked spectra, the use of a full thermo-chemical code with an expanded CO model, and concrete predictions for ELT-METIS. However, the central physical mechanism—the absence of CO at the inner dust rim—is not self-consistently produced by the DALI models, and the quantitative gas-to-dust thresholds are effectively selected grid values rather than fitted constraints. The paper therefore currently has the status of a well-motivated but partially assumption-driven interpretation, rather than a fully established structural dichotomy.","major_comments":[{"comment":"The low-v2/v1 branch, which is a central result of the paper, is not reproduced by the fiducial DALI models; Sec. 4.2 states that 'clearly none of these models reproduce the trends in the data.' The agreement in Fig. 12 is obtained only after manually removing the inner-rim contribution to the spectra (Sec. 4.3), which carries ≈ 40% of the v1 flux and ≈ 90% of the v2 flux. This subtraction is imposed rather than derived: Appendix F explicitly concedes that the CO abundance in the inner disk rim is probably overestimated by DALI and that the gas temperature and chemistry there are 'very uncertain,' and Fig. F.1 shows that the DALI kinetic network keeps CO abundant to higher temperatures than equilibrium chemistry. Because the absence of CO at the inner rim is the physical basis for the entire low-v2/v1 interpretation, the paper's central dichotomy currently rests on an unmodeled assumption rather than on a self-consistent model prediction.","section":"Sec. 4.2–4.3, Figs. 9 and 12"},{"comment":"The quantitative thresholds in the conclusions ('gas-to-dust ratios below 1000' and 'greater than 10000', 'cavity of at least 5 AU') are not obtained from a fit with uncertainties. The DALI grid samples gas-to-dust ratios of 10, 100, 1000, and 10000 and inner radii of 0.4–15 AU, and the models that match the data are simply those grid points that fall in the observed region. Moreover, the high-v2/v1 models in Appendix E require a specially shaped cavity wall with a Gaussian density profile inside Rin and gas-to-dust ratios of 2×104–105, parameters introduced ad hoc to lower the v1 flux by about a factor of 50 and to cool the wall. The paper should present these as proof-of-concept models and refrain from claiming tight physical thresholds unless a fitting or marginalization is performed.","section":"Sec. 4.1 and Appendix E"},{"comment":"The unique attribution of the low-v2/v1 inner-disk branch to thermal dissociation of CO near the sublimation radius is not established against competing mechanisms. In Sec. 3.1.3 the authors note that dust emission and absorption lower the v2 line more than the v1 line, so dust opacity can reduce v2/v1 by itself; in Sec. 5.1.1 they argue that even a small inner hole would leave atomic gas at the inner edge, but this is not modeled. The present data and models therefore cannot exclude alternatives such as a small dust-free gap, a slightly different inner-rim temperature structure, or dust opacity masking the rim. I would ask the authors to test at least one concrete alternative (e.g., rim models with dust extinction) and to strengthen the atomic-line observational test that they propose, before claiming that the dichotomy originates from CO dissociation.","section":"Sec. 5.1.1 and Sec. 3.1.3"}],"minor_comments":[{"comment":"The text says that none of the fiducial models reproduce the observed trends, but later in the same section and in the caption of Fig. 9 the LTE models with small cavities are described as consistent with low v2/v1 sources; this apparent contradiction should be clarified.","section":"Sec. 4.2.1"},{"comment":"The expression '1− exp−[ hν/kT ]' is missing a closing parenthesis or bracket; the equation is hard to read as typeset.","section":"Eq. (2)"},{"comment":"The description 'area between the blue and white lines' in Sec. 3.4.1 is not self-explanatory; the figure should label the blue lines or the caption should explain how the v1-flux constraint is encoded.","section":"Fig. 4"},{"comment":"The two solution families in Fig. 6 are called 'Solution #1' and 'Solution #2' in the text but are not labeled in the figure; please add labels or a legend.","section":"Sec. 3.4.2"},{"comment":"The statement that noise is added to achieve a signal-to-noise ratio of about 200 lacks details (noise distribution, seed); this is not essential but would improve reproducibility.","section":"Sec. 4.1"}],"recommendation":"major_revision","confidential_remarks":"The paper is within the scope of A&A and addresses an interesting observational dichotomy, and the slab-model analysis is clean. My main concern is that the central interpretation leans on a manual subtraction of the inner rim that the thermo-chemical model itself does not produce; I would support publication after the authors either add a self-consistent treatment of the inner-rim chemistry/cooling or substantially temper the claims, and after the model-fitting nature of the gas-to-dust thresholds is acknowledged."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Quick take: this is a serious modeling paper with a plausible two-family explanation for the puzzling v2/v1–radius trend in Herbig disks. It deserves a serious referee, but the central scenario is not yet established in detail.\n\nWhat is actually new: the explicit identification of two physical regimes—a warm, low-column inner disk surface with modest gas-to-dust ratios for compact CO emission, and a dust-depleted cavity wall with very high gas-to-dust ratios for extended emission—and the packaging of v2/v1 as a practical diagnostic for gas cavities and dust traps. The slab-model analysis is careful, the DALI grid is systematic, and the comparison to observed line profiles and fluxes is done with the same extraction pipeline. The simulated METIS images are a concrete and useful bonus. Credit also goes to the authors for being honest: Appendix F explicitly flags that DALI likely overestimates CO at the inner rim and that the gas temperature and chemistry there are very uncertain.\n\nThe soft spots are real and concentrated in two load-bearing steps. First, the fiducial DALI models fail to reproduce the low-v2/v1 sources, and the fix in Sec. 4.3 is to subtract the inner-rim contribution by hand. The physical justification—thermal dissociation of CO above ~3000 K at the dust sublimation front—is plausible but is not a prediction of the model; Fig. F.1 shows the kinetic network produces the molecular-to-atomic transition more slowly and at higher temperatures than equilibrium chemistry. If CO survives at the rim, the dichotomy is less clean and alternative excitation mechanisms might mimic the low ratios. Second, the high-v2/v1 branch requires gas-to-dust ratios above 10000 and specially shaped cavity walls (Appendix E). Those are fitted parameters, not independent predictions. The paper frames the threshold as a prediction, but it is post-hoc. That does not kill the paper, but it does mean the central scenario is not yet nailed down.\n\nWho this is for: anyone working on inner disk chemistry, CO diagnostics, or transition disk structure. The broad interpretation will likely survive, but the details need source-specific modeling and direct tests (atomic gas observations, METIS imaging). Bottom line: it deserves a serious referee and a conditional accept—the authors should be asked to turn the rim-dissociation assumption into a testable prediction and to show that the gas-to-dust thresholds are robust to model geometry.","headline":"A plausible two-family explanation for the CO v2/v1–radius trend, but the low-ratio branch rests on an inner-rim dissociation that the model itself does not predict.","tokens_in":36939,"tokens_out":2171,"would_cite":true,"duration_ms":25779,"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":"Two inner-disk configurations explain the CO vibrational dichotomy in Herbig Ae disks.","keywords":["protoplanetary disks","Herbig Ae stars","CO rovibrational emission","inner disk structure","gas-to-dust ratio","dust traps","disk cavities","thermo-chemical disk models"],"falsifier":"Search for CO rovibrational emission from the inner dust rim of a low-$v_2/v_1$ Herbig disk using high-angular-resolution 4.7 micron interferometry or spatially resolved spectroscopy: a detection of rim CO above the model's upper limits would break the thermal-dissociation premise.","tokens_in":2101,"feed_emoji":"🪐","tokens_out":6789,"duration_ms":130642,"temperature":0.7,"pith_summary":"The paper sets out to explain a counterintuitive observational pattern in planet-forming disks around Herbig Ae stars: CO gas emitting close to the star (within about 5 AU) shows a low vibrational excitation ratio, while CO gas emitting only at larger radii shows a high ratio. Using slab models and a grid of thermo-chemical disk models compared directly to observed line fluxes, profiles, and emitting radii, the authors argue that the two regimes require different physical structures. Low-ratio, small-radius emission lives in a warm (400–1300 K) inner disk surface with gas-to-dust ratios below 1000 and CO columns below $10^{18}$ cm$^{-2}$, with no CO at the dusty sublimation rim because the gas there is heated above 3000 K and dissociated. High-ratio, large-radius emission is explained by an inner cavity of at least 5 AU in both gas and dust, with a cool (100–300 K), dense CO reservoir at the cavity wall requiring gas-to-dust ratios above 10,000. If right, the line ratio becomes a practical probe of inner-disk substructures and of dust trapping that may be caused by giant planets.","feed_headline":"Two inner-disk layouts explain Herbig CO line split","feed_subtitle":"A vibration-line ratio in CO traces gas-to-dust ratios and planet-carved cavities in planet-forming disks.","key_machinery":"The load-bearing machinery is a two-step modeling chain. First, analytic slab models and non-LTE radiative-transfer slab models map the vibrational ratio $v_2/v_1$ (the flux ratio between CO lines from the second and first vibrationally excited states) as a function of temperature, CO column, density, and infrared radiation field; this isolates the allowed parameter regions. Second, a grid of thermo-chemical disk models with varying inner radius (0.4–15 AU) and gas-to-dust ratio (10–10000) produces synthetic CO line profiles, fluxes, and ratios that are analysed exactly as the observations, including removal of the inner-rim contribution to isolate disk-surface emission. The central identity used throughout is the peak line-surface-brightness relation for a slab, which shows how optical depth in the $v_1$ line pushes the $v_2/v_1$ ratio up with column, and the requirement that gas and dust temperatures remain coupled in the emitting layer.","core_discovery":"The central discovery claim is that the observed anti-correlation between the CO vibrational ratio $v_2/v_1$ and the emitting radius is not a smooth radial excitation gradient but a dichotomy between two inner-disk geometries. The paper argues that every low-$v_2/v_1$ source with CO inside 5 AU is best reproduced by emission from the disk surface at 400–1300 K, with moderate CO columns and gas-to-dust ratios below 1000, and that the inner dust rim contributes essentially no CO because gas at the sublimation radius is hot enough to be atomic. Every high-$v_2/v_1$ source with CO outside 5 AU requires a cavity in both gas and dust at least 5 AU wide, CO columns above $10^{18}$ cm$^{-2}$ at the cavity wall with gas-to-dust ratios above 10000, and a strong drop in gas surface density inside the cavity. In both regimes the CO excitation is nearly thermalized with the dust (roughly 20–50% temperature differences), and the high gas-to-dust ratios at cavity walls are interpreted as dust traps, consistent with planet-carved cavities.","pith_inferences":["Extending the paper's logic, the radius at which CO rovibrational emission switches on could serve as a thermometer for the atomic-to-molecular transition in the inner disk, mapping gas temperatures at the sublimation front across a sample of sources.","The observed dichotomy may be an evolutionary sequence: as a growing planet evacuates the inner disk, a source would move from the low-ratio surface-emission branch to the high-ratio cavity-wall branch, making narrow high-ratio lines a late-stage signature of planet carving.","A similar modeling exercise applied to T Tauri disks could test whether the difference in UV spectral shape (continuum-dominated versus Lyman-alpha-dominated irradiation) is what prevents the same dichotomy there, since the paper argues Herbigs dissociate CO at small radii while T Tauris do not."],"forward_implications":["For low-$v_2/v_1$ disks, CO rovibrational emission traces the disk surface rather than the inner rim, so the near-infrared continuum and CO lines probe different regions of the inner disk.","For high-$v_2/v_1$ disks, the measured CO radius directly marks the inner edge of the molecular outer disk, because the cavity is almost devoid of CO and the steep line profiles come from a cavity wall.","High gas-to-dust ratios above 10000 at cavity walls are naturally produced by dust traps, so the high-$v_2/v_1$ sources are prime candidates for giant planets carving the cavity.","Group II disks with broad low-ratio lines must be geometrically flat or radially confined by inner substructures; smooth flared disks are ruled out for most of them.","Observing CO rovibrational ratios can identify inner cavities and residual inner dust belts even without spatially resolving the disk, and future 3–5 micron integral-field observations should directly image the predicted rings and cavity walls."],"supporting_citations":[{"why":"Defines the three Herbig groups and provides the observed $v_2/v_1$, CO emitting radii, and NIR excess values that the models are anchored to.","marker":"Banzatti et al. (2018)"},{"why":"Establishes the measurement of $v_2/v_1$ from stacked $v_2$ P(4) and $v_1$ P(10) lines and the Keplerian HWHM-based emitting radius.","marker":"Banzatti & Pontoppidan (2015)"},{"why":"Reports 13CO rovibrational observations that indicate large CO columns around $10^{19}$ cm$^{-2}$, supporting the high-column cavity-wall solution.","marker":"van der Plas et al. (2015)"},{"why":"Describes the thermo-chemical disk model and CO chemistry and photodissociation used to compute the full-disk model grid.","marker":"Bruderer (2013)"},{"why":"Provides the non-LTE radiative-transfer slab code used to map excitation conditions for the vibrational ratio.","marker":"van der Tak et al. (2007)"},{"why":"Connects low stellar Fe abundances in low-NIR group I disks to accretion of dust-depleted gas, supporting the dust-trap interpretation.","marker":"Kama et al. (2015)"},{"why":"Resolves a sub-millimeter dust ring at about 6 AU in HD 142666; the paper shows that the CO emission is confined inside this ring.","marker":"Huang et al. (2018)"},{"why":"Documents gas cavities smaller than dust cavities in transition disks and attributes them to giant planets, the preferred explanation for the high-ratio sources.","marker":"van der Marel et al. (2016)"},{"why":"Provides the NIR excess measurements and imaging context used to separate group I and group II disks.","marker":"Garufi et al. (2017)"}],"fun_headline_variants":["CO line ratio reveals two inner disk architectures","Herbig disks split by CO vibration line ratios","Planet-shaped cavities and warm surfaces explain CO lines","Inner disk dichotomy: warm surface vs. dust-trap cavity","CO lines fingerprint two inner disk types around Herbig stars"],"cache_read_input_tokens":38912,"weakest_assumption_plain":"The argument depends on gas at the dust sublimation radius being hot enough (above roughly 3000 K) to destroy CO, and the paper itself says its model likely overestimates CO abundance there and that the temperature and chemistry are very uncertain.","fun_headline_variants_meta":{"raw":{"variants":["CO line ratio reveals two inner disk architectures","Herbig disks split by CO vibration line ratios","Planet-shaped cavities and warm surfaces explain CO lines","Inner disk dichotomy: warm surface vs. dust-trap cavity","CO lines fingerprint two inner disk types around Herbig stars"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000792,"raw_usage":{"total_tokens":3591,"prompt_tokens":1150,"completion_tokens":2441,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":766,"completion_tokens_details":{"reasoning_tokens":2365}},"tokens_in":766,"tokens_out":2441,"duration_ms":17475,"temperature":1.0,"reasoning_tokens":2365,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T05:01:34.894022+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Search for CO rovibrational emission from the inner dust rim of a low-$v_2/v_1$ Herbig disk using high-angular-resolution 4.7 micron interferometry or spatially resolved spectroscopy: a detection of rim CO above the model's upper limits would break the thermal-dissociation premise.","supporting_citations":[{"cited_title":"E., Waters , L","cited_arxiv_id":null,"evidence_quote":"Reports 13CO rovibrational observations that indicate large CO columns around $10^{19}$ cm$^{-2}$, supporting the high-column cavity-wall solution."},{"cited_title":"2013, , 559, A46","cited_arxiv_id":null,"evidence_quote":"Describes the thermo-chemical disk model and CO chemistry and photodissociation used to compute the full-disk model grid."}],"review_version":1}