{"id":"7c39a7de-58e6-4e3c-9fff-e23dad690ecb","arxiv_id":"1908.03267","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":5,"one_line_summary":"Spatially resolved CO isotopologue observations reveal order-of-magnitude radial variations in CO gas abundance in four protoplanetary disks, inconsistent with chemical processing alone.","lead":"This paper used ALMA observations of four planet-forming disks to show that the amount of gaseous CO varies with distance from the star by about a factor of ten. It argues that chemical reactions alone cannot explain the weak CO signals and that drifting, icy dust grains are likely involved.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Radial CO abundance pattern may partly reflect assumed gas surface density shape: R_c and outer taper are not tested, yet they set the model brightness profile against which abundance variations are inferred.","rationale":"The paper's headline claim is a radial variation of at least an order of magnitude in CO abundance. The derivation pipeline in Section 4.4 matches observed brightness to a single forward model with a prescribed, smooth gas surface density. Any error in the assumed density shape translates directly into the inferred abundance pattern, because the model's constant-abundance brightness profile is the baseline against which depletion factors are measured. The paper tests only gamma +/- 0.2 and shows the line brightness is not very sensitive to that parameter, but this does not cover the exponential taper radius R_c or the gas scale height. R_c is set from the CO emission itself for DM Tau, creating a circularity: the observed CO extent sets the density, and then the CO brightness is interpreted as abundance. A smaller R_c or a steeper taper would make the model fainter in the outer disk, potentially flattening the derived abundance gradient. The reader's concern about optical depth and fixed isotopologue ratios is valid and related (both are model-fidelity issues), but the density-shape degeneracy is more direct: it can alter the qualitative radial pattern, not just the absolute level. The proposed test, re-deriving depletion profiles across a grid of plausible density models, would settle whether the order-of-magnitude variation is robust. Since the paper is a first attempt and already conditional, I recommend keeping the CONDITIONAL verdict and adding this density-shape test as a necessary condition for acceptance.","tokens_in":20373,"tokens_out":14686,"duration_ms":164634,"concrete_test":"For each disk, recompute the C18O (and TW Hya 13CO) radial intensity profiles for a grid of gas surface density models with R_c multiplied by 0.5, 0.75, 1.5, and 2.0, and with alternative outer tapers (e.g., a steeper exponential exp[-(R/R_c)^3] or a hard cutoff at 0.8 R_c), keeping total gas mass fixed and re-fitting h0 and psi to the SED. Repeat the Section 4.4 depletion-factor fit for each model. If the inferred maximum-to-minimum CO abundance ratio within any disk falls below roughly a factor of 3 for any gas density profile still consistent with continuum and scattered-light constraints, the order-of-magnitude claim is not robust.","verdict_should_be":"UNCHANGED","load_bearing_attack":"Section 4.4 derives CO abundance versus radius by comparing observed C18O/13CO radial profiles to a single thermo-chemical model per disk with a fixed parametric gas surface density (eq. 1), then rescaling model CO abundance by a constant factor at each radius to match. The derived radial variation is therefore a residual after removing the model's assumed density profile. The paper tests only the power-law index gamma (+/-0.2, Section 4.3) and finds the line brightness changes by less than 30%, but the outer taper radius R_c and the gas scale height/flaring are not varied. R_c is often chosen directly from the CO emission extent (e.g., DM Tau R_c = 270 AU 'based on the extent of C18O (3-2) emission'; Table 4). If the true gas surface density falls more steeply or has a smaller R_c, the constant-abundance model is fainter in the outer disk, and the inferred 'increase of CO abundance with radius' in the outer regions would be reduced or vanish. The paper's statement that total gas mass normalization is not crucial is reasonable, but the shape of the density profile is crucial. Unless the gas surface density is independently constrained, the order-of-magnitude radial variation could be a systematic effect of the assumed density structure.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper uses spatially resolved ALMA C18O (and, for TW Hya, 13CO) observations of four protoplanetary disks — DM Tau, TW Hya, HD 163296, and IM Lup — together with the RAC2D thermo-chemical code and parametric disk structures to infer how the CO gas abundance in the warm molecular layer varies with radius. For each disk, the authors compute a standard constant-abundance chemical model, generate synthetic line images, and then rescale the model CO abundance by a radius-dependent constant factor to match the observed radial line brightness profiles. They report that CO is depleted by an order of magnitude or more relative to the interstellar ratio, that the depletion varies by at least an order of magnitude within individual disks, and that the radial depletion patterns in three of the four disks qualitatively resemble predictions from dust-evolution models that include pebble formation, settling, and drift. They also argue that chemical processing alone, even with a high cosmic-ray ionization rate, cannot explain the weak CO isotopologue emission because the depletion occurs too deep in the disk.","tokens_in":20659,"tokens_out":4382,"duration_ms":50856,"significance":"If the central claim holds, it would be an important result: constant-abundance CO-based estimates of disk mass, disk radius, and gas-to-dust ratio would be systematically biased, and the derived radial depletion profiles would provide a new observational constraint on dust evolution and volatile transport. The paper has notable strengths: a homogeneous four-disk framework, detailed physical-chemical modeling, explicit tests of the cosmic-ray ionization rate and of the surface-density power-law index, and a clear statement that the total gas mass normalization is not the dominant uncertainty. The use of optically thin C18O and 13CO lines is appropriate in principle, and the comparison to Krijt et al. (2018) is useful even if qualitative. However, the derivation rests on several assumptions — the shape of the gas surface density profile, optically thin emission, and fixed isotopologue ratios — that are not fully stress-tested, and the key Figure 8 presents depletion factors without propagated uncertainties. These issues are addressable, but they are load-bearing for the order-of-magnitude claim.","major_comments":[{"comment":"The radial CO abundance profile is derived by comparing observed line brightness to a single model per disk with a fixed parametric gas surface density shape. The paper varies only the power-law index γ by ±0.2 (§4.3) and keeps the characteristic radius R_c and the outer exponential taper fixed. This is a concern because the inversion compares I_CO to the model's assumed gas column: any mismatch in the assumed density shape is absorbed into the inferred abundance variation. The concern is concrete for DM Tau, where R_c = 270 AU is set 'based on the extent of C18O (3-2) emission' (Table 4). If the true gas surface density has a smaller R_c or a steeper outer taper, the constant-abundance model is fainter in the outer disk, and the inferred rise of CO abundance with radius would be reduced or disappear. The statement that total gas mass normalization is not crucial does not address the shape of the profile. Please vary R_c and the outer taper (or justify them with independent constraints) and quantify how much of the inferred order-of-magnitude variation in Figure 8 survives.","section":"§4.4, eq. (1), Table 4"},{"comment":"The inversion assumes that the observed C18O surface brightness is proportional to the CO column in the emitting layer, and that fixed interstellar isotopologue ratios (CO/13CO = 69, CO/C18O = 570) apply at all radii. If C18O emission is not fully optically thin in the regions where depletion factors are derived, or if isotope-selective photodissociation changes the local C18O/CO ratio, the derived abundance variation would partly be a brightness variation. The text cites previous work supporting the optical thinness of TW Hya 13CO beyond 70 AU, but for the other disks the C18O optical depth is asserted rather than demonstrated. Please include a per-radius check — for example, model optical depth profiles or a comparison of C18O J = 2-1 and 3-2 where available — to verify that I_line ∝ N_CO holds at the radii used in Figure 8.","section":"§3.3 and §4.4"},{"comment":"The central quantitative claim — at least an order-of-magnitude radial variation in CO abundance — is presented without propagated uncertainties. The observed radial profiles in Figure 7 have 1σ error bars and the flux calibration uncertainty is stated as 10%, but the depletion factors in Figure 8 have no error bars or shaded ranges. The γ test in §4.3 shows that line brightness changes by less than 30%, but a 30% brightness change can translate into a substantially larger change in depletion factor because the mapping is nonlinear and because the depletion factor is normalized to a model that itself depends on other disk parameters. Please add uncertainty ranges to the four panels of Figure 8, or otherwise quantify how robust the radial gradient is to noise, calibration, and the adopted disk structure parameters.","section":"Figure 8 and §4.4"}],"minor_comments":[{"comment":"The caption contains placeholder text ('Lorem ipsum') and a stray fragment ('ebea'); these should be removed before publication.","section":"Figure 8 caption"},{"comment":"There are typos: 'V ariations' in the title line should be 'Variations', and 'Shakura-Sunyeav' in §5.1.3 should be 'Shakura-Sunyaev'.","section":"Abstract and §5.1.3"},{"comment":"The row labeled 'H100 (AU)' lists no numerical values; the adopted gas scale heights at 100 AU should be given explicitly for all four disks.","section":"Table 4"},{"comment":"The text says 'φ from 1 to 1.3' when describing the flaring-parameter grid; this should be ψ to match eq. (3).","section":"§3.3"},{"comment":"The statement that observed depletions are 'a factor of few higher' than the Krijt et al. (2018) simulations is vague; please give the actual ranges (for example, factor of 10 or larger observed versus factor of 2-6 predicted).","section":"§5.1.2"}],"recommendation":"major_revision","confidential_remarks":"The paper is a solid observational-theoretical study and the central claim is plausible, but the inversion's sensitivity to the assumed gas surface-density shape is a real correctness risk that should be tested before publication. The lack of uncertainties in Figure 8 is also important for a quantitative claim of this kind. I do not see grounds for rejection; the requested checks are within the scope of a revision."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Two things to know up front. This is the first paper to derive radially resolved CO abundance profiles for four disks in one homogeneous framework, and the central finding—order-of-magnitude radial variation—is likely real in outline, though the exact profiles depend on assumptions that are not fully tested. Second, the draft is unfinished: the Figure 8 caption contains \"Lorem ipsum\" placeholder text.\n\nWhat's genuinely new: earlier work established disk-averaged CO depletion and predicted radial patterns from chemistry and dust evolution. Zhang et al. go beyond averages. They fit C18O (and 13CO for TW Hya) radial brightness profiles with thermo-chemical models and show that chemical processing alone, even with an ISM-level cosmic-ray rate, depletes mostly the deep layer and barely changes the observable line surface brightness. The comparison with Krijt et al. (2018) dust-evolution predictions is qualitative and involves co-authors, but it is not circular: the predictions are not used to fit the observations. The paper gives the community a genuinely new observational constraint on volatile transport and planet-forming material.\n\nWhere I'm less comfortable. The inversion in Section 4.4 rescales model CO by a constant factor at each radius and assumes the line brightness traces CO column in the optically thin layer. That is a reasonable starting point for C18O beyond roughly 30 AU, but the fixed isotopologue ratios and the neglect of isotope-selective photodissociation could mimic part of the apparent outer-disk depletion. More importantly, the gas surface density shape is not varied. They test gamma by ±0.2, but not R_c or the outer taper; for DM Tau, R_c is set from the C18O extent, which is the same emission used to infer the abundance profile. A more compact gas disk would reduce the inferred outward abundance rise. I don't think this kills the central claim—the mid-disk depletion is large and common to all four disks—but the amplitude of the order-of-magnitude statement is uncertain until that degeneracy is explored. The depletion factors in Figure 8 also lack propagated uncertainties, so it's hard to know how significant the radial variation really is.\n\nBottom line: serious, useful work with a real new measurement and several systematics that need airing. I'd send it to a good referee rather than desk-reject, mainly to force the R_c/taper and uncertainty discussion. The placeholder caption has to go.","headline":"First radially resolved CO abundance profiles in four disks show order-of-magnitude variation, but the size of the claim depends on an untested gas surface density shape and an unfinished plot caption.","tokens_in":21155,"tokens_out":3716,"would_cite":true,"duration_ms":39264,"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":"This paper finds that CO gas abundance in planet-forming disks varies by at least an order of magnitude with radius, so constant-abundance CO-based disk mass and radius estimates are biased.","keywords":["CO abundance","protoplanetary disks","CO isotopologues","radial abundance variation","dust evolution","pebble drift","astrochemistry","disk gas mass"],"falsifier":"Resolved maps of a rarer CO isotopologue such as C17O, using a known isotopologue ratio, would settle the claim: if C17O emission falls as steeply as the rescaled C18O models require, the abundance variation is real; if C17O tracks a constant interstellar ratio, the depletion is an optical-depth artifact. Alternatively, resolved HD line observations of true gas surface density at the same annuli would directly compare gas column to CO-derived column.","tokens_in":2057,"feed_emoji":"🪐","tokens_out":3744,"duration_ms":103831,"temperature":0.7,"pith_summary":"Carbon monoxide (CO) is the standard tracer for gas in planet-forming disks, but its abundance is usually assumed constant at the interstellar ratio. This study compares spatially resolved CO-isotopologue images of four disks with thermo-chemical models to derive how the CO gas abundance changes with distance from the star. It finds a variation of at least an order of magnitude within each disk, with a common pattern of strong depletion in an intermediate zone and a recovery farther out. It also shows that chemistry alone puts the depletion too deep to explain the faint lines, while dust growth and the inward drift of icy grains naturally produce the observed radial shape. If correct, constant-abundance CO measurements of disk mass, radius, and gas-to-dust ratio are systematically biased.","feed_headline":"CO gas in planet-forming disks swings tenfold with radius","feed_subtitle":"Constant interstellar CO ratios mislead disk mass and radius estimates; dust evolution may drive the pattern.","key_machinery":"The method is an abundance-rescaling comparison: for each disk a two-dimensional thermo-chemical model with time-dependent chemistry and line radiative transfer is calibrated to known disk and stellar parameters, then the model's CO abundance is multiplied by a radius-independent rescaling factor for each annulus until the predicted C18O line surface brightness matches the observed deprojected radial profile; that factor defines the depletion. The interpretive machinery is pebble transport: CO freezes outside the mid-plane snowline, icy grains grow, settle, and drift inward, releasing vapor inside the snowline and leaving a depleted intermediate zone that recovers in the outer disk. A second candidate, cosmic-ray-driven chemical conversion, is tested but fails because the depletion occurs too deep.","core_discovery":"The central claim is that the CO gas abundance in the warm molecular layer is not constant but varies strongly with radius in all four disks: DM Tau, TW Hya, HD 163296, and IM Lup. The derived profiles show an order-of-magnitude range within each disk, with TW Hya's outer disk depleted by factors of 100-300. The paper compares these profiles with two candidate mechanisms. Chemical processing with an interstellar cosmic-ray ionization rate can convert enough CO, but mostly in the deep disk, leaving the observable layer nearly unchanged. Dust evolution, in which CO-ice-coated grains grow, settle, and drift inward, predicts the right radial pattern - modest enhancement inside the snowline, peak depletion just beyond it, and recovery outward - but underestimates the depletion magnitude. The paper concludes that a coupled chemical-dynamical treatment is needed and that the CO abundance profile is a record of volatile transport.","pith_inferences":["If the derived profiles are correct, resolved images of a rarer isotopologue such as C17O should show the same radial swing; if not, the C18O depletion would be partly an optical-depth artifact.","A testable extension is that disks with more evolved dust, inferred from stronger millimeter continuum or larger grain sizes, should show deeper CO depletion in the intermediate zone with the depletion peak shifting outward.","Comparing other carbon-bearing molecules, such as C2H, at the same radii could reveal whether the pattern reflects a general elemental carbon depletion rather than a CO-specific process.","Resolved hydrogen-deuteride gas surface density measurements would convert brightness-derived depletion factors into true abundances."],"forward_implications":["Disk masses and radii derived from CO lines under a constant abundance will carry a radius-dependent systematic error, not just a uniform offset.","Radial CO depletion profiles become a practical probe of dust evolution: the location and depth of the depletion zone locate where pebble growth and inward drift have stripped volatile ice.","Predictions for planet-forming volatile inventories must couple chemistry with the dynamical transport of icy grains, since either alone underpredicts the depletion.","The extreme outer CO deficit in TW Hya implies either severe CO removal or actual gas depletion beyond 100 au, which changes interpretations of its outer disk.","In HD 163296 the CO enhancement just inside its snowline is a signpost for icy pebbles crossing the snowline and vaporizing there."],"supporting_citations":[{"why":"Supplies the thermo-chemical disk model used to compute CO abundance structures and synthetic line emission.","marker":"Du & Bergin 2014"},{"why":"Shows cosmic-ray-driven CO processing and its dependence on grain surface area; provides the depletion levels compared with observations.","marker":"Schwarz et al. 2018"},{"why":"Calculates CO processing timescales and the temperature range where depletion is most efficient.","marker":"Bosman et al. 2018"},{"why":"Establishes the vertical settling and mixing mechanism that depletes CO vapor in the warm molecular layer.","marker":"Krijt et al. 2016"},{"why":"Provides the two-dimensional dust-evolution simulation whose predicted radial CO depletion pattern is compared with the observed profiles.","marker":"Krijt et al. 2018"},{"why":"Supplies the local ISM isotopologue ratios used to rescale model CO abundances to 13CO and C18O.","marker":"Wilson 1999"},{"why":"Gives the hydrogen-deuteride based gas mass estimate for TW Hya showing CO is depleted while hydrogen gas is abundant.","marker":"Bergin et al. 2013"},{"why":"Provides an independent gas mass and CO depletion measurement for DM Tau used in the sample context.","marker":"McClure et al. 2016"},{"why":"Supplies the comprehensive model parameters for IM Lup and a low cosmic-ray ionization constraint used in the modeling.","marker":"Cleeves et al. 2016"},{"why":"Shows that TW Hya 13CO (3-2) is optically thin beyond 70 AU, justifying the use of 13CO emission there.","marker":"Schwarz et al. 2016"}],"fun_headline_variants":["CO gas abundance swings tenfold across protoplanetary disks","Disk CO radial profile: order-of-magnitude variations","CO abundance varies 10x with radius in planet-forming disks","Radial CO abundance changes: tenfold swings in disks","CO gas not uniform: tenfold radial variations in disks"],"cache_read_input_tokens":23296,"weakest_assumption_plain":"The derivation assumes that C18O emission (and TW Hya 13CO beyond 70 AU) is optically thin and that multiplying the model's CO abundance by a constant factor at each radius accurately represents the true abundance; if optical depth or isotope-selective photodissociation contributes to the faint emission, the derived order-of-magnitude radial variation could be partly a brightness effect.","fun_headline_variants_meta":{"raw":{"variants":["CO gas abundance swings tenfold across protoplanetary disks","Disk CO radial profile: order-of-magnitude variations","CO abundance varies 10x with radius in planet-forming disks","Radial CO abundance changes: tenfold swings in disks","CO gas not uniform: tenfold radial variations in disks"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000587,"raw_usage":{"total_tokens":2774,"prompt_tokens":979,"completion_tokens":1795,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":595,"completion_tokens_details":{"reasoning_tokens":1723}},"tokens_in":595,"tokens_out":1795,"duration_ms":12662,"temperature":1.0,"reasoning_tokens":1723,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T14:19:29.211517+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Resolved maps of a rarer CO isotopologue such as C17O, using a known isotopologue ratio, would settle the claim: if C17O emission falls as steeply as the rescaled C18O models require, the abundance variation is real; if C17O tracks a constant interstellar ratio, the depletion is an optical-depth artifact. Alternatively, resolved HD line observations of true gas surface density at the same annuli would directly compare gas column to CO-derived column.","supporting_citations":[{"cited_title":"A., Cleeves, L","cited_arxiv_id":null,"evidence_quote":"Supplies the thermo-chemical disk model used to compute CO abundance structures and synthetic line emission."},{"cited_title":"R., Bergin, E","cited_arxiv_id":null,"evidence_quote":"Provides the two-dimensional dust-evolution simulation whose predicted radial CO depletion pattern is compared with the observed profiles."},{"cited_title":"I., ¨Oberg, K","cited_arxiv_id":null,"evidence_quote":"Supplies the comprehensive model parameters for IM Lup and a low cosmic-ray ionization constraint used in the modeling."},{"cited_title":"R., Bergin, E","cited_arxiv_id":null,"evidence_quote":"Shows that TW Hya 13CO (3-2) is optically thin beyond 70 AU, justifying the use of 13CO emission there."}],"review_version":1}